First 3 patches are more-or-less cleanups/preparations.
Patches 4/5 are fixes for netns file descriptors leaks/open.
Patch 6 was sent to me/contributed off-list by Mohammad, who wants 32-bit
kernels to run TCP-AO.
Patch 7 is a workaround/fix for slow VMs. Albeit, I can't reproduce
the issue, but I hope it will fix netdev flakes for connect-deny-*
tests.
And the biggest change is adding TCP-AO tracepoints to selftests.
I think it's a good addition by the following reasons:
- The related tracepoints are now tested;
- It allows tcp-ao selftests to raise expectations on the kernel
behavior - up from the syscalls exit statuses + net counters.
- Provides tracepoints usage samples.
As tracepoints are not a stable ABI, any kernel changes done to them
will be reflected to the selftests, which also will allow users
to see how to change their code. It's quite better than parsing dmesg
(what BGP was doing pre-tracepoints, ugh).
Somewhat arguably, the code parses trace_pipe, rather than uses
libtraceevent (which any sane user should do). The reason behind that is
the same as for rt-netlink macros instead of libmnl: I'm trying
to minimize the library dependencies of the selftests. And the
performance of formatting text in kernel and parsing it again in a test
is not critical.
Current output sample:
> ok 73 Trace events matched expectations: 13 tcp_hash_md5_required[2] tcp_hash_md5_unexpected[4] tcp_hash_ao_required[3] tcp_ao_key_not_found[4]
Previously, tracepoints selftests were part of kernel tcp tracepoints
submission [1], but since then the code was quite changed:
- Now generic tracing setup is in lib/ftrace.c, separate from
lib/ftrace-tcp.c which utilizes TCP trace points. This separation
allows future selftests to trace non-TCP events, i.e. to find out
an skb's drop reason, which was useful in the creation of TCP-CLOSE
stress-test (not in this patch set, but used in attempt to reproduce
the issue from [2]).
- Another change is that in the previous submission the trace events
where used only to detect unexpected TCP-AO/TCP-MD5 events. In this
version the selftests will fail if an expected trace event didn't
appear.
Let's see how reliable this is on the netdev bot - it obviously passes
on my testing, but potentially may require a temporary XFAIL patch
if it misbehaves on a slow VM.
[1] https://lore.kernel.org/lkml/20240224-tcp-ao-tracepoints-v1-0-15f31b7f30a7@…
[2] https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net.git/commit/?id=3…
Signed-off-by: Dmitry Safonov <0x7f454c46(a)gmail.com>
---
Changes in v3:
- Corrected the selftests printing of tcp header flags, parsed from
trace points
- Fixed an issue with VRF kconfig checks (and tests)
- Made check for unexpected trace events XFAIL, yet looking into the
reason behind the fail
- Link to v2: https://lore.kernel.org/r/20240802-tcp-ao-selftests-upd-6-12-v2-0-370c99358…
Changes in v2:
- Fixed two issues with parsing TCP-AO events: the socket state and TCP
segment flags. Hopefully, won't fail on netdev.
- Reword patch 1 & 2 messages to be more informative and at some degree
formal (Paolo)
- Since commit e33a02ed6a4f ("selftests: Add printf attribute to
kselftest prints") it's possible to use __printf instead of "raw" gcc
attribute - switch using that, as checkpatch suggests.
- Link to v1: https://lore.kernel.org/r/20240730-tcp-ao-selftests-upd-6-12-v1-0-ffd4bf15d…
---
Dmitry Safonov (7):
selftests/net: Clean-up double assignment
selftests/net: Provide test_snprintf() helper
selftests/net: Be consistent in kconfig checks
selftests/net: Open /proc/thread-self in open_netns()
selftests/net: Don't forget to close nsfd after switch_save_ns()
selftests/net: Synchronize client/server before counters checks
selftests/net: Add trace events matching to tcp_ao
Mohammad Nassiri (1):
selftests/tcp_ao: Fix printing format for uint64_t
tools/testing/selftests/net/tcp_ao/Makefile | 3 +-
tools/testing/selftests/net/tcp_ao/bench-lookups.c | 2 +-
tools/testing/selftests/net/tcp_ao/config | 1 +
tools/testing/selftests/net/tcp_ao/connect-deny.c | 25 +-
tools/testing/selftests/net/tcp_ao/connect.c | 6 +-
tools/testing/selftests/net/tcp_ao/icmps-discard.c | 2 +-
.../testing/selftests/net/tcp_ao/key-management.c | 18 +-
tools/testing/selftests/net/tcp_ao/lib/aolib.h | 176 ++++++-
.../testing/selftests/net/tcp_ao/lib/ftrace-tcp.c | 549 +++++++++++++++++++++
tools/testing/selftests/net/tcp_ao/lib/ftrace.c | 466 +++++++++++++++++
tools/testing/selftests/net/tcp_ao/lib/kconfig.c | 31 +-
tools/testing/selftests/net/tcp_ao/lib/setup.c | 17 +-
tools/testing/selftests/net/tcp_ao/lib/sock.c | 1 -
tools/testing/selftests/net/tcp_ao/lib/utils.c | 26 +
tools/testing/selftests/net/tcp_ao/restore.c | 30 +-
tools/testing/selftests/net/tcp_ao/rst.c | 2 +-
tools/testing/selftests/net/tcp_ao/self-connect.c | 19 +-
tools/testing/selftests/net/tcp_ao/seq-ext.c | 28 +-
.../selftests/net/tcp_ao/setsockopt-closed.c | 6 +-
tools/testing/selftests/net/tcp_ao/unsigned-md5.c | 35 +-
20 files changed, 1376 insertions(+), 67 deletions(-)
---
base-commit: a9c60712d71ff07197b2982899b9db28ed548ded
change-id: 20240730-tcp-ao-selftests-upd-6-12-4d3e53a74f3f
Best regards,
--
Dmitry Safonov <0x7f454c46(a)gmail.com>
From: Hao Ge <gehao(a)kylinos.cn>
Smatch reported the following warning:
./tools/testing/selftests/bpf/testing_helpers.c:455 get_xlated_program()
warn: variable dereferenced before check 'buf' (see line 454)
It seems correct,so let's modify it based on it's suggestion.
Actually,commit b23ed4d74c4d ("selftests/bpf: Fix invalid pointer
check in get_xlated_program()") fixed an issue in the test_verifier.c
once,but it was reverted this time.
Let's solve this issue with the minimal changes possible.
Reported-by: Dan Carpenter <dan.carpenter(a)linaro.org>
Closes: https://lore.kernel.org/all/1eb3732f-605a-479d-ba64-cd14250cbf91@stanley.mo…
Fixes: b4b7a4099b8c ("selftests/bpf: Factor out get_xlated_program() helper")
Signed-off-by: Hao Ge <gehao(a)kylinos.cn>
---
tools/testing/selftests/bpf/testing_helpers.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/testing/selftests/bpf/testing_helpers.c b/tools/testing/selftests/bpf/testing_helpers.c
index d5379a0e6da8..34dfea295c8e 100644
--- a/tools/testing/selftests/bpf/testing_helpers.c
+++ b/tools/testing/selftests/bpf/testing_helpers.c
@@ -451,7 +451,7 @@ int get_xlated_program(int fd_prog, struct bpf_insn **buf, __u32 *cnt)
*cnt = xlated_prog_len / buf_element_size;
*buf = calloc(*cnt, buf_element_size);
- if (!buf) {
+ if (!*buf) {
perror("can't allocate xlated program buffer");
return -ENOMEM;
}
--
2.25.1
This small series includes fixes for creation of veth pairs for
networkless kernels & adds tests for turning the different network
interface features on and off in selftests/net/netdevice.sh script.
Tested using vng and compiles for network as well as networkless kernel.
Changes in v9:
Removed veth1 to avoid redundant testing as per feedback in v8.
Changes in v8:
https://lore.kernel.org/all/20240819121235.39514-1-jain.abhinav177@gmail.co…
Remove redundant variable initial_state as per feedback in v7.
Changes in v7:
https://lore.kernel.org/all/20240815105924.1389290-1-jain.abhinav177@gmail.…
Create a third patch in the series to do SKIP -> XFAIL replacement.
Add logic to incorporate XFAIL on setting IP address for veth pair.
Changes in v6:
https://lore.kernel.org/all/20240814191517.50466-1-jain.abhinav177@gmail.com
Use XFAIL for ethtool operations that are unsupported instead of SKIP.
Changes in v5:
https://lore.kernel.org/all/20240808122452.25683-1-jain.abhinav177@gmail.com
Rectify the syntax for ip add link.
Fix the veth_created condition check.
Changes in v4:
https://lore.kernel.org/all/20240807175717.7775-1-jain.abhinav177@gmail.com
Move veth creation/removal to the main shell script.
Tested using vng on a networkless kernel and the script works, sample
output below the changes.
Changes in v3:
https://lore.kernel.org/all/20240614113240.41550-1-jain.abhinav177@gmail.com
Add a check for netdev, create veth pair for testing.
Restore feature to its initial state.
Changes in v2:
https://lore.kernel.org/all/20240609132124.51683-1-jain.abhinav177@gmail.com
Remove tail usage; use read to parse the features from temp file.
v1:
https://lore.kernel.org/all/20240606212714.27472-1-jain.abhinav177@gmail.com
```
# selftests: net: netdevice.sh
# No valid network device found, creating veth pair
# PASS: veth0: set interface up
# PASS: veth0: set MAC address
# XFAIL: veth0: set IP address unsupported for veth*
# PASS: veth0: ethtool list features
# PASS: veth0: Turned off feature: rx-checksumming
# PASS: veth0: Turned on feature: rx-checksumming
# PASS: veth0: Restore feature rx-checksumming to initial state on
# Actual changes:
# tx-checksum-ip-generic: off
...
# PASS: veth0: Turned on feature: rx-udp-gro-forwarding
# PASS: veth0: Restore feature rx-udp-gro-forwarding to initial state off
# Cannot get register dump: Operation not supported
# XFAIL: veth0: ethtool dump not supported
# PASS: veth0: ethtool stats
# PASS: veth0: stop interface
```
Abhinav Jain (3):
selftests: net: Create veth pair for testing in networkless kernel
selftests: net: Add on/off checks for non-fixed features of interface
selftests: net: Use XFAIL for operations not supported by the driver
tools/testing/selftests/net/netdevice.sh | 60 ++++++++++++++++++++++--
1 file changed, 56 insertions(+), 4 deletions(-)
--
2.34.1
Uses 'loop_until' to wait for the atomic replace to unload all previous
livepatches, as on some machines with a large number of CPUs there is a
sizable delay between the atomic replace ocurring and when sysfs
updates accordingly.
Signed-off-by: Ryan Sullivan <rysulliv(a)redhat.com>
---
tools/testing/selftests/livepatch/test-livepatch.sh | 7 ++-----
1 file changed, 2 insertions(+), 5 deletions(-)
diff --git a/tools/testing/selftests/livepatch/test-livepatch.sh b/tools/testing/selftests/livepatch/test-livepatch.sh
index 65c9c058458d..bd13257bfdfe 100755
--- a/tools/testing/selftests/livepatch/test-livepatch.sh
+++ b/tools/testing/selftests/livepatch/test-livepatch.sh
@@ -139,11 +139,8 @@ load_lp $MOD_REPLACE replace=1
grep 'live patched' /proc/cmdline > /dev/kmsg
grep 'live patched' /proc/meminfo > /dev/kmsg
-mods=(/sys/kernel/livepatch/*)
-nmods=${#mods[@]}
-if [ "$nmods" -ne 1 ]; then
- die "Expecting only one moduled listed, found $nmods"
-fi
+loop_until 'mods=(/sys/kernel/livepatch/*); nmods=${#mods[@]}; [[ "$nmods" -eq 1 ]]' ||
+ die "Expecting only one moduled listed, found $nmods"
# These modules were disabled by the atomic replace
for mod in $MOD_LIVEPATCH3 $MOD_LIVEPATCH2 $MOD_LIVEPATCH1; do
--
2.44.0
v1: https://patchwork.kernel.org/project/linux-kselftest/list/?series=881867
v2: keep numerical values (David)
Cc: David Gow <davidgow(a)google.com>
Cc: Rae Moar <rmoar(a)google.com>
Michal Wajdeczko (3):
kunit: Improve format of the NOT_ERR_OR_NULL assertion
kunit: Improve format of the PTR_EQ|NE|NULL assertion
kunit: Improve format of the KUNIT_EXPECT_EQ assertion
lib/kunit/assert.c | 71 +++++++++++++++++++++++++++++++++++++---------
1 file changed, 57 insertions(+), 14 deletions(-)
--
2.43.0
This patch series is motivated by the following observation:
Raise a signal, jump to signal handler. The ucontext_t structure dumped
by kernel to userspace has a uc_sigmask field having the mask of blocked
signals. If you run a fresh minimalistic program doing this, this field
is empty, even if you block some signals while registering the handler
with sigaction().
Here is what the man-pages have to say:
sigaction(2): "sa_mask specifies a mask of signals which should be blocked
(i.e., added to the signal mask of the thread in which the signal handler
is invoked) during execution of the signal handler. In addition, the
signal which triggered the handler will be blocked, unless the SA_NODEFER
flag is used."
signal(7): Under "Execution of signal handlers", (1.3) implies:
"The thread's current signal mask is accessible via the ucontext_t
object that is pointed to by the third argument of the signal handler."
But, (1.4) states:
"Any signals specified in act->sa_mask when registering the handler with
sigprocmask(2) are added to the thread's signal mask. The signal being
delivered is also added to the signal mask, unless SA_NODEFER was
specified when registering the handler. These signals are thus blocked
while the handler executes."
There clearly is no distinction being made in the man pages between
"Thread's signal mask" and ucontext_t; this logically should imply
that a signal blocked by populating struct sigaction should be visible
in ucontext_t.
Here is what the kernel code does (for Aarch64):
do_signal() -> handle_signal() -> sigmask_to_save(), which returns
¤t->blocked, is passed to setup_rt_frame() -> setup_sigframe() ->
__copy_to_user(). Hence, ¤t->blocked is copied to ucontext_t
exposed to userspace. Returning back to handle_signal(),
signal_setup_done() -> signal_delivered() -> sigorsets() and
set_current_blocked() are responsible for using information from
struct ksignal ksig, which was populated through the sigaction()
system call in kernel/signal.c:
copy_from_user(&new_sa.sa, act, sizeof(new_sa.sa)),
to update ¤t->blocked; hence, the set of blocked signals for the
current thread is updated AFTER the kernel dumps ucontext_t to
userspace.
Assuming that the above is indeed the intended behaviour, because it
semantically makes sense, since the signals blocked using sigaction()
remain blocked only till the execution of the handler, and not in the
context present before jumping to the handler (but nothing can be
confirmed from the man-pages), the series introduces a test for
mangling with uc_sigmask. I will send a separate series to fix the
man-pages.
The proposed selftest has been tested out on Aarch32, Aarch64 and x86_64.
v5->v6:
- Drop renaming of sas.c
- Include the explanation from the cover letter in the changelog
for the second patch
v4->v5:
- Remove a redundant print statement
v3->v4:
- Allocate sigsets as automatic variables to avoid malloc()
v2->v3:
- ucontext describes current state -> ucontext describes interrupted context
- Add a comment for blockage of USR2 even after return from handler
- Describe blockage of signals in a better way
v1->v2:
- Replace all occurrences of SIGPIPE with SIGSEGV
- Fixed a mismatch between code comment and ksft log
- Add a testcase: Raise the same signal again; it must not be queued
- Remove unneeded <assert.h>, <unistd.h>
- Give a detailed test description in the comments; also describe the
exact meaning of delivered and blocked
- Handle errors for all libc functions/syscalls
- Mention tests in Makefile and .gitignore in alphabetical order
v1:
- https://lore.kernel.org/all/20240607122319.768640-1-dev.jain@arm.com/
Dev Jain (2):
selftests: Rename sigaltstack to generic signal
selftests: Add a test mangling with uc_sigmask
tools/testing/selftests/Makefile | 2 +-
.../{sigaltstack => signal}/.gitignore | 1 +
.../{sigaltstack => signal}/Makefile | 3 +-
.../current_stack_pointer.h | 0
.../selftests/signal/mangle_uc_sigmask.c | 184 ++++++++++++++++++
.../selftests/{sigaltstack => signal}/sas.c | 0
6 files changed, 188 insertions(+), 2 deletions(-)
rename tools/testing/selftests/{sigaltstack => signal}/.gitignore (70%)
rename tools/testing/selftests/{sigaltstack => signal}/Makefile (56%)
rename tools/testing/selftests/{sigaltstack => signal}/current_stack_pointer.h (100%)
create mode 100644 tools/testing/selftests/signal/mangle_uc_sigmask.c
rename tools/testing/selftests/{sigaltstack => signal}/sas.c (100%)
--
2.30.2
A few tests check if nettest exists in the $PATH before adding
$PWD to $PATH and re-checking. They don't discard stderr on
the first check (and nettest is built as part of selftests,
so it's pretty normal for it to not be available in system $PATH).
This leads to output noise:
which: no nettest in (/home/virtme/tools/fs/bin:/home/virtme/tools/fs/sbin:/home/virtme/tools/fs/usr/bin:/home/virtme/tools/fs/usr/sbin:/bin:/sbin:/usr/bin:/usr/sbin:/usr/local/bin)
Add a common helper for the check which does silence stderr.
There is another small functional change hiding here, because pmtu.sh
and fib_rule_tests.sh used to return from the test case rather than
completely exit. Building nettest is not hard, there should be no need
to maintain the ability to selectively skip cases in its absence.
Reviewed-by: Ido Schimmel <idosch(a)nvidia.com>
Signed-off-by: Jakub Kicinski <kuba(a)kernel.org>
---
v3:
- delete accidentally added profiling
v2: https://lore.kernel.org/20240820004217.1087392-1-kuba@kernel.org
- fold in the changes from Ido
v1: https://lore.kernel.org/20240817183848.658443-1-kuba@kernel.org
CC: shuah(a)kernel.org
CC: linux-kselftest(a)vger.kernel.org
CC: Ido Schimmel <idosch(a)idosch.org>
---
tools/testing/selftests/net/fcnal-test.sh | 9 +----
tools/testing/selftests/net/fib_rule_tests.sh | 37 +------------------
tools/testing/selftests/net/lib.sh | 15 ++++++++
tools/testing/selftests/net/pmtu.sh | 8 +---
.../selftests/net/unicast_extensions.sh | 9 +----
.../selftests/net/vrf_route_leaking.sh | 3 +-
6 files changed, 22 insertions(+), 59 deletions(-)
diff --git a/tools/testing/selftests/net/fcnal-test.sh b/tools/testing/selftests/net/fcnal-test.sh
index 386ebd829df5..899dbad0104b 100755
--- a/tools/testing/selftests/net/fcnal-test.sh
+++ b/tools/testing/selftests/net/fcnal-test.sh
@@ -4304,14 +4304,7 @@ elif [ "$TESTS" = "ipv6" ]; then
TESTS="$TESTS_IPV6"
fi
-# nettest can be run from PATH or from same directory as this selftest
-if ! which nettest >/dev/null; then
- PATH=$PWD:$PATH
- if ! which nettest >/dev/null; then
- echo "'nettest' command not found; skipping tests"
- exit $ksft_skip
- fi
-fi
+check_gen_prog "nettest"
declare -i nfail=0
declare -i nsuccess=0
diff --git a/tools/testing/selftests/net/fib_rule_tests.sh b/tools/testing/selftests/net/fib_rule_tests.sh
index 89034c5b69dc..53c5c1ad437e 100755
--- a/tools/testing/selftests/net/fib_rule_tests.sh
+++ b/tools/testing/selftests/net/fib_rule_tests.sh
@@ -51,31 +51,6 @@ log_test()
fi
}
-check_nettest()
-{
- if which nettest > /dev/null 2>&1; then
- return 0
- fi
-
- # Add the selftest directory to PATH if not already done
- if [ "${SELFTEST_PATH}" = "" ]; then
- SELFTEST_PATH="$(dirname $0)"
- PATH="${PATH}:${SELFTEST_PATH}"
-
- # Now retry with the new path
- if which nettest > /dev/null 2>&1; then
- return 0
- fi
-
- if [ "${ret}" -eq 0 ]; then
- ret="${ksft_skip}"
- fi
- echo "nettest not found (try 'make -C ${SELFTEST_PATH} nettest')"
- fi
-
- return 1
-}
-
setup()
{
set -e
@@ -317,11 +292,6 @@ fib_rule6_connect_test()
echo
echo "IPv6 FIB rule connect tests"
- if ! check_nettest; then
- echo "SKIP: Could not run test without nettest tool"
- return
- fi
-
setup_peer
$IP -6 rule add dsfield 0x04 table $RTABLE_PEER
@@ -516,11 +486,6 @@ fib_rule4_connect_test()
echo
echo "IPv4 FIB rule connect tests"
- if ! check_nettest; then
- echo "SKIP: Could not run test without nettest tool"
- return
- fi
-
setup_peer
$IP -4 rule add dsfield 0x04 table $RTABLE_PEER
@@ -584,6 +549,8 @@ if [ ! -x "$(command -v ip)" ]; then
exit $ksft_skip
fi
+check_gen_prog "nettest"
+
# start clean
cleanup &> /dev/null
setup
diff --git a/tools/testing/selftests/net/lib.sh b/tools/testing/selftests/net/lib.sh
index 8ee4489238ca..be8707bfb46e 100644
--- a/tools/testing/selftests/net/lib.sh
+++ b/tools/testing/selftests/net/lib.sh
@@ -125,6 +125,21 @@ slowwait_for_counter()
slowwait "$timeout" until_counter_is ">= $((base + delta))" "$@"
}
+# Check for existence of tools which are built as part of selftests
+# but may also already exist in $PATH
+check_gen_prog()
+{
+ local prog_name=$1; shift
+
+ if ! which $prog_name >/dev/null 2>/dev/null; then
+ PATH=$PWD:$PATH
+ if ! which $prog_name >/dev/null; then
+ echo "'$prog_name' command not found; skipping tests"
+ exit $ksft_skip
+ fi
+ fi
+}
+
remove_ns_list()
{
local item=$1
diff --git a/tools/testing/selftests/net/pmtu.sh b/tools/testing/selftests/net/pmtu.sh
index 24a50622406c..569bce8b6383 100755
--- a/tools/testing/selftests/net/pmtu.sh
+++ b/tools/testing/selftests/net/pmtu.sh
@@ -681,13 +681,7 @@ setup_xfrm() {
}
setup_nettest_xfrm() {
- if ! which nettest >/dev/null; then
- PATH=$PWD:$PATH
- if ! which nettest >/dev/null; then
- echo "'nettest' command not found; skipping tests"
- return 1
- fi
- fi
+ check_gen_prog "nettest"
[ ${1} -eq 6 ] && proto="-6" || proto=""
port=${2}
diff --git a/tools/testing/selftests/net/unicast_extensions.sh b/tools/testing/selftests/net/unicast_extensions.sh
index f52aa5f7da52..3e751234ccfe 100755
--- a/tools/testing/selftests/net/unicast_extensions.sh
+++ b/tools/testing/selftests/net/unicast_extensions.sh
@@ -30,14 +30,7 @@
source lib.sh
-# nettest can be run from PATH or from same directory as this selftest
-if ! which nettest >/dev/null; then
- PATH=$PWD:$PATH
- if ! which nettest >/dev/null; then
- echo "'nettest' command not found; skipping tests"
- exit $ksft_skip
- fi
-fi
+check_gen_prog "nettest"
result=0
diff --git a/tools/testing/selftests/net/vrf_route_leaking.sh b/tools/testing/selftests/net/vrf_route_leaking.sh
index 152171fb1fc8..e9c2f71da207 100755
--- a/tools/testing/selftests/net/vrf_route_leaking.sh
+++ b/tools/testing/selftests/net/vrf_route_leaking.sh
@@ -59,7 +59,6 @@
# while it is forwarded between different vrfs.
source lib.sh
-PATH=$PWD:$PWD/tools/testing/selftests/net:$PATH
VERBOSE=0
PAUSE_ON_FAIL=no
DEFAULT_TTYPE=sym
@@ -636,6 +635,8 @@ EOF
# Some systems don't have a ping6 binary anymore
command -v ping6 > /dev/null 2>&1 && ping6=$(command -v ping6) || ping6=$(command -v ping)
+check_gen_prog "nettest"
+
TESTS_IPV4="ipv4_ping_ttl ipv4_traceroute ipv4_ping_frag ipv4_ping_local ipv4_tcp_local
ipv4_udp_local ipv4_ping_ttl_asym ipv4_traceroute_asym"
TESTS_IPV6="ipv6_ping_ttl ipv6_traceroute ipv6_ping_local ipv6_tcp_local ipv6_udp_local
--
2.46.0
Hello all,
This patch series targets a long-standing BPF usability issue - the lack
of general cross-compilation support - by enabling cross-endian usage of
libbpf and cross-endian build targets for selftests/bpf. Use cases range
from better BPF support for embedded systems based on e.g. big-endian
MIPS, to more build/test options for s390x systems.
Initial development and testing used mips64, since this arch makes
switching the build byte-order trivial and is thus very handy for A/B
testing. However, it lacks some key features (bpf2bpf call, kfuncs, etc)
making for poor selftests/bpf coverage.
Final testing takes the kernel and selftests/bpf cross-built from x86_64
to s390x, and runs the result under QEMU/s390x. That same configuration
could also be used on kernel-patches/bpf CI for regression testing endian
support or perhaps load-sharing s390x builds across x86_64 systems.
This thread includes some background regarding testing on QEMU/s390x and
the generally favourable results (3 failures running test_progs):
https://lore.kernel.org/bpf/ZsEcsaa3juxxQBUf@kodidev-ubuntu/
Feedback and suggestions are welcome!
Best regards,
Tony
Tony Ambardar (8):
libbpf: Improve log message formatting
libbpf: Fix header comment typos for BTF.ext
libbpf: Fix output .symtab byte-order during linking
libbpf: Support BTF.ext loading and output in either endianness
libbpf: Support opening bpf objects of either endianness
libbpf: Support linking bpf objects of either endianness
libbpf: Support creating light skeleton of either endianness
selftests/bpf: Support cross-endian building
tools/lib/bpf/bpf_gen_internal.h | 1 +
tools/lib/bpf/btf.c | 167 +++++++++++++++++++++++--
tools/lib/bpf/btf.h | 2 +
tools/lib/bpf/btf_dump.c | 2 +-
tools/lib/bpf/btf_relocate.c | 2 +-
tools/lib/bpf/gen_loader.c | 179 +++++++++++++++++++++------
tools/lib/bpf/libbpf.c | 26 +++-
tools/lib/bpf/libbpf.map | 2 +
tools/lib/bpf/libbpf_internal.h | 17 ++-
tools/lib/bpf/linker.c | 108 +++++++++++++---
tools/lib/bpf/relo_core.c | 2 +-
tools/lib/bpf/skel_internal.h | 3 +-
tools/testing/selftests/bpf/Makefile | 7 +-
13 files changed, 438 insertions(+), 80 deletions(-)
--
2.34.1
In this series from Geliang, modifying MPTCP BPF selftests, we have:
- A new MPTCP subflow BPF program setting socket options per subflow: it
looks better to have this old test program in the BPF selftests to
track regressions and to serve as example.
Note: Nicolas is no longer working for Tessares, but he did this work
while working for them, and his email address is no longer available.
- A new MPTCP BPF subtest validating the new BPF program added in the
first patch.
Signed-off-by: Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
---
Changes in v4:
- Drop former patch 2/3: MPTCP's pm_nl_ctl requires a new header file:
- I will check later if it is possible to avoid having duplicated
header files in tools/include/uapi, but no need to block this series
for that. Patch 2/3 can be added later if needed.
- Patch 2/2: skip the test if 'ip mptcp' is not available.
- Link to v3: https://lore.kernel.org/r/20240703-upstream-bpf-next-20240506-mptcp-subflow…
Changes in v3:
- Sorry for the delay between v2 and v3, this series was conflicting
with the "add netns helpers", but it looks like it is on hold:
https://lore.kernel.org/cover.1715821541.git.tanggeliang@kylinos.cn
- Patch 1/3 includes "bpf_tracing_net.h", introduced in between.
- New patch 2/3: "selftests/bpf: Add mptcp pm_nl_ctl link".
- Patch 3/3: use the tool introduced in patch 2/3 + SYS_NOFAIL() helper.
- Link to v2: https://lore.kernel.org/r/20240509-upstream-bpf-next-20240506-mptcp-subflow…
Changes in v2:
- Previous patches 1/4 and 2/4 have been dropped from this series:
- 1/4: "selftests/bpf: Handle SIGINT when creating netns":
- A new version, more generic and no longer specific to MPTCP BPF
selftest will be sent later, as part of a new series. (Alexei)
- 2/4: "selftests/bpf: Add RUN_MPTCP_TEST macro":
- Removed, not to hide helper functions in macros. (Alexei)
- The commit message of patch 1/2 has been clarified to avoid some
possible confusions spot by Alexei.
- Link to v1: https://lore.kernel.org/r/20240507-upstream-bpf-next-20240506-mptcp-subflow…
---
Geliang Tang (1):
selftests/bpf: Add mptcp subflow subtest
Nicolas Rybowski (1):
selftests/bpf: Add mptcp subflow example
tools/testing/selftests/bpf/prog_tests/mptcp.c | 105 ++++++++++++++++++++++
tools/testing/selftests/bpf/progs/mptcp_subflow.c | 59 ++++++++++++
2 files changed, 164 insertions(+)
---
base-commit: 3d650ab5e7d9c4d7306e4c116f8aa9980bf13295
change-id: 20240506-upstream-bpf-next-20240506-mptcp-subflow-test-faef6654bfa3
Best regards,
--
Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
First, generalize resctrl selftest non-contiguous CAT check to not
assume non-AMD vendor implies Intel. Second, improve kselftest common
parts and resctrl selftest such that the use of __cpuid_count() does
not lead into a build failure (happens at least on ARM).
The last patch might still require some work on which symbol the
conditional in kselftest.h is implemented. I could not find any
pre-existing one that could be used. Perhaps somebody who's more
familiar with the kselftest build system has a better suggestion on
which symbol the logic should be based at?
Ilpo Järvinen (3):
selftests/resctrl: Generalize non-contiguous CAT check
selftests/resctrl: Always initialize ecx to avoid build warnings
[RFC] kselftest: Provide __cpuid_count() stub on non-x86 archs
tools/testing/selftests/kselftest.h | 6 +++++
tools/testing/selftests/lib.mk | 4 ++++
tools/testing/selftests/resctrl/cat_test.c | 28 +++++++++++++---------
3 files changed, 27 insertions(+), 11 deletions(-)
--
2.39.2
This is mostly a re-post of a series [1] that was apparently lost last year.
[1] https://lore.kernel.org/all/20230928133821.1467-1-michal.wajdeczko@intel.co…
Cc: David Gow <davidgow(a)google.com>
Cc: Rae Moar <rmoar(a)google.com>
Michal Wajdeczko (3):
kunit: Improve format of the NOT_ERR_OR_NULL assertion
kunit: Improve format of the PTR_EQ|NE|NULL assertion
kunit: Improve format of the KUNIT_EXPECT_EQ assertion
lib/kunit/assert.c | 64 ++++++++++++++++++++++++++++++++++++----------
1 file changed, 50 insertions(+), 14 deletions(-)
--
2.43.0
Adds a selftest that creates two virtual interfaces, assigns one to a
new namespace, and assigns IP addresses to both.
It listens on the destination interface using socat and configures a
dynamic target on netconsole, pointing to the destination IP address.
The test then checks if the message was received properly on the
destination interface.
Signed-off-by: Breno Leitao <leitao(a)debian.org>
---
Changelog:
v6:
* Check for SRC and DST ip before starting the test (Jakub)
* Revert the printk configuration at the end of the test (Jakub)
* Fix the modprobe stderr redirection (Jakub)
v5:
* Replace check_file_size() by "test -s" (Matthieu)
* https://lore.kernel.org/all/20240819090406.1441297-1-leitao@debian.org/#t
v4:
* Avoid sleeping in waiting for sockets and files (Matthieu Baerts)
* Some other improvements (Matthieu Baerts)
* Add configfs as a dependency (Jakub)
* https://lore.kernel.org/all/20240816132450.346744-1-leitao@debian.org/
v3:
* Defined CONFIGs in config file (Jakub)
* Identention fixes (Petr Machata)
* Use setup_ns in a better way (Matthieu Baerts)
* Add dependencies in TEST_INCLUDES (Hangbin Liu)
* https://lore.kernel.org/all/20240815095157.3064722-1-leitao@debian.org/
v2:
* Change the location of the path (Jakub)
* Move from veth to netdevsim
* Other small changes in dependency checks and cleanup
* https://lore.kernel.org/all/20240813183825.837091-1-leitao@debian.org/
v1:
* https://lore.kernel.org/all/ZqyUHN770pjSofTC@gmail.com/
MAINTAINERS | 1 +
tools/testing/selftests/drivers/net/Makefile | 5 +-
tools/testing/selftests/drivers/net/config | 4 +
.../selftests/drivers/net/netcons_basic.sh | 234 ++++++++++++++++++
4 files changed, 243 insertions(+), 1 deletion(-)
create mode 100755 tools/testing/selftests/drivers/net/netcons_basic.sh
diff --git a/MAINTAINERS b/MAINTAINERS
index 5dbf23cf11c8..9a371ddd8719 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -15772,6 +15772,7 @@ M: Breno Leitao <leitao(a)debian.org>
S: Maintained
F: Documentation/networking/netconsole.rst
F: drivers/net/netconsole.c
+F: tools/testing/selftests/drivers/net/netcons_basic.sh
NETDEVSIM
M: Jakub Kicinski <kuba(a)kernel.org>
diff --git a/tools/testing/selftests/drivers/net/Makefile b/tools/testing/selftests/drivers/net/Makefile
index e54f382bcb02..39fb97a8c1df 100644
--- a/tools/testing/selftests/drivers/net/Makefile
+++ b/tools/testing/selftests/drivers/net/Makefile
@@ -1,8 +1,11 @@
# SPDX-License-Identifier: GPL-2.0
-TEST_INCLUDES := $(wildcard lib/py/*.py)
+TEST_INCLUDES := $(wildcard lib/py/*.py) \
+ ../../net/net_helper.sh \
+ ../../net/lib.sh \
TEST_PROGS := \
+ netcons_basic.sh \
ping.py \
queues.py \
stats.py \
diff --git a/tools/testing/selftests/drivers/net/config b/tools/testing/selftests/drivers/net/config
index f6a58ce8a230..a2d8af60876d 100644
--- a/tools/testing/selftests/drivers/net/config
+++ b/tools/testing/selftests/drivers/net/config
@@ -1,2 +1,6 @@
CONFIG_IPV6=y
CONFIG_NETDEVSIM=m
+CONFIG_CONFIGFS_FS=y
+CONFIG_NETCONSOLE=m
+CONFIG_NETCONSOLE_DYNAMIC=y
+CONFIG_NETCONSOLE_EXTENDED_LOG=y
diff --git a/tools/testing/selftests/drivers/net/netcons_basic.sh b/tools/testing/selftests/drivers/net/netcons_basic.sh
new file mode 100755
index 000000000000..b4bfb451ccb6
--- /dev/null
+++ b/tools/testing/selftests/drivers/net/netcons_basic.sh
@@ -0,0 +1,234 @@
+#!/usr/bin/env bash
+# SPDX-License-Identifier: GPL-2.0
+
+# This test creates two netdevsim virtual interfaces, assigns one of them (the
+# "destination interface") to a new namespace, and assigns IP addresses to both
+# interfaces.
+#
+# It listens on the destination interface using socat and configures a dynamic
+# target on netconsole, pointing to the destination IP address.
+#
+# Finally, it checks whether the message was received properly on the
+# destination interface. Note that this test may pollute the kernel log buffer
+# (dmesg) and relies on dynamic configuration and namespaces being configured.
+#
+# Author: Breno Leitao <leitao(a)debian.org>
+
+set -euo pipefail
+
+SCRIPTDIR=$(dirname "$(readlink -e "${BASH_SOURCE[0]}")")
+
+# Simple script to test dynamic targets in netconsole
+SRCIF="" # to be populated later
+SRCIP=192.168.1.1
+DSTIF="" # to be populated later
+DSTIP=192.168.1.2
+
+PORT="6666"
+MSG="netconsole selftest"
+TARGET=$(mktemp -u netcons_XXXXX)
+DEFAULT_PRINTK_VALUES=$(cat /proc/sys/kernel/printk)
+NETCONS_CONFIGFS="/sys/kernel/config/netconsole"
+NETCONS_PATH="${NETCONS_CONFIGFS}"/"${TARGET}"
+# NAMESPACE will be populated by setup_ns with a random value
+NAMESPACE=""
+
+# IDs for netdevsim
+NSIM_DEV_1_ID=$((256 + RANDOM % 256))
+NSIM_DEV_2_ID=$((512 + RANDOM % 256))
+
+# Used to create and delete namespaces
+source "${SCRIPTDIR}"/../../net/lib.sh
+source "${SCRIPTDIR}"/../../net/net_helper.sh
+
+# Create netdevsim interfaces
+create_ifaces() {
+ local NSIM_DEV_SYS_NEW=/sys/bus/netdevsim/new_device
+
+ echo "$NSIM_DEV_2_ID" > "$NSIM_DEV_SYS_NEW"
+ echo "$NSIM_DEV_1_ID" > "$NSIM_DEV_SYS_NEW"
+ udevadm settle 2> /dev/null || true
+
+ local NSIM1=/sys/bus/netdevsim/devices/netdevsim"$NSIM_DEV_1_ID"
+ local NSIM2=/sys/bus/netdevsim/devices/netdevsim"$NSIM_DEV_2_ID"
+
+ # These are global variables
+ SRCIF=$(find "$NSIM1"/net -maxdepth 1 -type d ! \
+ -path "$NSIM1"/net -exec basename {} \;)
+ DSTIF=$(find "$NSIM2"/net -maxdepth 1 -type d ! \
+ -path "$NSIM2"/net -exec basename {} \;)
+}
+
+link_ifaces() {
+ local NSIM_DEV_SYS_LINK="/sys/bus/netdevsim/link_device"
+ local SRCIF_IFIDX=$(cat /sys/class/net/"$SRCIF"/ifindex)
+ local DSTIF_IFIDX=$(cat /sys/class/net/"$DSTIF"/ifindex)
+
+ exec {NAMESPACE_FD}</var/run/netns/"${NAMESPACE}"
+ exec {INITNS_FD}</proc/self/ns/net
+
+ # Bind the dst interface to namespace
+ ip link set "${DSTIF}" netns "${NAMESPACE}"
+
+ # Linking one device to the other one (on the other namespace}
+ if ! echo "${INITNS_FD}:$SRCIF_IFIDX $NAMESPACE_FD:$DSTIF_IFIDX" > $NSIM_DEV_SYS_LINK
+ then
+ echo "linking netdevsim1 with netdevsim2 should succeed"
+ cleanup
+ exit "${ksft_skip}"
+ fi
+}
+
+function configure_ip() {
+ # Configure the IPs for both interfaces
+ ip netns exec "${NAMESPACE}" ip addr add "${DSTIP}"/24 dev "${DSTIF}"
+ ip netns exec "${NAMESPACE}" ip link set "${DSTIF}" up
+
+ ip addr add "${SRCIP}"/24 dev "${SRCIF}"
+ ip link set "${SRCIF}" up
+}
+
+function set_network() {
+ # setup_ns function is coming from lib.sh
+ setup_ns NAMESPACE
+
+ # Create both interfaces, and assign the destination to a different
+ # namespace
+ create_ifaces
+
+ # Link both interfaces back to back
+ link_ifaces
+
+ configure_ip
+}
+
+function create_dynamic_target() {
+ DSTMAC=$(ip netns exec "${NAMESPACE}" \
+ ip link show "${DSTIF}" | awk '/ether/ {print $2}')
+
+ # Create a dynamic target
+ mkdir "${NETCONS_PATH}"
+
+ echo "${DSTIP}" > "${NETCONS_PATH}"/remote_ip
+ echo "${SRCIP}" > "${NETCONS_PATH}"/local_ip
+ echo "${DSTMAC}" > "${NETCONS_PATH}"/remote_mac
+ echo "${SRCIF}" > "${NETCONS_PATH}"/dev_name
+
+ echo 1 > "${NETCONS_PATH}"/enabled
+}
+
+function cleanup() {
+ local NSIM_DEV_SYS_DEL="/sys/bus/netdevsim/del_device"
+
+ # delete netconsole dynamic reconfiguration
+ echo 0 > "${NETCONS_PATH}"/enabled
+ # Remove the configfs entry
+ rmdir "${NETCONS_PATH}"
+
+ # Delete netdevsim devices
+ echo "$NSIM_DEV_2_ID" > "$NSIM_DEV_SYS_DEL"
+ echo "$NSIM_DEV_1_ID" > "$NSIM_DEV_SYS_DEL"
+
+ # this is coming from lib.sh
+ cleanup_all_ns
+
+ # Restoring printk configurations
+ echo "${DEFAULT_PRINTK_VALUES}" > /proc/sys/kernel/printk
+}
+
+function listen_port_and_save_to() {
+ local OUTPUT=${1}
+ # Just wait for 2 seconds
+ timeout 2 ip netns exec "${NAMESPACE}" \
+ socat UDP-LISTEN:"${PORT}",fork "${OUTPUT}"
+}
+
+function validate_result() {
+ local TMPFILENAME="$1"
+
+ # Check if the file exists
+ if [ ! -f "$TMPFILENAME" ]; then
+ echo "FAIL: File was not generated." >&2
+ exit "${ksft_fail}"
+ fi
+
+ if ! grep -q "${MSG}" "${TMPFILENAME}"; then
+ echo "FAIL: ${MSG} not found in ${TMPFILENAME}" >&2
+ cat "${TMPFILENAME}" >&2
+ exit "${ksft_fail}"
+ fi
+
+ # Delete the file once it is validated, otherwise keep it
+ # for debugging purposes
+ rm "${TMPFILENAME}"
+ exit "${ksft_pass}"
+}
+
+function check_for_dependencies() {
+ if [ "$(id -u)" -ne 0 ]; then
+ echo "This test must be run as root" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which socat > /dev/null ; then
+ echo "SKIP: socat(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which ip > /dev/null ; then
+ echo "SKIP: ip(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which udevadm > /dev/null ; then
+ echo "SKIP: udevadm(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if [ ! -d "${NETCONS_CONFIGFS}" ]; then
+ echo "SKIP: directory ${NETCONS_CONFIGFS} does not exist. Check if NETCONSOLE_DYNAMIC is enabled" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ip link show "${DSTIF}" 2> /dev/null; then
+ echo "SKIP: interface ${DSTIF} exists in the system. Not overwriting it." >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ip addr list | grep -E "inet.*(${SRCIP}|${DSTIP})" 2> /dev/null; then
+ echo "SKIP: IPs already in use. Skippig it" >&2
+ exit "${ksft_skip}"
+ fi
+}
+
+# ========== #
+# Start here #
+# ========== #
+modprobe netdevsim 2> /dev/null || true
+modprobe netconsole 2> /dev/null || true
+
+# The content of kmsg will be save to the following file
+OUTPUT_FILE="/tmp/${TARGET}"
+
+# Check for basic system dependency and exit if not found
+check_for_dependencies
+# Set current loglevel to KERN_INFO(6), and default to KERN_NOTICE(5)
+echo "6 5" > /proc/sys/kernel/printk
+# Remove the namespace, interfaces and netconsole target on exit
+trap cleanup EXIT
+# Create one namespace and two interfaces
+set_network
+# Create a dynamic target for netconsole
+create_dynamic_target
+# Listed for netconsole port inside the namespace and destination interface
+listen_port_and_save_to "${OUTPUT_FILE}" &
+# Wait for socat to start and listen to the port.
+wait_local_port_listen "${NAMESPACE}" "${PORT}" udp
+# Send the message
+echo "${MSG}: ${TARGET}" > /dev/kmsg
+# Wait until socat saves the file to disk
+busywait "${BUSYWAIT_TIMEOUT}" test -s "${OUTPUT_FILE}"
+
+# Make sure the message was received in the dst part
+# and exit
+validate_result "${OUTPUT_FILE}"
--
2.43.5
[based on mm-unstable, 651c8c1d7359]
Optimize mseal checks by removing the separate can_modify_mm() step, and
just doing checks on the individual vmas, when various operations are
themselves iterating through the tree. This provides a nice speedup and restores
performance parity with pre-mseal[3].
This series also depends on the powerpc series that removes arch_unmap[2]. This
series is already in mm-unstable.
will-it-scale mmap1_process[1] -t 1 results:
commit 3450fe2b574b4345e4296ccae395149e1a357fee:
min:277605 max:277605 total:277605
min:281784 max:281784 total:281784
min:277238 max:277238 total:277238
min:281761 max:281761 total:281761
min:274279 max:274279 total:274279
min:254854 max:254854 total:254854
measurement
min:269143 max:269143 total:269143
min:270454 max:270454 total:270454
min:243523 max:243523 total:243523
min:251148 max:251148 total:251148
min:209669 max:209669 total:209669
min:190426 max:190426 total:190426
min:231219 max:231219 total:231219
min:275364 max:275364 total:275364
min:266540 max:266540 total:266540
min:242572 max:242572 total:242572
min:284469 max:284469 total:284469
min:278882 max:278882 total:278882
min:283269 max:283269 total:283269
min:281204 max:281204 total:281204
After this patch set:
min:280580 max:280580 total:280580
min:290514 max:290514 total:290514
min:291006 max:291006 total:291006
min:290352 max:290352 total:290352
min:294582 max:294582 total:294582
min:293075 max:293075 total:293075
measurement
min:295613 max:295613 total:295613
min:294070 max:294070 total:294070
min:293193 max:293193 total:293193
min:291631 max:291631 total:291631
min:295278 max:295278 total:295278
min:293782 max:293782 total:293782
min:290361 max:290361 total:290361
min:294517 max:294517 total:294517
min:293750 max:293750 total:293750
min:293572 max:293572 total:293572
min:295239 max:295239 total:295239
min:292932 max:292932 total:292932
min:293319 max:293319 total:293319
min:294954 max:294954 total:294954
This was a Completely Unscientific test but seems to show there were around 5-10% gains on ops per second.
Oliver performed his own tests and showed[3] a similar ~5% gain in them.
[1]: mmap1_process does mmap and munmap in a loop. I didn't bother testing multithreading cases.
[2]: https://lore.kernel.org/all/20240807124103.85644-1-mpe@ellerman.id.au/
[3]: https://lore.kernel.org/all/ZrMMJfe9aXSWxJz6@xsang-OptiPlex-9020/
Link: https://lore.kernel.org/all/202408041602.caa0372-oliver.sang@intel.com/
Changes in v3:
- Moved can_modify_vma to vma.h instead of internal.h (Lorenzo)
- Fixed a bug in munmap where we used the wrong VMA pointer
- Added tests for the previous munmap bug
- Moved the mremap source vma check upwards, to stop us from unmapping
dest while the source is sealed (Liam)
Changes in v2:
- Rebased on top of mm-unstable
- Removed a superfluous check in mremap (Jeff Xu)
Signed-off-by: Pedro Falcato <pedro.falcato(a)gmail.com>
---
Pedro Falcato (7):
mm: Move can_modify_vma to mm/vma.h
mm/munmap: Replace can_modify_mm with can_modify_vma
mm/mprotect: Replace can_modify_mm with can_modify_vma
mm/mremap: Replace can_modify_mm with can_modify_vma
mseal: Replace can_modify_mm_madv with a vma variant
mm: Remove can_modify_mm()
selftests/mm: add more mseal traversal tests
mm/internal.h | 16 -----
mm/madvise.c | 13 +---
mm/mmap.c | 11 +---
mm/mprotect.c | 12 +---
mm/mremap.c | 32 ++-------
mm/mseal.c | 55 ++--------------
mm/vma.c | 19 ++++--
mm/vma.h | 35 ++++++++++
tools/testing/selftests/mm/mseal_test.c | 111 +++++++++++++++++++++++++++++++-
9 files changed, 174 insertions(+), 130 deletions(-)
---
base-commit: 651c8c1d735983040bec4f71d0e2e690f3c0fc2d
change-id: 20240816-mseal-depessimize-f39d9f4c32c6
Best regards,
--
Pedro Falcato <pedro.falcato(a)gmail.com>
In this series from Geliang, modifying MPTCP BPF selftests, we have:
- A new MPTCP subflow BPF program setting socket options per subflow: it
looks better to have this old test program in the BPF selftests to
track regressions and to serve as example.
Note: Nicolas is no longer working for Tessares, but he did this work
while working for them, and his email address is no longer available.
- A new MPTCP BPF subtest validating this new BPF program.
Signed-off-by: Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
---
Changes in v2:
- Previous patches 1/4 and 2/4 have been dropped from this series:
- 1/4: "selftests/bpf: Handle SIGINT when creating netns":
- A new version, more generic and no longer specific to MPTCP BPF
selftest will be sent later, as part of a new series. (Alexei)
- 2/4: "selftests/bpf: Add RUN_MPTCP_TEST macro":
- Removed, not to hide helper functions in macros. (Alexei)
- The commit message of patch 1/2 has been clarified to avoid some
possible confusions spot by Alexei.
- Link to v1: https://lore.kernel.org/r/20240507-upstream-bpf-next-20240506-mptcp-subflow…
---
Geliang Tang (1):
selftests/bpf: Add mptcp subflow subtest
Nicolas Rybowski (1):
selftests/bpf: Add mptcp subflow example
tools/testing/selftests/bpf/prog_tests/mptcp.c | 109 ++++++++++++++++++++++
tools/testing/selftests/bpf/progs/mptcp_subflow.c | 70 ++++++++++++++
2 files changed, 179 insertions(+)
---
base-commit: 009367099eb61a4fc2af44d4eb06b6b4de7de6db
change-id: 20240506-upstream-bpf-next-20240506-mptcp-subflow-test-faef6654bfa3
Best regards,
--
Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
The kernel has recently added support for shadow stacks, currently
x86 only using their CET feature but both arm64 and RISC-V have
equivalent features (GCS and Zicfiss respectively), I am actively
working on GCS[1]. With shadow stacks the hardware maintains an
additional stack containing only the return addresses for branch
instructions which is not generally writeable by userspace and ensures
that any returns are to the recorded addresses. This provides some
protection against ROP attacks and making it easier to collect call
stacks. These shadow stacks are allocated in the address space of the
userspace process.
Our API for shadow stacks does not currently offer userspace any
flexiblity for managing the allocation of shadow stacks for newly
created threads, instead the kernel allocates a new shadow stack with
the same size as the normal stack whenever a thread is created with the
feature enabled. The stacks allocated in this way are freed by the
kernel when the thread exits or shadow stacks are disabled for the
thread. This lack of flexibility and control isn't ideal, in the vast
majority of cases the shadow stack will be over allocated and the
implicit allocation and deallocation is not consistent with other
interfaces. As far as I can tell the interface is done in this manner
mainly because the shadow stack patches were in development since before
clone3() was implemented.
Since clone3() is readily extensible let's add support for specifying a
shadow stack when creating a new thread or process in a similar manner
to how the normal stack is specified, keeping the current implicit
allocation behaviour if one is not specified either with clone3() or
through the use of clone(). The user must provide a shadow stack
address and size, this must point to memory mapped for use as a shadow
stackby map_shadow_stack() with a shadow stack token at the top of the
stack.
Please note that the x86 portions of this code are build tested only, I
don't appear to have a system that can run CET avaible to me, I have
done testing with an integration into my pending work for GCS. There is
some possibility that the arm64 implementation may require the use of
clone3() and explicit userspace allocation of shadow stacks, this is
still under discussion.
Please further note that the token consumption done by clone3() is not
currently implemented in an atomic fashion, Rick indicated that he would
look into fixing this if people are OK with the implementation.
A new architecture feature Kconfig option for shadow stacks is added as
here, this was suggested as part of the review comments for the arm64
GCS series and since we need to detect if shadow stacks are supported it
seemed sensible to roll it in here.
[1] https://lore.kernel.org/r/20231009-arm64-gcs-v6-0-78e55deaa4dd@kernel.org/
Signed-off-by: Mark Brown <broonie(a)kernel.org>
---
Changes in v10:
- Integrate fixes & improvements for the x86 implementation from Rick
Edgecombe.
- Require that the shadow stack be VM_WRITE.
- Require that the shadow stack base and size be sizeof(void *) aligned.
- Clean up trailing newline.
- Link to v9: https://lore.kernel.org/r/20240819-clone3-shadow-stack-v9-0-962d74f99464@ke…
Changes in v9:
- Pull token validation earlier and report problems with an error return
to parent rather than signal delivery to the child.
- Verify that the top of the supplied shadow stack is VM_SHADOW_STACK.
- Rework token validation to only do the page mapping once.
- Drop no longer needed support for testing for signals in selftest.
- Fix typo in comments.
- Link to v8: https://lore.kernel.org/r/20240808-clone3-shadow-stack-v8-0-0acf37caf14c@ke…
Changes in v8:
- Fix token verification with user specified shadow stack.
- Don't track user managed shadow stacks for child processes.
- Link to v7: https://lore.kernel.org/r/20240731-clone3-shadow-stack-v7-0-a9532eebfb1d@ke…
Changes in v7:
- Rebase onto v6.11-rc1.
- Typo fixes.
- Link to v6: https://lore.kernel.org/r/20240623-clone3-shadow-stack-v6-0-9ee7783b1fb9@ke…
Changes in v6:
- Rebase onto v6.10-rc3.
- Ensure we don't try to free the parent shadow stack in error paths of
x86 arch code.
- Spelling fixes in userspace API document.
- Additional cleanups and improvements to the clone3() tests to support
the shadow stack tests.
- Link to v5: https://lore.kernel.org/r/20240203-clone3-shadow-stack-v5-0-322c69598e4b@ke…
Changes in v5:
- Rebase onto v6.8-rc2.
- Rework ABI to have the user allocate the shadow stack memory with
map_shadow_stack() and a token.
- Force inlining of the x86 shadow stack enablement.
- Move shadow stack enablement out into a shared header for reuse by
other tests.
- Link to v4: https://lore.kernel.org/r/20231128-clone3-shadow-stack-v4-0-8b28ffe4f676@ke…
Changes in v4:
- Formatting changes.
- Use a define for minimum shadow stack size and move some basic
validation to fork.c.
- Link to v3: https://lore.kernel.org/r/20231120-clone3-shadow-stack-v3-0-a7b8ed3e2acc@ke…
Changes in v3:
- Rebase onto v6.7-rc2.
- Remove stale shadow_stack in internal kargs.
- If a shadow stack is specified unconditionally use it regardless of
CLONE_ parameters.
- Force enable shadow stacks in the selftest.
- Update changelogs for RISC-V feature rename.
- Link to v2: https://lore.kernel.org/r/20231114-clone3-shadow-stack-v2-0-b613f8681155@ke…
Changes in v2:
- Rebase onto v6.7-rc1.
- Remove ability to provide preallocated shadow stack, just specify the
desired size.
- Link to v1: https://lore.kernel.org/r/20231023-clone3-shadow-stack-v1-0-d867d0b5d4d0@ke…
---
Mark Brown (8):
Documentation: userspace-api: Add shadow stack API documentation
selftests: Provide helper header for shadow stack testing
mm: Introduce ARCH_HAS_USER_SHADOW_STACK
fork: Add shadow stack support to clone3()
selftests/clone3: Remove redundant flushes of output streams
selftests/clone3: Factor more of main loop into test_clone3()
selftests/clone3: Allow tests to flag if -E2BIG is a valid error code
selftests/clone3: Test shadow stack support
Documentation/userspace-api/index.rst | 1 +
Documentation/userspace-api/shadow_stack.rst | 41 ++++
arch/x86/Kconfig | 1 +
arch/x86/include/asm/shstk.h | 11 +-
arch/x86/kernel/process.c | 2 +-
arch/x86/kernel/shstk.c | 96 ++++++--
fs/proc/task_mmu.c | 2 +-
include/asm-generic/cacheflush.h | 11 +
include/linux/mm.h | 2 +-
include/linux/sched/task.h | 18 ++
include/uapi/linux/sched.h | 13 +-
kernel/fork.c | 121 ++++++++++-
mm/Kconfig | 6 +
tools/testing/selftests/clone3/clone3.c | 254 ++++++++++++++++++----
tools/testing/selftests/clone3/clone3_selftests.h | 40 +++-
tools/testing/selftests/ksft_shstk.h | 61 ++++++
16 files changed, 593 insertions(+), 87 deletions(-)
---
base-commit: 8400291e289ee6b2bf9779ff1c83a291501f017b
change-id: 20231019-clone3-shadow-stack-15d40d2bf536
Best regards,
--
Mark Brown <broonie(a)kernel.org>
The arm64 Guarded Control Stack (GCS) feature provides support for
hardware protected stacks of return addresses, intended to provide
hardening against return oriented programming (ROP) attacks and to make
it easier to gather call stacks for applications such as profiling.
When GCS is active a secondary stack called the Guarded Control Stack is
maintained, protected with a memory attribute which means that it can
only be written with specific GCS operations. The current GCS pointer
can not be directly written to by userspace. When a BL is executed the
value stored in LR is also pushed onto the GCS, and when a RET is
executed the top of the GCS is popped and compared to LR with a fault
being raised if the values do not match. GCS operations may only be
performed on GCS pages, a data abort is generated if they are not.
The combination of hardware enforcement and lack of extra instructions
in the function entry and exit paths should result in something which
has less overhead and is more difficult to attack than a purely software
implementation like clang's shadow stacks.
This series implements support for use of GCS by userspace, along with
support for use of GCS within KVM guests. It does not enable use of GCS
by either EL1 or EL2, this will be implemented separately. Executables
are started without GCS and must use a prctl() to enable it, it is
expected that this will be done very early in application execution by
the dynamic linker or other startup code. For dynamic linking this will
be done by checking that everything in the executable is marked as GCS
compatible.
x86 has an equivalent feature called shadow stacks, this series depends
on the x86 patches for generic memory management support for the new
guarded/shadow stack page type and shares APIs as much as possible. As
there has been extensive discussion with the wider community around the
ABI for shadow stacks I have as far as practical kept implementation
decisions close to those for x86, anticipating that review would lead to
similar conclusions in the absence of strong reasoning for divergence.
The main divergence I am concious of is that x86 allows shadow stack to
be enabled and disabled repeatedly, freeing the shadow stack for the
thread whenever disabled, while this implementation keeps the GCS
allocated after disable but refuses to reenable it. This is to avoid
races with things actively walking the GCS during a disable, we do
anticipate that some systems will wish to disable GCS at runtime but are
not aware of any demand for subsequently reenabling it.
x86 uses an arch_prctl() to manage enable and disable, since only x86
and S/390 use arch_prctl() a generic prctl() was proposed[1] as part of a
patch set for the equivalent RISC-V Zicfiss feature which I initially
adopted fairly directly but following review feedback has been revised
quite a bit.
We currently maintain the x86 pattern of implicitly allocating a shadow
stack for threads started with shadow stack enabled, there has been some
discussion of removing this support and requiring the use of clone3()
with explicit allocation of shadow stacks instead. I have no strong
feelings either way, implicit allocation is not really consistent with
anything else we do and creates the potential for errors around thread
exit but on the other hand it is existing ABI on x86 and minimises the
changes needed in userspace code.
glibc and bionic changes using this ABI have been implemented and
tested. Headless Android systems have been validated and Ross Burton
has used this code has been used to bring up a Yocto system with GCS
enabed as standard, a test implementation of V8 support has also been
done.
There is an open issue with support for CRIU, on x86 this required the
ability to set the GCS mode via ptrace. This series supports
configuring mode bits other than enable/disable via ptrace but it needs
to be confirmed if this is sufficient.
The series depends on support for shadow stacks in clone3(), that series
includes the addition of ARCH_HAS_USER_SHADOW_STACK.
https://lore.kernel.org/r/20240731-clone3-shadow-stack-v7-0-a9532eebfb1d@ke…
You can see a branch with the full set of dependencies against Linus'
tree at:
https://git.kernel.org/pub/scm/linux/kernel/git/broonie/misc.git arm64-gcs
[1] https://lore.kernel.org/lkml/20230213045351.3945824-1-debug@rivosinc.com/
Signed-off-by: Mark Brown <broonie(a)kernel.org>
---
Changes in v10:
- Fix issues with THP.
- Tighten up requirements for initialising GCSCR*.
- Only generate GCS signal frames for threads using GCS.
- Only context switch EL1 GCS registers if S1PIE is enabled.
- Move context switch of GCSCRE0_EL1 to EL0 context switch.
- Make GCS registers unconditionally visible to userspace.
- Use FHU infrastructure.
- Don't change writability of ID_AA64PFR1_EL1 for KVM.
- Remove unused arguments from alloc_gcs().
- Typo fixes.
- Link to v9: https://lore.kernel.org/r/20240625-arm64-gcs-v9-0-0f634469b8f0@kernel.org
Changes in v9:
- Rebase onto v6.10-rc3.
- Restructure and clarify memory management fault handling.
- Fix up basic-gcs for the latest clone3() changes.
- Convert to newly merged KVM ID register based feature configuration.
- Fixes for NV traps.
- Link to v8: https://lore.kernel.org/r/20240203-arm64-gcs-v8-0-c9fec77673ef@kernel.org
Changes in v8:
- Invalidate signal cap token on stack when consuming.
- Typo and other trivial fixes.
- Don't try to use process_vm_write() on GCS, it intentionally does not
work.
- Fix leak of thread GCSs.
- Rebase onto latest clone3() series.
- Link to v7: https://lore.kernel.org/r/20231122-arm64-gcs-v7-0-201c483bd775@kernel.org
Changes in v7:
- Rebase onto v6.7-rc2 via the clone3() patch series.
- Change the token used to cap the stack during signal handling to be
compatible with GCSPOPM.
- Fix flags for new page types.
- Fold in support for clone3().
- Replace copy_to_user_gcs() with put_user_gcs().
- Link to v6: https://lore.kernel.org/r/20231009-arm64-gcs-v6-0-78e55deaa4dd@kernel.org
Changes in v6:
- Rebase onto v6.6-rc3.
- Add some more gcsb_dsync() barriers following spec clarifications.
- Due to ongoing discussion around clone()/clone3() I've not updated
anything there, the behaviour is the same as on previous versions.
- Link to v5: https://lore.kernel.org/r/20230822-arm64-gcs-v5-0-9ef181dd6324@kernel.org
Changes in v5:
- Don't map any permissions for user GCSs, we always use EL0 accessors
or use a separate mapping of the page.
- Reduce the standard size of the GCS to RLIMIT_STACK/2.
- Enforce a PAGE_SIZE alignment requirement on map_shadow_stack().
- Clarifications and fixes to documentation.
- More tests.
- Link to v4: https://lore.kernel.org/r/20230807-arm64-gcs-v4-0-68cfa37f9069@kernel.org
Changes in v4:
- Implement flags for map_shadow_stack() allowing the cap and end of
stack marker to be enabled independently or not at all.
- Relax size and alignment requirements for map_shadow_stack().
- Add more blurb explaining the advantages of hardware enforcement.
- Link to v3: https://lore.kernel.org/r/20230731-arm64-gcs-v3-0-cddf9f980d98@kernel.org
Changes in v3:
- Rebase onto v6.5-rc4.
- Add a GCS barrier on context switch.
- Add a GCS stress test.
- Link to v2: https://lore.kernel.org/r/20230724-arm64-gcs-v2-0-dc2c1d44c2eb@kernel.org
Changes in v2:
- Rebase onto v6.5-rc3.
- Rework prctl() interface to allow each bit to be locked independently.
- map_shadow_stack() now places the cap token based on the size
requested by the caller not the actual space allocated.
- Mode changes other than enable via ptrace are now supported.
- Expand test coverage.
- Various smaller fixes and adjustments.
- Link to v1: https://lore.kernel.org/r/20230716-arm64-gcs-v1-0-bf567f93bba6@kernel.org
---
Mark Brown (40):
arm64/mm: Restructure arch_validate_flags() for extensibility
prctl: arch-agnostic prctl for shadow stack
mman: Add map_shadow_stack() flags
arm64: Document boot requirements for Guarded Control Stacks
arm64/gcs: Document the ABI for Guarded Control Stacks
arm64/sysreg: Add definitions for architected GCS caps
arm64/gcs: Add manual encodings of GCS instructions
arm64/gcs: Provide put_user_gcs()
arm64/gcs: Provide basic EL2 setup to allow GCS usage at EL0 and EL1
arm64/cpufeature: Runtime detection of Guarded Control Stack (GCS)
arm64/mm: Allocate PIE slots for EL0 guarded control stack
mm: Define VM_SHADOW_STACK for arm64 when we support GCS
arm64/mm: Map pages for guarded control stack
KVM: arm64: Manage GCS access and registers for guests
arm64/idreg: Add overrride for GCS
arm64/hwcap: Add hwcap for GCS
arm64/traps: Handle GCS exceptions
arm64/mm: Handle GCS data aborts
arm64/gcs: Context switch GCS state for EL0
arm64/gcs: Ensure that new threads have a GCS
arm64/gcs: Implement shadow stack prctl() interface
arm64/mm: Implement map_shadow_stack()
arm64/signal: Set up and restore the GCS context for signal handlers
arm64/signal: Expose GCS state in signal frames
arm64/ptrace: Expose GCS via ptrace and core files
arm64: Add Kconfig for Guarded Control Stack (GCS)
kselftest/arm64: Verify the GCS hwcap
kselftest: Provide shadow stack enable helpers for arm64
selftests/clone3: Enable arm64 shadow stack testing
kselftest/arm64: Add GCS as a detected feature in the signal tests
kselftest/arm64: Add framework support for GCS to signal handling tests
kselftest/arm64: Allow signals tests to specify an expected si_code
kselftest/arm64: Always run signals tests with GCS enabled
kselftest/arm64: Add very basic GCS test program
kselftest/arm64: Add a GCS test program built with the system libc
kselftest/arm64: Add test coverage for GCS mode locking
kselftest/arm64: Add GCS signal tests
kselftest/arm64: Add a GCS stress test
kselftest/arm64: Enable GCS for the FP stress tests
KVM: selftests: arm64: Add GCS registers to get-reg-list
Documentation/admin-guide/kernel-parameters.txt | 3 +
Documentation/arch/arm64/booting.rst | 30 +
Documentation/arch/arm64/elf_hwcaps.rst | 2 +
Documentation/arch/arm64/gcs.rst | 233 +++++++
Documentation/arch/arm64/index.rst | 1 +
Documentation/filesystems/proc.rst | 2 +-
arch/arm64/Kconfig | 20 +
arch/arm64/include/asm/cpufeature.h | 6 +
arch/arm64/include/asm/el2_setup.h | 29 +
arch/arm64/include/asm/esr.h | 28 +-
arch/arm64/include/asm/exception.h | 2 +
arch/arm64/include/asm/gcs.h | 107 +++
arch/arm64/include/asm/hwcap.h | 1 +
arch/arm64/include/asm/kvm_host.h | 8 +
arch/arm64/include/asm/mman.h | 23 +-
arch/arm64/include/asm/pgtable-prot.h | 14 +-
arch/arm64/include/asm/processor.h | 7 +
arch/arm64/include/asm/sysreg.h | 20 +
arch/arm64/include/asm/uaccess.h | 40 ++
arch/arm64/include/asm/vncr_mapping.h | 2 +
arch/arm64/include/uapi/asm/hwcap.h | 1 +
arch/arm64/include/uapi/asm/ptrace.h | 8 +
arch/arm64/include/uapi/asm/sigcontext.h | 9 +
arch/arm64/kernel/cpufeature.c | 12 +
arch/arm64/kernel/cpuinfo.c | 1 +
arch/arm64/kernel/entry-common.c | 23 +
arch/arm64/kernel/pi/idreg-override.c | 2 +
arch/arm64/kernel/process.c | 85 +++
arch/arm64/kernel/ptrace.c | 59 ++
arch/arm64/kernel/signal.c | 240 ++++++-
arch/arm64/kernel/traps.c | 11 +
arch/arm64/kvm/hyp/include/hyp/sysreg-sr.h | 49 +-
arch/arm64/kvm/sys_regs.c | 12 +
arch/arm64/mm/Makefile | 1 +
arch/arm64/mm/fault.c | 42 ++
arch/arm64/mm/gcs.c | 324 +++++++++
arch/arm64/mm/mmap.c | 10 +-
arch/arm64/tools/cpucaps | 1 +
arch/x86/include/uapi/asm/mman.h | 3 -
include/linux/mm.h | 16 +-
include/uapi/asm-generic/mman.h | 4 +
include/uapi/linux/elf.h | 1 +
include/uapi/linux/prctl.h | 22 +
kernel/sys.c | 30 +
tools/testing/selftests/arm64/Makefile | 2 +-
tools/testing/selftests/arm64/abi/hwcap.c | 19 +
tools/testing/selftests/arm64/fp/assembler.h | 15 +
tools/testing/selftests/arm64/fp/fpsimd-test.S | 2 +
tools/testing/selftests/arm64/fp/sve-test.S | 2 +
tools/testing/selftests/arm64/fp/za-test.S | 2 +
tools/testing/selftests/arm64/fp/zt-test.S | 2 +
tools/testing/selftests/arm64/gcs/.gitignore | 5 +
tools/testing/selftests/arm64/gcs/Makefile | 24 +
tools/testing/selftests/arm64/gcs/asm-offsets.h | 0
tools/testing/selftests/arm64/gcs/basic-gcs.c | 357 ++++++++++
tools/testing/selftests/arm64/gcs/gcs-locking.c | 200 ++++++
.../selftests/arm64/gcs/gcs-stress-thread.S | 311 +++++++++
tools/testing/selftests/arm64/gcs/gcs-stress.c | 530 +++++++++++++++
tools/testing/selftests/arm64/gcs/gcs-util.h | 100 +++
tools/testing/selftests/arm64/gcs/libc-gcs.c | 736 +++++++++++++++++++++
tools/testing/selftests/arm64/signal/.gitignore | 1 +
.../testing/selftests/arm64/signal/test_signals.c | 17 +-
.../testing/selftests/arm64/signal/test_signals.h | 6 +
.../selftests/arm64/signal/test_signals_utils.c | 32 +-
.../selftests/arm64/signal/test_signals_utils.h | 39 ++
.../arm64/signal/testcases/gcs_exception_fault.c | 62 ++
.../selftests/arm64/signal/testcases/gcs_frame.c | 88 +++
.../arm64/signal/testcases/gcs_write_fault.c | 67 ++
.../selftests/arm64/signal/testcases/testcases.c | 7 +
.../selftests/arm64/signal/testcases/testcases.h | 1 +
tools/testing/selftests/clone3/clone3_selftests.h | 26 +
tools/testing/selftests/ksft_shstk.h | 37 ++
tools/testing/selftests/kvm/aarch64/get-reg-list.c | 28 +
73 files changed, 4222 insertions(+), 40 deletions(-)
---
base-commit: 2d2c15fd64fcaba525a96e3198e4a4732680a49e
change-id: 20230303-arm64-gcs-e311ab0d8729
Best regards,
--
Mark Brown <broonie(a)kernel.org>
This small series includes fixes for creation of veth pairs for
networkless kernels & adds tests for turning the different network
interface features on and off in selftests/net/netdevice.sh script.
Tested using vng and compiles for network as well as networkless kernel.
Changes in v8:
Remove redundant variable initial_state as per feedback in v7.
Changes in v7:
https://lore.kernel.org/all/20240815105924.1389290-1-jain.abhinav177@gmail.…
Create a third patch in the series to do SKIP -> XFAIL replacement.
Add logic to incorporate XFAIL on setting IP address for veth pair.
Changes in v6:
https://lore.kernel.org/all/20240814191517.50466-1-jain.abhinav177@gmail.com
Use XFAIL for ethtool operations that are unsupported instead of SKIP.
Changes in v5:
https://lore.kernel.org/all/20240808122452.25683-1-jain.abhinav177@gmail.com
Rectify the syntax for ip add link.
Fix the veth_created condition check.
Changes in v4:
https://lore.kernel.org/all/20240807175717.7775-1-jain.abhinav177@gmail.com
Move veth creation/removal to the main shell script.
Tested using vng on a networkless kernel and the script works, sample
output below the changes.
Changes in v3:
https://lore.kernel.org/all/20240614113240.41550-1-jain.abhinav177@gmail.com
Add a check for netdev, create veth pair for testing.
Restore feature to its initial state.
Changes in v2:
https://lore.kernel.org/all/20240609132124.51683-1-jain.abhinav177@gmail.com
Remove tail usage; use read to parse the features from temp file.
v1:
https://lore.kernel.org/all/20240606212714.27472-1-jain.abhinav177@gmail.com
```
# selftests: net: netdevice.sh
# No valid network device found, creating veth pair
# PASS: veth0: set interface up
# PASS: veth0: set MAC address
# XFAIL: veth0: set IP address unsupported for veth*
# PASS: veth0: ethtool list features
# PASS: veth0: Turned off feature: rx-checksumming
# PASS: veth0: Turned on feature: rx-checksumming
# PASS: veth0: Restore feature rx-checksumming to initial state on
# Actual changes:
# tx-checksum-ip-generic: off
...
# PASS: veth0: Turned on feature: rx-udp-gro-forwarding
# PASS: veth0: Restore feature rx-udp-gro-forwarding to initial state off
# Cannot get register dump: Operation not supported
# XFAIL: veth0: ethtool dump not supported
# PASS: veth0: ethtool stats
# PASS: veth0: stop interface
```
Abhinav Jain (3):
selftests: net: Create veth pair for testing in networkless kernel
selftests: net: Add on/off checks for non-fixed features of interface
selftests: net: Use XFAIL for operations not supported by the driver
tools/testing/selftests/net/netdevice.sh | 61 ++++++++++++++++++++++--
1 file changed, 57 insertions(+), 4 deletions(-)
--
2.34.1
This patch series introduces a set of regression tests for various s390x
CPU subfunctions in KVM. The tests ensure that the KVM implementation accurately
reflects the behavior of actual CPU instructions for these subfunctions.
The series adds tests for a total of 15 instructions across five patches,
covering a range of operations including sorting, compression, and various
cryptographic functions. Each patch follows a consistent testing pattern:
1. Obtain the KVM_S390_VM_CPU_MACHINE_SUBFUNC attribute for the VM.
2. Execute the relevant asm instructions.
3. Compare KVM-reported results with direct instruction execution results.
Testing has been performed on s390x hardware with KVM support. All tests
pass successfully, verifying the correct implementation of these
subfunctions in KVM.
---
v2:
* Fix facility_bit type from bool to int
---
Hariharan Mari (5):
KVM: s390: selftests: Add regression tests for SORTL and DFLTCC CPU
subfunctions
KVM: s390: selftests: Add regression tests for PRNO, KDSA and KMA
crypto subfunctions
KVM: s390: selftests: Add regression tests for KMCTR, KMF, KMO and PCC
crypto subfunctions
KVM: s390: selftests: Add regression tests for KMAC, KMC, KM, KIMD and
KLMD crypto subfunctions
KVM: s390: selftests: Add regression tests for PLO subfunctions
tools/testing/selftests/kvm/Makefile | 1 +
.../selftests/kvm/include/s390x/facility.h | 50 +++
.../kvm/s390x/cpumodel_subfuncs_test.c | 343 ++++++++++++++++++
3 files changed, 394 insertions(+)
create mode 100644 tools/testing/selftests/kvm/include/s390x/facility.h
create mode 100644 tools/testing/selftests/kvm/s390x/cpumodel_subfuncs_test.c
--
2.45.2
From: Jeff Xu <jeffxu(a)chromium.org>
mremap doesn't allow relocate, expand, shrink across VMA boundaries,
refactor the code to check src address range before doing anything on
the destination, i.e. destination won't be unmapped, if src address
failed the boundaries check.
This also allows us to remove can_modify_mm from mremap.c, since
the src address must be single VMA, can_modify_vma is used.
It is likely this will improve the performance on mremap, previously
the code does sealing check using can_modify_mm for the src address range,
and the new code removed the loop (used by can_modify_mm).
In order to verify this patch doesn't regress on mremap, I added tests in
mseal_test, the test patch can be applied before mremap refactor patch or
checkin independently.
Also this patch doesn't change mseal's existing schematic: if sealing fail,
user can expect the src/dst address isn't updated. So this patch can be
applied regardless if we decided to go with current out-of-loop approach
or in-loop approach currently in discussion.
Regarding the perf test report by stress-ng [1] title:
8be7258aad: stress-ng.pagemove.page_remaps_per_sec -4.4% regression
The test is using below for testing:
stress-ng --timeout 60 --times --verify --metrics --no-rand-seed --pagemove 64
I can't repro this using ChromeOS, the pagemove test shows large value
of stddev and stderr, and can't reasonably refect the performance impact.
For example: I write a c program [2] to run the above pagemove test 10 times
and calculate the stddev, stderr, for 3 commits:
1> before mseal feature is added:
Ops/sec:
Mean : 3564.40
Std Dev : 2737.35 (76.80% of Mean)
Std Err : 865.63 (24.29% of Mean)
2> after mseal feature is added:
Ops/sec:
Mean : 2703.84
Std Dev : 2085.13 (77.12% of Mean)
Std Err : 659.38 (24.39% of Mean)
3> after current patch (mremap refactor)
Ops/sec:
Mean : 3603.67
Std Dev : 2422.22 (67.22% of Mean)
Std Err : 765.97 (21.26% of Mean)
The result shows 21%-24% stderr, this means whatever perf improvment/impact
there might be won't be measured correctly by this test.
This test machine has 32G memory, Intel(R) Celeron(R) 7305, 5 CPU.
And I reboot the machine before each test, and take the first 10 runs with
run_stress_ng 10
(I will run longer duration to see if test still shows large stdDev,StdErr)
[1] https://lore.kernel.org/lkml/202408041602.caa0372-oliver.sang@intel.com/
[2] https://github.com/peaktocreek/mmperf/blob/main/run_stress_ng.c
Jeff Xu (2):
mseal:selftest mremap across VMA boundaries.
mseal: refactor mremap to remove can_modify_mm
mm/internal.h | 24 ++
mm/mremap.c | 77 +++----
mm/mseal.c | 17 --
tools/testing/selftests/mm/mseal_test.c | 293 +++++++++++++++++++++++-
4 files changed, 353 insertions(+), 58 deletions(-)
--
2.46.0.76.ge559c4bf1a-goog
Adds a selftest that creates two virtual interfaces, assigns one to a
new namespace, and assigns IP addresses to both.
It listens on the destination interface using socat and configures a
dynamic target on netconsole, pointing to the destination IP address.
The test then checks if the message was received properly on the
destination interface.
Signed-off-by: Breno Leitao <leitao(a)debian.org>
Acked-by: Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
---
Changelog:
v5:
* Replace check_file_size() by "test -s" (Matthieu)
v4:
* Avoid sleeping in waiting for sockets and files (Matthieu Baerts)
* Some other improvements (Matthieu Baerts)
* Add configfs as a dependency (Jakub)
* https://lore.kernel.org/all/20240816132450.346744-1-leitao@debian.org/
v3:
* Defined CONFIGs in config file (Jakub)
* Identention fixes (Petr Machata)
* Use setup_ns in a better way (Matthieu Baerts)
* Add dependencies in TEST_INCLUDES (Hangbin Liu)
* https://lore.kernel.org/all/20240815095157.3064722-1-leitao@debian.org/
v2:
* Change the location of the path (Jakub)
* Move from veth to netdevsim
* Other small changes in dependency checks and cleanup
* https://lore.kernel.org/all/20240813183825.837091-1-leitao@debian.org/
v1:
* https://lore.kernel.org/all/ZqyUHN770pjSofTC@gmail.com/
MAINTAINERS | 1 +
tools/testing/selftests/drivers/net/Makefile | 5 +-
tools/testing/selftests/drivers/net/config | 4 +
.../selftests/drivers/net/netcons_basic.sh | 225 ++++++++++++++++++
4 files changed, 234 insertions(+), 1 deletion(-)
create mode 100755 tools/testing/selftests/drivers/net/netcons_basic.sh
diff --git a/MAINTAINERS b/MAINTAINERS
index 5dbf23cf11c8..9a371ddd8719 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -15772,6 +15772,7 @@ M: Breno Leitao <leitao(a)debian.org>
S: Maintained
F: Documentation/networking/netconsole.rst
F: drivers/net/netconsole.c
+F: tools/testing/selftests/drivers/net/netcons_basic.sh
NETDEVSIM
M: Jakub Kicinski <kuba(a)kernel.org>
diff --git a/tools/testing/selftests/drivers/net/Makefile b/tools/testing/selftests/drivers/net/Makefile
index e54f382bcb02..39fb97a8c1df 100644
--- a/tools/testing/selftests/drivers/net/Makefile
+++ b/tools/testing/selftests/drivers/net/Makefile
@@ -1,8 +1,11 @@
# SPDX-License-Identifier: GPL-2.0
-TEST_INCLUDES := $(wildcard lib/py/*.py)
+TEST_INCLUDES := $(wildcard lib/py/*.py) \
+ ../../net/net_helper.sh \
+ ../../net/lib.sh \
TEST_PROGS := \
+ netcons_basic.sh \
ping.py \
queues.py \
stats.py \
diff --git a/tools/testing/selftests/drivers/net/config b/tools/testing/selftests/drivers/net/config
index f6a58ce8a230..a2d8af60876d 100644
--- a/tools/testing/selftests/drivers/net/config
+++ b/tools/testing/selftests/drivers/net/config
@@ -1,2 +1,6 @@
CONFIG_IPV6=y
CONFIG_NETDEVSIM=m
+CONFIG_CONFIGFS_FS=y
+CONFIG_NETCONSOLE=m
+CONFIG_NETCONSOLE_DYNAMIC=y
+CONFIG_NETCONSOLE_EXTENDED_LOG=y
diff --git a/tools/testing/selftests/drivers/net/netcons_basic.sh b/tools/testing/selftests/drivers/net/netcons_basic.sh
new file mode 100755
index 000000000000..137875505663
--- /dev/null
+++ b/tools/testing/selftests/drivers/net/netcons_basic.sh
@@ -0,0 +1,225 @@
+#!/usr/bin/env bash
+# SPDX-License-Identifier: GPL-2.0
+
+# This test creates two netdevsim virtual interfaces, assigns one of them (the
+# "destination interface") to a new namespace, and assigns IP addresses to both
+# interfaces.
+#
+# It listens on the destination interface using socat and configures a dynamic
+# target on netconsole, pointing to the destination IP address.
+#
+# Finally, it checks whether the message was received properly on the
+# destination interface. Note that this test may pollute the kernel log buffer
+# (dmesg) and relies on dynamic configuration and namespaces being configured.
+#
+# Author: Breno Leitao <leitao(a)debian.org>
+
+set -euo pipefail
+
+SCRIPTDIR=$(dirname "$(readlink -e "${BASH_SOURCE[0]}")")
+
+# Simple script to test dynamic targets in netconsole
+SRCIF="" # to be populated later
+SRCIP=192.168.1.1
+DSTIF="" # to be populated later
+DSTIP=192.168.1.2
+
+PORT="6666"
+MSG="netconsole selftest"
+TARGET=$(mktemp -u netcons_XXXXX)
+NETCONS_CONFIGFS="/sys/kernel/config/netconsole"
+NETCONS_PATH="${NETCONS_CONFIGFS}"/"${TARGET}"
+# NAMESPACE will be populated by setup_ns with a random value
+NAMESPACE=""
+
+# IDs for netdevsim
+NSIM_DEV_1_ID=$((256 + RANDOM % 256))
+NSIM_DEV_2_ID=$((512 + RANDOM % 256))
+
+# Used to create and delete namespaces
+source "${SCRIPTDIR}"/../../net/lib.sh
+source "${SCRIPTDIR}"/../../net/net_helper.sh
+
+# Create netdevsim interfaces
+create_ifaces() {
+ local NSIM_DEV_SYS_NEW=/sys/bus/netdevsim/new_device
+
+ echo "$NSIM_DEV_2_ID" > "$NSIM_DEV_SYS_NEW"
+ echo "$NSIM_DEV_1_ID" > "$NSIM_DEV_SYS_NEW"
+ udevadm settle 2> /dev/null || true
+
+ local NSIM1=/sys/bus/netdevsim/devices/netdevsim"$NSIM_DEV_1_ID"
+ local NSIM2=/sys/bus/netdevsim/devices/netdevsim"$NSIM_DEV_2_ID"
+
+ # These are global variables
+ SRCIF=$(find "$NSIM1"/net -maxdepth 1 -type d ! \
+ -path "$NSIM1"/net -exec basename {} \;)
+ DSTIF=$(find "$NSIM2"/net -maxdepth 1 -type d ! \
+ -path "$NSIM2"/net -exec basename {} \;)
+}
+
+link_ifaces() {
+ local NSIM_DEV_SYS_LINK="/sys/bus/netdevsim/link_device"
+ local SRCIF_IFIDX=$(cat /sys/class/net/"$SRCIF"/ifindex)
+ local DSTIF_IFIDX=$(cat /sys/class/net/"$DSTIF"/ifindex)
+
+ exec {NAMESPACE_FD}</var/run/netns/"${NAMESPACE}"
+ exec {INITNS_FD}</proc/self/ns/net
+
+ # Bind the dst interface to namespace
+ ip link set "${DSTIF}" netns "${NAMESPACE}"
+
+ # Linking one device to the other one (on the other namespace}
+ if ! echo "${INITNS_FD}:$SRCIF_IFIDX $NAMESPACE_FD:$DSTIF_IFIDX" > $NSIM_DEV_SYS_LINK
+ then
+ echo "linking netdevsim1 with netdevsim2 should succeed"
+ cleanup
+ exit "${ksft_skip}"
+ fi
+}
+
+function configure_ip() {
+ # Configure the IPs for both interfaces
+ ip netns exec "${NAMESPACE}" ip addr add "${DSTIP}"/24 dev "${DSTIF}"
+ ip netns exec "${NAMESPACE}" ip link set "${DSTIF}" up
+
+ ip addr add "${SRCIP}"/24 dev "${SRCIF}"
+ ip link set "${SRCIF}" up
+}
+
+function set_network() {
+ # setup_ns function is coming from lib.sh
+ setup_ns NAMESPACE
+
+ # Create both interfaces, and assign the destination to a different
+ # namespace
+ create_ifaces
+
+ # Link both interfaces back to back
+ link_ifaces
+
+ configure_ip
+}
+
+function create_dynamic_target() {
+ DSTMAC=$(ip netns exec "${NAMESPACE}" \
+ ip link show "${DSTIF}" | awk '/ether/ {print $2}')
+
+ # Create a dynamic target
+ mkdir "${NETCONS_PATH}"
+
+ echo "${DSTIP}" > "${NETCONS_PATH}"/remote_ip
+ echo "${SRCIP}" > "${NETCONS_PATH}"/local_ip
+ echo "${DSTMAC}" > "${NETCONS_PATH}"/remote_mac
+ echo "${SRCIF}" > "${NETCONS_PATH}"/dev_name
+
+ echo 1 > "${NETCONS_PATH}"/enabled
+}
+
+function cleanup() {
+ local NSIM_DEV_SYS_DEL="/sys/bus/netdevsim/del_device"
+
+ # delete netconsole dynamic reconfiguration
+ echo 0 > "${NETCONS_PATH}"/enabled
+ # Remove the configfs entry
+ rmdir "${NETCONS_PATH}"
+
+ # Delete netdevsim devices
+ echo "$NSIM_DEV_2_ID" > "$NSIM_DEV_SYS_DEL"
+ echo "$NSIM_DEV_1_ID" > "$NSIM_DEV_SYS_DEL"
+
+ # this is coming from lib.sh
+ cleanup_all_ns
+}
+
+function listen_port_and_save_to() {
+ local OUTPUT=${1}
+ # Just wait for 2 seconds
+ timeout 2 ip netns exec "${NAMESPACE}" \
+ socat UDP-LISTEN:"${PORT}",fork "${OUTPUT}"
+}
+
+function validate_result() {
+ local TMPFILENAME="$1"
+
+ # Check if the file exists
+ if [ ! -f "$TMPFILENAME" ]; then
+ echo "FAIL: File was not generated." >&2
+ exit "${ksft_fail}"
+ fi
+
+ if ! grep -q "${MSG}" "${TMPFILENAME}"; then
+ echo "FAIL: ${MSG} not found in ${TMPFILENAME}" >&2
+ cat "${TMPFILENAME}" >&2
+ exit "${ksft_fail}"
+ fi
+
+ # Delete the file once it is validated, otherwise keep it
+ # for debugging purposes
+ rm "${TMPFILENAME}"
+ exit "${ksft_pass}"
+}
+
+function check_for_dependencies() {
+ if [ "$(id -u)" -ne 0 ]; then
+ echo "This test must be run as root" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which socat > /dev/null ; then
+ echo "SKIP: socat(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which ip > /dev/null ; then
+ echo "SKIP: ip(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which udevadm > /dev/null ; then
+ echo "SKIP: udevadm(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if [ ! -d "${NETCONS_CONFIGFS}" ]; then
+ echo "SKIP: directory ${NETCONS_CONFIGFS} does not exist. Check if NETCONSOLE_DYNAMIC is enabled" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ip link show "${DSTIF}" 2> /dev/null; then
+ echo "SKIP: interface ${DSTIF} exists in the system. Not overwriting it." >&2
+ exit "${ksft_skip}"
+ fi
+}
+
+# ========== #
+# Start here #
+# ========== #
+modprobe netdevsim 2> /dev/null || true
+modprobe netconsole 2 > /dev/null || true
+
+# The content of kmsg will be save to the following file
+OUTPUT_FILE="/tmp/${TARGET}"
+
+# Check for basic system dependency and exit if not found
+check_for_dependencies
+# Set current loglevel to KERN_INFO(6), and default to KERN_NOTICE(5)
+echo "6 5" > /proc/sys/kernel/printk
+# Remove the namespace, interfaces and netconsole target on exit
+trap cleanup EXIT
+# Create one namespace and two interfaces
+set_network
+# Create a dynamic target for netconsole
+create_dynamic_target
+# Listed for netconsole port inside the namespace and destination interface
+listen_port_and_save_to "${OUTPUT_FILE}" &
+# Wait for socat to start and listen to the port.
+wait_local_port_listen "${NAMESPACE}" "${PORT}" udp
+# Send the message
+echo "${MSG}: ${TARGET}" > /dev/kmsg
+# Wait until socat saves the file to disk
+busywait "${BUSYWAIT_TIMEOUT}" test -s "${OUTPUT_FILE}"
+
+# Make sure the message was received in the dst part
+# and exit
+validate_result "${OUTPUT_FILE}"
--
2.43.5
From: "Steven Rostedt (Google)" <rostedt(a)goodmis.org>
The function "scheduler_tick" was renamed to "sched_tick" and a selftest
that used that function for testing function trace filtering used that
function as part of the test.
But the change causes it to fail when run on older kernels. As tests
should not fail on older kernels, add a check to see which name is
available before testing.
Fixes: 86dd6c04ef9f2 ("sched/balancing: Rename scheduler_tick() => sched_tick()")
Signed-off-by: Steven Rostedt (Google) <rostedt(a)goodmis.org>
---
.../ftrace/test.d/ftrace/func_set_ftrace_file.tc | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/tools/testing/selftests/ftrace/test.d/ftrace/func_set_ftrace_file.tc b/tools/testing/selftests/ftrace/test.d/ftrace/func_set_ftrace_file.tc
index 073a748b9380..263f6b798c85 100644
--- a/tools/testing/selftests/ftrace/test.d/ftrace/func_set_ftrace_file.tc
+++ b/tools/testing/selftests/ftrace/test.d/ftrace/func_set_ftrace_file.tc
@@ -19,7 +19,14 @@ fail() { # mesg
FILTER=set_ftrace_filter
FUNC1="schedule"
-FUNC2="sched_tick"
+if grep '^sched_tick\b' available_filter_functions; then
+ FUNC2="sched_tick"
+elif grep '^scheduler_tick\b' available_filter_functions; then
+ FUNC2="scheduler_tick"
+else
+ exit_unresolved
+fi
+
ALL_FUNCS="#### all functions enabled ####"
--
2.43.0
A few tests check if nettest exists in the $PATH before adding
$PWD to $PATH and re-checking. They don't discard stderr on
the first check (and nettest is built as part of selftests,
so it's pretty normal for it to not be available in system $PATH).
This leads to output noise:
which: no nettest in (/home/virtme/tools/fs/bin:/home/virtme/tools/fs/sbin:/home/virtme/tools/fs/usr/bin:/home/virtme/tools/fs/usr/sbin:/bin:/sbin:/usr/bin:/usr/sbin:/usr/local/bin)
Add a common helper for the check which does silence stderr.
There is another small functional change hiding here, because pmtu.sh
and fib_rule_tests.sh used to return from the test case rather than
completely exit. Building nettest is not hard, there should be no need
to maintain the ability to selectively skip cases in its absence.
Signed-off-by: Jakub Kicinski <kuba(a)kernel.org>
---
v2:
- fold in the changes from Ido
v1: https://lore.kernel.org/20240817183848.658443-1-kuba@kernel.org
CC: shuah(a)kernel.org
CC: linux-kselftest(a)vger.kernel.org
CC: Ido Schimmel <idosch(a)idosch.org>
---
tools/testing/selftests/net/fcnal-test.sh | 9 +----
tools/testing/selftests/net/fib_rule_tests.sh | 37 +------------------
tools/testing/selftests/net/lib.sh | 15 ++++++++
tools/testing/selftests/net/pmtu.sh | 8 +---
tools/testing/selftests/net/settings | 1 +
.../selftests/net/unicast_extensions.sh | 9 +----
.../selftests/net/vrf_route_leaking.sh | 3 +-
7 files changed, 23 insertions(+), 59 deletions(-)
diff --git a/tools/testing/selftests/net/fcnal-test.sh b/tools/testing/selftests/net/fcnal-test.sh
index 386ebd829df5..899dbad0104b 100755
--- a/tools/testing/selftests/net/fcnal-test.sh
+++ b/tools/testing/selftests/net/fcnal-test.sh
@@ -4304,14 +4304,7 @@ elif [ "$TESTS" = "ipv6" ]; then
TESTS="$TESTS_IPV6"
fi
-# nettest can be run from PATH or from same directory as this selftest
-if ! which nettest >/dev/null; then
- PATH=$PWD:$PATH
- if ! which nettest >/dev/null; then
- echo "'nettest' command not found; skipping tests"
- exit $ksft_skip
- fi
-fi
+check_gen_prog "nettest"
declare -i nfail=0
declare -i nsuccess=0
diff --git a/tools/testing/selftests/net/fib_rule_tests.sh b/tools/testing/selftests/net/fib_rule_tests.sh
index 89034c5b69dc..53c5c1ad437e 100755
--- a/tools/testing/selftests/net/fib_rule_tests.sh
+++ b/tools/testing/selftests/net/fib_rule_tests.sh
@@ -51,31 +51,6 @@ log_test()
fi
}
-check_nettest()
-{
- if which nettest > /dev/null 2>&1; then
- return 0
- fi
-
- # Add the selftest directory to PATH if not already done
- if [ "${SELFTEST_PATH}" = "" ]; then
- SELFTEST_PATH="$(dirname $0)"
- PATH="${PATH}:${SELFTEST_PATH}"
-
- # Now retry with the new path
- if which nettest > /dev/null 2>&1; then
- return 0
- fi
-
- if [ "${ret}" -eq 0 ]; then
- ret="${ksft_skip}"
- fi
- echo "nettest not found (try 'make -C ${SELFTEST_PATH} nettest')"
- fi
-
- return 1
-}
-
setup()
{
set -e
@@ -317,11 +292,6 @@ fib_rule6_connect_test()
echo
echo "IPv6 FIB rule connect tests"
- if ! check_nettest; then
- echo "SKIP: Could not run test without nettest tool"
- return
- fi
-
setup_peer
$IP -6 rule add dsfield 0x04 table $RTABLE_PEER
@@ -516,11 +486,6 @@ fib_rule4_connect_test()
echo
echo "IPv4 FIB rule connect tests"
- if ! check_nettest; then
- echo "SKIP: Could not run test without nettest tool"
- return
- fi
-
setup_peer
$IP -4 rule add dsfield 0x04 table $RTABLE_PEER
@@ -584,6 +549,8 @@ if [ ! -x "$(command -v ip)" ]; then
exit $ksft_skip
fi
+check_gen_prog "nettest"
+
# start clean
cleanup &> /dev/null
setup
diff --git a/tools/testing/selftests/net/lib.sh b/tools/testing/selftests/net/lib.sh
index 8ee4489238ca..be8707bfb46e 100644
--- a/tools/testing/selftests/net/lib.sh
+++ b/tools/testing/selftests/net/lib.sh
@@ -125,6 +125,21 @@ slowwait_for_counter()
slowwait "$timeout" until_counter_is ">= $((base + delta))" "$@"
}
+# Check for existence of tools which are built as part of selftests
+# but may also already exist in $PATH
+check_gen_prog()
+{
+ local prog_name=$1; shift
+
+ if ! which $prog_name >/dev/null 2>/dev/null; then
+ PATH=$PWD:$PATH
+ if ! which $prog_name >/dev/null; then
+ echo "'$prog_name' command not found; skipping tests"
+ exit $ksft_skip
+ fi
+ fi
+}
+
remove_ns_list()
{
local item=$1
diff --git a/tools/testing/selftests/net/pmtu.sh b/tools/testing/selftests/net/pmtu.sh
index 24a50622406c..569bce8b6383 100755
--- a/tools/testing/selftests/net/pmtu.sh
+++ b/tools/testing/selftests/net/pmtu.sh
@@ -681,13 +681,7 @@ setup_xfrm() {
}
setup_nettest_xfrm() {
- if ! which nettest >/dev/null; then
- PATH=$PWD:$PATH
- if ! which nettest >/dev/null; then
- echo "'nettest' command not found; skipping tests"
- return 1
- fi
- fi
+ check_gen_prog "nettest"
[ ${1} -eq 6 ] && proto="-6" || proto=""
port=${2}
diff --git a/tools/testing/selftests/net/settings b/tools/testing/selftests/net/settings
index ed8418e8217a..a38764182822 100644
--- a/tools/testing/selftests/net/settings
+++ b/tools/testing/selftests/net/settings
@@ -1 +1,2 @@
timeout=3600
+profile=1
diff --git a/tools/testing/selftests/net/unicast_extensions.sh b/tools/testing/selftests/net/unicast_extensions.sh
index f52aa5f7da52..3e751234ccfe 100755
--- a/tools/testing/selftests/net/unicast_extensions.sh
+++ b/tools/testing/selftests/net/unicast_extensions.sh
@@ -30,14 +30,7 @@
source lib.sh
-# nettest can be run from PATH or from same directory as this selftest
-if ! which nettest >/dev/null; then
- PATH=$PWD:$PATH
- if ! which nettest >/dev/null; then
- echo "'nettest' command not found; skipping tests"
- exit $ksft_skip
- fi
-fi
+check_gen_prog "nettest"
result=0
diff --git a/tools/testing/selftests/net/vrf_route_leaking.sh b/tools/testing/selftests/net/vrf_route_leaking.sh
index 152171fb1fc8..e9c2f71da207 100755
--- a/tools/testing/selftests/net/vrf_route_leaking.sh
+++ b/tools/testing/selftests/net/vrf_route_leaking.sh
@@ -59,7 +59,6 @@
# while it is forwarded between different vrfs.
source lib.sh
-PATH=$PWD:$PWD/tools/testing/selftests/net:$PATH
VERBOSE=0
PAUSE_ON_FAIL=no
DEFAULT_TTYPE=sym
@@ -636,6 +635,8 @@ EOF
# Some systems don't have a ping6 binary anymore
command -v ping6 > /dev/null 2>&1 && ping6=$(command -v ping6) || ping6=$(command -v ping)
+check_gen_prog "nettest"
+
TESTS_IPV4="ipv4_ping_ttl ipv4_traceroute ipv4_ping_frag ipv4_ping_local ipv4_tcp_local
ipv4_udp_local ipv4_ping_ttl_asym ipv4_traceroute_asym"
TESTS_IPV6="ipv6_ping_ttl ipv6_traceroute ipv6_ping_local ipv6_tcp_local ipv6_udp_local
--
2.46.0
Here are more fixes for the MPTCP in-kernel path-manager. In this
series, the fixes are around the endpoint IDs not being reusable for
on-going connections when re-creating endpoints with previously used IDs.
- Patch 1 fixes this case for endpoints being used to send ADD_ADDR.
Patch 2 validates this fix. The issue is present since v5.10.
- Patch 3 fixes this case for endpoints being used to establish new
subflows. Patch 4 validates this fix. The issue is present since v5.10.
- Patch 5 fixes this case when all endpoints are flushed. Patch 6
validates this fix. The issue is present since v5.13.
- Patch 7 removes a helper that is confusing, and introduced in v5.10.
It helps simplifying the next patches.
- Patch 8 makes sure a 'subflow' counter is only decremented when
removing a 'subflow' endpoint. Can be backported up to v5.13.
- Patch 9 is similar, but for a 'signal' counter. Can be backported up
to v5.10.
- Patch 10 checks the last max accepted ADD_ADDR limit before accepting
new ADD_ADDR. For v5.10 as well.
- Patch 11 removes a wrong restriction for the userspace PM, added
during a refactoring in v6.5.
- Patch 12 makes sure the fullmesh mode sets the ID 0 when a new subflow
using the source address of the initial subflow is created. Patch 13
covers this case. This issue is present since v5.15.
- Patch 14 avoid possible UaF when selecting an address from the
endpoints list.
Signed-off-by: Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
---
Matthieu Baerts (NGI0) (14):
mptcp: pm: re-using ID of unused removed ADD_ADDR
selftests: mptcp: join: check re-using ID of unused ADD_ADDR
mptcp: pm: re-using ID of unused removed subflows
selftests: mptcp: join: check re-using ID of closed subflow
mptcp: pm: re-using ID of unused flushed subflows
selftests: mptcp: join: test for flush/re-add endpoints
mptcp: pm: remove mptcp_pm_remove_subflow()
mptcp: pm: only mark 'subflow' endp as available
mptcp: pm: only decrement add_addr_accepted for MPJ req
mptcp: pm: check add_addr_accept_max before accepting new ADD_ADDR
mptcp: pm: only in-kernel cannot have entries with ID 0
mptcp: pm: fullmesh: select the right ID later
selftests: mptcp: join: validate fullmesh endp on 1st sf
mptcp: pm: avoid possible UaF when selecting endp
net/mptcp/pm.c | 13 ---
net/mptcp/pm_netlink.c | 142 ++++++++++++++++--------
net/mptcp/protocol.h | 3 -
tools/testing/selftests/net/mptcp/mptcp_join.sh | 76 +++++++++++--
4 files changed, 160 insertions(+), 74 deletions(-)
---
base-commit: 565d121b69980637f040eb4d84289869cdaabedf
change-id: 20240819-net-mptcp-pm-reusing-id-eb08827b7be6
Best regards,
--
Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
Looking at timestamped output of netdev CI reveals that
most of the time in forwarding tests for custom route
hashing is spent on a single case, namely the test which
uses ping (mausezahn does not support flow labels).
On a non-debug kernel we spend 714 of 730 total test
runtime (97%) on this test case. While having flow label
support in a traffic gen tool / mausezahn would be best,
we can significantly speed up the loop by putting ip vrf exec
outside of the iteration.
In a test of 1000 pings using a normal loop takes 50 seconds
to finish. While using:
ip vrf exec $vrf sh -c "$loop-body"
takes 12 seconds (1/4 of the time).
Some of the slowness is likely due to our inefficient virtualization
setup, but even on my laptop running "ip link help" 16k times takes
25-30 seconds, so I think it's worth optimizing even for fastest
setups.
Signed-off-by: Jakub Kicinski <kuba(a)kernel.org>
---
CC: shuah(a)kernel.org
CC: idosch(a)nvidia.com
CC: linux-kselftest(a)vger.kernel.org
---
.../selftests/net/forwarding/custom_multipath_hash.sh | 8 ++++----
.../selftests/net/forwarding/gre_custom_multipath_hash.sh | 8 ++++----
.../net/forwarding/ip6gre_custom_multipath_hash.sh | 8 ++++----
3 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/tools/testing/selftests/net/forwarding/custom_multipath_hash.sh b/tools/testing/selftests/net/forwarding/custom_multipath_hash.sh
index 1783c10215e5..7d531f7091e6 100755
--- a/tools/testing/selftests/net/forwarding/custom_multipath_hash.sh
+++ b/tools/testing/selftests/net/forwarding/custom_multipath_hash.sh
@@ -224,10 +224,10 @@ send_dst_ipv6()
send_flowlabel()
{
# Generate 16384 echo requests, each with a random flow label.
- for _ in $(seq 1 16384); do
- ip vrf exec v$h1 \
- $PING6 2001:db8:4::2 -F 0 -c 1 -q >/dev/null 2>&1
- done
+ ip vrf exec v$h1 sh -c \
+ "for _ in {1..16384}; do \
+ $PING6 2001:db8:4::2 -F 0 -c 1 -q >/dev/null 2>&1; \
+ done"
}
send_src_udp6()
diff --git a/tools/testing/selftests/net/forwarding/gre_custom_multipath_hash.sh b/tools/testing/selftests/net/forwarding/gre_custom_multipath_hash.sh
index 9788bd0f6e8b..dda11a4a9450 100755
--- a/tools/testing/selftests/net/forwarding/gre_custom_multipath_hash.sh
+++ b/tools/testing/selftests/net/forwarding/gre_custom_multipath_hash.sh
@@ -319,10 +319,10 @@ send_dst_ipv6()
send_flowlabel()
{
# Generate 16384 echo requests, each with a random flow label.
- for _ in $(seq 1 16384); do
- ip vrf exec v$h1 \
- $PING6 2001:db8:2::2 -F 0 -c 1 -q >/dev/null 2>&1
- done
+ ip vrf exec v$h1 sh -c \
+ "for _ in {1..16384}; do \
+ $PING6 2001:db8:2::2 -F 0 -c 1 -q >/dev/null 2>&1; \
+ done"
}
send_src_udp6()
diff --git a/tools/testing/selftests/net/forwarding/ip6gre_custom_multipath_hash.sh b/tools/testing/selftests/net/forwarding/ip6gre_custom_multipath_hash.sh
index 2ab9eaaa5532..e28b4a079e52 100755
--- a/tools/testing/selftests/net/forwarding/ip6gre_custom_multipath_hash.sh
+++ b/tools/testing/selftests/net/forwarding/ip6gre_custom_multipath_hash.sh
@@ -321,10 +321,10 @@ send_dst_ipv6()
send_flowlabel()
{
# Generate 16384 echo requests, each with a random flow label.
- for _ in $(seq 1 16384); do
- ip vrf exec v$h1 \
- $PING6 2001:db8:2::2 -F 0 -c 1 -q >/dev/null 2>&1
- done
+ ip vrf exec v$h1 sh -c \
+ "for _ in {1..16384}; do \
+ $PING6 2001:db8:2::2 -F 0 -c 1 -q >/dev/null 2>&1; \
+ done"
}
send_src_udp6()
--
2.46.0
kunit_driver_create() accepts a name for the driver, but does not copy
it, so if that name is either on the stack, or otherwise freed, we end
up with a use-after-free when the driver is cleaned up.
Instead, strdup() the name, and manage it as another KUnit allocation.
As there was no existing kunit_kstrdup(), we add one. Further, add a
kunit_ variant of strdup_const() and kfree_const(), so we don't need to
allocate and manage the string in the majority of cases where it's a
constant.
However, these are inline functions, and is_kernel_rodata() only works
for built-in code. This causes problems in two cases:
- If kunit is built as a module, __{start,end}_rodata is not defined.
- If a kunit test using these functions is built as a module, it will
suffer the same fate.
This fixes a KASAN splat with overflow.overflow_allocation_test, when
built as a module.
Restrict the is_kernel_rodata() case to when KUnit is built as a module,
which fixes the first case, at the cost of losing the optimisation.
Also, make kunit_{kstrdup,kfree}_const non-inline, so that other modules
using them will not accidentally depend on is_kernel_rodata(). If KUnit
is built-in, they'll benefit from the optimisation, if KUnit is not,
they won't, but the string will be properly duplicated.
Fixes: d03c720e03bd ("kunit: Add APIs for managing devices")
Reported-by: Nico Pache <npache(a)redhat.com>
Closes: https://groups.google.com/g/kunit-dev/c/81V9b9QYON0
Reviewed-by: Kees Cook <kees(a)kernel.org>
Reviewed-by: Maxime Ripard <mripard(a)kernel.org>
Reviewed-by: Rae Moar <rmoar(a)google.com>
Signed-off-by: David Gow <davidgow(a)google.com>
---
This is a combination of the previous version of this patch with the
follow-up fix "kunit: Fix kunit_kstrdup_const() with modules".
kunit_kstrdup_const() now falls back to kstrdup() if KUnit is built as a
module, and is no longer inlined. This should fix the issues we'd seen
before.
I've not tried doing something fancy by looking at module rodata
sections: it might be a possible optimisation, but it seems like it'd
overcomplicate things for this initial change. If we hit a KUnit test
where this is a bottleneck (or if I have some more spare time), we can
look into it.
The overflow_kunit test has been fixed independently to not rely on this
anyway, so there shouldn't be any current cases of this causing issues,
but it's worth making the API robust regardless.
Changes since previous version:
https://lore.kernel.org/linux-kselftest/20240731070207.3918687-1-davidgow@g…
- Fix module support by integrating:
https://lore.kernel.org/linux-kselftest/20240806020136.3481593-1-davidgow@g…
---
include/kunit/test.h | 48 ++++++++++++++++++++++++++++++++++++++++++++
lib/kunit/device.c | 7 +++++--
lib/kunit/test.c | 19 ++++++++++++++++++
3 files changed, 72 insertions(+), 2 deletions(-)
diff --git a/include/kunit/test.h b/include/kunit/test.h
index e2a1f0928e8b..5ac237c949a0 100644
--- a/include/kunit/test.h
+++ b/include/kunit/test.h
@@ -28,6 +28,7 @@
#include <linux/types.h>
#include <asm/rwonce.h>
+#include <asm/sections.h>
/* Static key: true if any KUnit tests are currently running */
DECLARE_STATIC_KEY_FALSE(kunit_running);
@@ -480,6 +481,53 @@ static inline void *kunit_kcalloc(struct kunit *test, size_t n, size_t size, gfp
return kunit_kmalloc_array(test, n, size, gfp | __GFP_ZERO);
}
+
+/**
+ * kunit_kfree_const() - conditionally free test managed memory
+ * @x: pointer to the memory
+ *
+ * Calls kunit_kfree() only if @x is not in .rodata section.
+ * See kunit_kstrdup_const() for more information.
+ */
+void kunit_kfree_const(struct kunit *test, const void *x);
+
+/**
+ * kunit_kstrdup() - Duplicates a string into a test managed allocation.
+ *
+ * @test: The test context object.
+ * @str: The NULL-terminated string to duplicate.
+ * @gfp: flags passed to underlying kmalloc().
+ *
+ * See kstrdup() and kunit_kmalloc_array() for more information.
+ */
+static inline char *kunit_kstrdup(struct kunit *test, const char *str, gfp_t gfp)
+{
+ size_t len;
+ char *buf;
+
+ if (!str)
+ return NULL;
+
+ len = strlen(str) + 1;
+ buf = kunit_kmalloc(test, len, gfp);
+ if (buf)
+ memcpy(buf, str, len);
+ return buf;
+}
+
+/**
+ * kunit_kstrdup_const() - Conditionally duplicates a string into a test managed allocation.
+ *
+ * @test: The test context object.
+ * @str: The NULL-terminated string to duplicate.
+ * @gfp: flags passed to underlying kmalloc().
+ *
+ * Calls kunit_kstrdup() only if @str is not in the rodata section. Must be freed with
+ * kunit_kfree_const() -- not kunit_kfree().
+ * See kstrdup_const() and kunit_kmalloc_array() for more information.
+ */
+const char *kunit_kstrdup_const(struct kunit *test, const char *str, gfp_t gfp);
+
/**
* kunit_vm_mmap() - Allocate KUnit-tracked vm_mmap() area
* @test: The test context object.
diff --git a/lib/kunit/device.c b/lib/kunit/device.c
index 25c81ed465fb..520c1fccee8a 100644
--- a/lib/kunit/device.c
+++ b/lib/kunit/device.c
@@ -89,7 +89,7 @@ struct device_driver *kunit_driver_create(struct kunit *test, const char *name)
if (!driver)
return ERR_PTR(err);
- driver->name = name;
+ driver->name = kunit_kstrdup_const(test, name, GFP_KERNEL);
driver->bus = &kunit_bus_type;
driver->owner = THIS_MODULE;
@@ -192,8 +192,11 @@ void kunit_device_unregister(struct kunit *test, struct device *dev)
const struct device_driver *driver = to_kunit_device(dev)->driver;
kunit_release_action(test, device_unregister_wrapper, dev);
- if (driver)
+ if (driver) {
+ const char *driver_name = driver->name;
kunit_release_action(test, driver_unregister_wrapper, (void *)driver);
+ kunit_kfree_const(test, driver_name);
+ }
}
EXPORT_SYMBOL_GPL(kunit_device_unregister);
diff --git a/lib/kunit/test.c b/lib/kunit/test.c
index e8b1b52a19ab..089c832e3cdb 100644
--- a/lib/kunit/test.c
+++ b/lib/kunit/test.c
@@ -874,6 +874,25 @@ void kunit_kfree(struct kunit *test, const void *ptr)
}
EXPORT_SYMBOL_GPL(kunit_kfree);
+void kunit_kfree_const(struct kunit *test, const void *x)
+{
+#if !IS_MODULE(CONFIG_KUNIT)
+ if (!is_kernel_rodata((unsigned long)x))
+#endif
+ kunit_kfree(test, x);
+}
+EXPORT_SYMBOL_GPL(kunit_kfree_const);
+
+const char *kunit_kstrdup_const(struct kunit *test, const char *str, gfp_t gfp)
+{
+#if !IS_MODULE(CONFIG_KUNIT)
+ if (is_kernel_rodata((unsigned long)str))
+ return str;
+#endif
+ return kunit_kstrdup(test, str, gfp);
+}
+EXPORT_SYMBOL_GPL(kunit_kstrdup_const);
+
void kunit_cleanup(struct kunit *test)
{
struct kunit_resource *res;
--
2.46.0.184.g6999bdac58-goog
xtheadvector is a custom extension that is based upon riscv vector
version 0.7.1 [1]. All of the vector routines have been modified to
support this alternative vector version based upon whether xtheadvector
was determined to be supported at boot.
vlenb is not supported on the existing xtheadvector hardware, so a
devicetree property thead,vlenb is added to provide the vlenb to Linux.
There is a new hwprobe key RISCV_HWPROBE_KEY_VENDOR_EXT_THEAD_0 that is
used to request which thead vendor extensions are supported on the
current platform. This allows future vendors to allocate hwprobe keys
for their vendor.
Support for xtheadvector is also added to the vector kselftests.
Signed-off-by: Charlie Jenkins <charlie(a)rivosinc.com>
[1] https://github.com/T-head-Semi/thead-extension-spec/blob/95358cb2cca9489361…
---
This series is a continuation of a different series that was fragmented
into two other series in an attempt to get part of it merged in the 6.10
merge window. The split-off series did not get merged due to a NAK on
the series that added the generic riscv,vlenb devicetree entry. This
series has converted riscv,vlenb to thead,vlenb to remedy this issue.
The original series is titled "riscv: Support vendor extensions and
xtheadvector" [3].
The series titled "riscv: Extend cpufeature.c to detect vendor
extensions" is still under development and this series is based on that
series! [4]
I have tested this with an Allwinner Nezha board. I ran into issues
booting the board after 6.9-rc1 so I applied these patches to 6.8. There
are a couple of minor merge conflicts that do arrise when doing that, so
please let me know if you have been able to boot this board with a 6.9
kernel. I used SkiffOS [1] to manage building the image, but upgraded
the U-Boot version to Samuel Holland's more up-to-date version [2] and
changed out the device tree used by U-Boot with the device trees that
are present in upstream linux and this series. Thank you Samuel for all
of the work you did to make this task possible.
[1] https://github.com/skiffos/SkiffOS/tree/master/configs/allwinner/nezha
[2] https://github.com/smaeul/u-boot/commit/2e89b706f5c956a70c989cd31665f1429e9…
[3] https://lore.kernel.org/all/20240503-dev-charlie-support_thead_vector_6_9-v…
[4] https://lore.kernel.org/lkml/20240719-support_vendor_extensions-v3-4-0af758…
---
Changes in v9:
- Rebase onto palmer's for-next
- Fix sparse error in arch/riscv/kernel/vendor_extensions/thead.c
- Fix maybe-uninitialized warning in arch/riscv/include/asm/vendor_extensions/vendor_hwprobe.h
- Wrap some long lines
- Link to v8: https://lore.kernel.org/r/20240724-xtheadvector-v8-0-cf043168e137@rivosinc.…
Changes in v8:
- Rebase onto palmer's for-next
- Link to v7: https://lore.kernel.org/r/20240724-xtheadvector-v7-0-b741910ada3e@rivosinc.…
Changes in v7:
- Add defs for has_xtheadvector_no_alternatives() and has_xtheadvector()
when vector disabled. (Palmer)
- Link to v6: https://lore.kernel.org/r/20240722-xtheadvector-v6-0-c9af0130fa00@rivosinc.…
Changes in v6:
- Fix return type of is_vector_supported()/is_xthead_supported() to be bool
- Link to v5: https://lore.kernel.org/r/20240719-xtheadvector-v5-0-4b485fc7d55f@rivosinc.…
Changes in v5:
- Rebase on for-next
- Link to v4: https://lore.kernel.org/r/20240702-xtheadvector-v4-0-2bad6820db11@rivosinc.…
Changes in v4:
- Replace inline asm with C (Samuel)
- Rename VCSRs to CSRs (Samuel)
- Replace .insn directives with .4byte directives
- Link to v3: https://lore.kernel.org/r/20240619-xtheadvector-v3-0-bff39eb9668e@rivosinc.…
Changes in v3:
- Add back Heiko's signed-off-by (Conor)
- Mark RISCV_HWPROBE_KEY_VENDOR_EXT_THEAD_0 as a bitmask
- Link to v2: https://lore.kernel.org/r/20240610-xtheadvector-v2-0-97a48613ad64@rivosinc.…
Changes in v2:
- Removed extraneous references to "riscv,vlenb" (Jess)
- Moved declaration of "thead,vlenb" into cpus.yaml and added
restriction that it's only applicable to thead cores (Conor)
- Check CONFIG_RISCV_ISA_XTHEADVECTOR instead of CONFIG_RISCV_ISA_V for
thead,vlenb (Jess)
- Fix naming of hwprobe variables (Evan)
- Link to v1: https://lore.kernel.org/r/20240609-xtheadvector-v1-0-3fe591d7f109@rivosinc.…
---
Charlie Jenkins (12):
dt-bindings: riscv: Add xtheadvector ISA extension description
dt-bindings: cpus: add a thead vlen register length property
riscv: dts: allwinner: Add xtheadvector to the D1/D1s devicetree
riscv: Add thead and xtheadvector as a vendor extension
riscv: vector: Use vlenb from DT for thead
riscv: csr: Add CSR encodings for CSR_VXRM/CSR_VXSAT
riscv: Add xtheadvector instruction definitions
riscv: vector: Support xtheadvector save/restore
riscv: hwprobe: Add thead vendor extension probing
riscv: hwprobe: Document thead vendor extensions and xtheadvector extension
selftests: riscv: Fix vector tests
selftests: riscv: Support xtheadvector in vector tests
Heiko Stuebner (1):
RISC-V: define the elements of the VCSR vector CSR
Documentation/arch/riscv/hwprobe.rst | 10 +
Documentation/devicetree/bindings/riscv/cpus.yaml | 19 ++
.../devicetree/bindings/riscv/extensions.yaml | 10 +
arch/riscv/Kconfig.vendor | 26 ++
arch/riscv/boot/dts/allwinner/sun20i-d1s.dtsi | 3 +-
arch/riscv/include/asm/cpufeature.h | 2 +
arch/riscv/include/asm/csr.h | 15 +
arch/riscv/include/asm/hwprobe.h | 3 +-
arch/riscv/include/asm/switch_to.h | 2 +-
arch/riscv/include/asm/vector.h | 225 +++++++++++----
arch/riscv/include/asm/vendor_extensions/thead.h | 42 +++
.../include/asm/vendor_extensions/thead_hwprobe.h | 19 ++
.../include/asm/vendor_extensions/vendor_hwprobe.h | 37 +++
arch/riscv/include/uapi/asm/hwprobe.h | 3 +-
arch/riscv/include/uapi/asm/vendor/thead.h | 3 +
arch/riscv/kernel/cpufeature.c | 52 +++-
arch/riscv/kernel/kernel_mode_vector.c | 8 +-
arch/riscv/kernel/process.c | 4 +-
arch/riscv/kernel/signal.c | 6 +-
arch/riscv/kernel/sys_hwprobe.c | 5 +
arch/riscv/kernel/vector.c | 24 +-
arch/riscv/kernel/vendor_extensions.c | 10 +
arch/riscv/kernel/vendor_extensions/Makefile | 2 +
arch/riscv/kernel/vendor_extensions/thead.c | 18 ++
.../riscv/kernel/vendor_extensions/thead_hwprobe.c | 19 ++
tools/testing/selftests/riscv/vector/.gitignore | 3 +-
tools/testing/selftests/riscv/vector/Makefile | 17 +-
.../selftests/riscv/vector/v_exec_initval_nolibc.c | 94 +++++++
tools/testing/selftests/riscv/vector/v_helpers.c | 68 +++++
tools/testing/selftests/riscv/vector/v_helpers.h | 8 +
tools/testing/selftests/riscv/vector/v_initval.c | 22 ++
.../selftests/riscv/vector/v_initval_nolibc.c | 68 -----
.../selftests/riscv/vector/vstate_exec_nolibc.c | 20 +-
.../testing/selftests/riscv/vector/vstate_prctl.c | 305 +++++++++++++--------
34 files changed, 901 insertions(+), 271 deletions(-)
---
base-commit: 7c08a2615f149f64fb1bb4660997e152fb3a11a7
change-id: 20240530-xtheadvector-833d3d17b423
--
- Charlie
The tests are built on per architecture basis. When unsupported
architecture is specified, it has no tests and TEST_GEN_PROGS is empty.
The lib.mk has support for not building anything for such case. But KVM
makefile doesn't handle such case correctly. It doesn't check if
TEST_GEN_PROGS is empty or not and try to create directory by mkdir.
Hence mkdir generates the error.
mkdir: missing operand
Try 'mkdir --help' for more information.
This can be easily fixed by checking if TEST_GEN_PROGS isn't empty
before calling mkdir.
Cc: Paolo Bonzini <pbonzini(a)redhat.com>
Cc: Sean Christopherson <seanjc(a)google.com>
Signed-off-by: Muhammad Usama Anjum <usama.anjum(a)collabora.com>
---
Changes since v1:
- Instead of ignoring error, check TEST_GEN_PROGS's validity first
---
tools/testing/selftests/kvm/Makefile | 2 ++
1 file changed, 2 insertions(+)
diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
index 48d32c5aa3eb7..9f8ed82ff1d65 100644
--- a/tools/testing/selftests/kvm/Makefile
+++ b/tools/testing/selftests/kvm/Makefile
@@ -317,7 +317,9 @@ $(LIBKVM_S_OBJ): $(OUTPUT)/%.o: %.S $(GEN_HDRS)
$(LIBKVM_STRING_OBJ): $(OUTPUT)/%.o: %.c
$(CC) $(CFLAGS) $(CPPFLAGS) $(TARGET_ARCH) -c -ffreestanding $< -o $@
+ifneq ($(strip $(TEST_GEN_PROGS)),)
$(shell mkdir -p $(sort $(dir $(TEST_GEN_PROGS))))
+endif
$(SPLIT_TEST_GEN_OBJ): $(GEN_HDRS)
$(TEST_GEN_PROGS): $(LIBKVM_OBJS)
$(TEST_GEN_PROGS_EXTENDED): $(LIBKVM_OBJS)
--
2.39.2
v21: https://patchwork.kernel.org/project/netdevbpf/list/?series=880735&state=*
====
v20 addressed some comments and resolved a test failure, but introduced
an unfortunate build error with a config edge case I wasn't testing. v21
simply resolves that error.
Major Changes:
- Resolve build error with CONFIG_PAGE_POOL=n && CONFIG_NET=y
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v21/
v20: https://patchwork.kernel.org/project/netdevbpf/list/?series=879373&state=*
====
v20 aims to resolve a couple of bug reports against v19, and addresses
some review comments around the page_pool_check_memory_provider
mechanism.
Major changes:
- Test edge cases such as header split disabled in selftest.
- Change `offset = 0` back to `offset = offset - start` to resolve issue
found in RX path by Taehee (thanks!)
- Address a few comments around page_pool_check_memory_provider() from
Pavel & Jakub.
- Removed some unnecessary includes across various patches in the
series.
- Removed unnecessary EXPORT_SYMBOL(page_pool_mem_providers) (Jakub).
- Fix regression caused by incorrect dev_get_max_mp_channel check, along
with rename (Jakub).
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v20/
v19: https://patchwork.kernel.org/project/netdevbpf/list/?series=876852&state=*
====
v18 got a thorough review (thanks!), and this iteration addresses the
feedback.
Major changes:
- Prevent deactivating mp bound queues.
- Prevent installing xdp on mp bound netdevs, or installing mps on xdp
installed netdevs.
- Fix corner cases in netlink API vis-a-vis missing attributes.
- Iron out the unreadable netmem driver support story. To be honest, the
conversation with Jakub & Pavel got a bit confusing for me. I've
implemented an approach in this set that makes sense to me, and
AFAICT, addresses the requirements. It may be good as-is, or it
may be a conversation starter/continuer. To be honest IMO there
are many ways to skin this cat and I don't see an extremely strong
reason to go for one approach over another. Here is one approach you
may like.
- Don't reset niov dma_addr on allocation & free.
- Add some tests to the selftest that catches some of the issues around
missing netlink attributes or deactivating mp-bound queues.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v19/
v18: https://patchwork.kernel.org/project/netdevbpf/list/?series=874848&state=*
====
v17 got minor feedback: (a) to beef up the description on patch 1 and (b)
to remove the leading underscores in the header definition.
I applied (a). (b) seems to be against current conventions so I did not
apply before further discussion.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v17/
v17: https://patchwork.kernel.org/project/netdevbpf/list/?series=869900&state=*
====
v16 also got a very thorough review and some testing (thanks again!).
Thes version addresses all the concerns reported on v15, in terms of
feedback and issues reported.
Major changes:
- Use ASSERT_RTNL.
- Moved around some of the page_pool helpers definitions so I can hide
some netmem helpers in private files as Jakub suggested.
- Don't make every net_iov hold a ref on the binding as Jakub suggested.
- Fix issue reported by Taehee where we access queues after they have
been freed.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v17/
v16: https://patchwork.kernel.org/project/netdevbpf/list/?series=866353&state=*
====
v15 got a thorough review and some testing, and this version addresses almost
all the feedback. Some more minor comments where the authors said it
could be done later, I left out.
Major changes:
- Addition of dma-buf introspection to page-pool-get and queue-get.
- Fixes to selftests suggested by Taehee.
- Fixes to documentation suggested by Donald.
- A couple of suggestions and fixes to TCP patches by Eric and David.
- Fixes to number assignements suggested by Arnd.
- Use rtnl_lock()ing to guard against queue reconfiguration while the
page_pool initialization is happening. (Jakub).
- Fixes to a few warnings reproduced by Taehee.
- Fixes to dma-buf binding suggested by Taehee and Jakub.
- Fixes to netlink UAPI suggested by Jakub
- Applied a number of Reviewed-bys and Acked-bys (including ones I lost
from v13+).
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v16/
One caveat: Taehee reproduced a KASAN warning and reported it here:
https://lore.kernel.org/netdev/CAMArcTUdCxOBYGF3vpbq=eBvqZfnc44KBaQTN7H-wqd…
I estimate the issue to be minor and easily fixable:
https://lore.kernel.org/netdev/CAHS8izNgaqC--GGE2xd85QB=utUnOHmioCsDd1TNxJW…
I hope to be able to follow up with a fix to net tree as net-next closes
imminently, but if this iteration doesn't make it in, I will repost with
a fix squashed after net-next reopens, no problem.
v15: https://patchwork.kernel.org/project/netdevbpf/list/?series=865481&state=*
====
No material changes in this version, only a fix to linking against
libynl.a from the last version. Per Jakub's instructions I've pulled one
of his patches into this series, and now use the new libynl.a correctly,
I hope.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v15/
v14: https://patchwork.kernel.org/project/netdevbpf/list/?series=865135&archive=…
====
No material changes in this version. Only rebase and re-verification on
top of net-next. v13, I think, raced with commit ebad6d0334793
("net/ipv4: Use nested-BH locking for ipv4_tcp_sk.") being merged to
net-next that caused a patchwork failure to apply. This series should
apply cleanly on commit c4532232fa2a4 ("selftests: net: remove unneeded
IP_GRE config").
I did not wait the customary 24hr as Jakub said it's OK to repost as soon
as I build test the rebased version:
https://lore.kernel.org/netdev/20240625075926.146d769d@kernel.org/
v13: https://patchwork.kernel.org/project/netdevbpf/list/?series=861406&archive=…
====
Major changes:
--------------
This iteration addresses Pavel's review comments, applies his
reviewed-by's, and seeks to fix the patchwork build error (sorry!).
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v13/
v12: https://patchwork.kernel.org/project/netdevbpf/list/?series=859747&state=*
====
Major changes:
--------------
This iteration only addresses one minor comment from Pavel with regards
to the trace printing of netmem, and the patchwork build error
introduced in v11 because I missed doing an allmodconfig build, sorry.
Other than that v11, AFAICT, received no feedback. There is one
discussion about how the specifics of plugging io uring memory through
the page pool, but not relevant to content in this particular patchset,
AFAICT.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v12/
v11: https://patchwork.kernel.org/project/netdevbpf/list/?series=857457&state=*
====
Major Changes:
--------------
v11 addresses feedback received in v10. The major change is the removal
of the memory provider ops as requested by Christoph. We still
accomplish the same thing, but utilizing direct function calls with if
statements rather than generic ops.
Additionally address sparse warnings, bugs and review comments from
folks that reviewed.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v11/
Detailed changelog:
-------------------
- Fixes in netdev_rx_queue_restart() from Pavel & David.
- Remove commit e650e8c3a36f5 ("net: page_pool: create hooks for
custom page providers") from the series to address Christoph's
feedback and rebased other patches on the series on this change.
- Fixed build errors with CONFIG_DMA_SHARED_BUFFER &&
!CONFIG_GENERIC_ALLOCATOR build.
- Fixed sparse warnings pointed out by Paolo.
- Drop unnecessary gro_pull_from_frag0 checks.
- Added Bagas reviewed-by to docs.
Cc: Bagas Sanjaya <bagasdotme(a)gmail.com>
Cc: Steven Rostedt <rostedt(a)goodmis.org>
Cc: Christoph Hellwig <hch(a)infradead.org>
Cc: Nikolay Aleksandrov <razor(a)blackwall.org>
Cc: Taehee Yoo <ap420073(a)gmail.com>
Cc: Donald Hunter <donald.hunter(a)gmail.com>
v10: https://patchwork.kernel.org/project/netdevbpf/list/?series=852422&state=*
====
Major Changes:
--------------
v9 was sent right before the merge window closed (sorry!). v10 is almost
a re-send of the series now that the merge window re-opened. Only
rebased to latest net-next and addressed some minor iterative comments
received on v9.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v10/
Detailed changelog:
-------------------
- Fixed tokens leaking in DONTNEED setsockopt (Nikolay).
- Moved net_iov_dma_addr() to devmem.c and made it a devmem specific
helpers (David).
- Rename hook alloc_pages to alloc_netmems as alloc_pages is now
preprocessor macro defined and causes a build error.
v9:
===
Major Changes:
--------------
GVE queue API has been merged. Submitting this version as non-RFC after
rebasing on top of the merged API, and dropped the out of tree queue API
I was carrying on github. Addressed the little feedback v8 has received.
Detailed changelog:
------------------
- Added new patch from David Wei to this series for
netdev_rx_queue_restart()
- Fixed sparse error.
- Removed CONFIG_ checks in netmem_is_net_iov()
- Flipped skb->readable to skb->unreadable
- Minor fixes to selftests & docs.
RFC v8:
=======
Major Changes:
--------------
- Fixed build error generated by patch-by-patch build.
- Applied docs suggestions from Randy.
RFC v7:
=======
Major Changes:
--------------
This revision largely rebases on top of net-next and addresses the feedback
RFCv6 received from folks, namely Jakub, Yunsheng, Arnd, David, & Pavel.
The series remains in RFC because the queue-API ndos defined in this
series are not yet implemented. I have a GVE implementation I carry out
of tree for my testing. A upstreamable GVE implementation is in the
works. Aside from that, in my estimation all the patches are ready for
review/merge. Please do take a look.
As usual the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v7/
Detailed changelog:
- Use admin-perm in netlink API.
- Addressed feedback from Jakub with regards to netlink API
implementation.
- Renamed devmem.c functions to something more appropriate for that
file.
- Improve the performance seen through the page_pool benchmark.
- Fix the value definition of all the SO_DEVMEM_* uapi.
- Various fixes to documentation.
Perf - page-pool benchmark:
---------------------------
Improved performance of bench_page_pool_simple.ko tests compared to v6:
https://pastebin.com/raw/v5dYRg8L
net-next base: 8 cycle fast path.
RFC v6: 10 cycle fast path.
RFC v7: 9 cycle fast path.
RFC v7 with CONFIG_DMA_SHARED_BUFFER disabled: 8 cycle fast path,
same as baseline.
Perf - Devmem TCP benchmark:
---------------------
Perf is about the same regardless of the changes in v7, namely the
removal of the static_branch_unlikely to improve the page_pool benchmark
performance:
189/200gbps bi-directional throughput with RX devmem TCP and regular TCP
TX i.e. ~95% line rate.
RFC v6:
=======
Major Changes:
--------------
This revision largely rebases on top of net-next and addresses the little
feedback RFCv5 received.
The series remains in RFC because the queue-API ndos defined in this
series are not yet implemented. I have a GVE implementation I carry out
of tree for my testing. A upstreamable GVE implementation is in the
works. Aside from that, in my estimation all the patches are ready for
review/merge. Please do take a look.
As usual the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v6/
This version also comes with some performance data recorded in the cover
letter (see below changelog).
Detailed changelog:
- Rebased on top of the merged netmem_ref changes.
- Converted skb->dmabuf to skb->readable (Pavel). Pavel's original
suggestion was to remove the skb->dmabuf flag entirely, but when I
looked into it closely, I found the issue that if we remove the flag
we have to dereference the shinfo(skb) pointer to obtain the first
frag to tell whether an skb is readable or not. This can cause a
performance regression if it dirties the cache line when the
shinfo(skb) was not really needed. Instead, I converted the skb->dmabuf
flag into a generic skb->readable flag which can be re-used by io_uring
0-copy RX.
- Squashed a few locking optimizations from Eric Dumazet in the RX path
and the DEVMEM_DONTNEED setsockopt.
- Expanded the tests a bit. Added validation for invalid scenarios and
added some more coverage.
Perf - page-pool benchmark:
---------------------------
bench_page_pool_simple.ko tests with and without these changes:
https://pastebin.com/raw/ncHDwAbn
AFAIK the number that really matters in the perf tests is the
'tasklet_page_pool01_fast_path Per elem'. This one measures at about 8
cycles without the changes but there is some 1 cycle noise in some
results.
With the patches this regresses to 9 cycles with the changes but there
is 1 cycle noise occasionally running this test repeatedly.
Lastly I tried disable the static_branch_unlikely() in
netmem_is_net_iov() check. To my surprise disabling the
static_branch_unlikely() check reduces the fast path back to 8 cycles,
but the 1 cycle noise remains.
Perf - Devmem TCP benchmark:
---------------------
189/200gbps bi-directional throughput with RX devmem TCP and regular TCP
TX i.e. ~95% line rate.
Major changes in RFC v5:
========================
1. Rebased on top of 'Abstract page from net stack' series and used the
new netmem type to refer to LSB set pointers instead of re-using
struct page.
2. Downgraded this series back to RFC and called it RFC v5. This is
because this series is now dependent on 'Abstract page from net
stack'[1] and the queue API. Both are removed from the series to
reduce the patch # and those bits are fairly independent or
pre-requisite work.
3. Reworked the page_pool devmem support to use netmem and for some
more unified handling.
4. Reworked the reference counting of net_iov (renamed from
page_pool_iov) to use pp_ref_count for refcounting.
The full changes including the dependent series and GVE page pool
support is here:
https://github.com/mina/linux/commits/tcpdevmem-rfcv5/
[1] https://patchwork.kernel.org/project/netdevbpf/list/?series=810774
Major changes in v1:
====================
1. Implemented MVP queue API ndos to remove the userspace-visible
driver reset.
2. Fixed issues in the napi_pp_put_page() devmem frag unref path.
3. Removed RFC tag.
Many smaller addressed comments across all the patches (patches have
individual change log).
Full tree including the rest of the GVE driver changes:
https://github.com/mina/linux/commits/tcpdevmem-v1
Changes in RFC v3:
==================
1. Pulled in the memory-provider dependency from Jakub's RFC[1] to make the
series reviewable and mergeable.
2. Implemented multi-rx-queue binding which was a todo in v2.
3. Fix to cmsg handling.
The sticking point in RFC v2[2] was the device reset required to refill
the device rx-queues after the dmabuf bind/unbind. The solution
suggested as I understand is a subset of the per-queue management ops
Jakub suggested or similar:
https://lore.kernel.org/netdev/20230815171638.4c057dcd@kernel.org/
This is not addressed in this revision, because:
1. This point was discussed at netconf & netdev and there is openness to
using the current approach of requiring a device reset.
2. Implementing individual queue resetting seems to be difficult for my
test bed with GVE. My prototype to test this ran into issues with the
rx-queues not coming back up properly if reset individually. At the
moment I'm unsure if it's a mistake in the POC or a genuine issue in
the virtualization stack behind GVE, which currently doesn't test
individual rx-queue restart.
3. Our usecases are not bothered by requiring a device reset to refill
the buffer queues, and we'd like to support NICs that run into this
limitation with resetting individual queues.
My thought is that drivers that have trouble with per-queue configs can
use the support in this series, while drivers that support new netdev
ops to reset individual queues can automatically reset the queue as
part of the dma-buf bind/unbind.
The same approach with device resets is presented again for consideration
with other sticking points addressed.
This proposal includes the rx devmem path only proposed for merge. For a
snapshot of my entire tree which includes the GVE POC page pool support &
device memory support:
https://github.com/torvalds/linux/compare/master...mina:linux:tcpdevmem-v3
[1] https://lore.kernel.org/netdev/f8270765-a27b-6ccf-33ea-cda097168d79@redhat.…
[2] https://lore.kernel.org/netdev/CAHS8izOVJGJH5WF68OsRWFKJid1_huzzUK+hpKbLcL4…
Changes in RFC v2:
==================
The sticking point in RFC v1[1] was the dma-buf pages approach we used to
deliver the device memory to the TCP stack. RFC v2 is a proof-of-concept
that attempts to resolve this by implementing scatterlist support in the
networking stack, such that we can import the dma-buf scatterlist
directly. This is the approach proposed at a high level here[2].
Detailed changes:
1. Replaced dma-buf pages approach with importing scatterlist into the
page pool.
2. Replace the dma-buf pages centric API with a netlink API.
3. Removed the TX path implementation - there is no issue with
implementing the TX path with scatterlist approach, but leaving
out the TX path makes it easier to review.
4. Functionality is tested with this proposal, but I have not conducted
perf testing yet. I'm not sure there are regressions, but I removed
perf claims from the cover letter until they can be re-confirmed.
5. Added Signed-off-by: contributors to the implementation.
6. Fixed some bugs with the RX path since RFC v1.
Any feedback welcome, but specifically the biggest pending questions
needing feedback IMO are:
1. Feedback on the scatterlist-based approach in general.
2. Netlink API (Patch 1 & 2).
3. Approach to handle all the drivers that expect to receive pages from
the page pool (Patch 6).
[1] https://lore.kernel.org/netdev/dfe4bae7-13a0-3c5d-d671-f61b375cb0b4@gmail.c…
[2] https://lore.kernel.org/netdev/CAHS8izPm6XRS54LdCDZVd0C75tA1zHSu6jLVO8nzTLX…
==================
* TL;DR:
Device memory TCP (devmem TCP) is a proposal for transferring data to and/or
from device memory efficiently, without bouncing the data to a host memory
buffer.
* Problem:
A large amount of data transfers have device memory as the source and/or
destination. Accelerators drastically increased the volume of such transfers.
Some examples include:
- ML accelerators transferring large amounts of training data from storage into
GPU/TPU memory. In some cases ML training setup time can be as long as 50% of
TPU compute time, improving data transfer throughput & efficiency can help
improving GPU/TPU utilization.
- Distributed training, where ML accelerators, such as GPUs on different hosts,
exchange data among them.
- Distributed raw block storage applications transfer large amounts of data with
remote SSDs, much of this data does not require host processing.
Today, the majority of the Device-to-Device data transfers the network are
implemented as the following low level operations: Device-to-Host copy,
Host-to-Host network transfer, and Host-to-Device copy.
The implementation is suboptimal, especially for bulk data transfers, and can
put significant strains on system resources, such as host memory bandwidth,
PCIe bandwidth, etc. One important reason behind the current state is the
kernel’s lack of semantics to express device to network transfers.
* Proposal:
In this patch series we attempt to optimize this use case by implementing
socket APIs that enable the user to:
1. send device memory across the network directly, and
2. receive incoming network packets directly into device memory.
Packet _payloads_ go directly from the NIC to device memory for receive and from
device memory to NIC for transmit.
Packet _headers_ go to/from host memory and are processed by the TCP/IP stack
normally. The NIC _must_ support header split to achieve this.
Advantages:
- Alleviate host memory bandwidth pressure, compared to existing
network-transfer + device-copy semantics.
- Alleviate PCIe BW pressure, by limiting data transfer to the lowest level
of the PCIe tree, compared to traditional path which sends data through the
root complex.
* Patch overview:
** Part 1: netlink API
Gives user ability to bind dma-buf to an RX queue.
** Part 2: scatterlist support
Currently the standard for device memory sharing is DMABUF, which doesn't
generate struct pages. On the other hand, networking stack (skbs, drivers, and
page pool) operate on pages. We have 2 options:
1. Generate struct pages for dmabuf device memory, or,
2. Modify the networking stack to process scatterlist.
Approach #1 was attempted in RFC v1. RFC v2 implements approach #2.
** part 3: page pool support
We piggy back on page pool memory providers proposal:
https://github.com/kuba-moo/linux/tree/pp-providers
It allows the page pool to define a memory provider that provides the
page allocation and freeing. It helps abstract most of the device memory
TCP changes from the driver.
** part 4: support for unreadable skb frags
Page pool iovs are not accessible by the host; we implement changes
throughput the networking stack to correctly handle skbs with unreadable
frags.
** Part 5: recvmsg() APIs
We define user APIs for the user to send and receive device memory.
Not included with this series is the GVE devmem TCP support, just to
simplify the review. Code available here if desired:
https://github.com/mina/linux/tree/tcpdevmem
This series is built on top of net-next with Jakub's pp-providers changes
cherry-picked.
* NIC dependencies:
1. (strict) Devmem TCP require the NIC to support header split, i.e. the
capability to split incoming packets into a header + payload and to put
each into a separate buffer. Devmem TCP works by using device memory
for the packet payload, and host memory for the packet headers.
2. (optional) Devmem TCP works better with flow steering support & RSS support,
i.e. the NIC's ability to steer flows into certain rx queues. This allows the
sysadmin to enable devmem TCP on a subset of the rx queues, and steer
devmem TCP traffic onto these queues and non devmem TCP elsewhere.
The NIC I have access to with these properties is the GVE with DQO support
running in Google Cloud, but any NIC that supports these features would suffice.
I may be able to help reviewers bring up devmem TCP on their NICs.
* Testing:
The series includes a udmabuf kselftest that show a simple use case of
devmem TCP and validates the entire data path end to end without
a dependency on a specific dmabuf provider.
** Test Setup
Kernel: net-next with this series and memory provider API cherry-picked
locally.
Hardware: Google Cloud A3 VMs.
NIC: GVE with header split & RSS & flow steering support.
Cc: Pavel Begunkov <asml.silence(a)gmail.com>
Cc: David Wei <dw(a)davidwei.uk>
Cc: Jason Gunthorpe <jgg(a)ziepe.ca>
Cc: Yunsheng Lin <linyunsheng(a)huawei.com>
Cc: Shailend Chand <shailend(a)google.com>
Cc: Harshitha Ramamurthy <hramamurthy(a)google.com>
Cc: Shakeel Butt <shakeel.butt(a)linux.dev>
Cc: Jeroen de Borst <jeroendb(a)google.com>
Cc: Praveen Kaligineedi <pkaligineedi(a)google.com>
Mina Almasry (13):
netdev: add netdev_rx_queue_restart()
net: netdev netlink api to bind dma-buf to a net device
netdev: support binding dma-buf to netdevice
netdev: netdevice devmem allocator
page_pool: devmem support
memory-provider: dmabuf devmem memory provider
net: support non paged skb frags
net: add support for skbs with unreadable frags
tcp: RX path for devmem TCP
net: add SO_DEVMEM_DONTNEED setsockopt to release RX frags
net: add devmem TCP documentation
selftests: add ncdevmem, netcat for devmem TCP
netdev: add dmabuf introspection
Documentation/netlink/specs/netdev.yaml | 61 +++
Documentation/networking/devmem.rst | 269 +++++++++++
Documentation/networking/index.rst | 1 +
arch/alpha/include/uapi/asm/socket.h | 6 +
arch/mips/include/uapi/asm/socket.h | 6 +
arch/parisc/include/uapi/asm/socket.h | 6 +
arch/sparc/include/uapi/asm/socket.h | 6 +
include/linux/netdevice.h | 2 +
include/linux/skbuff.h | 61 ++-
include/linux/skbuff_ref.h | 9 +-
include/linux/socket.h | 1 +
include/net/devmem.h | 128 ++++++
include/net/mp_dmabuf_devmem.h | 44 ++
include/net/netdev_rx_queue.h | 5 +
include/net/netmem.h | 169 ++++++-
include/net/page_pool/helpers.h | 39 +-
include/net/page_pool/types.h | 22 +-
include/net/sock.h | 2 +
include/net/tcp.h | 5 +-
include/trace/events/page_pool.h | 12 +-
include/uapi/asm-generic/socket.h | 6 +
include/uapi/linux/netdev.h | 13 +
include/uapi/linux/uio.h | 17 +
net/core/Makefile | 3 +-
net/core/datagram.c | 6 +
net/core/dev.c | 22 +-
net/core/devmem.c | 374 +++++++++++++++
net/core/gro.c | 3 +-
net/core/netdev-genl-gen.c | 23 +
net/core/netdev-genl-gen.h | 6 +
net/core/netdev-genl.c | 118 +++++
net/core/netdev_rx_queue.c | 81 ++++
net/core/netmem_priv.h | 31 ++
net/core/page_pool.c | 117 +++--
net/core/page_pool_priv.h | 46 ++
net/core/page_pool_user.c | 29 ++
net/core/skbuff.c | 77 +++-
net/core/sock.c | 68 +++
net/ethtool/common.c | 8 +
net/ipv4/esp4.c | 3 +-
net/ipv4/tcp.c | 261 ++++++++++-
net/ipv4/tcp_input.c | 13 +-
net/ipv4/tcp_ipv4.c | 16 +
net/ipv4/tcp_minisocks.c | 2 +
net/ipv4/tcp_output.c | 5 +-
net/ipv6/esp6.c | 3 +-
net/packet/af_packet.c | 4 +-
tools/include/uapi/linux/netdev.h | 13 +
tools/testing/selftests/net/.gitignore | 1 +
tools/testing/selftests/net/Makefile | 9 +
tools/testing/selftests/net/ncdevmem.c | 588 ++++++++++++++++++++++++
51 files changed, 2699 insertions(+), 121 deletions(-)
create mode 100644 Documentation/networking/devmem.rst
create mode 100644 include/net/devmem.h
create mode 100644 include/net/mp_dmabuf_devmem.h
create mode 100644 net/core/devmem.c
create mode 100644 net/core/netdev_rx_queue.c
create mode 100644 net/core/netmem_priv.h
create mode 100644 tools/testing/selftests/net/ncdevmem.c
--
2.46.0.184.g6999bdac58-goog
On Tue, Aug 20, 2024 at 09:48:50AM +0800, Levi Zim wrote:
> On 2024-08-20 01:00, Charlie Jenkins wrote:
> > On Mon, Aug 19, 2024 at 01:55:57PM +0800, Levi Zim wrote:
> > > On 2024-03-22 22:06, Palmer Dabbelt wrote:
> > > > On Thu, 01 Feb 2024 18:28:06 PST (-0800), Charlie Jenkins wrote:
> > > > > On Wed, Jan 31, 2024 at 11:59:43PM +0800, Yangyu Chen wrote:
> > > > > > On Wed, 2024-01-31 at 22:41 +0800, Yangyu Chen wrote:
> > > > > > > On Tue, 2024-01-30 at 17:07 -0800, Charlie Jenkins wrote:
> > > > > > > > On riscv it is guaranteed that the address returned by mmap is less
> > > > > > > > than
> > > > > > > > the hint address. Allow mmap to return an address all the way up to
> > > > > > > > addr, if provided, rather than just up to the lower address space.
> > > > > > > > > > This provides a performance benefit as well, allowing
> > > > > > mmap to exit
> > > > > > > > after
> > > > > > > > checking that the address is in range rather than searching for a
> > > > > > > > valid
> > > > > > > > address.
> > > > > > > > > > It is possible to provide an address that uses at most the same
> > > > > > > > number
> > > > > > > > of bits, however it is significantly more computationally expensive
> > > > > > > > to
> > > > > > > > provide that number rather than setting the max to be the hint
> > > > > > > > address.
> > > > > > > > There is the instruction clz/clzw in Zbb that returns the highest
> > > > > > > > set
> > > > > > > > bit
> > > > > > > > which could be used to performantly implement this, but it would
> > > > > > > > still
> > > > > > > > be slower than the current implementation. At worst case, half of
> > > > > > > > the
> > > > > > > > address would not be able to be allocated when a hint address is
> > > > > > > > provided.
> > > > > > > > > > Signed-off-by: Charlie Jenkins<charlie(a)rivosinc.com>
> > > > > > > > ---
> > > > > > > > arch/riscv/include/asm/processor.h | 27 +++++++++++---------------
> > > > > > > > -
> > > > > > > > 1 file changed, 11 insertions(+), 16 deletions(-)
> > > > > > > > > > diff --git a/arch/riscv/include/asm/processor.h
> > > > > > > > b/arch/riscv/include/asm/processor.h
> > > > > > > > index f19f861cda54..8ece7a8f0e18 100644
> > > > > > > > --- a/arch/riscv/include/asm/processor.h
> > > > > > > > +++ b/arch/riscv/include/asm/processor.h
> > > > > > > > @@ -14,22 +14,16 @@
> > > > > > > >
> > > > > > > > #include <asm/ptrace.h>
> > > > > > > >
> > > > > > > > -#ifdef CONFIG_64BIT
> > > > > > > > -#define DEFAULT_MAP_WINDOW (UL(1) << (MMAP_VA_BITS - 1))
> > > > > > > > -#define STACK_TOP_MAX TASK_SIZE_64
> > > > > > > > -
> > > > > > > > #define arch_get_mmap_end(addr, len, flags) \
> > > > > > > > ({ \
> > > > > > > > unsigned long
> > > > > > > > mmap_end; \
> > > > > > > > typeof(addr) _addr = (addr); \
> > > > > > > > - if ((_addr) == 0 || (IS_ENABLED(CONFIG_COMPAT) &&
> > > > > > > > is_compat_task())) \
> > > > > > > > + if ((_addr) == 0 || \
> > > > > > > > + (IS_ENABLED(CONFIG_COMPAT) && is_compat_task()) || \
> > > > > > > > + ((_addr + len) > BIT(VA_BITS -
> > > > > > > > 1))) \
> > > > > > > > mmap_end = STACK_TOP_MAX; \
> > > > > > > > - else if ((_addr) >= VA_USER_SV57) \
> > > > > > > > - mmap_end = STACK_TOP_MAX; \
> > > > > > > > - else if ((((_addr) >= VA_USER_SV48)) && (VA_BITS >=
> > > > > > > > VA_BITS_SV48)) \
> > > > > > > > - mmap_end = VA_USER_SV48; \
> > > > > > > > else \
> > > > > > > > - mmap_end = VA_USER_SV39; \
> > > > > > > > + mmap_end = (_addr + len); \
> > > > > > > > mmap_end; \
> > > > > > > > })
> > > > > > > >
> > > > > > > > @@ -39,17 +33,18 @@
> > > > > > > > typeof(addr) _addr = (addr); \
> > > > > > > > typeof(base) _base = (base); \
> > > > > > > > unsigned long rnd_gap = DEFAULT_MAP_WINDOW - (_base); \
> > > > > > > > - if ((_addr) == 0 || (IS_ENABLED(CONFIG_COMPAT) &&
> > > > > > > > is_compat_task())) \
> > > > > > > > + if ((_addr) == 0 || \
> > > > > > > > + (IS_ENABLED(CONFIG_COMPAT) && is_compat_task()) || \
> > > > > > > > + ((_addr + len) > BIT(VA_BITS -
> > > > > > > > 1))) \
> > > > > > > > mmap_base = (_base); \
> > > > > > > > - else if (((_addr) >= VA_USER_SV57) && (VA_BITS >=
> > > > > > > > VA_BITS_SV57)) \
> > > > > > > > - mmap_base = VA_USER_SV57 - rnd_gap; \
> > > > > > > > - else if ((((_addr) >= VA_USER_SV48)) && (VA_BITS >=
> > > > > > > > VA_BITS_SV48)) \
> > > > > > > > - mmap_base = VA_USER_SV48 - rnd_gap; \
> > > > > > > > else \
> > > > > > > > - mmap_base = VA_USER_SV39 - rnd_gap; \
> > > > > > > > + mmap_base = (_addr + len) - rnd_gap; \
> > > > > > > > mmap_base; \
> > > > > > > > })
> > > > > > > >
> > > > > > > > +#ifdef CONFIG_64BIT
> > > > > > > > +#define DEFAULT_MAP_WINDOW (UL(1) << (MMAP_VA_BITS - 1))
> > > > > > > > +#define STACK_TOP_MAX TASK_SIZE_64
> > > > > > > > #else
> > > > > > > > #define DEFAULT_MAP_WINDOW TASK_SIZE
> > > > > > > > #define STACK_TOP_MAX TASK_SIZE
> > > > > > > > > > I have carefully tested your patch on qemu with sv57. A
> > > > > > bug that
> > > > > > > needs
> > > > > > > to be solved is that mmap with the same hint address without
> > > > > > > MAP_FIXED
> > > > > > > set will fail the second time.
> > > > > > > > Userspace code to reproduce the bug:
> > > > > > > > #include <sys/mman.h>
> > > > > > > #include <stdio.h>
> > > > > > > #include <stdint.h>
> > > > > > > > void test(char *addr) {
> > > > > > > char *res = mmap(addr, 4096, PROT_READ | PROT_WRITE,
> > > > > > > MAP_ANONYMOUS
> > > > > > > > MAP_PRIVATE, -1, 0);
> > > > > > > printf("hint %p got %p.\n", addr, res);
> > > > > > > }
> > > > > > > > int main (void) {
> > > > > > > test(1<<30);
> > > > > > > test(1<<30);
> > > > > > > test(1<<30);
> > > > > > > return 0;
> > > > > > > }
> > > > > > > > output:
> > > > > > > > hint 0x40000000 got 0x40000000.
> > > > > > > hint 0x40000000 got 0xffffffffffffffff.
> > > > > > > hint 0x40000000 got 0xffffffffffffffff.
> > > > > > > > output on x86:
> > > > > > > > hint 0x40000000 got 0x40000000.
> > > > > > > hint 0x40000000 got 0x7f9171363000.
> > > > > > > hint 0x40000000 got 0x7f9171362000.
> > > > > > > > It may need to implement a special arch_get_unmapped_area and
> > > > > > > arch_get_unmapped_area_topdown function.
> > > > > > >
> > > > > > This is because hint address < rnd_gap. I have tried to let mmap_base =
> > > > > > min((_addr + len), (base) + TASK_SIZE - DEFAULT_MAP_WINDOW). However it
> > > > > > does not work for bottom-up while ulimit -s is unlimited. You said this
> > > > > > behavior is expected from patch v2 review. However it brings a new
> > > > > > regression even on sv39 systems.
> > > > > >
> > > > > > I still don't know the reason why use addr+len as the upper-bound. I
> > > > > > think solution like x86/arm64/powerpc provide two address space switch
> > > > > > based on whether hint address above the default map window is enough.
> > > > > >
> > > > > Yep this is expected. It is up to the maintainers to decide.
> > > > Sorry I forgot to reply to this, I had a buffer sitting around somewhere
> > > > but I must have lost it.
> > > >
> > > > I think Charlie's approach is the right way to go. Putting my userspace
> > > > hat on, I'd much rather have my allocations fail rather than silently
> > > > ignore the hint when there's memory pressure.
> > > >
> > > > If there's some real use case that needs these low hints to be silently
> > > > ignored under VA pressure then we can try and figure something out that
> > > > makes those applications work.
> > > I could confirm that this patch has broken chromium's partition allocator on
> > > riscv64. The minimal reproduction I use is chromium-mmap.c:
> > >
> > > #include <stdio.h>
> > > #include <sys/mman.h>
> > >
> > > int main() {
> > > void* expected = (void*)0x400000000;
> > > void* addr = mmap(expected, 17179869184, PROT_NONE,
> > > MAP_PRIVATE|MAP_ANONYMOUS, -1, 0);
> > > if (addr != expected) {
> > It is not valid to assume that the address returned by mmap will be the
> > hint address. If the hint address is not available, mmap will return a
> > different address.
>
> Oh, sorry I didn't make it clear what is the expected behavior.
> The printf here is solely for debugging purpose and I don't mean that
> chromium expect it will get the hint address. The expected behavior is that
> both the two mmap calls will succeed.
>
> > > printf("Not expected address: %p != %p\n", addr, expected);
> > > }
> > > expected = (void*)0x3fffff000;
> > > addr = mmap(expected, 17179873280, PROT_NONE, MAP_PRIVATE|MAP_ANONYMOUS,
> > > -1, 0);
> > > if (addr != expected) {
> > > printf("Not expected address: %p != %p\n", addr, expected);
> > > }
> > > return 0;
> > > }
> > >
> > > The second mmap fails with ENOMEM. Manually reverting this commit fixes the
> > > issue for me. So I think it's clearly a regression and breaks userspace.
> > >
> > The issue here is that overlapping memory is being requested. This
> > second mmap will never be able to provide an address at 0x3fffff000 with
> > a size of 0x400001000 since mmap just provided an address at 0x400000000
> > with a size of 0x400000000.
> >
> > Before this patch, this request causes mmap to return a completely
> > arbitrary value. There is no reason to use a hint address in this manner
> > because the hint can never be respected. Since an arbitrary address is
> > desired, a hint of zero should be used.
> >
> > This patch causes the behavior to be more deterministic. Instead of
> > providing an arbitrary address, it causes the address to be less than or
> > equal to the hint address. This allows for applications to make
> > assumptions about the returned address.
>
> About the overlap, of course the partition allocator's request for
> overlapped vma seems unreasonable.
>
> But I still don't quite understand why mmap cannot use an address higher
> than the hint address.
> The hint address, after all, is a hint, not a requirement.
Yes that is fair. A "hint" that does not guarantee anything is
useless so architectures have abused the term quite a bit.
>
> Quoting the man page:
>
> > If another mapping already exists there, the kernel picks
> > a new address that may or may not depend on the hint. The
> > address of the new mapping is returned as the result of the call.
> So for casual programmers that only reads man page but not architecture
> specific kernel
> documentation, the current behavior of mmap on riscv64 failing on overlapped
> address ranges
> are quite surprising IMO.
The man pages for riscv are in desperate need of attention. I have
submitted a couple of updates to them recently, but there is a lot more
work to be done to help developers.
>
> And quoting the man page again about the errno:
>
> > ENOMEM No memory is available.
> >
> > ENOMEM The process's maximum number of mappings would have been
> > exceeded. This error can also occur for munmap(), when
> > unmapping a region in the middle of an existing mapping,
> > since this results in two smaller mappings on either side
> > of the region being unmapped.
> >
> > ENOMEM (since Linux 4.7) The process's RLIMIT_DATA limit,
> > described in getrlimit(2), would have been exceeded.
> >
> > ENOMEM We don't like addr, because it exceeds the virtual address
> > space of the CPU.
> >
>
> There's no matching description for the ENOMEM returned here.
> I would suggest removing "because it exceeds the virtual address
> space of the CPU." from the last item if the ENOMEM behavior here
> is expected.
This ENOMEM means something like "no memory available in the requested
region".
>
> > This code is unfortunately relying on the previously mostly undefined
> > behavior of the hint address in mmap.
> Although I haven't read the code of chromium's partition allocator to judge
> whether it should
> be improved or fixed for riscv64, I do know that the kernel "don't break
> userspace" and
> "never EVER blame the user programs".
The hint address design of mmap is a tricky one because it is largely
implementation defined and what the man pages say is not how it is
implemented in most architectures!
> > The goal of this patch is to help
> > developers have more consistent mmap behavior, but maybe it is necessary
> > to hide this behavior behind an mmap flag.
> Thank you for helping to shape a more consistent mmap behavior.
> I think this should be fixed ASAP either by allowing the hint address to be
> ignored
> (as suggested by the Linux man page), or hide this behavior behind an mmap
> flag as you said.
Having a flag could also lead to a generic way of defining this
behavior. Other architectures do not provide a way for applications to
guarantee that some number of bits are left unused in a virtual address,
and that was one of the motivating design goals here.
- Charlie
>
> > - Charlie
> >
> > > See alsohttps://github.com/riscv-forks/electron/issues/4
> > >
> > > > > - Charlie
> > > Sincerely,
> > > Levi
> > >
A few tests check if nettest exists in the $PATH before adding
$PWD to $PATH and re-checking. They don't discard stderr on
the first check (and nettest is built as part of selftests,
so it's pretty normal for it to not be available in system $PATH).
This leads to output noise:
which: no nettest in (/home/virtme/tools/fs/bin:/home/virtme/tools/fs/sbin:/home/virtme/tools/fs/usr/bin:/home/virtme/tools/fs/usr/sbin:/bin:/sbin:/usr/bin:/usr/sbin:/usr/local/bin)
Add a common helper for the check which does silence stderr.
There is another small functional change hiding here, because pmtu.sh
used to return from the test case rather than completely exit.
Building nettest is not hard, there should be no need to maintain
the ability to selectively skip cases in its absence.
Signed-off-by: Jakub Kicinski <kuba(a)kernel.org>
---
CC: shuah(a)kernel.org
CC: linux-kselftest(a)vger.kernel.org
---
tools/testing/selftests/net/fcnal-test.sh | 9 +--------
tools/testing/selftests/net/lib.sh | 15 +++++++++++++++
tools/testing/selftests/net/pmtu.sh | 8 +-------
tools/testing/selftests/net/settings | 1 +
tools/testing/selftests/net/unicast_extensions.sh | 9 +--------
5 files changed, 19 insertions(+), 23 deletions(-)
diff --git a/tools/testing/selftests/net/fcnal-test.sh b/tools/testing/selftests/net/fcnal-test.sh
index 386ebd829df5..899dbad0104b 100755
--- a/tools/testing/selftests/net/fcnal-test.sh
+++ b/tools/testing/selftests/net/fcnal-test.sh
@@ -4304,14 +4304,7 @@ elif [ "$TESTS" = "ipv6" ]; then
TESTS="$TESTS_IPV6"
fi
-# nettest can be run from PATH or from same directory as this selftest
-if ! which nettest >/dev/null; then
- PATH=$PWD:$PATH
- if ! which nettest >/dev/null; then
- echo "'nettest' command not found; skipping tests"
- exit $ksft_skip
- fi
-fi
+check_gen_prog "nettest"
declare -i nfail=0
declare -i nsuccess=0
diff --git a/tools/testing/selftests/net/lib.sh b/tools/testing/selftests/net/lib.sh
index 8ee4489238ca..be8707bfb46e 100644
--- a/tools/testing/selftests/net/lib.sh
+++ b/tools/testing/selftests/net/lib.sh
@@ -125,6 +125,21 @@ slowwait_for_counter()
slowwait "$timeout" until_counter_is ">= $((base + delta))" "$@"
}
+# Check for existence of tools which are built as part of selftests
+# but may also already exist in $PATH
+check_gen_prog()
+{
+ local prog_name=$1; shift
+
+ if ! which $prog_name >/dev/null 2>/dev/null; then
+ PATH=$PWD:$PATH
+ if ! which $prog_name >/dev/null; then
+ echo "'$prog_name' command not found; skipping tests"
+ exit $ksft_skip
+ fi
+ fi
+}
+
remove_ns_list()
{
local item=$1
diff --git a/tools/testing/selftests/net/pmtu.sh b/tools/testing/selftests/net/pmtu.sh
index 24a50622406c..569bce8b6383 100755
--- a/tools/testing/selftests/net/pmtu.sh
+++ b/tools/testing/selftests/net/pmtu.sh
@@ -681,13 +681,7 @@ setup_xfrm() {
}
setup_nettest_xfrm() {
- if ! which nettest >/dev/null; then
- PATH=$PWD:$PATH
- if ! which nettest >/dev/null; then
- echo "'nettest' command not found; skipping tests"
- return 1
- fi
- fi
+ check_gen_prog "nettest"
[ ${1} -eq 6 ] && proto="-6" || proto=""
port=${2}
diff --git a/tools/testing/selftests/net/settings b/tools/testing/selftests/net/settings
index ed8418e8217a..a38764182822 100644
--- a/tools/testing/selftests/net/settings
+++ b/tools/testing/selftests/net/settings
@@ -1 +1,2 @@
timeout=3600
+profile=1
diff --git a/tools/testing/selftests/net/unicast_extensions.sh b/tools/testing/selftests/net/unicast_extensions.sh
index f52aa5f7da52..3e751234ccfe 100755
--- a/tools/testing/selftests/net/unicast_extensions.sh
+++ b/tools/testing/selftests/net/unicast_extensions.sh
@@ -30,14 +30,7 @@
source lib.sh
-# nettest can be run from PATH or from same directory as this selftest
-if ! which nettest >/dev/null; then
- PATH=$PWD:$PATH
- if ! which nettest >/dev/null; then
- echo "'nettest' command not found; skipping tests"
- exit $ksft_skip
- fi
-fi
+check_gen_prog "nettest"
result=0
--
2.46.0
v20: https://patchwork.kernel.org/project/netdevbpf/list/?series=879373&state=*
====
v20 aims to resolve a couple of bug reports against v19, and addresses
some review comments around the page_pool_check_memory_provider
mechanism.
Major changes:
- Test edge cases such as header split disabled in selftest.
- Change `offset = 0` back to `offset = offset - start` to resolve issue
found in RX path by Taehee (thanks!)
- Address a few comments around page_pool_check_memory_provider() from
Pavel & Jakub.
- Removed some unnecessary includes across various patches in the
series.
- Removed unnecessary EXPORT_SYMBOL(page_pool_mem_providers) (Jakub).
- Fix regression caused by incorrect dev_get_max_mp_channel check, along
with rename (Jakub).
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v20/
v19: https://patchwork.kernel.org/project/netdevbpf/list/?series=876852&state=*
====
v18 got a thorough review (thanks!), and this iteration addresses the
feedback.
Major changes:
- Prevent deactivating mp bound queues.
- Prevent installing xdp on mp bound netdevs, or installing mps on xdp
installed netdevs.
- Fix corner cases in netlink API vis-a-vis missing attributes.
- Iron out the unreadable netmem driver support story. To be honest, the
conversation with Jakub & Pavel got a bit confusing for me. I've
implemented an approach in this set that makes sense to me, and
AFAICT, addresses the requirements. It may be good as-is, or it
may be a conversation starter/continuer. To be honest IMO there
are many ways to skin this cat and I don't see an extremely strong
reason to go for one approach over another. Here is one approach you
may like.
- Don't reset niov dma_addr on allocation & free.
- Add some tests to the selftest that catches some of the issues around
missing netlink attributes or deactivating mp-bound queues.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v19/
v18: https://patchwork.kernel.org/project/netdevbpf/list/?series=874848&state=*
====
v17 got minor feedback: (a) to beef up the description on patch 1 and (b)
to remove the leading underscores in the header definition.
I applied (a). (b) seems to be against current conventions so I did not
apply before further discussion.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v17/
v17: https://patchwork.kernel.org/project/netdevbpf/list/?series=869900&state=*
====
v16 also got a very thorough review and some testing (thanks again!).
Thes version addresses all the concerns reported on v15, in terms of
feedback and issues reported.
Major changes:
- Use ASSERT_RTNL.
- Moved around some of the page_pool helpers definitions so I can hide
some netmem helpers in private files as Jakub suggested.
- Don't make every net_iov hold a ref on the binding as Jakub suggested.
- Fix issue reported by Taehee where we access queues after they have
been freed.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v17/
v16: https://patchwork.kernel.org/project/netdevbpf/list/?series=866353&state=*
====
v15 got a thorough review and some testing, and this version addresses almost
all the feedback. Some more minor comments where the authors said it
could be done later, I left out.
Major changes:
- Addition of dma-buf introspection to page-pool-get and queue-get.
- Fixes to selftests suggested by Taehee.
- Fixes to documentation suggested by Donald.
- A couple of suggestions and fixes to TCP patches by Eric and David.
- Fixes to number assignements suggested by Arnd.
- Use rtnl_lock()ing to guard against queue reconfiguration while the
page_pool initialization is happening. (Jakub).
- Fixes to a few warnings reproduced by Taehee.
- Fixes to dma-buf binding suggested by Taehee and Jakub.
- Fixes to netlink UAPI suggested by Jakub
- Applied a number of Reviewed-bys and Acked-bys (including ones I lost
from v13+).
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v16/
One caveat: Taehee reproduced a KASAN warning and reported it here:
https://lore.kernel.org/netdev/CAMArcTUdCxOBYGF3vpbq=eBvqZfnc44KBaQTN7H-wqd…
I estimate the issue to be minor and easily fixable:
https://lore.kernel.org/netdev/CAHS8izNgaqC--GGE2xd85QB=utUnOHmioCsDd1TNxJW…
I hope to be able to follow up with a fix to net tree as net-next closes
imminently, but if this iteration doesn't make it in, I will repost with
a fix squashed after net-next reopens, no problem.
v15: https://patchwork.kernel.org/project/netdevbpf/list/?series=865481&state=*
====
No material changes in this version, only a fix to linking against
libynl.a from the last version. Per Jakub's instructions I've pulled one
of his patches into this series, and now use the new libynl.a correctly,
I hope.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v15/
v14: https://patchwork.kernel.org/project/netdevbpf/list/?series=865135&archive=…
====
No material changes in this version. Only rebase and re-verification on
top of net-next. v13, I think, raced with commit ebad6d0334793
("net/ipv4: Use nested-BH locking for ipv4_tcp_sk.") being merged to
net-next that caused a patchwork failure to apply. This series should
apply cleanly on commit c4532232fa2a4 ("selftests: net: remove unneeded
IP_GRE config").
I did not wait the customary 24hr as Jakub said it's OK to repost as soon
as I build test the rebased version:
https://lore.kernel.org/netdev/20240625075926.146d769d@kernel.org/
v13: https://patchwork.kernel.org/project/netdevbpf/list/?series=861406&archive=…
====
Major changes:
--------------
This iteration addresses Pavel's review comments, applies his
reviewed-by's, and seeks to fix the patchwork build error (sorry!).
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v13/
v12: https://patchwork.kernel.org/project/netdevbpf/list/?series=859747&state=*
====
Major changes:
--------------
This iteration only addresses one minor comment from Pavel with regards
to the trace printing of netmem, and the patchwork build error
introduced in v11 because I missed doing an allmodconfig build, sorry.
Other than that v11, AFAICT, received no feedback. There is one
discussion about how the specifics of plugging io uring memory through
the page pool, but not relevant to content in this particular patchset,
AFAICT.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v12/
v11: https://patchwork.kernel.org/project/netdevbpf/list/?series=857457&state=*
====
Major Changes:
--------------
v11 addresses feedback received in v10. The major change is the removal
of the memory provider ops as requested by Christoph. We still
accomplish the same thing, but utilizing direct function calls with if
statements rather than generic ops.
Additionally address sparse warnings, bugs and review comments from
folks that reviewed.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v11/
Detailed changelog:
-------------------
- Fixes in netdev_rx_queue_restart() from Pavel & David.
- Remove commit e650e8c3a36f5 ("net: page_pool: create hooks for
custom page providers") from the series to address Christoph's
feedback and rebased other patches on the series on this change.
- Fixed build errors with CONFIG_DMA_SHARED_BUFFER &&
!CONFIG_GENERIC_ALLOCATOR build.
- Fixed sparse warnings pointed out by Paolo.
- Drop unnecessary gro_pull_from_frag0 checks.
- Added Bagas reviewed-by to docs.
Cc: Bagas Sanjaya <bagasdotme(a)gmail.com>
Cc: Steven Rostedt <rostedt(a)goodmis.org>
Cc: Christoph Hellwig <hch(a)infradead.org>
Cc: Nikolay Aleksandrov <razor(a)blackwall.org>
Cc: Taehee Yoo <ap420073(a)gmail.com>
Cc: Donald Hunter <donald.hunter(a)gmail.com>
v10: https://patchwork.kernel.org/project/netdevbpf/list/?series=852422&state=*
====
Major Changes:
--------------
v9 was sent right before the merge window closed (sorry!). v10 is almost
a re-send of the series now that the merge window re-opened. Only
rebased to latest net-next and addressed some minor iterative comments
received on v9.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v10/
Detailed changelog:
-------------------
- Fixed tokens leaking in DONTNEED setsockopt (Nikolay).
- Moved net_iov_dma_addr() to devmem.c and made it a devmem specific
helpers (David).
- Rename hook alloc_pages to alloc_netmems as alloc_pages is now
preprocessor macro defined and causes a build error.
v9:
===
Major Changes:
--------------
GVE queue API has been merged. Submitting this version as non-RFC after
rebasing on top of the merged API, and dropped the out of tree queue API
I was carrying on github. Addressed the little feedback v8 has received.
Detailed changelog:
------------------
- Added new patch from David Wei to this series for
netdev_rx_queue_restart()
- Fixed sparse error.
- Removed CONFIG_ checks in netmem_is_net_iov()
- Flipped skb->readable to skb->unreadable
- Minor fixes to selftests & docs.
RFC v8:
=======
Major Changes:
--------------
- Fixed build error generated by patch-by-patch build.
- Applied docs suggestions from Randy.
RFC v7:
=======
Major Changes:
--------------
This revision largely rebases on top of net-next and addresses the feedback
RFCv6 received from folks, namely Jakub, Yunsheng, Arnd, David, & Pavel.
The series remains in RFC because the queue-API ndos defined in this
series are not yet implemented. I have a GVE implementation I carry out
of tree for my testing. A upstreamable GVE implementation is in the
works. Aside from that, in my estimation all the patches are ready for
review/merge. Please do take a look.
As usual the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v7/
Detailed changelog:
- Use admin-perm in netlink API.
- Addressed feedback from Jakub with regards to netlink API
implementation.
- Renamed devmem.c functions to something more appropriate for that
file.
- Improve the performance seen through the page_pool benchmark.
- Fix the value definition of all the SO_DEVMEM_* uapi.
- Various fixes to documentation.
Perf - page-pool benchmark:
---------------------------
Improved performance of bench_page_pool_simple.ko tests compared to v6:
https://pastebin.com/raw/v5dYRg8L
net-next base: 8 cycle fast path.
RFC v6: 10 cycle fast path.
RFC v7: 9 cycle fast path.
RFC v7 with CONFIG_DMA_SHARED_BUFFER disabled: 8 cycle fast path,
same as baseline.
Perf - Devmem TCP benchmark:
---------------------
Perf is about the same regardless of the changes in v7, namely the
removal of the static_branch_unlikely to improve the page_pool benchmark
performance:
189/200gbps bi-directional throughput with RX devmem TCP and regular TCP
TX i.e. ~95% line rate.
RFC v6:
=======
Major Changes:
--------------
This revision largely rebases on top of net-next and addresses the little
feedback RFCv5 received.
The series remains in RFC because the queue-API ndos defined in this
series are not yet implemented. I have a GVE implementation I carry out
of tree for my testing. A upstreamable GVE implementation is in the
works. Aside from that, in my estimation all the patches are ready for
review/merge. Please do take a look.
As usual the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v6/
This version also comes with some performance data recorded in the cover
letter (see below changelog).
Detailed changelog:
- Rebased on top of the merged netmem_ref changes.
- Converted skb->dmabuf to skb->readable (Pavel). Pavel's original
suggestion was to remove the skb->dmabuf flag entirely, but when I
looked into it closely, I found the issue that if we remove the flag
we have to dereference the shinfo(skb) pointer to obtain the first
frag to tell whether an skb is readable or not. This can cause a
performance regression if it dirties the cache line when the
shinfo(skb) was not really needed. Instead, I converted the skb->dmabuf
flag into a generic skb->readable flag which can be re-used by io_uring
0-copy RX.
- Squashed a few locking optimizations from Eric Dumazet in the RX path
and the DEVMEM_DONTNEED setsockopt.
- Expanded the tests a bit. Added validation for invalid scenarios and
added some more coverage.
Perf - page-pool benchmark:
---------------------------
bench_page_pool_simple.ko tests with and without these changes:
https://pastebin.com/raw/ncHDwAbn
AFAIK the number that really matters in the perf tests is the
'tasklet_page_pool01_fast_path Per elem'. This one measures at about 8
cycles without the changes but there is some 1 cycle noise in some
results.
With the patches this regresses to 9 cycles with the changes but there
is 1 cycle noise occasionally running this test repeatedly.
Lastly I tried disable the static_branch_unlikely() in
netmem_is_net_iov() check. To my surprise disabling the
static_branch_unlikely() check reduces the fast path back to 8 cycles,
but the 1 cycle noise remains.
Perf - Devmem TCP benchmark:
---------------------
189/200gbps bi-directional throughput with RX devmem TCP and regular TCP
TX i.e. ~95% line rate.
Major changes in RFC v5:
========================
1. Rebased on top of 'Abstract page from net stack' series and used the
new netmem type to refer to LSB set pointers instead of re-using
struct page.
2. Downgraded this series back to RFC and called it RFC v5. This is
because this series is now dependent on 'Abstract page from net
stack'[1] and the queue API. Both are removed from the series to
reduce the patch # and those bits are fairly independent or
pre-requisite work.
3. Reworked the page_pool devmem support to use netmem and for some
more unified handling.
4. Reworked the reference counting of net_iov (renamed from
page_pool_iov) to use pp_ref_count for refcounting.
The full changes including the dependent series and GVE page pool
support is here:
https://github.com/mina/linux/commits/tcpdevmem-rfcv5/
[1] https://patchwork.kernel.org/project/netdevbpf/list/?series=810774
Major changes in v1:
====================
1. Implemented MVP queue API ndos to remove the userspace-visible
driver reset.
2. Fixed issues in the napi_pp_put_page() devmem frag unref path.
3. Removed RFC tag.
Many smaller addressed comments across all the patches (patches have
individual change log).
Full tree including the rest of the GVE driver changes:
https://github.com/mina/linux/commits/tcpdevmem-v1
Changes in RFC v3:
==================
1. Pulled in the memory-provider dependency from Jakub's RFC[1] to make the
series reviewable and mergeable.
2. Implemented multi-rx-queue binding which was a todo in v2.
3. Fix to cmsg handling.
The sticking point in RFC v2[2] was the device reset required to refill
the device rx-queues after the dmabuf bind/unbind. The solution
suggested as I understand is a subset of the per-queue management ops
Jakub suggested or similar:
https://lore.kernel.org/netdev/20230815171638.4c057dcd@kernel.org/
This is not addressed in this revision, because:
1. This point was discussed at netconf & netdev and there is openness to
using the current approach of requiring a device reset.
2. Implementing individual queue resetting seems to be difficult for my
test bed with GVE. My prototype to test this ran into issues with the
rx-queues not coming back up properly if reset individually. At the
moment I'm unsure if it's a mistake in the POC or a genuine issue in
the virtualization stack behind GVE, which currently doesn't test
individual rx-queue restart.
3. Our usecases are not bothered by requiring a device reset to refill
the buffer queues, and we'd like to support NICs that run into this
limitation with resetting individual queues.
My thought is that drivers that have trouble with per-queue configs can
use the support in this series, while drivers that support new netdev
ops to reset individual queues can automatically reset the queue as
part of the dma-buf bind/unbind.
The same approach with device resets is presented again for consideration
with other sticking points addressed.
This proposal includes the rx devmem path only proposed for merge. For a
snapshot of my entire tree which includes the GVE POC page pool support &
device memory support:
https://github.com/torvalds/linux/compare/master...mina:linux:tcpdevmem-v3
[1] https://lore.kernel.org/netdev/f8270765-a27b-6ccf-33ea-cda097168d79@redhat.…
[2] https://lore.kernel.org/netdev/CAHS8izOVJGJH5WF68OsRWFKJid1_huzzUK+hpKbLcL4…
Changes in RFC v2:
==================
The sticking point in RFC v1[1] was the dma-buf pages approach we used to
deliver the device memory to the TCP stack. RFC v2 is a proof-of-concept
that attempts to resolve this by implementing scatterlist support in the
networking stack, such that we can import the dma-buf scatterlist
directly. This is the approach proposed at a high level here[2].
Detailed changes:
1. Replaced dma-buf pages approach with importing scatterlist into the
page pool.
2. Replace the dma-buf pages centric API with a netlink API.
3. Removed the TX path implementation - there is no issue with
implementing the TX path with scatterlist approach, but leaving
out the TX path makes it easier to review.
4. Functionality is tested with this proposal, but I have not conducted
perf testing yet. I'm not sure there are regressions, but I removed
perf claims from the cover letter until they can be re-confirmed.
5. Added Signed-off-by: contributors to the implementation.
6. Fixed some bugs with the RX path since RFC v1.
Any feedback welcome, but specifically the biggest pending questions
needing feedback IMO are:
1. Feedback on the scatterlist-based approach in general.
2. Netlink API (Patch 1 & 2).
3. Approach to handle all the drivers that expect to receive pages from
the page pool (Patch 6).
[1] https://lore.kernel.org/netdev/dfe4bae7-13a0-3c5d-d671-f61b375cb0b4@gmail.c…
[2] https://lore.kernel.org/netdev/CAHS8izPm6XRS54LdCDZVd0C75tA1zHSu6jLVO8nzTLX…
==================
* TL;DR:
Device memory TCP (devmem TCP) is a proposal for transferring data to and/or
from device memory efficiently, without bouncing the data to a host memory
buffer.
* Problem:
A large amount of data transfers have device memory as the source and/or
destination. Accelerators drastically increased the volume of such transfers.
Some examples include:
- ML accelerators transferring large amounts of training data from storage into
GPU/TPU memory. In some cases ML training setup time can be as long as 50% of
TPU compute time, improving data transfer throughput & efficiency can help
improving GPU/TPU utilization.
- Distributed training, where ML accelerators, such as GPUs on different hosts,
exchange data among them.
- Distributed raw block storage applications transfer large amounts of data with
remote SSDs, much of this data does not require host processing.
Today, the majority of the Device-to-Device data transfers the network are
implemented as the following low level operations: Device-to-Host copy,
Host-to-Host network transfer, and Host-to-Device copy.
The implementation is suboptimal, especially for bulk data transfers, and can
put significant strains on system resources, such as host memory bandwidth,
PCIe bandwidth, etc. One important reason behind the current state is the
kernel’s lack of semantics to express device to network transfers.
* Proposal:
In this patch series we attempt to optimize this use case by implementing
socket APIs that enable the user to:
1. send device memory across the network directly, and
2. receive incoming network packets directly into device memory.
Packet _payloads_ go directly from the NIC to device memory for receive and from
device memory to NIC for transmit.
Packet _headers_ go to/from host memory and are processed by the TCP/IP stack
normally. The NIC _must_ support header split to achieve this.
Advantages:
- Alleviate host memory bandwidth pressure, compared to existing
network-transfer + device-copy semantics.
- Alleviate PCIe BW pressure, by limiting data transfer to the lowest level
of the PCIe tree, compared to traditional path which sends data through the
root complex.
* Patch overview:
** Part 1: netlink API
Gives user ability to bind dma-buf to an RX queue.
** Part 2: scatterlist support
Currently the standard for device memory sharing is DMABUF, which doesn't
generate struct pages. On the other hand, networking stack (skbs, drivers, and
page pool) operate on pages. We have 2 options:
1. Generate struct pages for dmabuf device memory, or,
2. Modify the networking stack to process scatterlist.
Approach #1 was attempted in RFC v1. RFC v2 implements approach #2.
** part 3: page pool support
We piggy back on page pool memory providers proposal:
https://github.com/kuba-moo/linux/tree/pp-providers
It allows the page pool to define a memory provider that provides the
page allocation and freeing. It helps abstract most of the device memory
TCP changes from the driver.
** part 4: support for unreadable skb frags
Page pool iovs are not accessible by the host; we implement changes
throughput the networking stack to correctly handle skbs with unreadable
frags.
** Part 5: recvmsg() APIs
We define user APIs for the user to send and receive device memory.
Not included with this series is the GVE devmem TCP support, just to
simplify the review. Code available here if desired:
https://github.com/mina/linux/tree/tcpdevmem
This series is built on top of net-next with Jakub's pp-providers changes
cherry-picked.
* NIC dependencies:
1. (strict) Devmem TCP require the NIC to support header split, i.e. the
capability to split incoming packets into a header + payload and to put
each into a separate buffer. Devmem TCP works by using device memory
for the packet payload, and host memory for the packet headers.
2. (optional) Devmem TCP works better with flow steering support & RSS support,
i.e. the NIC's ability to steer flows into certain rx queues. This allows the
sysadmin to enable devmem TCP on a subset of the rx queues, and steer
devmem TCP traffic onto these queues and non devmem TCP elsewhere.
The NIC I have access to with these properties is the GVE with DQO support
running in Google Cloud, but any NIC that supports these features would suffice.
I may be able to help reviewers bring up devmem TCP on their NICs.
* Testing:
The series includes a udmabuf kselftest that show a simple use case of
devmem TCP and validates the entire data path end to end without
a dependency on a specific dmabuf provider.
** Test Setup
Kernel: net-next with this series and memory provider API cherry-picked
locally.
Hardware: Google Cloud A3 VMs.
NIC: GVE with header split & RSS & flow steering support.
Cc: Pavel Begunkov <asml.silence(a)gmail.com>
Cc: David Wei <dw(a)davidwei.uk>
Cc: Jason Gunthorpe <jgg(a)ziepe.ca>
Cc: Yunsheng Lin <linyunsheng(a)huawei.com>
Cc: Shailend Chand <shailend(a)google.com>
Cc: Harshitha Ramamurthy <hramamurthy(a)google.com>
Cc: Shakeel Butt <shakeel.butt(a)linux.dev>
Cc: Jeroen de Borst <jeroendb(a)google.com>
Cc: Praveen Kaligineedi <pkaligineedi(a)google.com>
Mina Almasry (13):
netdev: add netdev_rx_queue_restart()
net: netdev netlink api to bind dma-buf to a net device
netdev: support binding dma-buf to netdevice
netdev: netdevice devmem allocator
page_pool: devmem support
memory-provider: dmabuf devmem memory provider
net: support non paged skb frags
net: add support for skbs with unreadable frags
tcp: RX path for devmem TCP
net: add SO_DEVMEM_DONTNEED setsockopt to release RX frags
net: add devmem TCP documentation
selftests: add ncdevmem, netcat for devmem TCP
netdev: add dmabuf introspection
Documentation/netlink/specs/netdev.yaml | 61 +++
Documentation/networking/devmem.rst | 269 +++++++++++
Documentation/networking/index.rst | 1 +
arch/alpha/include/uapi/asm/socket.h | 6 +
arch/mips/include/uapi/asm/socket.h | 6 +
arch/parisc/include/uapi/asm/socket.h | 6 +
arch/sparc/include/uapi/asm/socket.h | 6 +
include/linux/netdevice.h | 2 +
include/linux/skbuff.h | 61 ++-
include/linux/skbuff_ref.h | 9 +-
include/linux/socket.h | 1 +
include/net/devmem.h | 128 ++++++
include/net/mp_dmabuf_devmem.h | 44 ++
include/net/netdev_rx_queue.h | 5 +
include/net/netmem.h | 169 ++++++-
include/net/page_pool/helpers.h | 39 +-
include/net/page_pool/types.h | 22 +-
include/net/sock.h | 2 +
include/net/tcp.h | 5 +-
include/trace/events/page_pool.h | 12 +-
include/uapi/asm-generic/socket.h | 6 +
include/uapi/linux/netdev.h | 13 +
include/uapi/linux/uio.h | 17 +
net/core/Makefile | 3 +-
net/core/datagram.c | 6 +
net/core/dev.c | 22 +-
net/core/devmem.c | 374 +++++++++++++++
net/core/gro.c | 3 +-
net/core/netdev-genl-gen.c | 23 +
net/core/netdev-genl-gen.h | 6 +
net/core/netdev-genl.c | 118 +++++
net/core/netdev_rx_queue.c | 81 ++++
net/core/netmem_priv.h | 31 ++
net/core/page_pool.c | 117 +++--
net/core/page_pool_priv.h | 31 ++
net/core/page_pool_user.c | 29 ++
net/core/skbuff.c | 77 +++-
net/core/sock.c | 68 +++
net/ethtool/common.c | 8 +
net/ipv4/esp4.c | 3 +-
net/ipv4/tcp.c | 261 ++++++++++-
net/ipv4/tcp_input.c | 13 +-
net/ipv4/tcp_ipv4.c | 16 +
net/ipv4/tcp_minisocks.c | 2 +
net/ipv4/tcp_output.c | 5 +-
net/ipv6/esp6.c | 3 +-
net/packet/af_packet.c | 4 +-
tools/include/uapi/linux/netdev.h | 13 +
tools/testing/selftests/net/.gitignore | 1 +
tools/testing/selftests/net/Makefile | 9 +
tools/testing/selftests/net/ncdevmem.c | 587 ++++++++++++++++++++++++
51 files changed, 2683 insertions(+), 121 deletions(-)
create mode 100644 Documentation/networking/devmem.rst
create mode 100644 include/net/devmem.h
create mode 100644 include/net/mp_dmabuf_devmem.h
create mode 100644 net/core/devmem.c
create mode 100644 net/core/netdev_rx_queue.c
create mode 100644 net/core/netmem_priv.h
create mode 100644 tools/testing/selftests/net/ncdevmem.c
--
2.46.0.184.g6999bdac58-goog
Hi,
This attempts to implement PT_LOAD p_align support for static PIE builds.
I intend this to go into -next after the coming merge window so we can
maximize bake time. In the past we've had regressions with both the
selftests and the ELF loader. Hopefully we can shake everything out over
a few months. :)
Thanks!
-Kees
Kees Cook (3):
selftests/exec: Build both static and non-static load_address tests
binfmt_elf: Calculate total_size earlier
binfmt_elf: Honor PT_LOAD alignment for static PIE
fs/binfmt_elf.c | 94 ++++++++++++++-------
tools/testing/selftests/exec/Makefile | 19 +++--
tools/testing/selftests/exec/load_address.c | 67 ++++++++++++---
3 files changed, 130 insertions(+), 50 deletions(-)
--
2.34.1
Hello Hou Tao,
This is a semi-automatic email about new static checker warnings.
Commit b4b7a4099b8c ("selftests/bpf: Factor out get_xlated_program()
helper") from Jan 5, 2024, leads to the following Smatch complaint:
./tools/testing/selftests/bpf/testing_helpers.c:455 get_xlated_program()
warn: variable dereferenced before check 'buf' (see line 454)
./tools/testing/selftests/bpf/testing_helpers.c
453 *cnt = xlated_prog_len / buf_element_size;
454 *buf = calloc(*cnt, buf_element_size);
455 if (!buf) {
^^^
This should be *buf.
456 perror("can't allocate xlated program buffer");
457 return -ENOMEM;
regards,
dan carpenter
This series introduces a new VIOMMU infrastructure and related ioctls.
IOMMUFD has been using the HWPT infrastructure for all cases, including a
nested IO page table support. Yet, there're limitations for an HWPT-based
structure to support some advanced HW-accelerated features, such as CMDQV
on NVIDIA Grace, and HW-accelerated vIOMMU on AMD. Even for a multi-IOMMU
environment, it is not straightforward for nested HWPTs to share the same
parent HWPT (stage-2 IO pagetable), with the HWPT infrastructure alone.
The new VIOMMU object is an additional layer, between the nested HWPT and
its parent HWPT, to give to both the IOMMUFD core and an IOMMU driver an
additional structure to support HW-accelerated feature:
----------------------------
---------------- | | paging_hwpt0 |
| hwpt_nested0 |--->| viommu0 ------------------
---------------- | | HW-accel feats |
----------------------------
On a multi-IOMMU system, the VIOMMU object can be instanced to the number
of vIOMMUs in a guest VM, while holding the same parent HWPT to share the
stage-2 IO pagetable. Each VIOMMU then just need to only allocate its own
VMID to attach the shared stage-2 IO pagetable to the physical IOMMU:
----------------------------
---------------- | | paging_hwpt0 |
| hwpt_nested0 |--->| viommu0 ------------------
---------------- | | VMID0 |
----------------------------
----------------------------
---------------- | | paging_hwpt0 |
| hwpt_nested1 |--->| viommu1 ------------------
---------------- | | VMID1 |
----------------------------
As an initial part-1, add ioctls to support a VIOMMU-based invalidation:
IOMMUFD_CMD_VIOMMU_ALLOC to allocate a VIOMMU object
IOMMUFD_CMD_VIOMMU_SET/UNSET_VDEV_ID to set/clear device's virtual ID
IOMMUFD_CMD_VIOMMU_INVALIDATE to flush cache by a given driver data
Worth noting that the VDEV_ID is for a per-VIOMMU device list for drivers
to look up the device's physical instance from its virtual ID in a VM. It
is essential for a VIOMMU-based invalidation where the request contains a
device's virtual ID for its device cache flush, e.g. ATC invalidation.
As for the implementation of the series, add an IOMMU_VIOMMU_TYPE_DEFAULT
type for a core-allocated-core-managed VIOMMU object, allowing drivers to
simply hook a default viommu ops for viommu-based invalidation alone. And
provide some viommu helpers to drivers for VDEV_ID translation and parent
domain lookup. Introduce an IOMMU_VIOMMU_INVALIDATE_DATA_ARM_SMMUV3 for a
real world use case. This adds supports of arm-smmuv-v3's CMDQ_OP_ATC_INV
and CMDQ_OP_CFGI_CD/ALL commands, supplementing HWPT-based invalidations.
In the future, drivers will also be able to choose a driver-managed type
to hold its own structure by adding a new type to enum iommu_viommu_type.
More VIOMMU-based structures and ioctls will be introduced in part-2/3 to
support a driver-managed VIOMMU, e.g. VQUEUE object for a HW accelerated
queue, VIRQ (or VEVENT) object for IRQ injections. Although we repurposed
the VIOMMU object from an earlier RFC discussion, for a referece:
https://lore.kernel.org/all/cover.1712978212.git.nicolinc@nvidia.com/
This series is on Github:
https://github.com/nicolinc/iommufd/commits/iommufd_viommu_p1-v1
Thanks!
Nicolin
Jason Gunthorpe (1):
iommu/arm-smmu-v3: Allow ATS for IOMMU_DOMAIN_NESTED
Nicolin Chen (15):
iommufd/viommu: Add IOMMUFD_OBJ_VIOMMU and IOMMU_VIOMMU_ALLOC ioctl
iommu: Pass in a viommu pointer to domain_alloc_user op
iommufd: Allow pt_id to carry viommu_id for IOMMU_HWPT_ALLOC
iommufd/selftest: Add IOMMU_VIOMMU_ALLOC test coverage
iommufd/viommu: Add IOMMU_VIOMMU_SET/UNSET_VDEV_ID ioctl
iommufd/selftest: Add IOMMU_VIOMMU_SET/UNSET_VDEV_ID test coverage
iommufd/viommu: Add cache_invalidate for IOMMU_VIOMMU_TYPE_DEFAULT
iommufd/viommu: Add IOMMU_VIOMMU_INVALIDATE ioctl
iommufd/viommu: Make iommufd_viommu_find_device a public API
iommufd/selftest: Add mock_viommu_invalidate_user op
iommufd/selftest: Add IOMMU_TEST_OP_DEV_CHECK_CACHE test command
iommufd/selftest: Add coverage for IOMMU_VIOMMU_INVALIDATE ioctl
iommufd/viommu: Add iommufd_viommu_to_parent_domain helper
iommu/arm-smmu-v3: Extract an __arm_smmu_cache_invalidate_user helper
iommu/arm-smmu-v3: Add viommu cache invalidation support
drivers/iommu/amd/iommu.c | 1 +
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 90 +++++-
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.h | 2 +
drivers/iommu/intel/iommu.c | 1 +
drivers/iommu/iommufd/Makefile | 3 +-
drivers/iommu/iommufd/device.c | 9 +
drivers/iommu/iommufd/hw_pagetable.c | 27 +-
drivers/iommu/iommufd/iommufd_private.h | 37 +++
drivers/iommu/iommufd/iommufd_test.h | 30 ++
drivers/iommu/iommufd/main.c | 15 +
drivers/iommu/iommufd/selftest.c | 88 +++++-
drivers/iommu/iommufd/viommu.c | 249 +++++++++++++++++
include/linux/iommu.h | 6 +
include/linux/iommufd.h | 35 +++
include/uapi/linux/iommufd.h | 139 ++++++++-
tools/testing/selftests/iommu/iommufd.c | 263 +++++++++++++++++-
tools/testing/selftests/iommu/iommufd_utils.h | 126 +++++++++
17 files changed, 1095 insertions(+), 26 deletions(-)
create mode 100644 drivers/iommu/iommufd/viommu.c
--
2.43.0
Make timespec pointers, pointers to const in checklist function. As a
consequence, make list parameter in checklist function pointer to const
as well. Const-correctness increases readability.
Improvement was found by running cppcheck tool on the patched file as
follows:
```
cppcheck --enable=all \
tools/testing/selftests/timers/threadtest.c \
--suppress=missingIncludeSystem \
--suppress=unusedFunction
```
Reviewed-by: Shuah Khan <skhan(a)linuxfoundation.org>
Signed-off-by: Piotr Zalewski <pZ010001011111(a)proton.me>
---
tools/testing/selftests/timers/threadtest.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tools/testing/selftests/timers/threadtest.c b/tools/testing/selftests/timers/threadtest.c
index 76b38e41d9c7..d5564bbf0e50 100644
--- a/tools/testing/selftests/timers/threadtest.c
+++ b/tools/testing/selftests/timers/threadtest.c
@@ -38,10 +38,10 @@ struct timespec global_list[LISTSIZE];
int listcount = 0;
-void checklist(struct timespec *list, int size)
+void checklist(const struct timespec *list, int size)
{
int i, j;
- struct timespec *a, *b;
+ const struct timespec *a, *b;
/* scan the list */
for (i = 0; i < size-1; i++) {
--
2.46.0
This patch series introduces a set of regression tests for various s390x
CPU subfunctions in KVM. The tests ensure that the KVM implementation accurately
reflects the behavior of actual CPU instructions for these subfunctions.
The series adds tests for a total of 15 instructions across five patches,
covering a range of operations including sorting, compression, and various
cryptographic functions. Each patch follows a consistent testing pattern:
1. Obtain the KVM_S390_VM_CPU_MACHINE_SUBFUNC attribute for the VM.
2. Execute the relevant asm instructions.
3. Compare KVM-reported results with direct instruction execution results.
Testing has been performed on s390x hardware with KVM support. All tests
pass successfully, verifying the correct implementation of these
subfunctions in KVM.
Hariharan Mari (5):
KVM: s390: selftests: Add regression tests for SORTL and DFLTCC CPU
subfunctions
KVM: s390: selftests: Add regression tests for PRNO, KDSA and KMA
crypto subfunctions
KVM: s390: selftests: Add regression tests for KMCTR, KMF, KMO and PCC
crypto subfunctions
KVM: s390: selftests: Add regression tests for KMAC, KMC, KM, KIMD and
KLMD crypto subfunctions
KVM: s390: selftests: Add regression tests for PLO subfunctions
tools/testing/selftests/kvm/Makefile | 1 +
.../selftests/kvm/include/s390x/facility.h | 50 +++
.../kvm/s390x/cpumodel_subfuncs_test.c | 343 ++++++++++++++++++
3 files changed, 394 insertions(+)
create mode 100644 tools/testing/selftests/kvm/include/s390x/facility.h
create mode 100644 tools/testing/selftests/kvm/s390x/cpumodel_subfuncs_test.c
--
2.45.2
This test validates that the mapping between a mm_cid and a NUMA node id
remains invariant for the process lifetime for a process with a number of
threads >= number of allowed CPUs. In other words, it validates that if
any thread within the process running on behalf of a mm_cid N observes a
NUMA node id M, all threads within this process will always observe the
same NUMA node id value when running on behalf of that same mm_cid.
This characteristic is important for NUMA locality.
On all architectures except Power, the NUMA topology is never
reconfigured after a CPU has been associated with a NUMA node in the
system lifetime. Even on Power, we can assume that NUMA topology
reconfiguration happens rarely, and therefore we do not expect it to
happen while the NUMA test is running.
As a result the NUMA node id associated with a mm_cid should be
invariant as long as:
- A process has a number of threads >= number of allowed CPUs,
- The allowed CPUs mask is unchanged, and
- The NUMA configuration is unchanged.
This test is skipped on architectures that do not implement
rseq_load_u32_u32.
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers(a)efficios.com>
Cc: Peter Zijlstra <peterz(a)infradead.org>
Cc: Ingo Molnar <mingo(a)redhat.com>
Cc: Shuah Khan <skhan(a)linuxfoundation.org>
Cc: linux-kselftest(a)vger.kernel.org
---
tools/testing/selftests/rseq/.gitignore | 1 +
tools/testing/selftests/rseq/Makefile | 2 +-
.../testing/selftests/rseq/basic_numa_test.c | 144 ++++++++++++++++++
3 files changed, 146 insertions(+), 1 deletion(-)
create mode 100644 tools/testing/selftests/rseq/basic_numa_test.c
diff --git a/tools/testing/selftests/rseq/.gitignore b/tools/testing/selftests/rseq/.gitignore
index 16496de5f6ce..8a8d163cbb9f 100644
--- a/tools/testing/selftests/rseq/.gitignore
+++ b/tools/testing/selftests/rseq/.gitignore
@@ -1,4 +1,5 @@
# SPDX-License-Identifier: GPL-2.0-only
+basic_numa_test
basic_percpu_ops_test
basic_percpu_ops_mm_cid_test
basic_test
diff --git a/tools/testing/selftests/rseq/Makefile b/tools/testing/selftests/rseq/Makefile
index 5a3432fceb58..9ef1c949114a 100644
--- a/tools/testing/selftests/rseq/Makefile
+++ b/tools/testing/selftests/rseq/Makefile
@@ -14,7 +14,7 @@ LDLIBS += -lpthread -ldl
# still track changes to header files and depend on shared object.
OVERRIDE_TARGETS = 1
-TEST_GEN_PROGS = basic_test basic_percpu_ops_test basic_percpu_ops_mm_cid_test param_test \
+TEST_GEN_PROGS = basic_test basic_numa_test basic_percpu_ops_test basic_percpu_ops_mm_cid_test param_test \
param_test_benchmark param_test_compare_twice param_test_mm_cid \
param_test_mm_cid_benchmark param_test_mm_cid_compare_twice
diff --git a/tools/testing/selftests/rseq/basic_numa_test.c b/tools/testing/selftests/rseq/basic_numa_test.c
new file mode 100644
index 000000000000..8e51c662057d
--- /dev/null
+++ b/tools/testing/selftests/rseq/basic_numa_test.c
@@ -0,0 +1,144 @@
+// SPDX-License-Identifier: LGPL-2.1
+/*
+ * Basic rseq NUMA test. Validate that (mm_cid, numa_node_id) pairs are
+ * invariant when the number of threads >= number of allowed CPUs, as
+ * long as those preconditions are respected:
+ *
+ * - A process has a number of threads >= number of allowed CPUs,
+ * - The allowed CPUs mask is unchanged, and
+ * - The NUMA configuration is unchanged.
+ */
+#define _GNU_SOURCE
+#include <assert.h>
+#include <sched.h>
+#include <signal.h>
+#include <stdio.h>
+#include <string.h>
+#include <poll.h>
+#include <sys/time.h>
+
+#include "rseq.h"
+
+#define NR_LOOPS 100
+
+static int nr_threads, nr_active_threads, test_go, test_stop;
+
+#ifdef RSEQ_ARCH_HAS_LOAD_U32_U32
+
+static int cpu_numa_id[CPU_SETSIZE];
+
+static int get_affinity_weight(void)
+{
+ cpu_set_t allowed_cpus;
+
+ if (sched_getaffinity(0, sizeof(allowed_cpus), &allowed_cpus)) {
+ perror("sched_getaffinity");
+ abort();
+ }
+ return CPU_COUNT(&allowed_cpus);
+}
+
+static void numa_id_init(void)
+{
+ int i;
+
+ for (i = 0; i < CPU_SETSIZE; i++)
+ cpu_numa_id[i] = -1;
+}
+
+static void *test_thread(void *arg)
+{
+ int i;
+
+ if (rseq_register_current_thread()) {
+ fprintf(stderr, "Error: rseq_register_current_thread(...) failed(%d): %s\n",
+ errno, strerror(errno));
+ abort();
+ }
+ /*
+ * Rendez-vous across all threads to make sure the number of
+ * threads >= number of possible CPUs for the entire test duration.
+ */
+ if (__atomic_add_fetch(&nr_active_threads, 1, __ATOMIC_RELAXED) == nr_threads)
+ __atomic_store_n(&test_go, 1, __ATOMIC_RELAXED);
+ while (!__atomic_load_n(&test_go, __ATOMIC_RELAXED))
+ rseq_barrier();
+
+ for (i = 0; i < NR_LOOPS; i++) {
+ uint32_t mm_cid, node;
+ int cached_node_id;
+
+ while (rseq_load_u32_u32(RSEQ_MO_RELAXED, &mm_cid,
+ &rseq_get_abi()->mm_cid,
+ &node, &rseq_get_abi()->node_id) != 0) {
+ /* Retry. */
+ }
+ cached_node_id = RSEQ_READ_ONCE(cpu_numa_id[mm_cid]);
+ if (cached_node_id == -1) {
+ RSEQ_WRITE_ONCE(cpu_numa_id[mm_cid], node);
+ } else {
+ if (node != cached_node_id) {
+ fprintf(stderr, "Error: NUMA node id discrepancy: mm_cid %u cached node id %d node id %u.\n",
+ mm_cid, cached_node_id, node);
+ fprintf(stderr, "This is likely a kernel bug, or caused by a concurrent NUMA topology reconfiguration.\n");
+ abort();
+ }
+ }
+ (void) poll(NULL, 0, 10); /* wait 10ms */
+ }
+ /*
+ * Rendez-vous before exiting all threads to make sure the
+ * number of threads >= number of possible CPUs for the entire
+ * test duration.
+ */
+ if (__atomic_sub_fetch(&nr_active_threads, 1, __ATOMIC_RELAXED) == 0)
+ __atomic_store_n(&test_stop, 1, __ATOMIC_RELAXED);
+ while (!__atomic_load_n(&test_stop, __ATOMIC_RELAXED))
+ rseq_barrier();
+
+ if (rseq_unregister_current_thread()) {
+ fprintf(stderr, "Error: rseq_unregister_current_thread(...) failed(%d): %s\n",
+ errno, strerror(errno));
+ abort();
+ }
+ return NULL;
+}
+
+static int test_numa(void)
+{
+ pthread_t tid[nr_threads];
+ int err, i;
+ void *tret;
+
+ numa_id_init();
+
+ printf("testing rseq (mm_cid, numa_node_id) invariant, multi-threaded (%d threads)\n",
+ nr_threads);
+
+ for (i = 0; i < nr_threads; i++) {
+ err = pthread_create(&tid[i], NULL, test_thread, NULL);
+ if (err != 0)
+ abort();
+ }
+
+ for (i = 0; i < nr_threads; i++) {
+ err = pthread_join(tid[i], &tret);
+ if (err != 0)
+ abort();
+ }
+
+ return 0;
+}
+#else
+static int test_numa(void)
+{
+ fprintf(stderr, "rseq_load_u32_u32 is not implemented on this architecture. Skipping numa test.\n");
+ return 0;
+}
+#endif
+
+int main(int argc, char **argv)
+{
+ nr_threads = get_affinity_weight();
+ return test_numa();
+}
--
2.39.2
There are 2 issues for the current udpgro test. The first one is the testing
doesn't record all the failures, which may report pass but the test actually
failed. e.g.
https://netdev-3.bots.linux.dev/vmksft-net/results/725661/45-udpgro-sh/stdo…
The other one is after commit d7db7775ea2e ("net: veth: do not manipulate
GRO when using XDP"), there is no need to load xdp program to enable GRO
on veth device.
Hangbin Liu (2):
selftests: udpgro: report error when receive failed
selftests: udpgro: no need to load xdp for gro
tools/testing/selftests/net/udpgro.sh | 53 ++++++++++++++-------------
1 file changed, 28 insertions(+), 25 deletions(-)
--
2.45.0
Hi,
Here is the v4 patch to support polling on event 'hist' file.
The previous version is here;
https://lore.kernel.org/all/172359427367.323666.6446548762874507863.stgit@d…
This version uses getopt() in poll test command in [3/3] according to
Shuah's comment in the previous thread.
Background
----------
There has been interest in allowing user programs to monitor kernel
events in real time. Ftrace provides `trace_pipe` interface to wait
on events in the ring buffer, but it is needed to wait until filling
up a page with events in the ring buffer. We can also peek the
`trace` file periodically, but that is inefficient way to monitor
a randomely happening event.
Overview
--------
This patch set allows user to `poll`(or `select`, `epoll`) on event
histogram interface. As you know each event has its own `hist` file
which shows histograms generated by trigger action. So user can set
a new hist trigger on any event you want to monitor, and poll on the
`hist` file until it is updated.
There are 2 poll events are supported, POLLIN and POLLPRI. POLLIN
means that there are any readable update on `hist` file and this
event will be flashed only when you call read(). So, this is
useful if you want to read the histogram periodically.
The other POLLPRI event is for monitoring trace event. Like the
POLLIN, this will be returned when the histogram is updated, but
you don't need to read() the file and use poll() again.
Note that this waits for histogram update (not event arrival), thus
you must set a histogram on the event at first.
Usage
-----
Here is an example usage:
----
TRACEFS=/sys/kernel/tracing
EVENT=$TRACEFS/events/sched/sched_process_free
# setup histogram trigger and enable event
echo "hist:key=comm" >> $EVENT/trigger
echo 1 > $EVENT/enable
# Wait for update
poll pri $EVENT/hist
# Event arrived.
echo "process free event is comming"
tail $TRACEFS/trace
----
The 'poll' command is in the selftest patch.
You can take this series also from here;
https://git.kernel.org/pub/scm/linux/kernel/git/mhiramat/linux.git/log/?h=t…
Thank you,
---
Masami Hiramatsu (Google) (3):
tracing/hist: Add poll(POLLIN) support on hist file
tracing/hist: Support POLLPRI event for poll on histogram
selftests/tracing: Add hist poll() support test
include/linux/trace_events.h | 5 +
kernel/trace/trace_events.c | 18 ++++
kernel/trace/trace_events_hist.c | 101 +++++++++++++++++++-
tools/testing/selftests/ftrace/Makefile | 2
tools/testing/selftests/ftrace/poll.c | 74 +++++++++++++++
.../ftrace/test.d/trigger/trigger-hist-poll.tc | 74 +++++++++++++++
6 files changed, 271 insertions(+), 3 deletions(-)
create mode 100644 tools/testing/selftests/ftrace/poll.c
create mode 100644 tools/testing/selftests/ftrace/test.d/trigger/trigger-hist-poll.tc
--
Masami Hiramatsu (Google) <mhiramat(a)kernel.org>
There are multiple possible timer sources which could be useful for
the sound stream synchronization: hrtimers, hardware clocks (e.g. PTP),
timer wheels (jiffies). Currently, using one of them to synchronize
the audio stream of snd-aloop module would require writing a
kernel-space driver which exports an ALSA timer through the
snd_timer interface.
However, it is not really convenient for application developers, who may
want to define their custom timer sources for audio synchronization.
For instance, we could have a network application which receives frames
and sends them to snd-aloop pcm device, and another application
listening on the other end of snd-aloop. It makes sense to transfer a
new period of data only when certain amount of frames is received
through the network, but definitely not when a certain amount of jiffies
on a local system elapses. Since all of the devices are purely virtual
it won't introduce any glitches and will help the application developers
to avoid using sample-rate conversion.
This patch series introduces userspace-driven ALSA timers: virtual
timers which are created and controlled from userspace. The timer can
be created from the userspace using the new ioctl SNDRV_TIMER_IOCTL_CREATE.
After creating a timer, it becomes available for use system-wide, so it
can be passed to snd-aloop as a timer source (timer_source parameter
would be "-1.SNDRV_TIMER_GLOBAL_UDRIVEN.{timer_id}"). When the userspace
app decides to trigger a timer, it calls another ioctl
SNDRV_TIMER_IOCTL_TRIGGER on the file descriptor of a timer. It
initiates a transfer of a new period of data.
Userspace-driven timers are associated with file descriptors. If the
application wishes to destroy the timer, it can simply release the file
descriptor of a virtual timer.
I believe introducing new ioctl calls is quite inconvenient (as we have
a limited amount of them), but other possible ways of app <-> kernel
communication (like virtual FS) seem completely inappropriate for this
task (but I'd love to discuss alternative solutions).
This patch series also updates the snd-aloop module so the global timers
can be used as a timer_source for it (it allows using userspace-driven
timers as timer source).
V1 -> V2:
- Fix some problems found by Christophe Jaillet
<christophe.jaillet(a)wanadoo.fr>
V2 -> V3:
- Add improvements suggested by Takashi Iwai <tiwai(a)suse.de>
V3 -> V4:
- Address comments from Jaroslav Kysela <perex(a)perex.cz> and Mark Brown
<broonie(a)kernel.org>
V4 -> V5:
- Add missing error processing noticed by Takashi Iwai <tiwai(a)suse.de>
- Return timer file descriptor as part of the snd_timer_uinfo structure.
This is a more standard way of using ioctl interface, where the return
value of the ioctl is either 0 or an error code.
Please, find the patch-specific changelog in the following patches.
Ivan Orlov (4):
ALSA: aloop: Allow using global timers
Docs/sound: Add documentation for userspace-driven ALSA timers
ALSA: timer: Introduce virtual userspace-driven timers
selftests: ALSA: Cover userspace-driven timers with test
Documentation/sound/index.rst | 1 +
Documentation/sound/utimers.rst | 126 +++++++++++
include/uapi/sound/asound.h | 17 +-
sound/core/Kconfig | 10 +
sound/core/timer.c | 225 ++++++++++++++++++++
sound/drivers/aloop.c | 2 +
tools/testing/selftests/alsa/Makefile | 4 +-
tools/testing/selftests/alsa/global-timer.c | 87 ++++++++
tools/testing/selftests/alsa/utimer-test.c | 164 ++++++++++++++
9 files changed, 633 insertions(+), 3 deletions(-)
create mode 100644 Documentation/sound/utimers.rst
create mode 100644 tools/testing/selftests/alsa/global-timer.c
create mode 100644 tools/testing/selftests/alsa/utimer-test.c
--
2.34.1
Since commit 255c1c7279ab ("tc-testing: Allow test cases to be skipped")
the variable test_ordinal doesn't exist in call_pre_case().
So it should not be accessed when an exception occurs.
This resolves the following splat:
...
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File ".../tdc.py", line 1028, in <module>
main()
File ".../tdc.py", line 1022, in main
set_operation_mode(pm, parser, args, remaining)
File ".../tdc.py", line 966, in set_operation_mode
catresults = test_runner_serial(pm, args, alltests)
File ".../tdc.py", line 642, in test_runner_serial
(index, tsr) = test_runner(pm, args, alltests)
File ".../tdc.py", line 536, in test_runner
res = run_one_test(pm, args, index, tidx)
File ".../tdc.py", line 419, in run_one_test
pm.call_pre_case(tidx)
File ".../tdc.py", line 146, in call_pre_case
print('test_ordinal is {}'.format(test_ordinal))
NameError: name 'test_ordinal' is not defined
Fixes: 255c1c7279ab ("tc-testing: Allow test cases to be skipped")
Signed-off-by: Simon Horman <horms(a)kernel.org>
---
tools/testing/selftests/tc-testing/tdc.py | 1 -
1 file changed, 1 deletion(-)
diff --git a/tools/testing/selftests/tc-testing/tdc.py b/tools/testing/selftests/tc-testing/tdc.py
index ee349187636f..4f255cec0c22 100755
--- a/tools/testing/selftests/tc-testing/tdc.py
+++ b/tools/testing/selftests/tc-testing/tdc.py
@@ -143,7 +143,6 @@ class PluginMgr:
except Exception as ee:
print('exception {} in call to pre_case for {} plugin'.
format(ee, pgn_inst.__class__))
- print('test_ordinal is {}'.format(test_ordinal))
print('testid is {}'.format(caseinfo['id']))
raise
The kernel has recently added support for shadow stacks, currently
x86 only using their CET feature but both arm64 and RISC-V have
equivalent features (GCS and Zicfiss respectively), I am actively
working on GCS[1]. With shadow stacks the hardware maintains an
additional stack containing only the return addresses for branch
instructions which is not generally writeable by userspace and ensures
that any returns are to the recorded addresses. This provides some
protection against ROP attacks and making it easier to collect call
stacks. These shadow stacks are allocated in the address space of the
userspace process.
Our API for shadow stacks does not currently offer userspace any
flexiblity for managing the allocation of shadow stacks for newly
created threads, instead the kernel allocates a new shadow stack with
the same size as the normal stack whenever a thread is created with the
feature enabled. The stacks allocated in this way are freed by the
kernel when the thread exits or shadow stacks are disabled for the
thread. This lack of flexibility and control isn't ideal, in the vast
majority of cases the shadow stack will be over allocated and the
implicit allocation and deallocation is not consistent with other
interfaces. As far as I can tell the interface is done in this manner
mainly because the shadow stack patches were in development since before
clone3() was implemented.
Since clone3() is readily extensible let's add support for specifying a
shadow stack when creating a new thread or process in a similar manner
to how the normal stack is specified, keeping the current implicit
allocation behaviour if one is not specified either with clone3() or
through the use of clone(). The user must provide a shadow stack
address and size, this must point to memory mapped for use as a shadow
stackby map_shadow_stack() with a shadow stack token at the top of the
stack.
Please note that the x86 portions of this code are build tested only, I
don't appear to have a system that can run CET avaible to me, I have
done testing with an integration into my pending work for GCS. There is
some possibility that the arm64 implementation may require the use of
clone3() and explicit userspace allocation of shadow stacks, this is
still under discussion.
Please further note that the token consumption done by clone3() is not
currently implemented in an atomic fashion, Rick indicated that he would
look into fixing this if people are OK with the implementation.
A new architecture feature Kconfig option for shadow stacks is added as
here, this was suggested as part of the review comments for the arm64
GCS series and since we need to detect if shadow stacks are supported it
seemed sensible to roll it in here.
[1] https://lore.kernel.org/r/20231009-arm64-gcs-v6-0-78e55deaa4dd@kernel.org/
Signed-off-by: Mark Brown <broonie(a)kernel.org>
---
Changes in v8:
- Fix token verification with user specified shadow stack.
- Don't track user managed shadow stacks for child processes.
- Link to v7: https://lore.kernel.org/r/20240731-clone3-shadow-stack-v7-0-a9532eebfb1d@ke…
Changes in v7:
- Rebase onto v6.11-rc1.
- Typo fixes.
- Link to v6: https://lore.kernel.org/r/20240623-clone3-shadow-stack-v6-0-9ee7783b1fb9@ke…
Changes in v6:
- Rebase onto v6.10-rc3.
- Ensure we don't try to free the parent shadow stack in error paths of
x86 arch code.
- Spelling fixes in userspace API document.
- Additional cleanups and improvements to the clone3() tests to support
the shadow stack tests.
- Link to v5: https://lore.kernel.org/r/20240203-clone3-shadow-stack-v5-0-322c69598e4b@ke…
Changes in v5:
- Rebase onto v6.8-rc2.
- Rework ABI to have the user allocate the shadow stack memory with
map_shadow_stack() and a token.
- Force inlining of the x86 shadow stack enablement.
- Move shadow stack enablement out into a shared header for reuse by
other tests.
- Link to v4: https://lore.kernel.org/r/20231128-clone3-shadow-stack-v4-0-8b28ffe4f676@ke…
Changes in v4:
- Formatting changes.
- Use a define for minimum shadow stack size and move some basic
validation to fork.c.
- Link to v3: https://lore.kernel.org/r/20231120-clone3-shadow-stack-v3-0-a7b8ed3e2acc@ke…
Changes in v3:
- Rebase onto v6.7-rc2.
- Remove stale shadow_stack in internal kargs.
- If a shadow stack is specified unconditionally use it regardless of
CLONE_ parameters.
- Force enable shadow stacks in the selftest.
- Update changelogs for RISC-V feature rename.
- Link to v2: https://lore.kernel.org/r/20231114-clone3-shadow-stack-v2-0-b613f8681155@ke…
Changes in v2:
- Rebase onto v6.7-rc1.
- Remove ability to provide preallocated shadow stack, just specify the
desired size.
- Link to v1: https://lore.kernel.org/r/20231023-clone3-shadow-stack-v1-0-d867d0b5d4d0@ke…
---
Mark Brown (9):
Documentation: userspace-api: Add shadow stack API documentation
selftests: Provide helper header for shadow stack testing
mm: Introduce ARCH_HAS_USER_SHADOW_STACK
fork: Add shadow stack support to clone3()
selftests/clone3: Remove redundant flushes of output streams
selftests/clone3: Factor more of main loop into test_clone3()
selftests/clone3: Explicitly handle child exits due to signals
selftests/clone3: Allow tests to flag if -E2BIG is a valid error code
selftests/clone3: Test shadow stack support
Documentation/userspace-api/index.rst | 1 +
Documentation/userspace-api/shadow_stack.rst | 41 ++++
arch/x86/Kconfig | 1 +
arch/x86/include/asm/shstk.h | 11 +-
arch/x86/kernel/process.c | 2 +-
arch/x86/kernel/shstk.c | 105 +++++++---
fs/proc/task_mmu.c | 2 +-
include/linux/mm.h | 2 +-
include/linux/sched/task.h | 13 ++
include/uapi/linux/sched.h | 13 +-
kernel/fork.c | 76 ++++++--
mm/Kconfig | 6 +
tools/testing/selftests/clone3/clone3.c | 224 ++++++++++++++++++----
tools/testing/selftests/clone3/clone3_selftests.h | 40 +++-
tools/testing/selftests/ksft_shstk.h | 63 ++++++
15 files changed, 513 insertions(+), 87 deletions(-)
---
base-commit: 8400291e289ee6b2bf9779ff1c83a291501f017b
change-id: 20231019-clone3-shadow-stack-15d40d2bf536
Best regards,
--
Mark Brown <broonie(a)kernel.org>
Adds a selftest that creates two virtual interfaces, assigns one to a
new namespace, and assigns IP addresses to both.
It listens on the destination interface using socat and configures a
dynamic target on netconsole, pointing to the destination IP address.
The test then checks if the message was received properly on the
destination interface.
Signed-off-by: Breno Leitao <leitao(a)debian.org>
---
Changelog:
v4:
* Avoid sleeping in waiting for sockets and files (Matthieu Baerts)
* Some other improvements (Matthieu Baerts)
* Add configfs as a dependency (Jakub)
v3:
* Defined CONFIGs in config file (Jakub)
* Identention fixes (Petr Machata)
* Use setup_ns in a better way (Matthieu Baerts)
* Add dependencies in TEST_INCLUDES (Hangbin Liu)
* https://lore.kernel.org/all/20240815095157.3064722-1-leitao@debian.org/
v2:
* Change the location of the path (Jakub)
* Move from veth to netdevsim
* Other small changes in dependency checks and cleanup
* https://lore.kernel.org/all/20240813183825.837091-1-leitao@debian.org/
v1:
* https://lore.kernel.org/all/ZqyUHN770pjSofTC@gmail.com/
MAINTAINERS | 1 +
tools/testing/selftests/drivers/net/Makefile | 4 +-
tools/testing/selftests/drivers/net/config | 4 +
.../selftests/drivers/net/netcons_basic.sh | 249 ++++++++++++++++++
4 files changed, 257 insertions(+), 1 deletion(-)
create mode 100755 tools/testing/selftests/drivers/net/netcons_basic.sh
diff --git a/MAINTAINERS b/MAINTAINERS
index 5dbf23cf11c8..9a371ddd8719 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -15772,6 +15772,7 @@ M: Breno Leitao <leitao(a)debian.org>
S: Maintained
F: Documentation/networking/netconsole.rst
F: drivers/net/netconsole.c
+F: tools/testing/selftests/drivers/net/netcons_basic.sh
NETDEVSIM
M: Jakub Kicinski <kuba(a)kernel.org>
diff --git a/tools/testing/selftests/drivers/net/Makefile b/tools/testing/selftests/drivers/net/Makefile
index e54f382bcb02..8bee2e94049b 100644
--- a/tools/testing/selftests/drivers/net/Makefile
+++ b/tools/testing/selftests/drivers/net/Makefile
@@ -1,8 +1,10 @@
# SPDX-License-Identifier: GPL-2.0
-TEST_INCLUDES := $(wildcard lib/py/*.py)
+TEST_INCLUDES := $(wildcard lib/py/*.py) \
+ ../../net/lib.sh \
TEST_PROGS := \
+ netcons_basic.sh \
ping.py \
queues.py \
stats.py \
diff --git a/tools/testing/selftests/drivers/net/config b/tools/testing/selftests/drivers/net/config
index f6a58ce8a230..a2d8af60876d 100644
--- a/tools/testing/selftests/drivers/net/config
+++ b/tools/testing/selftests/drivers/net/config
@@ -1,2 +1,6 @@
CONFIG_IPV6=y
CONFIG_NETDEVSIM=m
+CONFIG_CONFIGFS_FS=y
+CONFIG_NETCONSOLE=m
+CONFIG_NETCONSOLE_DYNAMIC=y
+CONFIG_NETCONSOLE_EXTENDED_LOG=y
diff --git a/tools/testing/selftests/drivers/net/netcons_basic.sh b/tools/testing/selftests/drivers/net/netcons_basic.sh
new file mode 100755
index 000000000000..5c3686af1fe8
--- /dev/null
+++ b/tools/testing/selftests/drivers/net/netcons_basic.sh
@@ -0,0 +1,249 @@
+#!/usr/bin/env bash
+# SPDX-License-Identifier: GPL-2.0
+
+# This test creates two netdevsim virtual interfaces, assigns one of them (the
+# "destination interface") to a new namespace, and assigns IP addresses to both
+# interfaces.
+#
+# It listens on the destination interface using socat and configures a dynamic
+# target on netconsole, pointing to the destination IP address.
+#
+# Finally, it checks whether the message was received properly on the
+# destination interface. Note that this test may pollute the kernel log buffer
+# (dmesg) and relies on dynamic configuration and namespaces being configured.
+#
+# Author: Breno Leitao <leitao(a)debian.org>
+
+set -euo pipefail
+
+SCRIPTDIR=$(dirname "$(readlink -e "${BASH_SOURCE[0]}")")
+
+# Simple script to test dynamic targets in netconsole
+SRCIF="" # to be populated later
+SRCIP=192.168.1.1
+DSTIF="" # to be populated later
+DSTIP=192.168.1.2
+
+PORT="6666"
+MSG="netconsole selftest"
+TARGET=$(mktemp -u netcons_XXXXX)
+NETCONS_CONFIGFS="/sys/kernel/config/netconsole"
+NETCONS_PATH="${NETCONS_CONFIGFS}"/"${TARGET}"
+# NAMESPACE will be populated by setup_ns with a random value
+NAMESPACE=""
+
+# IDs for netdevsim
+NSIM_DEV_1_ID=$((256 + RANDOM % 256))
+NSIM_DEV_2_ID=$((512 + RANDOM % 256))
+
+# Used to create and delete namespaces
+source "${SCRIPTDIR}"/../../net/lib.sh
+source "${SCRIPTDIR}"/../../net/net_helper.sh
+
+# Create netdevsim interfaces
+create_ifaces() {
+ local NSIM_DEV_SYS_NEW=/sys/bus/netdevsim/new_device
+
+ echo "$NSIM_DEV_2_ID" > "$NSIM_DEV_SYS_NEW"
+ echo "$NSIM_DEV_1_ID" > "$NSIM_DEV_SYS_NEW"
+ udevadm settle 2> /dev/null || true
+
+ local NSIM1=/sys/bus/netdevsim/devices/netdevsim"$NSIM_DEV_1_ID"
+ local NSIM2=/sys/bus/netdevsim/devices/netdevsim"$NSIM_DEV_2_ID"
+
+ # These are global variables
+ SRCIF=$(find "$NSIM1"/net -maxdepth 1 -type d ! \
+ -path "$NSIM1"/net -exec basename {} \;)
+ DSTIF=$(find "$NSIM2"/net -maxdepth 1 -type d ! \
+ -path "$NSIM2"/net -exec basename {} \;)
+}
+
+link_ifaces() {
+ local NSIM_DEV_SYS_LINK="/sys/bus/netdevsim/link_device"
+ local SRCIF_IFIDX=$(cat /sys/class/net/"$SRCIF"/ifindex)
+ local DSTIF_IFIDX=$(cat /sys/class/net/"$DSTIF"/ifindex)
+
+ exec {NAMESPACE_FD}</var/run/netns/"${NAMESPACE}"
+ exec {INITNS_FD}</proc/self/ns/net
+
+ # Bind the dst interface to namespace
+ ip link set "${DSTIF}" netns "${NAMESPACE}"
+
+ # Linking one device to the other one (on the other namespace}
+ if ! echo "${INITNS_FD}:$SRCIF_IFIDX $NAMESPACE_FD:$DSTIF_IFIDX" > $NSIM_DEV_SYS_LINK
+ then
+ echo "linking netdevsim1 with netdevsim2 should succeed"
+ cleanup
+ exit "${ksft_skip}"
+ fi
+}
+
+function configure_ip() {
+ # Configure the IPs for both interfaces
+ ip netns exec "${NAMESPACE}" ip addr add "${DSTIP}"/24 dev "${DSTIF}"
+ ip netns exec "${NAMESPACE}" ip link set "${DSTIF}" up
+
+ ip addr add "${SRCIP}"/24 dev "${SRCIF}"
+ ip link set "${SRCIF}" up
+}
+
+function set_network() {
+ # setup_ns function is coming from lib.sh
+ setup_ns NAMESPACE
+
+ # Create both interfaces, and assign the destination to a different
+ # namespace
+ create_ifaces
+
+ # Link both interfaces back to back
+ link_ifaces
+
+ configure_ip
+}
+
+function create_dynamic_target() {
+ DSTMAC=$(ip netns exec "${NAMESPACE}" \
+ ip link show "${DSTIF}" | awk '/ether/ {print $2}')
+
+ # Create a dynamic target
+ mkdir "${NETCONS_PATH}"
+
+ echo "${DSTIP}" > "${NETCONS_PATH}"/remote_ip
+ echo "${SRCIP}" > "${NETCONS_PATH}"/local_ip
+ echo "${DSTMAC}" > "${NETCONS_PATH}"/remote_mac
+ echo "${SRCIF}" > "${NETCONS_PATH}"/dev_name
+
+ echo 1 > "${NETCONS_PATH}"/enabled
+}
+
+function cleanup() {
+ local NSIM_DEV_SYS_DEL="/sys/bus/netdevsim/del_device"
+
+ # delete netconsole dynamic reconfiguration
+ echo 0 > "${NETCONS_PATH}"/enabled
+ # Remove the configfs entry
+ rmdir "${NETCONS_PATH}"
+
+ # Delete netdevsim devices
+ echo "$NSIM_DEV_2_ID" > "$NSIM_DEV_SYS_DEL"
+ echo "$NSIM_DEV_1_ID" > "$NSIM_DEV_SYS_DEL"
+
+ # this is coming from lib.sh
+ cleanup_all_ns
+}
+
+function listen_port_and_save_to() {
+ local OUTPUT=${1}
+ # Just wait for 2 seconds
+ timeout 2 ip netns exec "${NAMESPACE}" \
+ socat UDP-LISTEN:"${PORT}",fork "${OUTPUT}"
+}
+
+function validate_result() {
+ local TMPFILENAME="$1"
+
+ # Check if the file exists
+ if [ ! -f "$TMPFILENAME" ]; then
+ echo "FAIL: File was not generated." >&2
+ exit "${ksft_fail}"
+ fi
+
+ if ! grep -q "${MSG}" "${TMPFILENAME}"; then
+ echo "FAIL: ${MSG} not found in ${TMPFILENAME}" >&2
+ cat "${TMPFILENAME}" >&2
+ exit "${ksft_fail}"
+ fi
+
+ # Delete the file once it is validated, otherwise keep it
+ # for debugging purposes
+ rm "${TMPFILENAME}"
+ exit "${ksft_pass}"
+}
+
+function check_for_dependencies() {
+ if [ "$(id -u)" -ne 0 ]; then
+ echo "This script must be run as root" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which socat > /dev/null ; then
+ echo "SKIP: socat(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which ip > /dev/null ; then
+ echo "SKIP: ip(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which udevadm > /dev/null ; then
+ echo "SKIP: udevadm(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if [ ! -d "${NETCONS_CONFIGFS}" ]; then
+ echo "SKIP: directory ${NETCONS_CONFIGFS} does not exist. Check if NETCONSOLE_DYNAMIC is enabled" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ip link show "${DSTIF}" 2> /dev/null; then
+ echo "SKIP: interface ${DSTIF} exists in the system. Not overwriting it." >&2
+ exit "${ksft_skip}"
+ fi
+}
+
+check_file_size() {
+ local file="$1"
+
+ if [[ ! -f "$file" ]]; then
+ # File might not exist yet
+ return 1
+ fi
+
+ # Get file size
+ local size=$(stat -c %s "$file" 2>/dev/null)
+ # Check if stat command succeeded
+ if [[ $? -ne 0 ]]; then
+ return 1
+ fi
+
+ # Check if size is greater than zero
+ if [[ "$size" -gt 0 ]]; then
+ return 0 # file size > 0
+ else
+ return 1 # file size == 0
+ fi
+}
+
+
+# ========== #
+# Start here #
+# ========== #
+modprobe netdevsim 2> /dev/null || true
+modprobe netconsole 2 > /dev/null || true
+
+# The content of kmsg will be save to the following file
+OUTPUT_FILE="/tmp/${TARGET}"
+
+# Check for basic system dependency and exit if not found
+check_for_dependencies
+# Set current loglevel to KERN_INFO(6), and default to KERN_NOTICE(5)
+echo "6 5" > /proc/sys/kernel/printk
+# Remove the namespace, interfaces and netconsole target on exit
+trap cleanup EXIT
+# Create one namespace and two interfaces
+set_network
+# Create a dynamic target for netconsole
+create_dynamic_target
+# Listed for netconsole port inside the namespace and destination interface
+listen_port_and_save_to "${OUTPUT_FILE}" &
+# Wait for socat to start and listen to the port.
+wait_local_port_listen "${NAMESPACE}" "${PORT}" udp
+# Send the message
+echo "${MSG}: ${TARGET}" > /dev/kmsg
+# Wait until socat saves the file to disk
+busywait "${BUSYWAIT_TIMEOUT}" check_file_size "${OUTPUT_FILE}"
+
+# Make sure the message was received in the dst part
+# and exit
+validate_result "${OUTPUT_FILE}"
--
2.43.5
Adds a selftest that creates two virtual interfaces, assigns one to a
new namespace, and assigns IP addresses to both.
It listens on the destination interface using socat and configures a
dynamic target on netconsole, pointing to the destination IP address.
The test then checks if the message was received properly on the
destination interface.
Signed-off-by: Breno Leitao <leitao(a)debian.org>
---
Changelog:
v3:
* Defined CONFIGs in config file (Jakub)
* Identention fixes (Petr Machata)
* Use setup_ns in a better way (Matthieu Baerts)
* Add dependencies in TEST_INCLUDES (Hangbin Liu)
v2:
* Change the location of the path (Jakub)
* Move from veth to netdevsim
* Other small changes in dependency checks and cleanup
* https://lore.kernel.org/all/20240813183825.837091-1-leitao@debian.org/
v1:
* https://lore.kernel.org/all/ZqyUHN770pjSofTC@gmail.com/
MAINTAINERS | 1 +
tools/testing/selftests/drivers/net/Makefile | 4 +-
tools/testing/selftests/drivers/net/config | 2 +
.../selftests/drivers/net/netcons_basic.sh | 226 ++++++++++++++++++
4 files changed, 232 insertions(+), 1 deletion(-)
create mode 100755 tools/testing/selftests/drivers/net/netcons_basic.sh
diff --git a/MAINTAINERS b/MAINTAINERS
index 7b291c3a9aa4..9f0be36939f9 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -15772,6 +15772,7 @@ M: Breno Leitao <leitao(a)debian.org>
S: Maintained
F: Documentation/networking/netconsole.rst
F: drivers/net/netconsole.c
+F: tools/testing/selftests/drivers/net/netcons_basic.sh
NETDEVSIM
M: Jakub Kicinski <kuba(a)kernel.org>
diff --git a/tools/testing/selftests/drivers/net/Makefile b/tools/testing/selftests/drivers/net/Makefile
index e54f382bcb02..8bee2e94049b 100644
--- a/tools/testing/selftests/drivers/net/Makefile
+++ b/tools/testing/selftests/drivers/net/Makefile
@@ -1,8 +1,10 @@
# SPDX-License-Identifier: GPL-2.0
-TEST_INCLUDES := $(wildcard lib/py/*.py)
+TEST_INCLUDES := $(wildcard lib/py/*.py) \
+ ../../net/lib.sh \
TEST_PROGS := \
+ netcons_basic.sh \
ping.py \
queues.py \
stats.py \
diff --git a/tools/testing/selftests/drivers/net/config b/tools/testing/selftests/drivers/net/config
index f6a58ce8a230..8f46112263f0 100644
--- a/tools/testing/selftests/drivers/net/config
+++ b/tools/testing/selftests/drivers/net/config
@@ -1,2 +1,4 @@
CONFIG_IPV6=y
CONFIG_NETDEVSIM=m
+CONFIG_NETCONSOLE=m
+CONFIG_NETCONSOLE_DYNAMIC=y
diff --git a/tools/testing/selftests/drivers/net/netcons_basic.sh b/tools/testing/selftests/drivers/net/netcons_basic.sh
new file mode 100755
index 000000000000..929f27a0fd9c
--- /dev/null
+++ b/tools/testing/selftests/drivers/net/netcons_basic.sh
@@ -0,0 +1,226 @@
+#!/usr/bin/env bash
+# SPDX-License-Identifier: GPL-2.0
+
+# This test creates two netdevsim virtual interfaces, assigns one of them (the
+# "destination interface") to a new namespace, and assigns IP addresses to both
+# interfaces.
+#
+# It listens on the destination interface using socat and configures a dynamic
+# target on netconsole, pointing to the destination IP address.
+#
+# Finally, it checks whether the message was received properly on the
+# destination interface. Note that this test may pollute the kernel log buffer
+# (dmesg) and relies on dynamic configuration and namespaces being configured.
+#
+# Author: Breno Leitao <leitao(a)debian.org>
+
+set -euo pipefail
+
+SCRIPTDIR=$(dirname "$(readlink -e "${BASH_SOURCE[0]}")")
+
+# Simple script to test dynamic targets in netconsole
+SRCIF="" # to be populated later
+SRCIP=192.168.1.1
+DSTIF="" # to be populated later
+DSTIP=192.168.1.2
+
+PORT="6666"
+MSG="netconsole selftest"
+TARGET=$(mktemp -u netcons_XXXXX)
+NETCONS_CONFIGFS="/sys/kernel/config/netconsole"
+NETCONS_PATH="${NETCONS_CONFIGFS}"/"${TARGET}"
+# This will have some tmp values appended to it in set_network()
+NAMESPACE="netconsns_dst"
+
+# IDs for netdevsim
+NSIM_DEV_1_ID=$((256 + RANDOM % 256))
+NSIM_DEV_2_ID=$((512 + RANDOM % 256))
+
+# Used to create and delete namespaces
+source "${SCRIPTDIR}"/../../net/lib.sh
+
+# Create netdevsim interfaces
+create_ifaces() {
+ local NSIM_DEV_SYS_NEW=/sys/bus/netdevsim/new_device
+
+ echo "$NSIM_DEV_2_ID" > "$NSIM_DEV_SYS_NEW"
+ echo "$NSIM_DEV_1_ID" > "$NSIM_DEV_SYS_NEW"
+ udevadm settle || true
+
+ local NSIM1=/sys/bus/netdevsim/devices/netdevsim"$NSIM_DEV_1_ID"
+ local NSIM2=/sys/bus/netdevsim/devices/netdevsim"$NSIM_DEV_2_ID"
+
+ # These are global variables
+ SRCIF=$(find "$NSIM1"/net -maxdepth 1 -type d ! \
+ -path "$NSIM1"/net -exec basename {} \;)
+ DSTIF=$(find "$NSIM2"/net -maxdepth 1 -type d ! \
+ -path "$NSIM2"/net -exec basename {} \;)
+}
+
+link_ifaces() {
+ local NSIM_DEV_SYS_LINK="/sys/bus/netdevsim/link_device"
+ local SRCIF_IFIDX=$(cat /sys/class/net/"$SRCIF"/ifindex)
+ local DSTIF_IFIDX=$(cat /sys/class/net/"$DSTIF"/ifindex)
+
+ exec {NAMESPACE_FD}</var/run/netns/"${NAMESPACE}"
+ exec {INITNS_FD}</proc/self/ns/net
+
+ # Bind the dst interface to namespace
+ ip link set "${DSTIF}" netns "${NAMESPACE}"
+
+ # Linking one device to the other one (on the other namespace}
+ echo "${INITNS_FD}:$SRCIF_IFIDX $NAMESPACE_FD:$DSTIF_IFIDX" \
+ > $NSIM_DEV_SYS_LINK
+ if [ $? -ne 0 ]; then
+ echo "linking netdevsim1 with netdevsim2 should succeed"
+ cleanup
+ exit ${ksft_skip}
+ fi
+}
+
+function configure_ip() {
+ # Configure the IPs for both interfaces
+ ip netns exec "${NAMESPACE}" ip addr add "${DSTIP}"/24 dev "${DSTIF}"
+ ip netns exec "${NAMESPACE}" ip link set "${DSTIF}" up
+
+ ip addr add "${SRCIP}"/24 dev "${SRCIF}"
+ ip link set "${SRCIF}" up
+}
+
+function set_network() {
+ # This is coming from lib.sh
+ setup_ns NAMESPACE
+
+ # Create both interfaces, and assign the destination to a different
+ # namespace
+ create_ifaces
+
+ # Link both interfaces back to back
+ link_ifaces
+
+ configure_ip
+}
+
+function create_dynamic_target() {
+ DSTMAC=$(ip netns exec "${NAMESPACE}" \
+ ip link show "${DSTIF}" | awk '/ether/ {print $2}')
+
+ # Create a dynamic target
+ mkdir "${NETCONS_PATH}"
+
+ echo "${DSTIP}" > "${NETCONS_PATH}"/remote_ip
+ echo "${SRCIP}" > "${NETCONS_PATH}"/local_ip
+ echo "${DSTMAC}" > "${NETCONS_PATH}"/remote_mac
+ echo "${SRCIF}" > "${NETCONS_PATH}"/dev_name
+
+ echo 1 > "${NETCONS_PATH}"/enabled
+}
+
+function cleanup() {
+ local NSIM_DEV_SYS_DEL="/sys/bus/netdevsim/del_device"
+
+ # delete netconsole dynamic reconfiguration
+ echo 0 > "${NETCONS_PATH}"/enabled
+ # Remove the configfs entry
+ rmdir "${NETCONS_PATH}"
+
+ # Delete netdevsim devices
+ echo "$NSIM_DEV_2_ID" > "$NSIM_DEV_SYS_DEL"
+ echo "$NSIM_DEV_1_ID" > "$NSIM_DEV_SYS_DEL"
+
+ # this is coming from lib.sh
+ cleanup_all_ns
+}
+
+function listen_port_and_save_to() {
+ local OUTPUT=${1}
+ # Just wait for 2 seconds
+ timeout 2 ip netns exec "${NAMESPACE}" \
+ socat UDP-LISTEN:"${PORT}",fork "${OUTPUT}"
+}
+
+function validate_result() {
+ local TMPFILENAME="$1"
+
+ # Check if the file exists
+ if [ ! -f "$TMPFILENAME" ]; then
+ echo "FAIL: File was not generated." >&2
+ return ${ksft_fail}
+ fi
+
+ if ! grep -q "${MSG}" "${TMPFILENAME}"; then
+ echo "FAIL: ${MSG} not found in ${TMPFILENAME}" >&2
+ cat "${TMPFILENAME}" >&2
+ return ${ksft_fail}
+ fi
+
+ # Delete the file once it is validated, otherwise keep it
+ # for debugging purposes
+ rm "${TMPFILENAME}"
+ return ${ksft_pass}
+}
+
+function check_for_dependencies() {
+ if [ "$(id -u)" -ne 0 ]; then
+ echo "This script must be run as root" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which socat > /dev/null ; then
+ echo "SKIP: socat(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which ip > /dev/null ; then
+ echo "SKIP: ip(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which udevadm > /dev/null ; then
+ echo "SKIP: udevadm(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if [ ! -d "${NETCONS_CONFIGFS}" ]; then
+ echo "SKIP: directory ${NETCONS_CONFIGFS} does not exist. Check if NETCONSOLE_DYNAMIC is enabled" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ip link show "${DSTIF}" 2> /dev/null; then
+ echo "SKIP: interface ${DSTIF} exists in the system. Not overwriting it." >&2
+ exit "${ksft_skip}"
+ fi
+}
+
+# ========== #
+# Start here #
+# ========== #
+modprobe netdevsim || true
+modprobe netconsole || true
+
+# The content of kmsg will be save to the following file
+OUTPUT_FILE="/tmp/${TARGET}"
+
+# Check for basic system dependency and exit if not found
+check_for_dependencies
+# Remove the namespace, interfaces and netconsole target on exit
+trap cleanup EXIT
+# Create one namespace and two interfaces
+set_network
+# Create a dynamic target for netconsole
+create_dynamic_target
+# Listed for netconsole port inside the namespace and destination interface
+listen_port_and_save_to "${OUTPUT_FILE}" &
+
+# Wait for socat to start and listen to the port.
+sleep 1
+# Send the message
+echo "${MSG}: ${TARGET}" > /dev/kmsg
+# Wait until socat saves the file to disk
+sleep 1
+
+# Make sure the message was received in the dst part
+validate_result "${OUTPUT_FILE}"
+ret=$?
+
+exit ${ret}
--
2.43.5
This is a collection of patches I've gathered over the past several
months.
Patches 1-6/14 are supporting patches for selftests.
Patch 9/14 fixes PTP TX from a VLAN upper of a VLAN-aware bridge port
when using the "ocelot-8021q" tagging protocol. Patch 7/14 is its
supporting selftest.
Patch 10/14 fixes the QoS class used by PTP in the same case as above.
It is hard to quantify - there is no selftest.
Patch 11/14 fixes potential data corruption during PTP TX in the same
case as above. Again, there is no selftest.
Patch 13/14 fixes RX in the same case as above - 8021q upper of a
VLAN-aware bridge port, with the "ocelot-8021q" tagging protocol. Patch
12/14 is a supporting patch for this in the DSA core, and 7/14 is also
its selftest.
Patch 14/14 ensures that VLAN-aware bridges offloaded to Ocelot only
react to the ETH_P_8021Q TPID, and treat absolutely everything else as
VLAN-untagged, including ETH_P_8021AD. Patch 8/14 is the supporting
selftest.
Vladimir Oltean (14):
selftests: net: local_termination: refactor macvlan creation/deletion
selftests: net: local_termination: parameterize sending interface
selftests: net: local_termination: parameterize test name
selftests: net: local_termination: add one more test for VLAN-aware
bridges
selftests: net: local_termination: introduce new tests which capture
VLAN behavior
selftests: net: local_termination: don't use xfail_on_veth()
selftests: net: local_termination: add PTP frames to the mix
selftests: net: bridge_vlan_aware: test that other TPIDs are seen as
untagged
net: mscc: ocelot: use ocelot_xmit_get_vlan_info() also for FDMA and
register injection
net: mscc: ocelot: fix QoS class for injected packets with
"ocelot-8021q"
net: mscc: ocelot: serialize access to the injection/extraction groups
net: dsa: provide a software untagging function on RX for VLAN-aware
bridges
net: dsa: felix: fix VLAN tag loss on CPU reception with ocelot-8021q
net: mscc: ocelot: treat 802.1ad tagged traffic as 802.1Q-untagged
drivers/net/dsa/ocelot/felix.c | 126 ++++-
drivers/net/ethernet/mscc/ocelot.c | 279 +++++++++++-
drivers/net/ethernet/mscc/ocelot_fdma.c | 3 +-
drivers/net/ethernet/mscc/ocelot_vcap.c | 1 +
drivers/net/ethernet/mscc/ocelot_vsc7514.c | 4 +
include/linux/dsa/ocelot.h | 47 ++
include/net/dsa.h | 16 +-
include/soc/mscc/ocelot.h | 12 +-
include/soc/mscc/ocelot_vcap.h | 2 +
net/dsa/tag.c | 5 +-
net/dsa/tag.h | 135 ++++--
net/dsa/tag_ocelot.c | 37 +-
.../net/forwarding/bridge_vlan_aware.sh | 54 ++-
tools/testing/selftests/net/forwarding/lib.sh | 57 +++
.../net/forwarding/local_termination.sh | 431 +++++++++++++++---
15 files changed, 1036 insertions(+), 173 deletions(-)
--
2.34.1
This small series includes fixes for creation of veth pairs for
networkless kernels & adds tests for turning the different network
interface features on and off in selftests/net/netdevice.sh script.
Changes in v7:
Create a third patch in the series to do SKIP -> XFAIL replacement.
Add logic to incorporate XFAIL on setting IP address for veth pair.
Changes in v6:
https://lore.kernel.org/all/20240814191517.50466-1-jain.abhinav177@gmail.com
Use XFAIL for ethtool operations that are unsupported instead of SKIP.
Changes in v5:
https://lore.kernel.org/all/20240808122452.25683-1-jain.abhinav177@gmail.com
Rectify the syntax for ip add link.
Fix the veth_created condition check.
Changes in v4:
https://lore.kernel.org/all/20240807175717.7775-1-jain.abhinav177@gmail.com
Move veth creation/removal to the main shell script.
Tested using vng on a networkless kernel and the script works, sample
output below the changes.
Changes in v3:
https://lore.kernel.org/all/20240614113240.41550-1-jain.abhinav177@gmail.com
Add a check for netdev, create veth pair for testing.
Restore feature to its initial state.
Changes in v2:
https://lore.kernel.org/all/20240609132124.51683-1-jain.abhinav177@gmail.com
Remove tail usage; use read to parse the features from temp file.
v1:
https://lore.kernel.org/all/20240606212714.27472-1-jain.abhinav177@gmail.com
```
# selftests: net: netdevice.sh
# No valid network device found, creating veth pair
# PASS: veth0: set interface up
# PASS: veth0: set MAC address
# XFAIL: veth0: set IP address
# PASS: veth0: ethtool list features
# PASS: veth0: Turned off feature: rx-checksumming
# PASS: veth0: Turned on feature: rx-checksumming
# PASS: veth0: Restore feature rx-checksumming to initial state on
# Actual changes:
# tx-checksum-ip-generic: off
# tx-tcp-segmentation: off [not requested]
# PASS: veth0: Turned off feature: rx-udp-gro-forwarding
# PASS: veth0: Turned on feature: rx-udp-gro-forwarding
# PASS: veth0: Restore feature rx-udp-gro-forwarding to initial state off
# Cannot get register dump: Operation not supported
# XFAIL: veth0: ethtool dump not supported
# PASS: veth0: ethtool stats
# PASS: veth0: stop interface
```
Abhinav Jain (3):
selftests: net: Create veth pair for testing in networkless kernel
selftests: net: Add on/off checks for non-fixed features of interface
selftests: net: Use XFAIL for operations not supported by the driver
tools/testing/selftests/net/netdevice.sh | 61 ++++++++++++++++++++++--
1 file changed, 58 insertions(+), 3 deletions(-)
--
2.34.1
This small series includes fixes for creation of veth pairs for
networkless kernels & adds tests for turning the different network
interface features on and off in selftests/net/netdevice.sh script.
Changes in v6:
Use XFAIL for ethtool operations that are unsupported instead of SKIP.
Changes in v5:
https://lore.kernel.org/all/20240808122452.25683-1-jain.abhinav177@gmail.co…
Rectify the syntax for ip add link.
Fix the veth_created condition check.
Changes in v4:
https://lore.kernel.org/all/20240807175717.7775-1-jain.abhinav177@gmail.com/
Move veth creation/removal to the main shell script.
Tested using vng on a networkless kernel and the script works, sample
output below the changes.
Changes in v3:
https://lore.kernel.org/all/20240614113240.41550-1-jain.abhinav177@gmail.co…
Add a check for netdev, create veth pair for testing.
Restore feature to its initial state.
Changes in v2:
https://lore.kernel.org/all/20240609132124.51683-1-jain.abhinav177@gmail.co…
Remove tail usage; use read to parse the features from temp file.
v1:
https://lore.kernel.org/all/20240606212714.27472-1-jain.abhinav177@gmail.co…
```
# selftests: net: netdevice.sh
# No valid network device found, creating veth pair
# PASS: veth0: set interface up
# PASS: veth0: set MAC address
# SKIP: veth0: set IP address
# PASS: veth0: ethtool list features
# PASS: veth0: Turned off feature: rx-checksumming
# PASS: veth0: Turned on feature: rx-checksumming
# PASS: veth0: Restore feature rx-checksumming to initial state on
# Actual changes:
# PASS: veth0: Restore feature rx-gro-list to initial state off
# PASS: veth0: Turned off feature: rx-udp-gro-forwarding
# PASS: veth0: Turned on feature: rx-udp-gro-forwarding
# PASS: veth0: Restore feature rx-udp-gro-forwarding to initial state off
# Cannot get register dump: Operation not supported
# XFAIL: veth0: ethtool dump not supported
# PASS: veth0: ethtool stats
# PASS: veth0: stop interface
```
Abhinav Jain (2):
selftests: net: Create veth pair for testing in networkless kernel
selftests: net: Add on/off checks for non-fixed features of interface
tools/testing/selftests/net/netdevice.sh | 55 +++++++++++++++++++++++-
1 file changed, 53 insertions(+), 2 deletions(-)
--
2.34.1
When deleting netns, it is possible to still have some tasks running,
e.g. background tasks like tcpdump running in the background, not
stopped because the test has been interrupted.
Before deleting the netns, it is then safer to kill all attached PIDs,
if any. That should reduce some noises after the end of some tests, and
help with the debugging of some issues. That's why this modification is
seen as a "fix".
Fixes: 25ae948b4478 ("selftests/net: add lib.sh")
Acked-by: Mat Martineau <martineau(a)kernel.org>
Signed-off-by: Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
---
tools/testing/selftests/net/lib.sh | 1 +
1 file changed, 1 insertion(+)
diff --git a/tools/testing/selftests/net/lib.sh b/tools/testing/selftests/net/lib.sh
index d0219032f773..8ee4489238ca 100644
--- a/tools/testing/selftests/net/lib.sh
+++ b/tools/testing/selftests/net/lib.sh
@@ -146,6 +146,7 @@ cleanup_ns()
for ns in "$@"; do
[ -z "${ns}" ] && continue
+ ip netns pids "${ns}" 2> /dev/null | xargs -r kill || true
ip netns delete "${ns}" &> /dev/null || true
if ! busywait $BUSYWAIT_TIMEOUT ip netns list \| grep -vq "^$ns$" &> /dev/null; then
echo "Warn: Failed to remove namespace $ns"
---
base-commit: 58a63729c957621f1990c3494c702711188ca347
change-id: 20240813-upstream-net-20240813-selftests-net-lib-kill-f7964a3a58fe
Best regards,
--
Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
Hello,
this series brings a new set of test converted to the test_progs framework.
Since the tests are quite small, I chose to group three tests conversion in
the same series, but feel free to let me know if I should keep one series
per test. The series focuses on cgroup testing and converts the following
tests:
- get_cgroup_id_user
- cgroup_storage
- test_skb_cgroup_id_user
Signed-off-by: Alexis Lothoré (eBPF Foundation) <alexis.lothore(a)bootlin.com>
---
Changes in v4:
- Fix test after netns addition by making sure loopack interface is up
- Link to v3: https://lore.kernel.org/r/20240812-convert_cgroup_tests-v3-0-47ac6ce4e88b@b…
Changes in v3:
- Fixed multiple leaks on cgroup file descriptors and sockets
- Used dedicated network namespaces for tests involving network
- Link to v2: https://lore.kernel.org/r/20240806-convert_cgroup_tests-v2-0-180c57e5b710@b…
Changes in v2:
- Use global variables instead of maps when possible
- Collect review tags from Alan
- Link to v1: https://lore.kernel.org/r/20240731-convert_cgroup_tests-v1-0-14cbc51b6947@b…
---
Alexis Lothoré (eBPF Foundation) (4):
selftests/bpf: convert get_current_cgroup_id_user to test_progs
selftests/bpf: convert test_cgroup_storage to test_progs
selftests/bpf: add proper section name to bpf prog and rename it
selftests/bpf: convert test_skb_cgroup_id_user to test_progs
tools/testing/selftests/bpf/.gitignore | 3 -
tools/testing/selftests/bpf/Makefile | 8 +-
tools/testing/selftests/bpf/get_cgroup_id_user.c | 151 -----------------
.../selftests/bpf/prog_tests/cgroup_ancestor.c | 169 +++++++++++++++++++
.../bpf/prog_tests/cgroup_get_current_cgroup_id.c | 46 ++++++
.../selftests/bpf/prog_tests/cgroup_storage.c | 96 +++++++++++
...test_skb_cgroup_id_kern.c => cgroup_ancestor.c} | 14 +-
tools/testing/selftests/bpf/progs/cgroup_storage.c | 24 +++
.../selftests/bpf/progs/get_cgroup_id_kern.c | 26 +--
tools/testing/selftests/bpf/test_cgroup_storage.c | 174 --------------------
tools/testing/selftests/bpf/test_skb_cgroup_id.sh | 63 -------
.../selftests/bpf/test_skb_cgroup_id_user.c | 183 ---------------------
12 files changed, 344 insertions(+), 613 deletions(-)
---
base-commit: ab2c4aa104050a184c3411a973b165285549f732
change-id: 20240725-convert_cgroup_tests-d07c66053225
Best regards,
--
Alexis Lothoré, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
Adds a selftest that creates two virtual interfaces, assigns one to a
new namespace, and assigns IP addresses to both.
It listens on the destination interface using socat and configures a
dynamic target on netconsole, pointing to the destination IP address.
The test then checks if the message was received properly on the
destination interface.
Signed-off-by: Breno Leitao <leitao(a)debian.org>
---
Changelog:
v2:
* Change the location of the path (Jakub)
* Move from veth to netdevsim
* Other small changes in dependency checks and cleanup
v1:
* https://lore.kernel.org/all/ZqyUHN770pjSofTC@gmail.com/
MAINTAINERS | 1 +
tools/testing/selftests/drivers/net/Makefile | 1 +
.../selftests/drivers/net/netcons_basic.sh | 223 ++++++++++++++++++
3 files changed, 225 insertions(+)
create mode 100755 tools/testing/selftests/drivers/net/netcons_basic.sh
diff --git a/MAINTAINERS b/MAINTAINERS
index a9dace908305..ded45f1dff7e 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -15770,6 +15770,7 @@ M: Breno Leitao <leitao(a)debian.org>
S: Maintained
F: Documentation/networking/netconsole.rst
F: drivers/net/netconsole.c
+F: tools/testing/selftests/drivers/net/netcons_basic.sh
NETDEVSIM
M: Jakub Kicinski <kuba(a)kernel.org>
diff --git a/tools/testing/selftests/drivers/net/Makefile b/tools/testing/selftests/drivers/net/Makefile
index e54f382bcb02..928530b26abc 100644
--- a/tools/testing/selftests/drivers/net/Makefile
+++ b/tools/testing/selftests/drivers/net/Makefile
@@ -3,6 +3,7 @@
TEST_INCLUDES := $(wildcard lib/py/*.py)
TEST_PROGS := \
+ netcons_basic.sh \
ping.py \
queues.py \
stats.py \
diff --git a/tools/testing/selftests/drivers/net/netcons_basic.sh b/tools/testing/selftests/drivers/net/netcons_basic.sh
new file mode 100755
index 000000000000..e0e58fc7e89f
--- /dev/null
+++ b/tools/testing/selftests/drivers/net/netcons_basic.sh
@@ -0,0 +1,223 @@
+#!/usr/bin/env bash
+# SPDX-License-Identifier: GPL-2.0
+
+# This test creates two netdevsim virtual interfaces, assigns one of them (the
+# "destination interface") to a new namespace, and assigns IP addresses to both
+# interfaces.
+#
+# It listens on the destination interface using socat and configures a dynamic
+# target on netconsole, pointing to the destination IP address.
+#
+# Finally, it checks whether the message was received properly on the
+# destination interface. Note that this test may pollute the kernel log buffer
+# (dmesg) and relies on dynamic configuration and namespaces being configured.
+#
+# Author: Breno Leitao <leitao(a)debian.org>
+
+set -euo pipefail
+
+SCRIPTDIR=$(dirname "$(readlink -e "${BASH_SOURCE[0]}")")
+
+# Simple script to test dynamic targets in netconsole
+SRCIF="" # to be populated later
+SRCIP=192.168.1.1
+DSTIF="" # to be populated later
+DSTIP=192.168.1.2
+
+PORT="6666"
+MSG="netconsole selftest"
+TARGET=$(mktemp -u netcons_XXXXX)
+NETCONS_CONFIGFS="/sys/kernel/config/netconsole"
+NETCONS_PATH="${NETCONS_CONFIGFS}"/"${TARGET}"
+# This will have some tmp values appended to it in set_network()
+NAMESPACE="netconsns_dst"
+
+# IDs for netdevsim
+NSIM_DEV_1_ID=$((256 + RANDOM % 256))
+NSIM_DEV_2_ID=$((512 + RANDOM % 256))
+
+# Used to create and delete namespaces
+source "${SCRIPTDIR}"/../../net/lib.sh
+
+# Create netdevsim interfaces
+create_ifaces() {
+ local NSIM_DEV_SYS_NEW=/sys/bus/netdevsim/new_device
+
+ echo "$NSIM_DEV_2_ID" > "$NSIM_DEV_SYS_NEW"
+ echo "$NSIM_DEV_1_ID" > "$NSIM_DEV_SYS_NEW"
+ udevadm settle || true
+
+ local NSIM_DEV_1_SYS=/sys/bus/netdevsim/devices/netdevsim"$NSIM_DEV_1_ID"
+ local NSIM_DEV_2_SYS=/sys/bus/netdevsim/devices/netdevsim"$NSIM_DEV_2_ID"
+
+ # These are global variables
+ SRCIF=$(find "$NSIM_DEV_1_SYS"/net -maxdepth 1 -type d ! \
+ -path "$NSIM_DEV_1_SYS"/net -exec basename {} \;)
+ DSTIF=$(find "$NSIM_DEV_2_SYS"/net -maxdepth 1 -type d ! \
+ -path "$NSIM_DEV_2_SYS"/net -exec basename {} \;)
+}
+
+link_ifaces() {
+ local NSIM_DEV_SYS_LINK="/sys/bus/netdevsim/link_device"
+ local SRCIF_IFIDX=$(cat /sys/class/net/"$SRCIF"/ifindex)
+ local DSTIF_IFIDX=$(cat /sys/class/net/"$DSTIF"/ifindex)
+
+ exec {NAMESPACE_FD}</var/run/netns/"${NAMESPACE}"
+ exec {INITNS_FD}</proc/self/ns/net
+
+ # Bind the dst interface to namespace
+ ip link set "${DSTIF}" netns "${NAMESPACE}"
+
+ # Linking one device to the other one (on the other namespace}
+ echo "${INITNS_FD}:$SRCIF_IFIDX $NAMESPACE_FD:$DSTIF_IFIDX" > $NSIM_DEV_SYS_LINK
+ if [ $? -ne 0 ]; then
+ echo "linking netdevsim1 with netdevsim2 should succeed"
+ cleanup
+ exit ${ksft_skip}
+ fi
+}
+
+function configure_ip() {
+ # Configure the IPs for both interfaces
+ ip netns exec "${NAMESPACE}" ip addr add "${DSTIP}"/24 dev "${DSTIF}"
+ ip netns exec "${NAMESPACE}" ip link set "${DSTIF}" up
+
+ ip addr add "${SRCIP}"/24 dev "${SRCIF}"
+ ip link set "${SRCIF}" up
+}
+
+function set_network() {
+ # This is coming from lib.sh. And it does unbound variable access
+ set +u
+ setup_ns "${NAMESPACE}"
+ set -u
+ NAMESPACE=${NS_LIST[0]}
+
+ # Create both interfaces, and assign the destination to a different namespace
+ create_ifaces
+
+ # Link both interfaces back to back
+ link_ifaces
+
+ configure_ip
+}
+
+function create_dynamic_target() {
+ DSTMAC=$(ip netns exec "${NAMESPACE}" ip link show "${DSTIF}" | awk '/ether/ {print $2}')
+
+ # Create a dynamic target
+ mkdir "${NETCONS_PATH}"
+
+ echo "${DSTIP}" > "${NETCONS_PATH}"/remote_ip
+ echo "${SRCIP}" > "${NETCONS_PATH}"/local_ip
+ echo "${DSTMAC}" > "${NETCONS_PATH}"/remote_mac
+ echo "${SRCIF}" > "${NETCONS_PATH}"/dev_name
+
+ echo 1 > "${NETCONS_PATH}"/enabled
+}
+
+function cleanup() {
+ local NSIM_DEV_SYS_DEL="/sys/bus/netdevsim/del_device"
+
+ # delete netconsole dynamic reconfiguration
+ echo 0 > "${NETCONS_PATH}"/enabled
+ # Remove the configfs entry
+ rmdir "${NETCONS_PATH}"
+
+ # Delete netdevsim devices
+ echo "$NSIM_DEV_2_ID" > "$NSIM_DEV_SYS_DEL"
+ echo "$NSIM_DEV_1_ID" > "$NSIM_DEV_SYS_DEL"
+
+ # this is coming from lib.sh
+ cleanup_all_ns
+}
+
+function listen_port_and_save_to() {
+ OUTPUT=${1}
+ # Just wait for 2 seconds
+ timeout 2 ip netns exec "${NAMESPACE}" socat UDP-LISTEN:"${PORT}",fork "${OUTPUT}"
+}
+
+function validate_result() {
+ TMPFILENAME="$1"
+
+ # Check if the file exists
+ if [ ! -f "$TMPFILENAME" ]; then
+ echo "FAIL: File was not generated." >&2
+ return ${ksft_fail}
+ fi
+
+ if ! grep -q "${MSG}" "${TMPFILENAME}"; then
+ echo "FAIL: ${MSG} not found in ${TMPFILENAME}" >&2
+ cat "${TMPFILENAME}" >&2
+ return ${ksft_fail}
+ fi
+
+ # Delete the file once it is validated, otherwise keep it
+ # for debugging purposes
+ rm "${TMPFILENAME}"
+ return ${ksft_pass}
+}
+
+function check_for_dependencies() {
+ if [ "$(id -u)" -ne 0 ]; then
+ echo "This script must be run as root" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which socat > /dev/null ; then
+ echo "SKIP: socat(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which ip > /dev/null ; then
+ echo "SKIP: ip(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ! which udevadm > /dev/null ; then
+ echo "SKIP: udevadm(1) is not available" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if [ ! -d "${NETCONS_CONFIGFS}" ]; then
+ echo "SKIP: directory ${NETCONS_CONFIGFS} does not exist. Check if NETCONSOLE_DYNAMIC is enabled" >&2
+ exit "${ksft_skip}"
+ fi
+
+ if ip link show "${DSTIF}" 2> /dev/null; then
+ echo "SKIP: interface ${DSTIF} exists in the system. Not overwriting it."
+ exit "${ksft_skip}"
+ fi
+}
+
+# ========== #
+# Start here #
+# ========== #
+modprobe netdevsim || true
+# The content of kmsg will be save to the following file
+OUTPUT_FILE="/tmp/${TARGET}"
+
+# Check for basic system dependency and exit if not found
+check_for_dependencies
+# Remove the namespace, interfaces and netconsole target on exit
+trap cleanup EXIT
+# Create one namespace and two interfaces
+set_network
+# Create a dynamic target for netconsole
+create_dynamic_target
+# Listed for netconsole port inside the namespace and destination interface
+listen_port_and_save_to "${OUTPUT_FILE}" &
+
+# Wait for socat to start and listen to the port.
+sleep 1
+# Send the message
+echo "${MSG}: ${TARGET}" > /dev/kmsg
+# Wait until socat saves the file to disk
+sleep 1
+
+# Make sure the message was received in the dst part
+validate_result "${OUTPUT_FILE}"
+ret=$?
+
+exit ${ret}
--
2.43.5
Change expected_buf from (const void *) to (const char *)
in function __recvpair().
This change fixes the below warnings during test compilation:
```
In file included from msg_oob.c:14:
msg_oob.c: In function ‘__recvpair’:
../../kselftest_harness.h:106:40: warning: format ‘%s’ expects argument
of type ‘char *’,but argument 6 has type ‘const void *’ [-Wformat=]
../../kselftest_harness.h:101:17: note: in expansion of macro ‘__TH_LOG’
msg_oob.c:235:17: note: in expansion of macro ‘TH_LOG’
../../kselftest_harness.h:106:40: warning: format ‘%s’ expects argument
of type ‘char *’,but argument 6 has type ‘const void *’ [-Wformat=]
../../kselftest_harness.h:101:17: note: in expansion of macro ‘__TH_LOG’
msg_oob.c:259:25: note: in expansion of macro ‘TH_LOG’
```
Fixes: d098d77232c3 ("selftest: af_unix: Add msg_oob.c.")
Signed-off-by: Abhinav Jain <jain.abhinav177(a)gmail.com>
Reviewed-by: Kuniyuki Iwashima <kuniyu(a)amazon.com>
---
v2:
https://lore.kernel.org/all/20240812191122.1092806-1-jain.abhinav177@gmail.…
- Change the parameter expected_buf from (const void *) to (const char *)
in the function __recvpair() as per the feedback in v1.
- Add Fixes tag as per feedback in v1.
v1:
https://lore.kernel.org/netdev/20240810134037.669765-1-jain.abhinav177@gmai…
---
tools/testing/selftests/net/af_unix/msg_oob.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/testing/selftests/net/af_unix/msg_oob.c b/tools/testing/selftests/net/af_unix/msg_oob.c
index 16d0c172eaeb..535eb2c3d7d1 100644
--- a/tools/testing/selftests/net/af_unix/msg_oob.c
+++ b/tools/testing/selftests/net/af_unix/msg_oob.c
@@ -209,7 +209,7 @@ static void __sendpair(struct __test_metadata *_metadata,
static void __recvpair(struct __test_metadata *_metadata,
FIXTURE_DATA(msg_oob) *self,
- const void *expected_buf, int expected_len,
+ const char *expected_buf, int expected_len,
int buf_len, int flags)
{
int i, ret[2], recv_errno[2], expected_errno = 0;
--
2.34.1
On Thu, 2024-07-18 at 10:34 -0700, Fangrui Song wrote:
> glibc added support for .gnu.hash in 2006 and .hash has been obsoleted
> for more than one decade in many Linux distributions. Using
> --hash-style=sysv might imply unaddressed issues and confuse readers.
>
> Just drop the option and rely on the linker default, which is likely
> "both", or "gnu" when the distribution really wants to eliminate sysv
> hash overhead.
>
> Similar to commit 6b7e26547fad ("x86/vdso: Emit a GNU hash").
>
> Signed-off-by: Fangrui Song <maskray(a)google.com>
Hi Fangrui,
If I read tools/testing/selftests/vDSO/parse_vdso.c correctly, it does
know DT_GNU_HASH as at now. Thus after this change the vDSO selftests
are skipped with "Couldn't find __vdso_gettimeofday" etc if the distro
enables --hash-style=gnu by default.
So it seems we need to add DT_GNU_HASH support for parse_vdso.c to keep
test coverage.
> ---
> arch/arm64/kernel/vdso/Makefile | 2 +-
> arch/arm64/kernel/vdso32/Makefile | 2 +-
> 2 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/arch/arm64/kernel/vdso/Makefile b/arch/arm64/kernel/vdso/Makefile
> index d63930c82839..d11da6461278 100644
> --- a/arch/arm64/kernel/vdso/Makefile
> +++ b/arch/arm64/kernel/vdso/Makefile
> @@ -21,7 +21,7 @@ btildflags-$(CONFIG_ARM64_BTI_KERNEL) += -z force-bti
> # potential future proofing if we end up with internal calls to the exported
> # routines, as x86 does (see 6f121e548f83 ("x86, vdso: Reimplement vdso.so
> # preparation in build-time C")).
> -ldflags-y := -shared -soname=linux-vdso.so.1 --hash-style=sysv \
> +ldflags-y := -shared -soname=linux-vdso.so.1 \
> -Bsymbolic --build-id=sha1 -n $(btildflags-y)
>
> ifdef CONFIG_LD_ORPHAN_WARN
> diff --git a/arch/arm64/kernel/vdso32/Makefile b/arch/arm64/kernel/vdso32/Makefile
> index cc4508c604b2..25a2cb6317f3 100644
> --- a/arch/arm64/kernel/vdso32/Makefile
> +++ b/arch/arm64/kernel/vdso32/Makefile
> @@ -98,7 +98,7 @@ VDSO_AFLAGS += -D__ASSEMBLY__
> # From arm vDSO Makefile
> VDSO_LDFLAGS += -Bsymbolic --no-undefined -soname=linux-vdso.so.1
> VDSO_LDFLAGS += -z max-page-size=4096 -z common-page-size=4096
> -VDSO_LDFLAGS += -shared --hash-style=sysv --build-id=sha1
> +VDSO_LDFLAGS += -shared --build-id=sha1
> VDSO_LDFLAGS += --orphan-handling=$(CONFIG_LD_ORPHAN_WARN_LEVEL)
>
>
--
Xi Ruoyao <xry111(a)xry111.site>
School of Aerospace Science and Technology, Xidian University
v19: https://patchwork.kernel.org/project/netdevbpf/list/?series=876852&state=*
====
v18 got a thorough review (thanks!), and this iteration addresses the
feedback.
Major changes:
- Prevent deactivating mp bound queues.
- Prevent installing xdp on mp bound netdevs, or installing mps on xdp
installed netdevs.
- Fix corner cases in netlink API vis-a-vis missing attributes.
- Iron out the unreadable netmem driver support story. To be honest, the
conversation with Jakub & Pavel got a bit confusing for me. I've
implemented an approach in this set that makes sense to me, and
AFAICT, addresses the requirements. It may be good as-is, or it
may be a conversation starter/continuer. To be honest IMO there
are many ways to skin this cat and I don't see an extremely strong
reason to go for one approach over another. Here is one approach you
may like.
- Don't reset niov dma_addr on allocation & free.
- Add some tests to the selftest that catches some of the issues around
missing netlink attributes or deactivating mp-bound queues.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v19/
v18: https://patchwork.kernel.org/project/netdevbpf/list/?series=874848&state=*
====
v17 got minor feedback: (a) to beef up the description on patch 1 and (b)
to remove the leading underscores in the header definition.
I applied (a). (b) seems to be against current conventions so I did not
apply before further discussion.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v17/
v17: https://patchwork.kernel.org/project/netdevbpf/list/?series=869900&state=*
====
v16 also got a very thorough review and some testing (thanks again!).
Thes version addresses all the concerns reported on v15, in terms of
feedback and issues reported.
Major changes:
- Use ASSERT_RTNL.
- Moved around some of the page_pool helpers definitions so I can hide
some netmem helpers in private files as Jakub suggested.
- Don't make every net_iov hold a ref on the binding as Jakub suggested.
- Fix issue reported by Taehee where we access queues after they have
been freed.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v17/
v16: https://patchwork.kernel.org/project/netdevbpf/list/?series=866353&state=*
====
v15 got a thorough review and some testing, and this version addresses almost
all the feedback. Some more minor comments where the authors said it
could be done later, I left out.
Major changes:
- Addition of dma-buf introspection to page-pool-get and queue-get.
- Fixes to selftests suggested by Taehee.
- Fixes to documentation suggested by Donald.
- A couple of suggestions and fixes to TCP patches by Eric and David.
- Fixes to number assignements suggested by Arnd.
- Use rtnl_lock()ing to guard against queue reconfiguration while the
page_pool initialization is happening. (Jakub).
- Fixes to a few warnings reproduced by Taehee.
- Fixes to dma-buf binding suggested by Taehee and Jakub.
- Fixes to netlink UAPI suggested by Jakub
- Applied a number of Reviewed-bys and Acked-bys (including ones I lost
from v13+).
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v16/
One caveat: Taehee reproduced a KASAN warning and reported it here:
https://lore.kernel.org/netdev/CAMArcTUdCxOBYGF3vpbq=eBvqZfnc44KBaQTN7H-wqd…
I estimate the issue to be minor and easily fixable:
https://lore.kernel.org/netdev/CAHS8izNgaqC--GGE2xd85QB=utUnOHmioCsDd1TNxJW…
I hope to be able to follow up with a fix to net tree as net-next closes
imminently, but if this iteration doesn't make it in, I will repost with
a fix squashed after net-next reopens, no problem.
v15: https://patchwork.kernel.org/project/netdevbpf/list/?series=865481&state=*
====
No material changes in this version, only a fix to linking against
libynl.a from the last version. Per Jakub's instructions I've pulled one
of his patches into this series, and now use the new libynl.a correctly,
I hope.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v15/
v14: https://patchwork.kernel.org/project/netdevbpf/list/?series=865135&archive=…
====
No material changes in this version. Only rebase and re-verification on
top of net-next. v13, I think, raced with commit ebad6d0334793
("net/ipv4: Use nested-BH locking for ipv4_tcp_sk.") being merged to
net-next that caused a patchwork failure to apply. This series should
apply cleanly on commit c4532232fa2a4 ("selftests: net: remove unneeded
IP_GRE config").
I did not wait the customary 24hr as Jakub said it's OK to repost as soon
as I build test the rebased version:
https://lore.kernel.org/netdev/20240625075926.146d769d@kernel.org/
v13: https://patchwork.kernel.org/project/netdevbpf/list/?series=861406&archive=…
====
Major changes:
--------------
This iteration addresses Pavel's review comments, applies his
reviewed-by's, and seeks to fix the patchwork build error (sorry!).
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v13/
v12: https://patchwork.kernel.org/project/netdevbpf/list/?series=859747&state=*
====
Major changes:
--------------
This iteration only addresses one minor comment from Pavel with regards
to the trace printing of netmem, and the patchwork build error
introduced in v11 because I missed doing an allmodconfig build, sorry.
Other than that v11, AFAICT, received no feedback. There is one
discussion about how the specifics of plugging io uring memory through
the page pool, but not relevant to content in this particular patchset,
AFAICT.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v12/
v11: https://patchwork.kernel.org/project/netdevbpf/list/?series=857457&state=*
====
Major Changes:
--------------
v11 addresses feedback received in v10. The major change is the removal
of the memory provider ops as requested by Christoph. We still
accomplish the same thing, but utilizing direct function calls with if
statements rather than generic ops.
Additionally address sparse warnings, bugs and review comments from
folks that reviewed.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v11/
Detailed changelog:
-------------------
- Fixes in netdev_rx_queue_restart() from Pavel & David.
- Remove commit e650e8c3a36f5 ("net: page_pool: create hooks for
custom page providers") from the series to address Christoph's
feedback and rebased other patches on the series on this change.
- Fixed build errors with CONFIG_DMA_SHARED_BUFFER &&
!CONFIG_GENERIC_ALLOCATOR build.
- Fixed sparse warnings pointed out by Paolo.
- Drop unnecessary gro_pull_from_frag0 checks.
- Added Bagas reviewed-by to docs.
Cc: Bagas Sanjaya <bagasdotme(a)gmail.com>
Cc: Steven Rostedt <rostedt(a)goodmis.org>
Cc: Christoph Hellwig <hch(a)infradead.org>
Cc: Nikolay Aleksandrov <razor(a)blackwall.org>
Cc: Taehee Yoo <ap420073(a)gmail.com>
Cc: Donald Hunter <donald.hunter(a)gmail.com>
v10: https://patchwork.kernel.org/project/netdevbpf/list/?series=852422&state=*
====
Major Changes:
--------------
v9 was sent right before the merge window closed (sorry!). v10 is almost
a re-send of the series now that the merge window re-opened. Only
rebased to latest net-next and addressed some minor iterative comments
received on v9.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v10/
Detailed changelog:
-------------------
- Fixed tokens leaking in DONTNEED setsockopt (Nikolay).
- Moved net_iov_dma_addr() to devmem.c and made it a devmem specific
helpers (David).
- Rename hook alloc_pages to alloc_netmems as alloc_pages is now
preprocessor macro defined and causes a build error.
v9:
===
Major Changes:
--------------
GVE queue API has been merged. Submitting this version as non-RFC after
rebasing on top of the merged API, and dropped the out of tree queue API
I was carrying on github. Addressed the little feedback v8 has received.
Detailed changelog:
------------------
- Added new patch from David Wei to this series for
netdev_rx_queue_restart()
- Fixed sparse error.
- Removed CONFIG_ checks in netmem_is_net_iov()
- Flipped skb->readable to skb->unreadable
- Minor fixes to selftests & docs.
RFC v8:
=======
Major Changes:
--------------
- Fixed build error generated by patch-by-patch build.
- Applied docs suggestions from Randy.
RFC v7:
=======
Major Changes:
--------------
This revision largely rebases on top of net-next and addresses the feedback
RFCv6 received from folks, namely Jakub, Yunsheng, Arnd, David, & Pavel.
The series remains in RFC because the queue-API ndos defined in this
series are not yet implemented. I have a GVE implementation I carry out
of tree for my testing. A upstreamable GVE implementation is in the
works. Aside from that, in my estimation all the patches are ready for
review/merge. Please do take a look.
As usual the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v7/
Detailed changelog:
- Use admin-perm in netlink API.
- Addressed feedback from Jakub with regards to netlink API
implementation.
- Renamed devmem.c functions to something more appropriate for that
file.
- Improve the performance seen through the page_pool benchmark.
- Fix the value definition of all the SO_DEVMEM_* uapi.
- Various fixes to documentation.
Perf - page-pool benchmark:
---------------------------
Improved performance of bench_page_pool_simple.ko tests compared to v6:
https://pastebin.com/raw/v5dYRg8L
net-next base: 8 cycle fast path.
RFC v6: 10 cycle fast path.
RFC v7: 9 cycle fast path.
RFC v7 with CONFIG_DMA_SHARED_BUFFER disabled: 8 cycle fast path,
same as baseline.
Perf - Devmem TCP benchmark:
---------------------
Perf is about the same regardless of the changes in v7, namely the
removal of the static_branch_unlikely to improve the page_pool benchmark
performance:
189/200gbps bi-directional throughput with RX devmem TCP and regular TCP
TX i.e. ~95% line rate.
RFC v6:
=======
Major Changes:
--------------
This revision largely rebases on top of net-next and addresses the little
feedback RFCv5 received.
The series remains in RFC because the queue-API ndos defined in this
series are not yet implemented. I have a GVE implementation I carry out
of tree for my testing. A upstreamable GVE implementation is in the
works. Aside from that, in my estimation all the patches are ready for
review/merge. Please do take a look.
As usual the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v6/
This version also comes with some performance data recorded in the cover
letter (see below changelog).
Detailed changelog:
- Rebased on top of the merged netmem_ref changes.
- Converted skb->dmabuf to skb->readable (Pavel). Pavel's original
suggestion was to remove the skb->dmabuf flag entirely, but when I
looked into it closely, I found the issue that if we remove the flag
we have to dereference the shinfo(skb) pointer to obtain the first
frag to tell whether an skb is readable or not. This can cause a
performance regression if it dirties the cache line when the
shinfo(skb) was not really needed. Instead, I converted the skb->dmabuf
flag into a generic skb->readable flag which can be re-used by io_uring
0-copy RX.
- Squashed a few locking optimizations from Eric Dumazet in the RX path
and the DEVMEM_DONTNEED setsockopt.
- Expanded the tests a bit. Added validation for invalid scenarios and
added some more coverage.
Perf - page-pool benchmark:
---------------------------
bench_page_pool_simple.ko tests with and without these changes:
https://pastebin.com/raw/ncHDwAbn
AFAIK the number that really matters in the perf tests is the
'tasklet_page_pool01_fast_path Per elem'. This one measures at about 8
cycles without the changes but there is some 1 cycle noise in some
results.
With the patches this regresses to 9 cycles with the changes but there
is 1 cycle noise occasionally running this test repeatedly.
Lastly I tried disable the static_branch_unlikely() in
netmem_is_net_iov() check. To my surprise disabling the
static_branch_unlikely() check reduces the fast path back to 8 cycles,
but the 1 cycle noise remains.
Perf - Devmem TCP benchmark:
---------------------
189/200gbps bi-directional throughput with RX devmem TCP and regular TCP
TX i.e. ~95% line rate.
Major changes in RFC v5:
========================
1. Rebased on top of 'Abstract page from net stack' series and used the
new netmem type to refer to LSB set pointers instead of re-using
struct page.
2. Downgraded this series back to RFC and called it RFC v5. This is
because this series is now dependent on 'Abstract page from net
stack'[1] and the queue API. Both are removed from the series to
reduce the patch # and those bits are fairly independent or
pre-requisite work.
3. Reworked the page_pool devmem support to use netmem and for some
more unified handling.
4. Reworked the reference counting of net_iov (renamed from
page_pool_iov) to use pp_ref_count for refcounting.
The full changes including the dependent series and GVE page pool
support is here:
https://github.com/mina/linux/commits/tcpdevmem-rfcv5/
[1] https://patchwork.kernel.org/project/netdevbpf/list/?series=810774
Major changes in v1:
====================
1. Implemented MVP queue API ndos to remove the userspace-visible
driver reset.
2. Fixed issues in the napi_pp_put_page() devmem frag unref path.
3. Removed RFC tag.
Many smaller addressed comments across all the patches (patches have
individual change log).
Full tree including the rest of the GVE driver changes:
https://github.com/mina/linux/commits/tcpdevmem-v1
Changes in RFC v3:
==================
1. Pulled in the memory-provider dependency from Jakub's RFC[1] to make the
series reviewable and mergeable.
2. Implemented multi-rx-queue binding which was a todo in v2.
3. Fix to cmsg handling.
The sticking point in RFC v2[2] was the device reset required to refill
the device rx-queues after the dmabuf bind/unbind. The solution
suggested as I understand is a subset of the per-queue management ops
Jakub suggested or similar:
https://lore.kernel.org/netdev/20230815171638.4c057dcd@kernel.org/
This is not addressed in this revision, because:
1. This point was discussed at netconf & netdev and there is openness to
using the current approach of requiring a device reset.
2. Implementing individual queue resetting seems to be difficult for my
test bed with GVE. My prototype to test this ran into issues with the
rx-queues not coming back up properly if reset individually. At the
moment I'm unsure if it's a mistake in the POC or a genuine issue in
the virtualization stack behind GVE, which currently doesn't test
individual rx-queue restart.
3. Our usecases are not bothered by requiring a device reset to refill
the buffer queues, and we'd like to support NICs that run into this
limitation with resetting individual queues.
My thought is that drivers that have trouble with per-queue configs can
use the support in this series, while drivers that support new netdev
ops to reset individual queues can automatically reset the queue as
part of the dma-buf bind/unbind.
The same approach with device resets is presented again for consideration
with other sticking points addressed.
This proposal includes the rx devmem path only proposed for merge. For a
snapshot of my entire tree which includes the GVE POC page pool support &
device memory support:
https://github.com/torvalds/linux/compare/master...mina:linux:tcpdevmem-v3
[1] https://lore.kernel.org/netdev/f8270765-a27b-6ccf-33ea-cda097168d79@redhat.…
[2] https://lore.kernel.org/netdev/CAHS8izOVJGJH5WF68OsRWFKJid1_huzzUK+hpKbLcL4…
Changes in RFC v2:
==================
The sticking point in RFC v1[1] was the dma-buf pages approach we used to
deliver the device memory to the TCP stack. RFC v2 is a proof-of-concept
that attempts to resolve this by implementing scatterlist support in the
networking stack, such that we can import the dma-buf scatterlist
directly. This is the approach proposed at a high level here[2].
Detailed changes:
1. Replaced dma-buf pages approach with importing scatterlist into the
page pool.
2. Replace the dma-buf pages centric API with a netlink API.
3. Removed the TX path implementation - there is no issue with
implementing the TX path with scatterlist approach, but leaving
out the TX path makes it easier to review.
4. Functionality is tested with this proposal, but I have not conducted
perf testing yet. I'm not sure there are regressions, but I removed
perf claims from the cover letter until they can be re-confirmed.
5. Added Signed-off-by: contributors to the implementation.
6. Fixed some bugs with the RX path since RFC v1.
Any feedback welcome, but specifically the biggest pending questions
needing feedback IMO are:
1. Feedback on the scatterlist-based approach in general.
2. Netlink API (Patch 1 & 2).
3. Approach to handle all the drivers that expect to receive pages from
the page pool (Patch 6).
[1] https://lore.kernel.org/netdev/dfe4bae7-13a0-3c5d-d671-f61b375cb0b4@gmail.c…
[2] https://lore.kernel.org/netdev/CAHS8izPm6XRS54LdCDZVd0C75tA1zHSu6jLVO8nzTLX…
==================
* TL;DR:
Device memory TCP (devmem TCP) is a proposal for transferring data to and/or
from device memory efficiently, without bouncing the data to a host memory
buffer.
* Problem:
A large amount of data transfers have device memory as the source and/or
destination. Accelerators drastically increased the volume of such transfers.
Some examples include:
- ML accelerators transferring large amounts of training data from storage into
GPU/TPU memory. In some cases ML training setup time can be as long as 50% of
TPU compute time, improving data transfer throughput & efficiency can help
improving GPU/TPU utilization.
- Distributed training, where ML accelerators, such as GPUs on different hosts,
exchange data among them.
- Distributed raw block storage applications transfer large amounts of data with
remote SSDs, much of this data does not require host processing.
Today, the majority of the Device-to-Device data transfers the network are
implemented as the following low level operations: Device-to-Host copy,
Host-to-Host network transfer, and Host-to-Device copy.
The implementation is suboptimal, especially for bulk data transfers, and can
put significant strains on system resources, such as host memory bandwidth,
PCIe bandwidth, etc. One important reason behind the current state is the
kernel’s lack of semantics to express device to network transfers.
* Proposal:
In this patch series we attempt to optimize this use case by implementing
socket APIs that enable the user to:
1. send device memory across the network directly, and
2. receive incoming network packets directly into device memory.
Packet _payloads_ go directly from the NIC to device memory for receive and from
device memory to NIC for transmit.
Packet _headers_ go to/from host memory and are processed by the TCP/IP stack
normally. The NIC _must_ support header split to achieve this.
Advantages:
- Alleviate host memory bandwidth pressure, compared to existing
network-transfer + device-copy semantics.
- Alleviate PCIe BW pressure, by limiting data transfer to the lowest level
of the PCIe tree, compared to traditional path which sends data through the
root complex.
* Patch overview:
** Part 1: netlink API
Gives user ability to bind dma-buf to an RX queue.
** Part 2: scatterlist support
Currently the standard for device memory sharing is DMABUF, which doesn't
generate struct pages. On the other hand, networking stack (skbs, drivers, and
page pool) operate on pages. We have 2 options:
1. Generate struct pages for dmabuf device memory, or,
2. Modify the networking stack to process scatterlist.
Approach #1 was attempted in RFC v1. RFC v2 implements approach #2.
** part 3: page pool support
We piggy back on page pool memory providers proposal:
https://github.com/kuba-moo/linux/tree/pp-providers
It allows the page pool to define a memory provider that provides the
page allocation and freeing. It helps abstract most of the device memory
TCP changes from the driver.
** part 4: support for unreadable skb frags
Page pool iovs are not accessible by the host; we implement changes
throughput the networking stack to correctly handle skbs with unreadable
frags.
** Part 5: recvmsg() APIs
We define user APIs for the user to send and receive device memory.
Not included with this series is the GVE devmem TCP support, just to
simplify the review. Code available here if desired:
https://github.com/mina/linux/tree/tcpdevmem
This series is built on top of net-next with Jakub's pp-providers changes
cherry-picked.
* NIC dependencies:
1. (strict) Devmem TCP require the NIC to support header split, i.e. the
capability to split incoming packets into a header + payload and to put
each into a separate buffer. Devmem TCP works by using device memory
for the packet payload, and host memory for the packet headers.
2. (optional) Devmem TCP works better with flow steering support & RSS support,
i.e. the NIC's ability to steer flows into certain rx queues. This allows the
sysadmin to enable devmem TCP on a subset of the rx queues, and steer
devmem TCP traffic onto these queues and non devmem TCP elsewhere.
The NIC I have access to with these properties is the GVE with DQO support
running in Google Cloud, but any NIC that supports these features would suffice.
I may be able to help reviewers bring up devmem TCP on their NICs.
* Testing:
The series includes a udmabuf kselftest that show a simple use case of
devmem TCP and validates the entire data path end to end without
a dependency on a specific dmabuf provider.
** Test Setup
Kernel: net-next with this series and memory provider API cherry-picked
locally.
Hardware: Google Cloud A3 VMs.
NIC: GVE with header split & RSS & flow steering support.
Cc: Pavel Begunkov <asml.silence(a)gmail.com>
Cc: David Wei <dw(a)davidwei.uk>
Cc: Jason Gunthorpe <jgg(a)ziepe.ca>
Cc: Yunsheng Lin <linyunsheng(a)huawei.com>
Cc: Shailend Chand <shailend(a)google.com>
Cc: Harshitha Ramamurthy <hramamurthy(a)google.com>
Cc: Shakeel Butt <shakeel.butt(a)linux.dev>
Cc: Jeroen de Borst <jeroendb(a)google.com>
Cc: Praveen Kaligineedi <pkaligineedi(a)google.com>
Mina Almasry (13):
netdev: add netdev_rx_queue_restart()
net: netdev netlink api to bind dma-buf to a net device
netdev: support binding dma-buf to netdevice
netdev: netdevice devmem allocator
page_pool: devmem support
memory-provider: dmabuf devmem memory provider
net: support non paged skb frags
net: add support for skbs with unreadable frags
tcp: RX path for devmem TCP
net: add SO_DEVMEM_DONTNEED setsockopt to release RX frags
net: add devmem TCP documentation
selftests: add ncdevmem, netcat for devmem TCP
netdev: add dmabuf introspection
Documentation/netlink/specs/netdev.yaml | 61 +++
Documentation/networking/devmem.rst | 269 ++++++++++++
Documentation/networking/index.rst | 1 +
arch/alpha/include/uapi/asm/socket.h | 6 +
arch/mips/include/uapi/asm/socket.h | 6 +
arch/parisc/include/uapi/asm/socket.h | 6 +
arch/sparc/include/uapi/asm/socket.h | 6 +
include/linux/netdevice.h | 2 +
include/linux/skbuff.h | 61 ++-
include/linux/skbuff_ref.h | 9 +-
include/linux/socket.h | 1 +
include/net/devmem.h | 128 ++++++
include/net/mp_dmabuf_devmem.h | 44 ++
include/net/netdev_rx_queue.h | 5 +
include/net/netmem.h | 169 ++++++-
include/net/page_pool/helpers.h | 39 +-
include/net/page_pool/types.h | 22 +-
include/net/sock.h | 2 +
include/net/tcp.h | 5 +-
include/trace/events/page_pool.h | 12 +-
include/uapi/asm-generic/socket.h | 6 +
include/uapi/linux/netdev.h | 13 +
include/uapi/linux/uio.h | 17 +
net/core/Makefile | 3 +-
net/core/datagram.c | 6 +
net/core/dev.c | 25 +-
net/core/devmem.c | 378 ++++++++++++++++
net/core/gro.c | 3 +-
net/core/netdev-genl-gen.c | 23 +
net/core/netdev-genl-gen.h | 6 +
net/core/netdev-genl.c | 129 ++++++
net/core/netdev_rx_queue.c | 74 ++++
net/core/netmem_priv.h | 31 ++
net/core/page_pool.c | 120 +++--
net/core/page_pool_priv.h | 30 ++
net/core/page_pool_user.c | 30 ++
net/core/skbuff.c | 77 +++-
net/core/sock.c | 68 +++
net/ethtool/common.c | 9 +
net/ipv4/esp4.c | 3 +-
net/ipv4/tcp.c | 261 ++++++++++-
net/ipv4/tcp_input.c | 13 +-
net/ipv4/tcp_ipv4.c | 16 +
net/ipv4/tcp_minisocks.c | 2 +
net/ipv4/tcp_output.c | 5 +-
net/ipv6/esp6.c | 3 +-
net/packet/af_packet.c | 4 +-
tools/include/uapi/linux/netdev.h | 13 +
tools/testing/selftests/net/.gitignore | 1 +
tools/testing/selftests/net/Makefile | 9 +
tools/testing/selftests/net/ncdevmem.c | 560 ++++++++++++++++++++++++
51 files changed, 2671 insertions(+), 121 deletions(-)
create mode 100644 Documentation/networking/devmem.rst
create mode 100644 include/net/devmem.h
create mode 100644 include/net/mp_dmabuf_devmem.h
create mode 100644 net/core/devmem.c
create mode 100644 net/core/netdev_rx_queue.c
create mode 100644 net/core/netmem_priv.h
create mode 100644 tools/testing/selftests/net/ncdevmem.c
--
2.46.0.76.ge559c4bf1a-goog
Hi,
Here is the v3 patch to support polling on event 'hist' file.
The previous version is here;
https://lore.kernel.org/all/171998475175.54594.7244942110355338359.stgit@de…
This version is rebased on the v6.11-rc3.
Background
----------
There has been interest in allowing user programs to monitor kernel
events in real time. Ftrace provides `trace_pipe` interface to wait
on events in the ring buffer, but it is needed to wait until filling
up a page with events in the ring buffer. We can also peek the
`trace` file periodically, but that is inefficient way to monitor
a randomely happening event.
Overview
--------
This patch set allows user to `poll`(or `select`, `epoll`) on event
histogram interface. As you know each event has its own `hist` file
which shows histograms generated by trigger action. So user can set
a new hist trigger on any event you want to monitor, and poll on the
`hist` file until it is updated.
There are 2 poll events are supported, POLLIN and POLLPRI. POLLIN
means that there are any readable update on `hist` file and this
event will be flashed only when you call read(). So, this is
useful if you want to read the histogram periodically.
The other POLLPRI event is for monitoring trace event. Like the
POLLIN, this will be returned when the histogram is updated, but
you don't need to read() the file and use poll() again.
Note that this waits for histogram update (not event arrival), thus
you must set a histogram on the event at first.
Usage
-----
Here is an example usage:
----
TRACEFS=/sys/kernel/tracing
EVENT=$TRACEFS/events/sched/sched_process_free
# setup histogram trigger and enable event
echo "hist:key=comm" >> $EVENT/trigger
echo 1 > $EVENT/enable
# Wait for update
poll pri $EVENT/hist
# Event arrived.
echo "process free event is comming"
tail $TRACEFS/trace
----
The 'poll' command is in the selftest patch.
You can take this series also from here;
https://git.kernel.org/pub/scm/linux/kernel/git/mhiramat/linux.git/log/?h=t…
Thank you,
---
Masami Hiramatsu (Google) (3):
tracing/hist: Add poll(POLLIN) support on hist file
tracing/hist: Support POLLPRI event for poll on histogram
selftests/tracing: Add hist poll() support test
include/linux/trace_events.h | 5 +
kernel/trace/trace_events.c | 18 ++++
kernel/trace/trace_events_hist.c | 101 +++++++++++++++++++-
tools/testing/selftests/ftrace/Makefile | 2
tools/testing/selftests/ftrace/poll.c | 62 ++++++++++++
.../ftrace/test.d/trigger/trigger-hist-poll.tc | 74 +++++++++++++++
6 files changed, 259 insertions(+), 3 deletions(-)
create mode 100644 tools/testing/selftests/ftrace/poll.c
create mode 100644 tools/testing/selftests/ftrace/test.d/trigger/trigger-hist-poll.tc
--
Masami Hiramatsu (Google) <mhiramat(a)kernel.org>
Change the parameter expected_buf from (const void *) to (const char *)
in the function __recvpair() as per the feedback in v1.
Add Fixes tag as per feedback in v1.
This change fixes the below warnings during test compilation:
```
In file included from msg_oob.c:14:
msg_oob.c: In function ‘__recvpair’:
../../kselftest_harness.h:106:40: warning: format ‘%s’ expects argument
of type ‘char *’,but argument 6 has type ‘const void *’ [-Wformat=]
../../kselftest_harness.h:101:17: note: in expansion of macro ‘__TH_LOG’
msg_oob.c:235:17: note: in expansion of macro ‘TH_LOG’
../../kselftest_harness.h:106:40: warning: format ‘%s’ expects argument
of type ‘char *’,but argument 6 has type ‘const void *’ [-Wformat=]
../../kselftest_harness.h:101:17: note: in expansion of macro ‘__TH_LOG’
msg_oob.c:259:25: note: in expansion of macro ‘TH_LOG’
```
v1:
lore.kernel.org/netdev/20240810134037.669765-1-jain.abhinav177@gmail.com
Fixes: d098d77232c3 ("selftest: af_unix: Add msg_oob.c.")
Signed-off-by: Abhinav Jain <jain.abhinav177(a)gmail.com>
---
tools/testing/selftests/net/af_unix/msg_oob.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/testing/selftests/net/af_unix/msg_oob.c b/tools/testing/selftests/net/af_unix/msg_oob.c
index 16d0c172eaeb..535eb2c3d7d1 100644
--- a/tools/testing/selftests/net/af_unix/msg_oob.c
+++ b/tools/testing/selftests/net/af_unix/msg_oob.c
@@ -209,7 +209,7 @@ static void __sendpair(struct __test_metadata *_metadata,
static void __recvpair(struct __test_metadata *_metadata,
FIXTURE_DATA(msg_oob) *self,
- const void *expected_buf, int expected_len,
+ const char *expected_buf, int expected_len,
int buf_len, int flags)
{
int i, ret[2], recv_errno[2], expected_errno = 0;
--
2.34.1
There are 2 issues for the current udpgro test. The first one is the testing
doesn't record all the failures, which may report pass but the test actually
failed. e.g.
https://netdev-3.bots.linux.dev/vmksft-net/results/725661/45-udpgro-sh/stdo…
The other one is after commit d7db7775ea2e ("net: veth: do not manipulate
GRO when using XDP"), there is no need to load xdp program to enable GRO
on veth device.
Hangbin Liu (2):
selftests: udpgro: report error when receive failed
selftests: udpgro: no need to load xdp for gro
tools/testing/selftests/net/udpgro.sh | 50 +++++++++++++--------------
1 file changed, 25 insertions(+), 25 deletions(-)
--
2.45.0
Add documentation outlining the usage and details of devmem TCP.
Signed-off-by: Mina Almasry <almasrymina(a)google.com>
Reviewed-by: Bagas Sanjaya <bagasdotme(a)gmail.com>
Reviewed-by: Donald Hunter <donald.hunter(a)gmail.com>
---
v16:
- Add documentation on unbinding the NIC from dmabuf (Donald).
- Add note that any dmabuf should work (Donald).
v9: https://lore.kernel.org/netdev/20240403002053.2376017-14-almasrymina@google…
- Bagas doc suggestions.
v8:
- Applied docs suggestions (Randy). Thanks!
v7:
- Applied docs suggestions (Jakub).
v2:
- Missing spdx (simon)
- add to index.rst (simon)
---
Documentation/networking/devmem.rst | 269 ++++++++++++++++++++++++++++
Documentation/networking/index.rst | 1 +
2 files changed, 270 insertions(+)
create mode 100644 Documentation/networking/devmem.rst
diff --git a/Documentation/networking/devmem.rst b/Documentation/networking/devmem.rst
new file mode 100644
index 000000000000..417fc977844e
--- /dev/null
+++ b/Documentation/networking/devmem.rst
@@ -0,0 +1,269 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+=================
+Device Memory TCP
+=================
+
+
+Intro
+=====
+
+Device memory TCP (devmem TCP) enables receiving data directly into device
+memory (dmabuf). The feature is currently implemented for TCP sockets.
+
+
+Opportunity
+-----------
+
+A large number of data transfers have device memory as the source and/or
+destination. Accelerators drastically increased the prevalence of such
+transfers. Some examples include:
+
+- Distributed training, where ML accelerators, such as GPUs on different hosts,
+ exchange data.
+
+- Distributed raw block storage applications transfer large amounts of data with
+ remote SSDs. Much of this data does not require host processing.
+
+Typically the Device-to-Device data transfers in the network are implemented as
+the following low-level operations: Device-to-Host copy, Host-to-Host network
+transfer, and Host-to-Device copy.
+
+The flow involving host copies is suboptimal, especially for bulk data transfers,
+and can put significant strains on system resources such as host memory
+bandwidth and PCIe bandwidth.
+
+Devmem TCP optimizes this use case by implementing socket APIs that enable
+the user to receive incoming network packets directly into device memory.
+
+Packet payloads go directly from the NIC to device memory.
+
+Packet headers go to host memory and are processed by the TCP/IP stack
+normally. The NIC must support header split to achieve this.
+
+Advantages:
+
+- Alleviate host memory bandwidth pressure, compared to existing
+ network-transfer + device-copy semantics.
+
+- Alleviate PCIe bandwidth pressure, by limiting data transfer to the lowest
+ level of the PCIe tree, compared to the traditional path which sends data
+ through the root complex.
+
+
+More Info
+---------
+
+ slides, video
+ https://netdevconf.org/0x17/sessions/talk/device-memory-tcp.html
+
+ patchset
+ [RFC PATCH v6 00/12] Device Memory TCP
+ https://lore.kernel.org/netdev/20240305020153.2787423-1-almasrymina@google.…
+
+
+Interface
+=========
+
+
+Example
+-------
+
+tools/testing/selftests/net/ncdevmem.c:do_server shows an example of setting up
+the RX path of this API.
+
+
+NIC Setup
+---------
+
+Header split, flow steering, & RSS are required features for devmem TCP.
+
+Header split is used to split incoming packets into a header buffer in host
+memory, and a payload buffer in device memory.
+
+Flow steering & RSS are used to ensure that only flows targeting devmem land on
+an RX queue bound to devmem.
+
+Enable header split & flow steering::
+
+ # enable header split
+ ethtool -G eth1 tcp-data-split on
+
+
+ # enable flow steering
+ ethtool -K eth1 ntuple on
+
+Configure RSS to steer all traffic away from the target RX queue (queue 15 in
+this example)::
+
+ ethtool --set-rxfh-indir eth1 equal 15
+
+
+The user must bind a dmabuf to any number of RX queues on a given NIC using
+the netlink API::
+
+ /* Bind dmabuf to NIC RX queue 15 */
+ struct netdev_queue *queues;
+ queues = malloc(sizeof(*queues) * 1);
+
+ queues[0]._present.type = 1;
+ queues[0]._present.idx = 1;
+ queues[0].type = NETDEV_RX_QUEUE_TYPE_RX;
+ queues[0].idx = 15;
+
+ *ys = ynl_sock_create(&ynl_netdev_family, &yerr);
+
+ req = netdev_bind_rx_req_alloc();
+ netdev_bind_rx_req_set_ifindex(req, 1 /* ifindex */);
+ netdev_bind_rx_req_set_dmabuf_fd(req, dmabuf_fd);
+ __netdev_bind_rx_req_set_queues(req, queues, n_queue_index);
+
+ rsp = netdev_bind_rx(*ys, req);
+
+ dmabuf_id = rsp->dmabuf_id;
+
+
+The netlink API returns a dmabuf_id: a unique ID that refers to this dmabuf
+that has been bound.
+
+The user can unbind the dmabuf from the netdevice by closing the netlink socket
+that established the binding. We do this so that the binding is automatically
+unbound even if the userspace process crashes.
+
+Note that any reasonably well-behaved dmabuf from any exporter should work with
+devmem TCP, even if the dmabuf is not actually backed by devmem. An example of
+this is udmabuf, which wraps user memory (non-devmem) in a dmabuf.
+
+
+Socket Setup
+------------
+
+The socket must be flow steered to the dmabuf bound RX queue::
+
+ ethtool -N eth1 flow-type tcp4 ... queue 15,
+
+
+Receiving data
+--------------
+
+The user application must signal to the kernel that it is capable of receiving
+devmem data by passing the MSG_SOCK_DEVMEM flag to recvmsg::
+
+ ret = recvmsg(fd, &msg, MSG_SOCK_DEVMEM);
+
+Applications that do not specify the MSG_SOCK_DEVMEM flag will receive an EFAULT
+on devmem data.
+
+Devmem data is received directly into the dmabuf bound to the NIC in 'NIC
+Setup', and the kernel signals such to the user via the SCM_DEVMEM_* cmsgs::
+
+ for (cm = CMSG_FIRSTHDR(&msg); cm; cm = CMSG_NXTHDR(&msg, cm)) {
+ if (cm->cmsg_level != SOL_SOCKET ||
+ (cm->cmsg_type != SCM_DEVMEM_DMABUF &&
+ cm->cmsg_type != SCM_DEVMEM_LINEAR))
+ continue;
+
+ dmabuf_cmsg = (struct dmabuf_cmsg *)CMSG_DATA(cm);
+
+ if (cm->cmsg_type == SCM_DEVMEM_DMABUF) {
+ /* Frag landed in dmabuf.
+ *
+ * dmabuf_cmsg->dmabuf_id is the dmabuf the
+ * frag landed on.
+ *
+ * dmabuf_cmsg->frag_offset is the offset into
+ * the dmabuf where the frag starts.
+ *
+ * dmabuf_cmsg->frag_size is the size of the
+ * frag.
+ *
+ * dmabuf_cmsg->frag_token is a token used to
+ * refer to this frag for later freeing.
+ */
+
+ struct dmabuf_token token;
+ token.token_start = dmabuf_cmsg->frag_token;
+ token.token_count = 1;
+ continue;
+ }
+
+ if (cm->cmsg_type == SCM_DEVMEM_LINEAR)
+ /* Frag landed in linear buffer.
+ *
+ * dmabuf_cmsg->frag_size is the size of the
+ * frag.
+ */
+ continue;
+
+ }
+
+Applications may receive 2 cmsgs:
+
+- SCM_DEVMEM_DMABUF: this indicates the fragment landed in the dmabuf indicated
+ by dmabuf_id.
+
+- SCM_DEVMEM_LINEAR: this indicates the fragment landed in the linear buffer.
+ This typically happens when the NIC is unable to split the packet at the
+ header boundary, such that part (or all) of the payload landed in host
+ memory.
+
+Applications may receive no SO_DEVMEM_* cmsgs. That indicates non-devmem,
+regular TCP data that landed on an RX queue not bound to a dmabuf.
+
+
+Freeing frags
+-------------
+
+Frags received via SCM_DEVMEM_DMABUF are pinned by the kernel while the user
+processes the frag. The user must return the frag to the kernel via
+SO_DEVMEM_DONTNEED::
+
+ ret = setsockopt(client_fd, SOL_SOCKET, SO_DEVMEM_DONTNEED, &token,
+ sizeof(token));
+
+The user must ensure the tokens are returned to the kernel in a timely manner.
+Failure to do so will exhaust the limited dmabuf that is bound to the RX queue
+and will lead to packet drops.
+
+
+Implementation & Caveats
+========================
+
+Unreadable skbs
+---------------
+
+Devmem payloads are inaccessible to the kernel processing the packets. This
+results in a few quirks for payloads of devmem skbs:
+
+- Loopback is not functional. Loopback relies on copying the payload, which is
+ not possible with devmem skbs.
+
+- Software checksum calculation fails.
+
+- TCP Dump and bpf can't access devmem packet payloads.
+
+
+Testing
+=======
+
+More realistic example code can be found in the kernel source under
+tools/testing/selftests/net/ncdevmem.c
+
+ncdevmem is a devmem TCP netcat. It works very similarly to netcat, but
+receives data directly into a udmabuf.
+
+To run ncdevmem, you need to run it on a server on the machine under test, and
+you need to run netcat on a peer to provide the TX data.
+
+ncdevmem has a validation mode as well that expects a repeating pattern of
+incoming data and validates it as such. For example, you can launch
+ncdevmem on the server by::
+
+ ncdevmem -s <server IP> -c <client IP> -f eth1 -d 3 -n 0000:06:00.0 -l \
+ -p 5201 -v 7
+
+On client side, use regular netcat to send TX data to ncdevmem process
+on the server::
+
+ yes $(echo -e \\x01\\x02\\x03\\x04\\x05\\x06) | \
+ tr \\n \\0 | head -c 5G | nc <server IP> 5201 -p 5201
diff --git a/Documentation/networking/index.rst b/Documentation/networking/index.rst
index d1af04b952f8..0be9924db642 100644
--- a/Documentation/networking/index.rst
+++ b/Documentation/networking/index.rst
@@ -49,6 +49,7 @@ Contents:
cdc_mbim
dccp
dctcp
+ devmem
dns_resolver
driver
eql
--
2.46.0.76.ge559c4bf1a-goog
Add netdev_rx_queue_restart(), which resets an rx queue using the
queue API recently merged[1].
The queue API was merged to enable the core net stack to reset individual
rx queues to actuate changes in the rx queue's configuration. In later
patches in this series, we will use netdev_rx_queue_restart() to reset
rx queues after binding or unbinding dmabuf configuration, which will
cause reallocation of the page_pool to repopulate its memory using the
new configuration.
[1] https://lore.kernel.org/netdev/20240430231420.699177-1-shailend@google.com/…
Signed-off-by: David Wei <dw(a)davidwei.uk>
Signed-off-by: Mina Almasry <almasrymina(a)google.com>
Reviewed-by: Pavel Begunkov <asml.silence(a)gmail.com>
Reviewed-by: Jakub Kicinski <kuba(a)kernel.org>
---
v18:
- Add more color to commit message (Xuan Zhuo).
v17:
- Use ASSERT_RTNL() (Jakub).
v13:
- Add reviewed-by from Pavel (thanks!)
- Fixed comment (Pavel)
v11:
- Fix not checking dev->queue_mgmt_ops (Pavel).
- Fix ndo_queue_mem_free call that passed the wrong pointer (David).
v9: https://lore.kernel.org/all/20240502045410.3524155-4-dw@davidwei.uk/
(submitted by David).
- fixed SPDX license identifier (Simon).
- Rebased on top of merged queue API definition, and changed
implementation to match that.
- Replace rtnl_lock() with rtnl_is_locked() to make it useable from my
netlink code where rtnl is already locked.
---
include/net/netdev_rx_queue.h | 3 ++
net/core/Makefile | 1 +
net/core/netdev_rx_queue.c | 74 +++++++++++++++++++++++++++++++++++
3 files changed, 78 insertions(+)
create mode 100644 net/core/netdev_rx_queue.c
diff --git a/include/net/netdev_rx_queue.h b/include/net/netdev_rx_queue.h
index aa1716fb0e53..e78ca52d67fb 100644
--- a/include/net/netdev_rx_queue.h
+++ b/include/net/netdev_rx_queue.h
@@ -54,4 +54,7 @@ get_netdev_rx_queue_index(struct netdev_rx_queue *queue)
return index;
}
#endif
+
+int netdev_rx_queue_restart(struct net_device *dev, unsigned int rxq);
+
#endif
diff --git a/net/core/Makefile b/net/core/Makefile
index 62be9aef2528..f82232b358a2 100644
--- a/net/core/Makefile
+++ b/net/core/Makefile
@@ -19,6 +19,7 @@ obj-$(CONFIG_NETDEV_ADDR_LIST_TEST) += dev_addr_lists_test.o
obj-y += net-sysfs.o
obj-y += hotdata.o
+obj-y += netdev_rx_queue.o
obj-$(CONFIG_PAGE_POOL) += page_pool.o page_pool_user.o
obj-$(CONFIG_PROC_FS) += net-procfs.o
obj-$(CONFIG_NET_PKTGEN) += pktgen.o
diff --git a/net/core/netdev_rx_queue.c b/net/core/netdev_rx_queue.c
new file mode 100644
index 000000000000..da11720a5983
--- /dev/null
+++ b/net/core/netdev_rx_queue.c
@@ -0,0 +1,74 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+
+#include <linux/netdevice.h>
+#include <net/netdev_queues.h>
+#include <net/netdev_rx_queue.h>
+
+int netdev_rx_queue_restart(struct net_device *dev, unsigned int rxq_idx)
+{
+ void *new_mem, *old_mem;
+ int err;
+
+ if (!dev->queue_mgmt_ops || !dev->queue_mgmt_ops->ndo_queue_stop ||
+ !dev->queue_mgmt_ops->ndo_queue_mem_free ||
+ !dev->queue_mgmt_ops->ndo_queue_mem_alloc ||
+ !dev->queue_mgmt_ops->ndo_queue_start)
+ return -EOPNOTSUPP;
+
+ ASSERT_RTNL();
+
+ new_mem = kvzalloc(dev->queue_mgmt_ops->ndo_queue_mem_size, GFP_KERNEL);
+ if (!new_mem)
+ return -ENOMEM;
+
+ old_mem = kvzalloc(dev->queue_mgmt_ops->ndo_queue_mem_size, GFP_KERNEL);
+ if (!old_mem) {
+ err = -ENOMEM;
+ goto err_free_new_mem;
+ }
+
+ err = dev->queue_mgmt_ops->ndo_queue_mem_alloc(dev, new_mem, rxq_idx);
+ if (err)
+ goto err_free_old_mem;
+
+ err = dev->queue_mgmt_ops->ndo_queue_stop(dev, old_mem, rxq_idx);
+ if (err)
+ goto err_free_new_queue_mem;
+
+ err = dev->queue_mgmt_ops->ndo_queue_start(dev, new_mem, rxq_idx);
+ if (err)
+ goto err_start_queue;
+
+ dev->queue_mgmt_ops->ndo_queue_mem_free(dev, old_mem);
+
+ kvfree(old_mem);
+ kvfree(new_mem);
+
+ return 0;
+
+err_start_queue:
+ /* Restarting the queue with old_mem should be successful as we haven't
+ * changed any of the queue configuration, and there is not much we can
+ * do to recover from a failure here.
+ *
+ * WARN if we fail to recover the old rx queue, and at least free
+ * old_mem so we don't also leak that.
+ */
+ if (dev->queue_mgmt_ops->ndo_queue_start(dev, old_mem, rxq_idx)) {
+ WARN(1,
+ "Failed to restart old queue in error path. RX queue %d may be unhealthy.",
+ rxq_idx);
+ dev->queue_mgmt_ops->ndo_queue_mem_free(dev, old_mem);
+ }
+
+err_free_new_queue_mem:
+ dev->queue_mgmt_ops->ndo_queue_mem_free(dev, new_mem);
+
+err_free_old_mem:
+ kvfree(old_mem);
+
+err_free_new_mem:
+ kvfree(new_mem);
+
+ return err;
+}
--
2.46.0.76.ge559c4bf1a-goog
GCC 13.2.0 reported warning about (void *) being used as a param where (char *)
is expected:
In file included from msg_oob.c:14:
msg_oob.c: In function ‘__recvpair’:
../../kselftest_harness.h:106:40: warning: format ‘%s’ expects argument of type ‘char *’, \
but argument 6 has type ‘const void *’ [-Wformat=]
106 | fprintf(TH_LOG_STREAM, "# %s:%d:%s:" fmt "\n", \
| ^~~~~~~~~~~~~
../../kselftest_harness.h:101:17: note: in expansion of macro ‘__TH_LOG’
101 | __TH_LOG(fmt, ##__VA_ARGS__); \
| ^~~~~~~~
msg_oob.c:235:17: note: in expansion of macro ‘TH_LOG’
235 | TH_LOG("Expected:%s", expected_errno ? strerror(expected_errno) : expected_buf);
| ^~~~~~
../../kselftest_harness.h:106:40: warning: format ‘%s’ expects argument of type ‘char *’, \
but argument 6 has type ‘const void *’ [-Wformat=]
106 | fprintf(TH_LOG_STREAM, "# %s:%d:%s:" fmt "\n", \
| ^~~~~~~~~~~~~
../../kselftest_harness.h:101:17: note: in expansion of macro ‘__TH_LOG’
101 | __TH_LOG(fmt, ##__VA_ARGS__); \
| ^~~~~~~~
msg_oob.c:259:25: note: in expansion of macro ‘TH_LOG’
259 | TH_LOG("Expected:%s", expected_errno ? strerror(expected_errno) : expected_buf);
| ^~~~~~
As Simon suggested, all calls to __recvpair() have char * as expected_buf param, so
it is safe to change param type from (const void *) to (const char *), which silences
the warning.
Fixes: d098d77232c37 ("selftest: af_unix: Add msg_oob.c.")
Reported-by: Mirsad Todorovac <mtodorovac69(a)gmail.com>
Cc: "David S. Miller" <davem(a)davemloft.net>
Cc: Eric Dumazet <edumazet(a)google.com>
Cc: Jakub Kicinski <kuba(a)kernel.org>
Cc: Paolo Abeni <pabeni(a)redhat.com>
Cc: Shuah Khan <shuah(a)kernel.org>
Cc: Kuniyuki Iwashima <kuniyu(a)amazon.com>
Cc: netdev(a)vger.kernel.org
Cc: linux-kselftest(a)vger.kernel.org
Suggested-by: Simon Horman <horms(a)kernel.org>
Signed-off-by: Mirsad Todorovac <mtodorovac69(a)gmail.com>
---
v1 -> v2:
fixed a typo.
change funct param type rather than making two casts, as Simon suggested.
changed Subject: line to reflect the modification.
minor formatting changes.
v1:
initial version to fix the compiler warning.
tools/testing/selftests/net/af_unix/msg_oob.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/testing/selftests/net/af_unix/msg_oob.c b/tools/testing/selftests/net/af_unix/msg_oob.c
index 16d0c172eaeb..535eb2c3d7d1 100644
--- a/tools/testing/selftests/net/af_unix/msg_oob.c
+++ b/tools/testing/selftests/net/af_unix/msg_oob.c
@@ -209,7 +209,7 @@ static void __sendpair(struct __test_metadata *_metadata,
static void __recvpair(struct __test_metadata *_metadata,
FIXTURE_DATA(msg_oob) *self,
- const void *expected_buf, int expected_len,
+ const char *expected_buf, int expected_len,
int buf_len, int flags)
{
int i, ret[2], recv_errno[2], expected_errno = 0;
--
2.43.0
The BPF tracing infrastructure has undergone significant evolution,
leading to the introduction of more robust and efficient APIs.
However, some of the existing tests in the samples/bpf directory have
not kept pace with these developments. These outdated tests not only
create confusion among users but also increase maintenance overhead.
For starter, this patchset focuses on cleaning up outdated 'tracing'
related tests within the BPF testing framework. The goal is to
modernize and streamline selftests by removing obsolete tests and
migrating necessaries to more appropriate locations.
Daniel T. Lee (3):
selftests/bpf: migrate tracepoint overhead test to prog_tests
selftests/bpf: add rename tracepoint bench test
samples/bpf: remove obsolete tracing related tests
samples/bpf/Makefile | 12 -
samples/bpf/test_overhead_kprobe.bpf.c | 41 ----
samples/bpf/test_overhead_raw_tp.bpf.c | 17 --
samples/bpf/test_overhead_tp.bpf.c | 23 --
samples/bpf/test_overhead_user.c | 225 ------------------
samples/bpf/test_override_return.sh | 16 --
samples/bpf/test_probe_write_user.bpf.c | 52 ----
samples/bpf/test_probe_write_user_user.c | 108 ---------
samples/bpf/tracex7.bpf.c | 15 --
samples/bpf/tracex7_user.c | 56 -----
tools/testing/selftests/bpf/bench.c | 2 +
.../selftests/bpf/benchs/bench_rename.c | 16 ++
.../selftests/bpf/benchs/run_bench_rename.sh | 2 +-
.../selftests/bpf/prog_tests/test_overhead.c | 14 +-
.../selftests/bpf/progs/test_overhead.c | 11 +-
15 files changed, 39 insertions(+), 571 deletions(-)
delete mode 100644 samples/bpf/test_overhead_kprobe.bpf.c
delete mode 100644 samples/bpf/test_overhead_raw_tp.bpf.c
delete mode 100644 samples/bpf/test_overhead_tp.bpf.c
delete mode 100644 samples/bpf/test_overhead_user.c
delete mode 100755 samples/bpf/test_override_return.sh
delete mode 100644 samples/bpf/test_probe_write_user.bpf.c
delete mode 100644 samples/bpf/test_probe_write_user_user.c
delete mode 100644 samples/bpf/tracex7.bpf.c
delete mode 100644 samples/bpf/tracex7_user.c
--
2.43.0
This patchset introduces SEV-SNP test to the kernel selftest framework.
It also adds negative testing of SEV, ES and SNP VM types.
Patch 1 - Extend the sev smoke tests to use the SNP specific ioctl
calls and sets up memory to boot a SNP guest VM
Patch 2 - Cleanup patch that decouples the ioctl calls from the sev
selftest library with its test assert and status counterparts. No
functional change introduced
Patch 3 - Introduce ioctl test for SEV, ES
Patch 4 - Introduce positive and negative ioctl test for SEV-SNP
Patch 5 - Adds the X86_FEATURE_SEV_SNP vm type test for KVM_SEV_INIT2
Any feedback/review on the approach and design is highly appreciated!
Pratik R. Sampat (5):
selftests: KVM: Add a basic SNP smoke test
selftests: KVM: Decouple SEV ioctls from asserts
selftests: KVM: SEV IOCTL test
selftests: KVM: SNP IOCTL test
selftests: KVM: SEV-SNP test for KVM_SEV_INIT2
.../selftests/kvm/include/x86_64/processor.h | 1 +
.../selftests/kvm/include/x86_64/sev.h | 39 ++-
tools/testing/selftests/kvm/lib/kvm_util.c | 7 +-
.../selftests/kvm/lib/x86_64/processor.c | 6 +-
tools/testing/selftests/kvm/lib/x86_64/sev.c | 181 +++++++++++---
.../selftests/kvm/x86_64/sev_init2_tests.c | 13 +
.../selftests/kvm/x86_64/sev_smoke_test.c | 223 +++++++++++++++++-
7 files changed, 418 insertions(+), 52 deletions(-)
--
2.34.1
v18: https://patchwork.kernel.org/project/netdevbpf/list/?series=874848&state=*
====
v17 got minor feedback: (a) to beef up the description on patch 1 and (b)
to remove the leading underscores in the header definition.
I applied (a). (b) seems to be against current conventions so I did not
apply before further discussion.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v17/
v17: https://patchwork.kernel.org/project/netdevbpf/list/?series=869900&state=*
====
v16 also got a very thorough review and some testing (thanks again!).
Thes version addresses all the concerns reported on v15, in terms of
feedback and issues reported.
Major changes:
- Use ASSERT_RTNL.
- Moved around some of the page_pool helpers definitions so I can hide
some netmem helpers in private files as Jakub suggested.
- Don't make every net_iov hold a ref on the binding as Jakub suggested.
- Fix issue reported by Taehee where we access queues after they have
been freed.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v17/
v16: https://patchwork.kernel.org/project/netdevbpf/list/?series=866353&state=*
====
v15 got a thorough review and some testing, and this version addresses almost
all the feedback. Some more minor comments where the authors said it
could be done later, I left out.
Major changes:
- Addition of dma-buf introspection to page-pool-get and queue-get.
- Fixes to selftests suggested by Taehee.
- Fixes to documentation suggested by Donald.
- A couple of suggestions and fixes to TCP patches by Eric and David.
- Fixes to number assignements suggested by Arnd.
- Use rtnl_lock()ing to guard against queue reconfiguration while the
page_pool initialization is happening. (Jakub).
- Fixes to a few warnings reproduced by Taehee.
- Fixes to dma-buf binding suggested by Taehee and Jakub.
- Fixes to netlink UAPI suggested by Jakub
- Applied a number of Reviewed-bys and Acked-bys (including ones I lost
from v13+).
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v16/
One caveat: Taehee reproduced a KASAN warning and reported it here:
https://lore.kernel.org/netdev/CAMArcTUdCxOBYGF3vpbq=eBvqZfnc44KBaQTN7H-wqd…
I estimate the issue to be minor and easily fixable:
https://lore.kernel.org/netdev/CAHS8izNgaqC--GGE2xd85QB=utUnOHmioCsDd1TNxJW…
I hope to be able to follow up with a fix to net tree as net-next closes
imminently, but if this iteration doesn't make it in, I will repost with
a fix squashed after net-next reopens, no problem.
v15: https://patchwork.kernel.org/project/netdevbpf/list/?series=865481&state=*
====
No material changes in this version, only a fix to linking against
libynl.a from the last version. Per Jakub's instructions I've pulled one
of his patches into this series, and now use the new libynl.a correctly,
I hope.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v15/
v14: https://patchwork.kernel.org/project/netdevbpf/list/?series=865135&archive=…
====
No material changes in this version. Only rebase and re-verification on
top of net-next. v13, I think, raced with commit ebad6d0334793
("net/ipv4: Use nested-BH locking for ipv4_tcp_sk.") being merged to
net-next that caused a patchwork failure to apply. This series should
apply cleanly on commit c4532232fa2a4 ("selftests: net: remove unneeded
IP_GRE config").
I did not wait the customary 24hr as Jakub said it's OK to repost as soon
as I build test the rebased version:
https://lore.kernel.org/netdev/20240625075926.146d769d@kernel.org/
v13: https://patchwork.kernel.org/project/netdevbpf/list/?series=861406&archive=…
====
Major changes:
--------------
This iteration addresses Pavel's review comments, applies his
reviewed-by's, and seeks to fix the patchwork build error (sorry!).
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v13/
v12: https://patchwork.kernel.org/project/netdevbpf/list/?series=859747&state=*
====
Major changes:
--------------
This iteration only addresses one minor comment from Pavel with regards
to the trace printing of netmem, and the patchwork build error
introduced in v11 because I missed doing an allmodconfig build, sorry.
Other than that v11, AFAICT, received no feedback. There is one
discussion about how the specifics of plugging io uring memory through
the page pool, but not relevant to content in this particular patchset,
AFAICT.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v12/
v11: https://patchwork.kernel.org/project/netdevbpf/list/?series=857457&state=*
====
Major Changes:
--------------
v11 addresses feedback received in v10. The major change is the removal
of the memory provider ops as requested by Christoph. We still
accomplish the same thing, but utilizing direct function calls with if
statements rather than generic ops.
Additionally address sparse warnings, bugs and review comments from
folks that reviewed.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v11/
Detailed changelog:
-------------------
- Fixes in netdev_rx_queue_restart() from Pavel & David.
- Remove commit e650e8c3a36f5 ("net: page_pool: create hooks for
custom page providers") from the series to address Christoph's
feedback and rebased other patches on the series on this change.
- Fixed build errors with CONFIG_DMA_SHARED_BUFFER &&
!CONFIG_GENERIC_ALLOCATOR build.
- Fixed sparse warnings pointed out by Paolo.
- Drop unnecessary gro_pull_from_frag0 checks.
- Added Bagas reviewed-by to docs.
Cc: Bagas Sanjaya <bagasdotme(a)gmail.com>
Cc: Steven Rostedt <rostedt(a)goodmis.org>
Cc: Christoph Hellwig <hch(a)infradead.org>
Cc: Nikolay Aleksandrov <razor(a)blackwall.org>
Cc: Taehee Yoo <ap420073(a)gmail.com>
Cc: Donald Hunter <donald.hunter(a)gmail.com>
v10: https://patchwork.kernel.org/project/netdevbpf/list/?series=852422&state=*
====
Major Changes:
--------------
v9 was sent right before the merge window closed (sorry!). v10 is almost
a re-send of the series now that the merge window re-opened. Only
rebased to latest net-next and addressed some minor iterative comments
received on v9.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v10/
Detailed changelog:
-------------------
- Fixed tokens leaking in DONTNEED setsockopt (Nikolay).
- Moved net_iov_dma_addr() to devmem.c and made it a devmem specific
helpers (David).
- Rename hook alloc_pages to alloc_netmems as alloc_pages is now
preprocessor macro defined and causes a build error.
v9:
===
Major Changes:
--------------
GVE queue API has been merged. Submitting this version as non-RFC after
rebasing on top of the merged API, and dropped the out of tree queue API
I was carrying on github. Addressed the little feedback v8 has received.
Detailed changelog:
------------------
- Added new patch from David Wei to this series for
netdev_rx_queue_restart()
- Fixed sparse error.
- Removed CONFIG_ checks in netmem_is_net_iov()
- Flipped skb->readable to skb->unreadable
- Minor fixes to selftests & docs.
RFC v8:
=======
Major Changes:
--------------
- Fixed build error generated by patch-by-patch build.
- Applied docs suggestions from Randy.
RFC v7:
=======
Major Changes:
--------------
This revision largely rebases on top of net-next and addresses the feedback
RFCv6 received from folks, namely Jakub, Yunsheng, Arnd, David, & Pavel.
The series remains in RFC because the queue-API ndos defined in this
series are not yet implemented. I have a GVE implementation I carry out
of tree for my testing. A upstreamable GVE implementation is in the
works. Aside from that, in my estimation all the patches are ready for
review/merge. Please do take a look.
As usual the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v7/
Detailed changelog:
- Use admin-perm in netlink API.
- Addressed feedback from Jakub with regards to netlink API
implementation.
- Renamed devmem.c functions to something more appropriate for that
file.
- Improve the performance seen through the page_pool benchmark.
- Fix the value definition of all the SO_DEVMEM_* uapi.
- Various fixes to documentation.
Perf - page-pool benchmark:
---------------------------
Improved performance of bench_page_pool_simple.ko tests compared to v6:
https://pastebin.com/raw/v5dYRg8L
net-next base: 8 cycle fast path.
RFC v6: 10 cycle fast path.
RFC v7: 9 cycle fast path.
RFC v7 with CONFIG_DMA_SHARED_BUFFER disabled: 8 cycle fast path,
same as baseline.
Perf - Devmem TCP benchmark:
---------------------
Perf is about the same regardless of the changes in v7, namely the
removal of the static_branch_unlikely to improve the page_pool benchmark
performance:
189/200gbps bi-directional throughput with RX devmem TCP and regular TCP
TX i.e. ~95% line rate.
RFC v6:
=======
Major Changes:
--------------
This revision largely rebases on top of net-next and addresses the little
feedback RFCv5 received.
The series remains in RFC because the queue-API ndos defined in this
series are not yet implemented. I have a GVE implementation I carry out
of tree for my testing. A upstreamable GVE implementation is in the
works. Aside from that, in my estimation all the patches are ready for
review/merge. Please do take a look.
As usual the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v6/
This version also comes with some performance data recorded in the cover
letter (see below changelog).
Detailed changelog:
- Rebased on top of the merged netmem_ref changes.
- Converted skb->dmabuf to skb->readable (Pavel). Pavel's original
suggestion was to remove the skb->dmabuf flag entirely, but when I
looked into it closely, I found the issue that if we remove the flag
we have to dereference the shinfo(skb) pointer to obtain the first
frag to tell whether an skb is readable or not. This can cause a
performance regression if it dirties the cache line when the
shinfo(skb) was not really needed. Instead, I converted the skb->dmabuf
flag into a generic skb->readable flag which can be re-used by io_uring
0-copy RX.
- Squashed a few locking optimizations from Eric Dumazet in the RX path
and the DEVMEM_DONTNEED setsockopt.
- Expanded the tests a bit. Added validation for invalid scenarios and
added some more coverage.
Perf - page-pool benchmark:
---------------------------
bench_page_pool_simple.ko tests with and without these changes:
https://pastebin.com/raw/ncHDwAbn
AFAIK the number that really matters in the perf tests is the
'tasklet_page_pool01_fast_path Per elem'. This one measures at about 8
cycles without the changes but there is some 1 cycle noise in some
results.
With the patches this regresses to 9 cycles with the changes but there
is 1 cycle noise occasionally running this test repeatedly.
Lastly I tried disable the static_branch_unlikely() in
netmem_is_net_iov() check. To my surprise disabling the
static_branch_unlikely() check reduces the fast path back to 8 cycles,
but the 1 cycle noise remains.
Perf - Devmem TCP benchmark:
---------------------
189/200gbps bi-directional throughput with RX devmem TCP and regular TCP
TX i.e. ~95% line rate.
Major changes in RFC v5:
========================
1. Rebased on top of 'Abstract page from net stack' series and used the
new netmem type to refer to LSB set pointers instead of re-using
struct page.
2. Downgraded this series back to RFC and called it RFC v5. This is
because this series is now dependent on 'Abstract page from net
stack'[1] and the queue API. Both are removed from the series to
reduce the patch # and those bits are fairly independent or
pre-requisite work.
3. Reworked the page_pool devmem support to use netmem and for some
more unified handling.
4. Reworked the reference counting of net_iov (renamed from
page_pool_iov) to use pp_ref_count for refcounting.
The full changes including the dependent series and GVE page pool
support is here:
https://github.com/mina/linux/commits/tcpdevmem-rfcv5/
[1] https://patchwork.kernel.org/project/netdevbpf/list/?series=810774
Major changes in v1:
====================
1. Implemented MVP queue API ndos to remove the userspace-visible
driver reset.
2. Fixed issues in the napi_pp_put_page() devmem frag unref path.
3. Removed RFC tag.
Many smaller addressed comments across all the patches (patches have
individual change log).
Full tree including the rest of the GVE driver changes:
https://github.com/mina/linux/commits/tcpdevmem-v1
Changes in RFC v3:
==================
1. Pulled in the memory-provider dependency from Jakub's RFC[1] to make the
series reviewable and mergeable.
2. Implemented multi-rx-queue binding which was a todo in v2.
3. Fix to cmsg handling.
The sticking point in RFC v2[2] was the device reset required to refill
the device rx-queues after the dmabuf bind/unbind. The solution
suggested as I understand is a subset of the per-queue management ops
Jakub suggested or similar:
https://lore.kernel.org/netdev/20230815171638.4c057dcd@kernel.org/
This is not addressed in this revision, because:
1. This point was discussed at netconf & netdev and there is openness to
using the current approach of requiring a device reset.
2. Implementing individual queue resetting seems to be difficult for my
test bed with GVE. My prototype to test this ran into issues with the
rx-queues not coming back up properly if reset individually. At the
moment I'm unsure if it's a mistake in the POC or a genuine issue in
the virtualization stack behind GVE, which currently doesn't test
individual rx-queue restart.
3. Our usecases are not bothered by requiring a device reset to refill
the buffer queues, and we'd like to support NICs that run into this
limitation with resetting individual queues.
My thought is that drivers that have trouble with per-queue configs can
use the support in this series, while drivers that support new netdev
ops to reset individual queues can automatically reset the queue as
part of the dma-buf bind/unbind.
The same approach with device resets is presented again for consideration
with other sticking points addressed.
This proposal includes the rx devmem path only proposed for merge. For a
snapshot of my entire tree which includes the GVE POC page pool support &
device memory support:
https://github.com/torvalds/linux/compare/master...mina:linux:tcpdevmem-v3
[1] https://lore.kernel.org/netdev/f8270765-a27b-6ccf-33ea-cda097168d79@redhat.…
[2] https://lore.kernel.org/netdev/CAHS8izOVJGJH5WF68OsRWFKJid1_huzzUK+hpKbLcL4…
Changes in RFC v2:
==================
The sticking point in RFC v1[1] was the dma-buf pages approach we used to
deliver the device memory to the TCP stack. RFC v2 is a proof-of-concept
that attempts to resolve this by implementing scatterlist support in the
networking stack, such that we can import the dma-buf scatterlist
directly. This is the approach proposed at a high level here[2].
Detailed changes:
1. Replaced dma-buf pages approach with importing scatterlist into the
page pool.
2. Replace the dma-buf pages centric API with a netlink API.
3. Removed the TX path implementation - there is no issue with
implementing the TX path with scatterlist approach, but leaving
out the TX path makes it easier to review.
4. Functionality is tested with this proposal, but I have not conducted
perf testing yet. I'm not sure there are regressions, but I removed
perf claims from the cover letter until they can be re-confirmed.
5. Added Signed-off-by: contributors to the implementation.
6. Fixed some bugs with the RX path since RFC v1.
Any feedback welcome, but specifically the biggest pending questions
needing feedback IMO are:
1. Feedback on the scatterlist-based approach in general.
2. Netlink API (Patch 1 & 2).
3. Approach to handle all the drivers that expect to receive pages from
the page pool (Patch 6).
[1] https://lore.kernel.org/netdev/dfe4bae7-13a0-3c5d-d671-f61b375cb0b4@gmail.c…
[2] https://lore.kernel.org/netdev/CAHS8izPm6XRS54LdCDZVd0C75tA1zHSu6jLVO8nzTLX…
==================
* TL;DR:
Device memory TCP (devmem TCP) is a proposal for transferring data to and/or
from device memory efficiently, without bouncing the data to a host memory
buffer.
* Problem:
A large amount of data transfers have device memory as the source and/or
destination. Accelerators drastically increased the volume of such transfers.
Some examples include:
- ML accelerators transferring large amounts of training data from storage into
GPU/TPU memory. In some cases ML training setup time can be as long as 50% of
TPU compute time, improving data transfer throughput & efficiency can help
improving GPU/TPU utilization.
- Distributed training, where ML accelerators, such as GPUs on different hosts,
exchange data among them.
- Distributed raw block storage applications transfer large amounts of data with
remote SSDs, much of this data does not require host processing.
Today, the majority of the Device-to-Device data transfers the network are
implemented as the following low level operations: Device-to-Host copy,
Host-to-Host network transfer, and Host-to-Device copy.
The implementation is suboptimal, especially for bulk data transfers, and can
put significant strains on system resources, such as host memory bandwidth,
PCIe bandwidth, etc. One important reason behind the current state is the
kernel’s lack of semantics to express device to network transfers.
* Proposal:
In this patch series we attempt to optimize this use case by implementing
socket APIs that enable the user to:
1. send device memory across the network directly, and
2. receive incoming network packets directly into device memory.
Packet _payloads_ go directly from the NIC to device memory for receive and from
device memory to NIC for transmit.
Packet _headers_ go to/from host memory and are processed by the TCP/IP stack
normally. The NIC _must_ support header split to achieve this.
Advantages:
- Alleviate host memory bandwidth pressure, compared to existing
network-transfer + device-copy semantics.
- Alleviate PCIe BW pressure, by limiting data transfer to the lowest level
of the PCIe tree, compared to traditional path which sends data through the
root complex.
* Patch overview:
** Part 1: netlink API
Gives user ability to bind dma-buf to an RX queue.
** Part 2: scatterlist support
Currently the standard for device memory sharing is DMABUF, which doesn't
generate struct pages. On the other hand, networking stack (skbs, drivers, and
page pool) operate on pages. We have 2 options:
1. Generate struct pages for dmabuf device memory, or,
2. Modify the networking stack to process scatterlist.
Approach #1 was attempted in RFC v1. RFC v2 implements approach #2.
** part 3: page pool support
We piggy back on page pool memory providers proposal:
https://github.com/kuba-moo/linux/tree/pp-providers
It allows the page pool to define a memory provider that provides the
page allocation and freeing. It helps abstract most of the device memory
TCP changes from the driver.
** part 4: support for unreadable skb frags
Page pool iovs are not accessible by the host; we implement changes
throughput the networking stack to correctly handle skbs with unreadable
frags.
** Part 5: recvmsg() APIs
We define user APIs for the user to send and receive device memory.
Not included with this series is the GVE devmem TCP support, just to
simplify the review. Code available here if desired:
https://github.com/mina/linux/tree/tcpdevmem
This series is built on top of net-next with Jakub's pp-providers changes
cherry-picked.
* NIC dependencies:
1. (strict) Devmem TCP require the NIC to support header split, i.e. the
capability to split incoming packets into a header + payload and to put
each into a separate buffer. Devmem TCP works by using device memory
for the packet payload, and host memory for the packet headers.
2. (optional) Devmem TCP works better with flow steering support & RSS support,
i.e. the NIC's ability to steer flows into certain rx queues. This allows the
sysadmin to enable devmem TCP on a subset of the rx queues, and steer
devmem TCP traffic onto these queues and non devmem TCP elsewhere.
The NIC I have access to with these properties is the GVE with DQO support
running in Google Cloud, but any NIC that supports these features would suffice.
I may be able to help reviewers bring up devmem TCP on their NICs.
* Testing:
The series includes a udmabuf kselftest that show a simple use case of
devmem TCP and validates the entire data path end to end without
a dependency on a specific dmabuf provider.
** Test Setup
Kernel: net-next with this series and memory provider API cherry-picked
locally.
Hardware: Google Cloud A3 VMs.
NIC: GVE with header split & RSS & flow steering support.
Cc: Pavel Begunkov <asml.silence(a)gmail.com>
Cc: David Wei <dw(a)davidwei.uk>
Cc: Jason Gunthorpe <jgg(a)ziepe.ca>
Cc: Yunsheng Lin <linyunsheng(a)huawei.com>
Cc: Shailend Chand <shailend(a)google.com>
Cc: Harshitha Ramamurthy <hramamurthy(a)google.com>
Cc: Shakeel Butt <shakeel.butt(a)linux.dev>
Cc: Jeroen de Borst <jeroendb(a)google.com>
Cc: Praveen Kaligineedi <pkaligineedi(a)google.com>
Mina Almasry (14):
netdev: add netdev_rx_queue_restart()
net: netdev netlink api to bind dma-buf to a net device
netdev: support binding dma-buf to netdevice
netdev: netdevice devmem allocator
page_pool: move dmaddr helpers to .c file
page_pool: devmem support
memory-provider: dmabuf devmem memory provider
net: support non paged skb frags
net: add support for skbs with unreadable frags
tcp: RX path for devmem TCP
net: add SO_DEVMEM_DONTNEED setsockopt to release RX frags
net: add devmem TCP documentation
selftests: add ncdevmem, netcat for devmem TCP
netdev: add dmabuf introspection
Documentation/netlink/specs/netdev.yaml | 61 +++
Documentation/networking/devmem.rst | 269 ++++++++++++
Documentation/networking/index.rst | 1 +
arch/alpha/include/uapi/asm/socket.h | 6 +
arch/mips/include/uapi/asm/socket.h | 6 +
arch/parisc/include/uapi/asm/socket.h | 6 +
arch/sparc/include/uapi/asm/socket.h | 6 +
include/linux/skbuff.h | 61 ++-
include/linux/skbuff_ref.h | 9 +-
include/linux/socket.h | 1 +
include/net/devmem.h | 128 ++++++
include/net/mp_dmabuf_devmem.h | 44 ++
include/net/netdev_rx_queue.h | 5 +
include/net/netmem.h | 164 +++++++-
include/net/page_pool/helpers.h | 42 +-
include/net/page_pool/types.h | 8 +
include/net/sock.h | 2 +
include/net/tcp.h | 5 +-
include/trace/events/page_pool.h | 12 +-
include/uapi/asm-generic/socket.h | 6 +
include/uapi/linux/netdev.h | 13 +
include/uapi/linux/uio.h | 17 +
net/core/Makefile | 3 +-
net/core/datagram.c | 6 +
net/core/dev.c | 7 +-
net/core/devmem.c | 378 +++++++++++++++++
net/core/gro.c | 3 +-
net/core/netdev-genl-gen.c | 23 +
net/core/netdev-genl-gen.h | 6 +
net/core/netdev-genl.c | 111 +++++
net/core/netdev_rx_queue.c | 74 ++++
net/core/netmem_priv.h | 36 ++
net/core/page_pool.c | 147 +++++--
net/core/page_pool_priv.h | 3 +
net/core/page_pool_user.c | 4 +
net/core/skbuff.c | 77 +++-
net/core/sock.c | 68 +++
net/ipv4/esp4.c | 3 +-
net/ipv4/tcp.c | 261 +++++++++++-
net/ipv4/tcp_input.c | 13 +-
net/ipv4/tcp_ipv4.c | 16 +
net/ipv4/tcp_minisocks.c | 2 +
net/ipv4/tcp_output.c | 5 +-
net/ipv6/esp6.c | 3 +-
net/packet/af_packet.c | 4 +-
tools/include/uapi/linux/netdev.h | 13 +
tools/testing/selftests/net/.gitignore | 1 +
tools/testing/selftests/net/Makefile | 9 +
tools/testing/selftests/net/ncdevmem.c | 536 ++++++++++++++++++++++++
49 files changed, 2563 insertions(+), 121 deletions(-)
create mode 100644 Documentation/networking/devmem.rst
create mode 100644 include/net/devmem.h
create mode 100644 include/net/mp_dmabuf_devmem.h
create mode 100644 net/core/devmem.c
create mode 100644 net/core/netdev_rx_queue.c
create mode 100644 net/core/netmem_priv.h
create mode 100644 tools/testing/selftests/net/ncdevmem.c
--
2.46.0.rc2.264.g509ed76dc8-goog
This patch series adds a selftest suite to validate the s390x
architecture specific ucontrol KVM interface.
When creating a VM on s390x it is possible to create it as userspace
controlled VM or in short ucontrol VM.
These VMs delegates the management of the VM to userspace instead
of handling most events within the kernel. Consequently the userspace
has to manage interrupts, memory allocation etc.
Before this patch set this functionality lacks any public test cases.
It is desirable to add test cases for this interface to be able to
reduce the risk of breaking changes in the future.
In order to provision a ucontrol VM the kernel needs to be compiled with
the CONFIG_KVM_S390_UCONTROL enabled. The users with sys_admin capability
can then create a new ucontrol VM providing the KVM_VM_S390_UCONTROL
parameter to the KVM_CREATE_VM ioctl.
The kernels existing selftest helper functions can only be partially be
reused for these tests.
The test cases cover existing special handling of ucontrol VMs within the
implementation and basic VM creation and handling cases:
* Reject setting HPAGE when VM is ucontrol
* Assert KVM_GET_DIRTY_LOG is rejected
* Assert KVM_S390_VM_MEM_LIMIT_SIZE is rejected
* Assert state of initial SIE flags setup by the kernel
* Run simple program in VM with and without DAT
* Assert KVM_EXIT_S390_UCONTROL exit on not mapped memory access
* Assert functionality of storage keys in ucontrol VM
* Assert that memory region operations are rejected for ucontrol VMs
Running the test cases requires sys_admin capabilities to start the
ucontrol VM.
This can be achieved by running as root or with a command like:
sudo setpriv --reuid nobody --inh-caps -all,+sys_admin \
--ambient-caps -all,+sys_admin --bounding-set -all,+sys_admin \
./ucontrol_test
The patch set does also contain some code cleanup / consolidation of
architecture specific defines that are now used in multiple test cases.
---
v5:
- PATCH 7: Simplify segment and page table generation
- PATCH 8: Remove irrelevant code in skey test
Add some comments to skey test
v4:
- PATCH 5: Remove not yet used include for debug print functions
- PATCH 6: Add include for debug print functions (removed from patch 5)
Remove no longer needed code since stopped but is reset
before starting since v3 (thanks Janosch)
Adjust test output to use leading zeros instead of spaces in sieic
- PATCH 7: Rename constant to PGM_SEGMENT_TRANSLATION (thanks Janosch)
Put comments on their own lines
v3:
- Remove stopped bit before starting the VM (no initial stop in multiple
test cases) (thanks Janosch)
- PATCH 2: Clarified SIE control block vs SIE instruction (thanks
Janosch)
- PATCH 3: Make use of CAP_TO_MASK(CAP_SYS_ADMIN) instead of custom
define (thanks Janosch)
Removed Reviewed-By: Claudio
- PATCH 4: Remove erroneous 1MB offset from self->base_hva (thanks
Janosch)
- PATCH 6-8: Change name of test program _pgm to _asm to prevent confusion
- PATCH 10: Move KVM_S390_UCONTROL default option to actual debug config
(thanks Christian)
v2:
- add ucontrol to s390 debug config (new patch)
- PATCH 2: changed atomic_t to __u32 (thanks Claudio)
- PATCH 4: reformatted comment in FIXTURE_SETUP(uc_kvm)
- PATCH 5: refactored to display 8 byte blocks + more internal reuse
(thanks Claudio)
- PATCH 7: make use of more declarative defines instead of magic values
- PATCH 8: make use of more declarative defines instead of magic values
(thanks Claudio)
- PATCH 9: add reference to fix verified by the test case
Christoph Schlameuss (10):
selftests: kvm: s390: Define page sizes in shared header
selftests: kvm: s390: Add kvm_s390_sie_block definition for userspace
tests
selftests: kvm: s390: Add s390x ucontrol test suite with hpage test
selftests: kvm: s390: Add test fixture and simple VM setup tests
selftests: kvm: s390: Add debug print functions
selftests: kvm: s390: Add VM run test case
selftests: kvm: s390: Add uc_map_unmap VM test case
selftests: kvm: s390: Add uc_skey VM test case
selftests: kvm: s390: Verify reject memory region operations for
ucontrol VMs
s390: Enable KVM_S390_UCONTROL config in debug_defconfig
arch/s390/configs/debug_defconfig | 1 +
tools/testing/selftests/kvm/.gitignore | 1 +
tools/testing/selftests/kvm/Makefile | 1 +
.../selftests/kvm/include/s390x/debug_print.h | 69 ++
.../selftests/kvm/include/s390x/processor.h | 5 +
.../testing/selftests/kvm/include/s390x/sie.h | 240 +++++++
.../selftests/kvm/lib/s390x/processor.c | 10 +-
tools/testing/selftests/kvm/s390x/cmma_test.c | 7 +-
tools/testing/selftests/kvm/s390x/config | 2 +
.../testing/selftests/kvm/s390x/debug_test.c | 4 +-
tools/testing/selftests/kvm/s390x/memop.c | 4 +-
tools/testing/selftests/kvm/s390x/tprot.c | 5 +-
.../selftests/kvm/s390x/ucontrol_test.c | 598 ++++++++++++++++++
13 files changed, 931 insertions(+), 16 deletions(-)
create mode 100644 tools/testing/selftests/kvm/include/s390x/debug_print.h
create mode 100644 tools/testing/selftests/kvm/include/s390x/sie.h
create mode 100644 tools/testing/selftests/kvm/s390x/config
create mode 100644 tools/testing/selftests/kvm/s390x/ucontrol_test.c
base-commit: de9c2c66ad8e787abec7c9d7eff4f8c3cdd28aed
--
2.45.2
This test doesn't have support for other architectures. Altough resctrl
is supported on x86 and ARM, but arch_supports_noncont_cat() shows that
only x86 for AMD and Intel are supported by the test. We get build
errors when built for ARM and ARM64.
Hence add support in the Makefile to build this suite only for x86 and
x86_64 architectures.
Fixes: b733143cc455 ("selftests/resctrl: Make resctrl_tests run using kselftest framework")
Signed-off-by: Muhammad Usama Anjum <usama.anjum(a)collabora.com>
---
tools/testing/selftests/resctrl/Makefile | 2 ++
1 file changed, 2 insertions(+)
diff --git a/tools/testing/selftests/resctrl/Makefile b/tools/testing/selftests/resctrl/Makefile
index f408bd6bfc3d4..d5cf96315ef9b 100644
--- a/tools/testing/selftests/resctrl/Makefile
+++ b/tools/testing/selftests/resctrl/Makefile
@@ -3,7 +3,9 @@
CFLAGS = -g -Wall -O2 -D_FORTIFY_SOURCE=2
CFLAGS += $(KHDR_INCLUDES)
+ifeq ($(ARCH),$(filter $(ARCH),x86 x86_64))
TEST_GEN_PROGS := resctrl_tests
+endif
LOCAL_HDRS += $(wildcard *.h)
--
2.39.2
There are multiple possible timer sources which could be useful for
the sound stream synchronization: hrtimers, hardware clocks (e.g. PTP),
timer wheels (jiffies). Currently, using one of them to synchronize
the audio stream of snd-aloop module would require writing a
kernel-space driver which exports an ALSA timer through the
snd_timer interface.
However, it is not really convenient for application developers, who may
want to define their custom timer sources for audio synchronization.
For instance, we could have a network application which receives frames
and sends them to snd-aloop pcm device, and another application
listening on the other end of snd-aloop. It makes sense to transfer a
new period of data only when certain amount of frames is received
through the network, but definitely not when a certain amount of jiffies
on a local system elapses. Since all of the devices are purely virtual
it won't introduce any glitches and will help the application developers
to avoid using sample-rate conversion.
This patch series introduces userspace-driven ALSA timers: virtual
timers which are created and controlled from userspace. The timer can
be created from the userspace using the new ioctl SNDRV_TIMER_IOCTL_CREATE.
After creating a timer, it becomes available for use system-wide, so it
can be passed to snd-aloop as a timer source (timer_source parameter
would be "-1.SNDRV_TIMER_GLOBAL_UDRIVEN.{timer_id}"). When the userspace
app decides to trigger a timer, it calls another ioctl
SNDRV_TIMER_IOCTL_TRIGGER on the file descriptor of a timer. It
initiates a transfer of a new period of data.
Userspace-driven timers are associated with file descriptors. If the
application wishes to destroy the timer, it can simply release the file
descriptor of a virtual timer.
I believe introducing new ioctl calls is quite inconvenient (as we have
a limited amount of them), but other possible ways of app <-> kernel
communication (like virtual FS) seem completely inappropriate for this
task (but I'd love to discuss alternative solutions).
This patch series also updates the snd-aloop module so the global timers
can be used as a timer_source for it (it allows using userspace-driven
timers as timer source).
V1 -> V2:
- Fix some problems found by Christophe Jaillet
<christophe.jaillet(a)wanadoo.fr>
V2 -> V3:
- Add improvements suggested by Takashi Iwai <tiwai(a)suse.de>
V3 -> V4:
- Address comments from Jaroslav Kysela <perex(a)perex.cz> and Mark Brown
<broonie(a)kernel.org>
Please, find the patch-specific changelog in the following patches.
Ivan Orlov (4):
ALSA: aloop: Allow using global timers
Docs/sound: Add documentation for userspace-driven ALSA timers
ALSA: timer: Introduce virtual userspace-driven timers
selftests: ALSA: Cover userspace-driven timers with test
Documentation/sound/index.rst | 1 +
Documentation/sound/utimers.rst | 125 ++++++++++++
include/uapi/sound/asound.h | 16 +-
sound/core/Kconfig | 10 +
sound/core/timer.c | 210 ++++++++++++++++++++
sound/drivers/aloop.c | 2 +
tools/testing/selftests/alsa/Makefile | 4 +-
tools/testing/selftests/alsa/global-timer.c | 87 ++++++++
tools/testing/selftests/alsa/utimer-test.c | 164 +++++++++++++++
9 files changed, 616 insertions(+), 3 deletions(-)
create mode 100644 Documentation/sound/utimers.rst
create mode 100644 tools/testing/selftests/alsa/global-timer.c
create mode 100644 tools/testing/selftests/alsa/utimer-test.c
--
2.34.1
This patch series implements a new char misc driver, /dev/ntsync, which is used
to implement Windows NT synchronization primitives.
NT synchronization primitives are unique in that the wait functions both are
vectored, operate on multiple types of object with different behaviour (mutex,
semaphore, event), and affect the state of the objects they wait on. This model
is not compatible with existing kernel synchronization objects or interfaces,
and therefore the ntsync driver implements its own wait queues and locking.
This patch series is rebased against the "char-misc-next" branch of
gregkh/char-misc.git.
== Background ==
The Wine project emulates the Windows API in user space. One particular part of
that API, namely the NT synchronization primitives, have historically been
implemented via RPC to a dedicated "kernel" process. However, more recent
applications use these APIs more strenuously, and the overhead of RPC has become
a bottleneck.
The NT synchronization APIs are too complex to implement on top of existing
primitives without sacrificing correctness. Certain operations, such as
NtPulseEvent() or the "wait-for-all" mode of NtWaitForMultipleObjects(), require
direct control over the underlying wait queue, and implementing a wait queue
sufficiently robust for Wine in user space is not possible. This proposed
driver, therefore, implements the problematic interfaces directly in the Linux
kernel.
This driver was presented at Linux Plumbers Conference 2023. For those further
interested in the history of synchronization in Wine and past attempts to solve
this problem in user space, a recording of the presentation can be viewed here:
https://www.youtube.com/watch?v=NjU4nyWyhU8
== Performance ==
The performance measurements described below are copied from earlier versions of
the patch set. While some of the code has changed, I do not currently anticipate
that it has changed drastically enough to affect those measurements.
The gain in performance varies wildly depending on the application in question
and the user's hardware. For some games NT synchronization is not a bottleneck
and no change can be observed, but for others frame rate improvements of 50 to
150 percent are not atypical. The following table lists frame rate measurements
from a variety of games on a variety of hardware, taken by users Dmitry
Skvortsov, FuzzyQuils, OnMars, and myself:
Game Upstream ntsync improvement
===========================================================================
Anger Foot 69 99 43%
Call of Juarez 99.8 224.1 125%
Dirt 3 110.6 860.7 678%
Forza Horizon 5 108 160 48%
Lara Croft: Temple of Osiris 141 326 131%
Metro 2033 164.4 199.2 21%
Resident Evil 2 26 77 196%
The Crew 26 51 96%
Tiny Tina's Wonderlands 130 360 177%
Total War Saga: Troy 109 146 34%
===========================================================================
== Patches ==
The intended semantics of the patches are broadly intended to match those of the
corresponding Windows functions. For those not already familiar with the Windows
functions (or their undocumented behaviour), patch 27/28 provides a detailed
specification, and individual patches also include a brief description of the
API they are implementing.
The patches making use of this driver in Wine can be retrieved or browsed here:
https://repo.or.cz/wine/zf.git/shortlog/refs/heads/ntsync5
== Previous versions ==
Changes from v4:
* Rework wait-all locking code to avoid taking more than one spinlock at a time,
and also to fix a race where the wait-all lock would not be not correctly
taken. The new locking mechanism involves taking a simple spinlock for normal
"any" waits, and taking a device-wide mutex for "all" waits or when locking
any object that is involved in an "all" wait. The mechanism was written by
Peter Zijlstra.
* Try to reword or clarify various parts of the documentation (patch 27), per
Peter Zijlstra.
* I did not rename NTSYNC_IOC_SEM_POST to RELEASE (like NT) although this was
suggested by Peter Zijlstra, mostly because it's not clear to me that renaming
an already committed ioctl would be fine. The API committed isn't actually
usable yet, though, so if altering it would be fine on those grounds, I can
revise this series to rename the function accordingly.
* Similarly, I did not change the create ioctls to return the fd directly,
although this was suggested and would be a bit simpler and cleaner, because
NTSYNC_IOC_CREATE_SEM already exists upstream and returns the fd through a
struct. I can make this change in the next revision if that'd be preferable. I
also still would appreciate a clarification on the advice in [1].
[1] https://docs.kernel.org/driver-api/ioctl.html#return-code
* Link to v4: https://lore.kernel.org/lkml/20240416010837.333694-1-zfigura@codeweavers.co…
* Link to v3: https://lore.kernel.org/lkml/20240329000621.148791-1-zfigura@codeweavers.co…
* Link to v2: https://lore.kernel.org/lkml/20240219223833.95710-1-zfigura@codeweavers.com/
* Link to v1: https://lore.kernel.org/lkml/20240214233645.9273-1-zfigura@codeweavers.com/
* Link to RFC v2: https://lore.kernel.org/lkml/20240131021356.10322-1-zfigura@codeweavers.com/
* Link to RFC v1: https://lore.kernel.org/lkml/20240124004028.16826-1-zfigura@codeweavers.com/
Elizabeth Figura (28):
ntsync: Introduce NTSYNC_IOC_WAIT_ANY.
ntsync: Introduce NTSYNC_IOC_WAIT_ALL.
ntsync: Introduce NTSYNC_IOC_CREATE_MUTEX.
ntsync: Introduce NTSYNC_IOC_MUTEX_UNLOCK.
ntsync: Introduce NTSYNC_IOC_MUTEX_KILL.
ntsync: Introduce NTSYNC_IOC_CREATE_EVENT.
ntsync: Introduce NTSYNC_IOC_EVENT_SET.
ntsync: Introduce NTSYNC_IOC_EVENT_RESET.
ntsync: Introduce NTSYNC_IOC_EVENT_PULSE.
ntsync: Introduce NTSYNC_IOC_SEM_READ.
ntsync: Introduce NTSYNC_IOC_MUTEX_READ.
ntsync: Introduce NTSYNC_IOC_EVENT_READ.
ntsync: Introduce alertable waits.
selftests: ntsync: Add some tests for semaphore state.
selftests: ntsync: Add some tests for mutex state.
selftests: ntsync: Add some tests for NTSYNC_IOC_WAIT_ANY.
selftests: ntsync: Add some tests for NTSYNC_IOC_WAIT_ALL.
selftests: ntsync: Add some tests for wakeup signaling with
WINESYNC_IOC_WAIT_ANY.
selftests: ntsync: Add some tests for wakeup signaling with
WINESYNC_IOC_WAIT_ALL.
selftests: ntsync: Add some tests for manual-reset event state.
selftests: ntsync: Add some tests for auto-reset event state.
selftests: ntsync: Add some tests for wakeup signaling with events.
selftests: ntsync: Add tests for alertable waits.
selftests: ntsync: Add some tests for wakeup signaling via alerts.
selftests: ntsync: Add a stress test for contended waits.
maintainers: Add an entry for ntsync.
docs: ntsync: Add documentation for the ntsync uAPI.
ntsync: No longer depend on BROKEN.
Documentation/userspace-api/index.rst | 1 +
Documentation/userspace-api/ntsync.rst | 398 +++++
MAINTAINERS | 9 +
drivers/misc/Kconfig | 1 -
drivers/misc/ntsync.c | 989 +++++++++++-
include/uapi/linux/ntsync.h | 39 +
tools/testing/selftests/Makefile | 1 +
.../selftests/drivers/ntsync/.gitignore | 1 +
.../testing/selftests/drivers/ntsync/Makefile | 7 +
tools/testing/selftests/drivers/ntsync/config | 1 +
.../testing/selftests/drivers/ntsync/ntsync.c | 1407 +++++++++++++++++
11 files changed, 2850 insertions(+), 4 deletions(-)
create mode 100644 Documentation/userspace-api/ntsync.rst
create mode 100644 tools/testing/selftests/drivers/ntsync/.gitignore
create mode 100644 tools/testing/selftests/drivers/ntsync/Makefile
create mode 100644 tools/testing/selftests/drivers/ntsync/config
create mode 100644 tools/testing/selftests/drivers/ntsync/ntsync.c
base-commit: f5b335dc025cfee90957efa90dc72fada0d5abb4
--
2.43.0
Hello,
this series brings a new set of test converted to the test_progs framework.
Since the tests are quite small, I chose to group three tests conversion in
the same series, but feel free to let me know if I should keep one series
per test. The series focuses on cgroup testing and converts the following
tests:
- get_cgroup_id_user
- cgroup_storage
- test_skb_cgroup_id_user
Signed-off-by: Alexis Lothoré (eBPF Foundation) <alexis.lothore(a)bootlin.com>
---
Changes in v3:
- Fixed multiple leaks on cgroup file descriptors and sockets
- Used dedicated network namespaces for tests involving network
- Link to v2: https://lore.kernel.org/r/20240806-convert_cgroup_tests-v2-0-180c57e5b710@b…
Changes in v2:
- Use global variables instead of maps when possible
- Collect review tags from Alan
- Link to v1: https://lore.kernel.org/r/20240731-convert_cgroup_tests-v1-0-14cbc51b6947@b…
---
Alexis Lothoré (eBPF Foundation) (4):
selftests/bpf: convert get_current_cgroup_id_user to test_progs
selftests/bpf: convert test_cgroup_storage to test_progs
selftests/bpf: add proper section name to bpf prog and rename it
selftests/bpf: convert test_skb_cgroup_id_user to test_progs
tools/testing/selftests/bpf/.gitignore | 3 -
tools/testing/selftests/bpf/Makefile | 8 +-
tools/testing/selftests/bpf/get_cgroup_id_user.c | 151 -----------------
.../selftests/bpf/prog_tests/cgroup_ancestor.c | 169 +++++++++++++++++++
.../bpf/prog_tests/cgroup_get_current_cgroup_id.c | 46 ++++++
.../selftests/bpf/prog_tests/cgroup_storage.c | 94 +++++++++++
...test_skb_cgroup_id_kern.c => cgroup_ancestor.c} | 14 +-
tools/testing/selftests/bpf/progs/cgroup_storage.c | 24 +++
.../selftests/bpf/progs/get_cgroup_id_kern.c | 26 +--
tools/testing/selftests/bpf/test_cgroup_storage.c | 174 --------------------
tools/testing/selftests/bpf/test_skb_cgroup_id.sh | 63 -------
.../selftests/bpf/test_skb_cgroup_id_user.c | 183 ---------------------
12 files changed, 342 insertions(+), 613 deletions(-)
---
base-commit: ab2c4aa104050a184c3411a973b165285549f732
change-id: 20240725-convert_cgroup_tests-d07c66053225
Best regards,
--
Alexis Lothoré, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
Hello,
this series brings a new set of test converted to the test_progs framework.
Since the tests are quite small, I chose to group three tests conversion in
the same series, but feel free to let me know if I should keep one series
per test. The series focuses on cgroup testing and converts the following
tests:
- get_cgroup_id_user
- cgroup_storage
- test_skb_cgroup_id_user
Signed-off-by: Alexis Lothoré (eBPF Foundation) <alexis.lothore(a)bootlin.com>
---
Changes in v2:
- Use global variables instead of maps when possible
- Collect review tags from Alan
- Link to v1: https://lore.kernel.org/r/20240731-convert_cgroup_tests-v1-0-14cbc51b6947@b…
---
Alexis Lothoré (eBPF Foundation) (4):
selftests/bpf: convert get_current_cgroup_id_user to test_progs
selftests/bpf: convert test_cgroup_storage to test_progs
selftests/bpf: add proper section name to bpf prog and rename it
selftests/bpf: convert test_skb_cgroup_id_user to test_progs
tools/testing/selftests/bpf/.gitignore | 3 -
tools/testing/selftests/bpf/Makefile | 8 +-
tools/testing/selftests/bpf/get_cgroup_id_user.c | 151 -----------------
.../selftests/bpf/prog_tests/cgroup_ancestor.c | 154 +++++++++++++++++
.../bpf/prog_tests/cgroup_get_current_cgroup_id.c | 45 +++++
.../selftests/bpf/prog_tests/cgroup_storage.c | 65 ++++++++
...test_skb_cgroup_id_kern.c => cgroup_ancestor.c} | 14 +-
tools/testing/selftests/bpf/progs/cgroup_storage.c | 24 +++
.../selftests/bpf/progs/get_cgroup_id_kern.c | 26 +--
tools/testing/selftests/bpf/test_cgroup_storage.c | 174 --------------------
tools/testing/selftests/bpf/test_skb_cgroup_id.sh | 63 -------
.../selftests/bpf/test_skb_cgroup_id_user.c | 183 ---------------------
12 files changed, 297 insertions(+), 613 deletions(-)
---
base-commit: 34dbece299dfc462db4504268a697f29750d2932
change-id: 20240725-convert_cgroup_tests-d07c66053225
Best regards,
--
Alexis Lothoré, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
[1] mentions that memfd_secret is only supported on arm64, riscv, x86
and x86_64 for now. It doesn't support other architectures. I found the
build error on arm and decided to send the fix as it was creating noise
on KernelCI:
memfd_secret.c: In function 'memfd_secret':
memfd_secret.c:42:24: error: '__NR_memfd_secret' undeclared (first use in this function);
did you mean 'memfd_secret'?
42 | return syscall(__NR_memfd_secret, flags);
| ^~~~~~~~~~~~~~~~~
| memfd_secret
Hence I'm adding condition that memfd_secret should only be compiled on
supported architectures.
Also check in run_vmtests script if memfd_secret binary is present
before executing it.
[1] https://lore.kernel.org/all/20210518072034.31572-7-rppt@kernel.org/
Cc: stable(a)vger.kernel.org
Fixes: 76fe17ef588a ("secretmem: test: add basic selftest for memfd_secret(2)")
Reviewed-by: Shuah Khan <skhan(a)linuxfoundation.org>
Acked-by: Mike Rapoport (Microsoft) <rppt(a)kernel.org>
Signed-off-by: Muhammad Usama Anjum <usama.anjum(a)collabora.com>
---
Changes since v1:
- Add error message to the patch and reviewed-by tags
---
tools/testing/selftests/mm/Makefile | 2 ++
tools/testing/selftests/mm/run_vmtests.sh | 3 +++
2 files changed, 5 insertions(+)
diff --git a/tools/testing/selftests/mm/Makefile b/tools/testing/selftests/mm/Makefile
index 1a83b70e84535..4ea188be0588a 100644
--- a/tools/testing/selftests/mm/Makefile
+++ b/tools/testing/selftests/mm/Makefile
@@ -53,7 +53,9 @@ TEST_GEN_FILES += madv_populate
TEST_GEN_FILES += map_fixed_noreplace
TEST_GEN_FILES += map_hugetlb
TEST_GEN_FILES += map_populate
+ifneq (,$(filter $(ARCH),arm64 riscv riscv64 x86 x86_64))
TEST_GEN_FILES += memfd_secret
+endif
TEST_GEN_FILES += migration
TEST_GEN_FILES += mkdirty
TEST_GEN_FILES += mlock-random-test
diff --git a/tools/testing/selftests/mm/run_vmtests.sh b/tools/testing/selftests/mm/run_vmtests.sh
index 03ac4f2e1cce6..36045edb10dea 100755
--- a/tools/testing/selftests/mm/run_vmtests.sh
+++ b/tools/testing/selftests/mm/run_vmtests.sh
@@ -374,8 +374,11 @@ CATEGORY="hmm" run_test bash ./test_hmm.sh smoke
# MADV_POPULATE_READ and MADV_POPULATE_WRITE tests
CATEGORY="madv_populate" run_test ./madv_populate
+if [ -x ./memfd_secret ]
+then
(echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope 2>&1) | tap_prefix
CATEGORY="memfd_secret" run_test ./memfd_secret
+fi
# KSM KSM_MERGE_TIME_HUGE_PAGES test with size of 100
CATEGORY="ksm" run_test ./ksm_tests -H -s 100
--
2.39.2
If adding multiple config files to the merge_config.sh script and
rust/config is the fist one, then the last config fragment in this file
and the first config fragment in the second file wont be set, since
there isn't a newline in this file, so those two fragements end up at
the same row like:
CONFIG_SAMPLE_RUST_PRINT=mCONFIG_FRAGMENT=y
And non of those will be enabled when running 'olddefconfig' after.
Fixing the issue by adding a newline to the file.
Signed-off-by: Anders Roxell <anders.roxell(a)linaro.org>
---
tools/testing/selftests/rust/config | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/testing/selftests/rust/config b/tools/testing/selftests/rust/config
index b4002acd40bc..fa06cebae232 100644
--- a/tools/testing/selftests/rust/config
+++ b/tools/testing/selftests/rust/config
@@ -2,4 +2,4 @@ CONFIG_RUST=y
CONFIG_SAMPLES=y
CONFIG_SAMPLES_RUST=y
CONFIG_SAMPLE_RUST_MINIMAL=m
-CONFIG_SAMPLE_RUST_PRINT=m
\ No newline at end of file
+CONFIG_SAMPLE_RUST_PRINT=m
--
2.43.0
[1] mentions that memfd_secret is only supported on arm64, riscv, x86
and x86_64 for now. It doesn't support other architectures. I found the
build error on arm and decided to send the fix as it was creating noise
on KernelCI. Hence I'm adding condition that memfd_secret should only be
compiled on supported architectures.
Also check in run_vmtests script if memfd_secret binary is present
before executing it.
[1] https://lore.kernel.org/all/20210518072034.31572-7-rppt@kernel.org/
Cc: stable(a)vger.kernel.org
Fixes: 76fe17ef588a ("secretmem: test: add basic selftest for memfd_secret(2)")
Signed-off-by: Muhammad Usama Anjum <usama.anjum(a)collabora.com>
---
tools/testing/selftests/mm/Makefile | 2 ++
tools/testing/selftests/mm/run_vmtests.sh | 3 +++
2 files changed, 5 insertions(+)
diff --git a/tools/testing/selftests/mm/Makefile b/tools/testing/selftests/mm/Makefile
index 1a83b70e84535..4ea188be0588a 100644
--- a/tools/testing/selftests/mm/Makefile
+++ b/tools/testing/selftests/mm/Makefile
@@ -53,7 +53,9 @@ TEST_GEN_FILES += madv_populate
TEST_GEN_FILES += map_fixed_noreplace
TEST_GEN_FILES += map_hugetlb
TEST_GEN_FILES += map_populate
+ifneq (,$(filter $(ARCH),arm64 riscv riscv64 x86 x86_64))
TEST_GEN_FILES += memfd_secret
+endif
TEST_GEN_FILES += migration
TEST_GEN_FILES += mkdirty
TEST_GEN_FILES += mlock-random-test
diff --git a/tools/testing/selftests/mm/run_vmtests.sh b/tools/testing/selftests/mm/run_vmtests.sh
index 03ac4f2e1cce6..36045edb10dea 100755
--- a/tools/testing/selftests/mm/run_vmtests.sh
+++ b/tools/testing/selftests/mm/run_vmtests.sh
@@ -374,8 +374,11 @@ CATEGORY="hmm" run_test bash ./test_hmm.sh smoke
# MADV_POPULATE_READ and MADV_POPULATE_WRITE tests
CATEGORY="madv_populate" run_test ./madv_populate
+if [ -x ./memfd_secret ]
+then
(echo 0 | sudo tee /proc/sys/kernel/yama/ptrace_scope 2>&1) | tap_prefix
CATEGORY="memfd_secret" run_test ./memfd_secret
+fi
# KSM KSM_MERGE_TIME_HUGE_PAGES test with size of 100
CATEGORY="ksm" run_test ./ksm_tests -H -s 100
--
2.39.2
This small series includes fixes for creation of veth pairs for
networkless kernels & adds tests for turning the different network
interface features on and off in selftests/net/netdevice.sh script.
Changes in v5:
Rectify the syntax for ip add link.
Fix the veth_created condition check.
Changes in v4:
https://lore.kernel.org/all/20240807175717.7775-1-jain.abhinav177@gmail.com/
Move veth creation/removal to the main shell script.
Tested using vng on a networkless kernel and the script works, sample
output below the changes.
Changes in v3:
https://lore.kernel.org/all/20240614113240.41550-1-jain.abhinav177@gmail.co…
Add a check for netdev, create veth pair for testing.
Restore feature to its initial state.
Changes in v2:
https://lore.kernel.org/all/20240609132124.51683-1-jain.abhinav177@gmail.co…
Remove tail usage; use read to parse the features from temp file.
v1:
https://lore.kernel.org/all/20240606212714.27472-1-jain.abhinav177@gmail.co…
```
# selftests: net: netdevice.sh
# No valid network device found, creating veth pair
# PASS: veth0: set interface up
# PASS: veth0: set MAC address
# SKIP: veth0: set IP address
# PASS: veth0: ethtool list features
# PASS: veth0: Turned off feature: rx-checksumming
# PASS: veth0: Turned on feature: rx-checksumming
# PASS: veth0: Restore feature rx-checksumming to initial state on
# Actual changes:
# tx-checksum-ip-generic: off
# tx-tcp-segmentation: off [not requested]
....
....
....
# PASS: veth1: Restore feature tx-nocache-copy to initial state off
# PASS: veth1: Turned off feature: tx-vlan-stag-hw-insert
# PASS: veth1: Turned on feature: tx-vlan-stag-hw-insert
# PASS: veth1: Restore feature tx-vlan-stag-hw-insert to initial state on
# PASS: veth1: Turned off feature: rx-vlan-stag-hw-parse
# PASS: veth1: Turned on feature: rx-vlan-stag-hw-parse
# PASS: veth1: Restore feature rx-vlan-stag-hw-parse to initial state on
# PASS: veth1: Turned off feature: rx-gro-list
# PASS: veth1: Turned on feature: rx-gro-list
# PASS: veth1: Restore feature rx-gro-list to initial state off
# PASS: veth1: Turned off feature: rx-udp-gro-forwarding
# PASS: veth1: Turned on feature: rx-udp-gro-forwarding
# PASS: veth1: Restore feature rx-udp-gro-forwarding to initial state off
# Cannot get register dump: Operation not supported
# SKIP: veth1: ethtool dump not supported
# PASS: veth1: ethtool stats
# PASS: veth1: stop interface
# Removed veth pair
ok 12 selftests: net: netdevice.sh
```
Abhinav Jain (2):
selftests: net: Create veth pair for testing in networkless kernel
selftests: net: Add on/off checks for non-fixed features of interface
tools/testing/selftests/net/netdevice.sh | 53 +++++++++++++++++++++++-
1 file changed, 52 insertions(+), 1 deletion(-)
--
2.34.1
The current support for LLVM and clang in nolibc and its testsuite is
very limited.
* Various architectures plain do not compile
* The user *has* to specify "-Os" otherwise the program crashes
* Cross-compilation of the tests does not work
* Using clang is not wired up in run-tests.sh
This series extends this support.
Signed-off-by: Thomas Weißschuh <linux(a)weissschuh.net>
---
Changes in v2:
- Add support for all architectures
- powerpc: "selftests/nolibc: don't use libgcc when building with clang"
- mips: "tools/nolibc: mips: load current function to $t9"
- s390: "selftests/nolibc: use correct clang target for s390/powerz"
- Expand commit messages
- Use __nolibc_ prefix for custom macros
- Link to v1: https://lore.kernel.org/r/20240728-nolibc-llvm-v1-0-bc384269bc35@weissschuh…
---
Thomas Weißschuh (15):
tools/nolibc: arm: use clang-compatible asm syntax
tools/nolibc: mips: load current function to $t9
tools/nolibc: powerpc: limit stack-protector workaround to GCC
tools/nolibc: compiler: introduce __nolibc_has_attribute()
tools/nolibc: move entrypoint specifics to compiler.h
tools/nolibc: compiler: use attribute((naked)) if available
selftests/nolibc: report failure if no testcase passed
selftests/nolibc: avoid passing NULL to printf("%s")
selftests/nolibc: determine $(srctree) first
selftests/nolibc: add support for LLVM= parameter
selftests/nolibc: add cc-option compatible with clang cross builds
selftests/nolibc: run-tests.sh: avoid overwriting CFLAGS_EXTRA
selftests/nolibc: don't use libgcc when building with clang
selftests/nolibc: use correct clang target for s390/powerz
selftests/nolibc: run-tests.sh: allow building through LLVM
tools/include/nolibc/arch-aarch64.h | 4 +--
tools/include/nolibc/arch-arm.h | 8 +++---
tools/include/nolibc/arch-i386.h | 4 +--
tools/include/nolibc/arch-loongarch.h | 4 +--
tools/include/nolibc/arch-mips.h | 8 ++++--
tools/include/nolibc/arch-powerpc.h | 6 ++--
tools/include/nolibc/arch-riscv.h | 4 +--
tools/include/nolibc/arch-s390.h | 4 +--
tools/include/nolibc/arch-x86_64.h | 4 +--
tools/include/nolibc/compiler.h | 24 +++++++++++-----
tools/testing/selftests/nolibc/Makefile | 41 +++++++++++++++++++---------
tools/testing/selftests/nolibc/nolibc-test.c | 4 +--
tools/testing/selftests/nolibc/run-tests.sh | 16 ++++++++---
13 files changed, 83 insertions(+), 48 deletions(-)
---
base-commit: ae1f550efc11eaf1496c431d9c6e784cb49124c5
change-id: 20240727-nolibc-llvm-3fad68590d4c
Best regards,
--
Thomas Weißschuh <linux(a)weissschuh.net>
Hello Andrii Nakryiko,
This is a semi-automatic email about new static checker warnings.
Commit 8863238993e2 ("selftests/bpf: BPF register range bounds
tester") from Nov 11, 2023, leads to the following Smatch complaint:
./tools/testing/selftests/bpf/prog_tests/reg_bounds.c:1121 parse_reg_state()
warn: variable dereferenced before check 'p' (see line 1119)
./tools/testing/selftests/bpf/prog_tests/reg_bounds.c
1118 p = strpbrk(p, ",)");
1119 if (*p == ')')
1120 break;
1121 if (p)
Was this NULL check supposed to be checking for the NUL terminator?
1122 p++;
1123 }
regards,
dan carpenter
The "initial_nr_hugepages" variable is unsigned long so it takes up to
20 characters to print, plus 1 more character for the NUL terminator.
Unfortunately, this buffer is not quite large enough for the terminator
to fit. Also use snprintf() for a belt and suspenders approach.
Fixes: fb9293b6b015 ("selftests/mm: compaction_test: fix bogus test success and reduce probability of OOM-killer invocation")
Signed-off-by: Dan Carpenter <dan.carpenter(a)linaro.org>
---
tools/testing/selftests/mm/compaction_test.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/tools/testing/selftests/mm/compaction_test.c b/tools/testing/selftests/mm/compaction_test.c
index e140558e6f53..2c3a0eb6b22d 100644
--- a/tools/testing/selftests/mm/compaction_test.c
+++ b/tools/testing/selftests/mm/compaction_test.c
@@ -89,9 +89,10 @@ int check_compaction(unsigned long mem_free, unsigned long hugepage_size,
int fd, ret = -1;
int compaction_index = 0;
char nr_hugepages[20] = {0};
- char init_nr_hugepages[20] = {0};
+ char init_nr_hugepages[24] = {0};
- sprintf(init_nr_hugepages, "%lu", initial_nr_hugepages);
+ snprintf(init_nr_hugepages, sizeof(init_nr_hugepages),
+ "%lu", initial_nr_hugepages);
/* We want to test with 80% of available memory. Else, OOM killer comes
in to play */
--
2.43.0
The relative RPATH ("./") supplied to linker options in CFLAGS is resolved
relative to current working directory and not the executable directory,
which will lead in incorrect resolution when the test executables are run
from elsewhere. Changing it to $ORIGIN makes it resolve relative
to the directory in which the executables reside, which is supposedly
the desired behaviour.
Discovered by the check-rpaths script[1][2] that checks for insecure
RPATH/RUNPATH[3], such as relative directories, during an attempt
to package BPF selftests for later use in CI:
ERROR 0004: file '/usr/libexec/kselftests/bpf/urandom_read' contains an insecure runpath '.' in [.]
[1] https://github.com/rpm-software-management/rpm/blob/master/scripts/check-rp…
[2] https://github.com/rpm-software-management/rpm/blob/master/scripts/check-rp…
[3] https://cwe.mitre.org/data/definitions/426.html
Signed-off-by: Eugene Syromiatnikov <esyr(a)redhat.com>
---
tools/testing/selftests/alsa/Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/testing/selftests/alsa/Makefile b/tools/testing/selftests/alsa/Makefile
index c1ce39874e2b..0f204da9ea8e 100644
--- a/tools/testing/selftests/alsa/Makefile
+++ b/tools/testing/selftests/alsa/Makefile
@@ -6,7 +6,7 @@ LDLIBS += $(shell pkg-config --libs alsa)
ifeq ($(LDLIBS),)
LDLIBS += -lasound
endif
-CFLAGS += -L$(OUTPUT) -Wl,-rpath=./
+CFLAGS += -L$(OUTPUT) -Wl,-rpath=\$$ORIGIN/
LDLIBS+=-lpthread
--
2.28.0
Hello Andrii Nakryiko,
This is a semi-automatic email about new static checker warnings.
Commit c381203eadb7 ("selftests/bpf: add trusted global subprog arg
tests") from Jan 29, 2024, leads to the following Smatch complaint:
./tools/testing/selftests/bpf/progs/verifier_global_ptr_args.c:88 trusted_task_arg_nonnull_fail2()
warn: variable dereferenced before check 'nullable' (see line 86)
./tools/testing/selftests/bpf/progs/verifier_global_ptr_args.c
85 /* should fail, PTR_TO_BTF_ID_OR_NULL */
86 res = subprog_trusted_task_nonnull(nullable);
^^^^^^^^
This is dereferenced
87
88 if (nullable)
^^^^^^^^
NULL check is too late
89 bpf_task_release(nullable);
90
regards,
dan carpenter
Hello,
KernelCI is hosting a bi-weekly call on Thursday to discuss improvements
to existing upstream tests, the development of new tests to increase
kernel testing coverage, and the enablement of these tests in KernelCI.
Below is a list of the tests the community has been working on and their
latest status updates, as discussed in the last meeting held on
2024-08-08:
*KTAP performance counters*
- Upcoming presentation @LPC2024 by Tim Bird on adding benchmark results
to KTAP: https://lpc.events/event/18/contributions/1791/
- Proposing new system to handle benchmark data, composed of 3 parts:
adding syntax to KTAP to support benchmark values, using a set of
external criteria for interpreting benchmark results, an automated tool
to determine and set the reference values used in these criteria.
- One related topic for discussion is where to store the reference files
and the test configuration, including all details that might impact the
results.
*Missing devices kselftest*
- Proposing new kselftest to report devices that go missing in the system:
https://lore.kernel.org/all/20240724-kselftest-dev-exist-v1-1-9bc21aa761b5@…
- Received feedback on the usability of the test and main unknowns about
management of reference files in tests
*Boot time test*
- RFC:
https://lore.kernel.org/all/20240725110622.96301-1-laura.nao@collabora.com/…
- Got feedback on the potential use of scripts from pm-graph. The
bootgraph.py script could be adapted for the test with some
modifications, such as adding support for machine-readable output and
different ftrace configurations.
*Suspend/resume in cpufreq kselftest*
- Sent v2 for patch adding RTC wakeup alarm in the cpufreq kselftest:
https://lore.kernel.org/lkml/20240715192634.19272-1-shreeya.patel@collabora…
- Looking into using the sleepgraph.py script from pm-graph to calculate
suspend/hibernation and resume time:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/too…
- Would require additional configs to be enabled in an automated
environment
*TAP conformance in kselftests and other fixes*
- Some dead selftests were removed:
https://lore.kernel.org/all/20240725110817.659099-1-usama.anjum@collabora.c…
and
https://lore.kernel.org/all/20240725121212.808206-1-usama.anjum@collabora.c…
- Bitmap test module conversion to KUnit was not accepted:
https://lore.kernel.org/lkml/49108735-c776-4b6f-8264-62a827dd7b26@collabora…
- Fixing error when kvm suite is built for unsupported architecure:
https://lore.kernel.org/kvm/c2aaa06e-e86d-4af9-bce4-6067e53cdf39@collabora.…
Please reply to this thread if you'd like to join the call or discuss any
of the topics further. We look forward to collaborating with the community
to improve upstream tests and expand coverage to more areas of interest
within the kernel.
Best regards,
Laura Nao
From: Allison Henderson <allison.henderson(a)oracle.com>
Hi All,
This series is a new selftest that Vegard, Chuck and myself have been
working on to provide some test coverage for rds. I've modified the
scripts to include the feedback from the last version, but let me know
if there's anything missed. Questions and comments appreciated.
Thanks everyone!
Allison
Changes in v2:
- Removed qemu vm creation and related code
- Updated README.txt with examples of running the test with virtme
- Removed init.sh. run.sh now directly calls test.py
- Some clean up done with the return code handling since there is no
vm between the scripts anymore
- Imported ip python function in
tools/testing/selftests/net/lib/py/utils.py into test.py
- Adapted test.py to use the imported ip function, and removed the
local ip wrapper
- Some line wrap clean up
- Link to v1:
https://lore.kernel.org/netdev/20240626012834.5678-3-allison.henderson@orac…
Vegard Nossum (3):
.gitignore: add .gcda files
net: rds: add option for GCOV profiling
selftests: rds: add testing infrastructure
.gitignore | 1 +
Documentation/dev-tools/gcov.rst | 11 +
MAINTAINERS | 1 +
net/rds/Kconfig | 9 +
net/rds/Makefile | 5 +
tools/testing/selftests/Makefile | 1 +
tools/testing/selftests/net/rds/Makefile | 12 +
tools/testing/selftests/net/rds/README.txt | 41 ++++
tools/testing/selftests/net/rds/config.sh | 53 +++++
tools/testing/selftests/net/rds/run.sh | 224 ++++++++++++++++++
tools/testing/selftests/net/rds/test.py | 262 +++++++++++++++++++++
11 files changed, 620 insertions(+)
create mode 100644 tools/testing/selftests/net/rds/Makefile
create mode 100644 tools/testing/selftests/net/rds/README.txt
create mode 100755 tools/testing/selftests/net/rds/config.sh
create mode 100755 tools/testing/selftests/net/rds/run.sh
create mode 100644 tools/testing/selftests/net/rds/test.py
--
2.25.1
v17: https://patchwork.kernel.org/project/netdevbpf/list/?series=869900&state=*
====
v16 also got a very thorough review and some testing (thanks again!).
Thes version addresses all the concerns reported on v15, in terms of
feedback and issues reported.
Major changes:
- Use ASSERT_RTNL.
- Moved around some of the page_pool helpers definitions so I can hide
some netmem helpers in private files as Jakub suggested.
- Don't make every net_iov hold a ref on the binding as Jakub suggested.
- Fix issue reported by Taehee where we access queues after they have
been freed.
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v17/
v16: https://patchwork.kernel.org/project/netdevbpf/list/?series=866353&state=*
====
v15 got a thorough review and some testing, and this version addresses almost
all the feedback. Some more minor comments where the authors said it
could be done later, I left out.
Major changes:
- Addition of dma-buf introspection to page-pool-get and queue-get.
- Fixes to selftests suggested by Taehee.
- Fixes to documentation suggested by Donald.
- A couple of suggestions and fixes to TCP patches by Eric and David.
- Fixes to number assignements suggested by Arnd.
- Use rtnl_lock()ing to guard against queue reconfiguration while the
page_pool initialization is happening. (Jakub).
- Fixes to a few warnings reproduced by Taehee.
- Fixes to dma-buf binding suggested by Taehee and Jakub.
- Fixes to netlink UAPI suggested by Jakub
- Applied a number of Reviewed-bys and Acked-bys (including ones I lost
from v13+).
Full devmem TCP changes including the full GVE driver implementation is
here:
https://github.com/mina/linux/commits/tcpdevmem-v16/
One caveat: Taehee reproduced a KASAN warning and reported it here:
https://lore.kernel.org/netdev/CAMArcTUdCxOBYGF3vpbq=eBvqZfnc44KBaQTN7H-wqd…
I estimate the issue to be minor and easily fixable:
https://lore.kernel.org/netdev/CAHS8izNgaqC--GGE2xd85QB=utUnOHmioCsDd1TNxJW…
I hope to be able to follow up with a fix to net tree as net-next closes
imminently, but if this iteration doesn't make it in, I will repost with
a fix squashed after net-next reopens, no problem.
v15: https://patchwork.kernel.org/project/netdevbpf/list/?series=865481&state=*
====
No material changes in this version, only a fix to linking against
libynl.a from the last version. Per Jakub's instructions I've pulled one
of his patches into this series, and now use the new libynl.a correctly,
I hope.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v15/
v14: https://patchwork.kernel.org/project/netdevbpf/list/?series=865135&archive=…
====
No material changes in this version. Only rebase and re-verification on
top of net-next. v13, I think, raced with commit ebad6d0334793
("net/ipv4: Use nested-BH locking for ipv4_tcp_sk.") being merged to
net-next that caused a patchwork failure to apply. This series should
apply cleanly on commit c4532232fa2a4 ("selftests: net: remove unneeded
IP_GRE config").
I did not wait the customary 24hr as Jakub said it's OK to repost as soon
as I build test the rebased version:
https://lore.kernel.org/netdev/20240625075926.146d769d@kernel.org/
v13: https://patchwork.kernel.org/project/netdevbpf/list/?series=861406&archive=…
====
Major changes:
--------------
This iteration addresses Pavel's review comments, applies his
reviewed-by's, and seeks to fix the patchwork build error (sorry!).
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v13/
v12: https://patchwork.kernel.org/project/netdevbpf/list/?series=859747&state=*
====
Major changes:
--------------
This iteration only addresses one minor comment from Pavel with regards
to the trace printing of netmem, and the patchwork build error
introduced in v11 because I missed doing an allmodconfig build, sorry.
Other than that v11, AFAICT, received no feedback. There is one
discussion about how the specifics of plugging io uring memory through
the page pool, but not relevant to content in this particular patchset,
AFAICT.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v12/
v11: https://patchwork.kernel.org/project/netdevbpf/list/?series=857457&state=*
====
Major Changes:
--------------
v11 addresses feedback received in v10. The major change is the removal
of the memory provider ops as requested by Christoph. We still
accomplish the same thing, but utilizing direct function calls with if
statements rather than generic ops.
Additionally address sparse warnings, bugs and review comments from
folks that reviewed.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v11/
Detailed changelog:
-------------------
- Fixes in netdev_rx_queue_restart() from Pavel & David.
- Remove commit e650e8c3a36f5 ("net: page_pool: create hooks for
custom page providers") from the series to address Christoph's
feedback and rebased other patches on the series on this change.
- Fixed build errors with CONFIG_DMA_SHARED_BUFFER &&
!CONFIG_GENERIC_ALLOCATOR build.
- Fixed sparse warnings pointed out by Paolo.
- Drop unnecessary gro_pull_from_frag0 checks.
- Added Bagas reviewed-by to docs.
Cc: Bagas Sanjaya <bagasdotme(a)gmail.com>
Cc: Steven Rostedt <rostedt(a)goodmis.org>
Cc: Christoph Hellwig <hch(a)infradead.org>
Cc: Nikolay Aleksandrov <razor(a)blackwall.org>
Cc: Taehee Yoo <ap420073(a)gmail.com>
Cc: Donald Hunter <donald.hunter(a)gmail.com>
v10: https://patchwork.kernel.org/project/netdevbpf/list/?series=852422&state=*
====
Major Changes:
--------------
v9 was sent right before the merge window closed (sorry!). v10 is almost
a re-send of the series now that the merge window re-opened. Only
rebased to latest net-next and addressed some minor iterative comments
received on v9.
As usual, the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v10/
Detailed changelog:
-------------------
- Fixed tokens leaking in DONTNEED setsockopt (Nikolay).
- Moved net_iov_dma_addr() to devmem.c and made it a devmem specific
helpers (David).
- Rename hook alloc_pages to alloc_netmems as alloc_pages is now
preprocessor macro defined and causes a build error.
v9:
===
Major Changes:
--------------
GVE queue API has been merged. Submitting this version as non-RFC after
rebasing on top of the merged API, and dropped the out of tree queue API
I was carrying on github. Addressed the little feedback v8 has received.
Detailed changelog:
------------------
- Added new patch from David Wei to this series for
netdev_rx_queue_restart()
- Fixed sparse error.
- Removed CONFIG_ checks in netmem_is_net_iov()
- Flipped skb->readable to skb->unreadable
- Minor fixes to selftests & docs.
RFC v8:
=======
Major Changes:
--------------
- Fixed build error generated by patch-by-patch build.
- Applied docs suggestions from Randy.
RFC v7:
=======
Major Changes:
--------------
This revision largely rebases on top of net-next and addresses the feedback
RFCv6 received from folks, namely Jakub, Yunsheng, Arnd, David, & Pavel.
The series remains in RFC because the queue-API ndos defined in this
series are not yet implemented. I have a GVE implementation I carry out
of tree for my testing. A upstreamable GVE implementation is in the
works. Aside from that, in my estimation all the patches are ready for
review/merge. Please do take a look.
As usual the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v7/
Detailed changelog:
- Use admin-perm in netlink API.
- Addressed feedback from Jakub with regards to netlink API
implementation.
- Renamed devmem.c functions to something more appropriate for that
file.
- Improve the performance seen through the page_pool benchmark.
- Fix the value definition of all the SO_DEVMEM_* uapi.
- Various fixes to documentation.
Perf - page-pool benchmark:
---------------------------
Improved performance of bench_page_pool_simple.ko tests compared to v6:
https://pastebin.com/raw/v5dYRg8L
net-next base: 8 cycle fast path.
RFC v6: 10 cycle fast path.
RFC v7: 9 cycle fast path.
RFC v7 with CONFIG_DMA_SHARED_BUFFER disabled: 8 cycle fast path,
same as baseline.
Perf - Devmem TCP benchmark:
---------------------
Perf is about the same regardless of the changes in v7, namely the
removal of the static_branch_unlikely to improve the page_pool benchmark
performance:
189/200gbps bi-directional throughput with RX devmem TCP and regular TCP
TX i.e. ~95% line rate.
RFC v6:
=======
Major Changes:
--------------
This revision largely rebases on top of net-next and addresses the little
feedback RFCv5 received.
The series remains in RFC because the queue-API ndos defined in this
series are not yet implemented. I have a GVE implementation I carry out
of tree for my testing. A upstreamable GVE implementation is in the
works. Aside from that, in my estimation all the patches are ready for
review/merge. Please do take a look.
As usual the full devmem TCP changes including the full GVE driver
implementation is here:
https://github.com/mina/linux/commits/tcpdevmem-v6/
This version also comes with some performance data recorded in the cover
letter (see below changelog).
Detailed changelog:
- Rebased on top of the merged netmem_ref changes.
- Converted skb->dmabuf to skb->readable (Pavel). Pavel's original
suggestion was to remove the skb->dmabuf flag entirely, but when I
looked into it closely, I found the issue that if we remove the flag
we have to dereference the shinfo(skb) pointer to obtain the first
frag to tell whether an skb is readable or not. This can cause a
performance regression if it dirties the cache line when the
shinfo(skb) was not really needed. Instead, I converted the skb->dmabuf
flag into a generic skb->readable flag which can be re-used by io_uring
0-copy RX.
- Squashed a few locking optimizations from Eric Dumazet in the RX path
and the DEVMEM_DONTNEED setsockopt.
- Expanded the tests a bit. Added validation for invalid scenarios and
added some more coverage.
Perf - page-pool benchmark:
---------------------------
bench_page_pool_simple.ko tests with and without these changes:
https://pastebin.com/raw/ncHDwAbn
AFAIK the number that really matters in the perf tests is the
'tasklet_page_pool01_fast_path Per elem'. This one measures at about 8
cycles without the changes but there is some 1 cycle noise in some
results.
With the patches this regresses to 9 cycles with the changes but there
is 1 cycle noise occasionally running this test repeatedly.
Lastly I tried disable the static_branch_unlikely() in
netmem_is_net_iov() check. To my surprise disabling the
static_branch_unlikely() check reduces the fast path back to 8 cycles,
but the 1 cycle noise remains.
Perf - Devmem TCP benchmark:
---------------------
189/200gbps bi-directional throughput with RX devmem TCP and regular TCP
TX i.e. ~95% line rate.
Major changes in RFC v5:
========================
1. Rebased on top of 'Abstract page from net stack' series and used the
new netmem type to refer to LSB set pointers instead of re-using
struct page.
2. Downgraded this series back to RFC and called it RFC v5. This is
because this series is now dependent on 'Abstract page from net
stack'[1] and the queue API. Both are removed from the series to
reduce the patch # and those bits are fairly independent or
pre-requisite work.
3. Reworked the page_pool devmem support to use netmem and for some
more unified handling.
4. Reworked the reference counting of net_iov (renamed from
page_pool_iov) to use pp_ref_count for refcounting.
The full changes including the dependent series and GVE page pool
support is here:
https://github.com/mina/linux/commits/tcpdevmem-rfcv5/
[1] https://patchwork.kernel.org/project/netdevbpf/list/?series=810774
Major changes in v1:
====================
1. Implemented MVP queue API ndos to remove the userspace-visible
driver reset.
2. Fixed issues in the napi_pp_put_page() devmem frag unref path.
3. Removed RFC tag.
Many smaller addressed comments across all the patches (patches have
individual change log).
Full tree including the rest of the GVE driver changes:
https://github.com/mina/linux/commits/tcpdevmem-v1
Changes in RFC v3:
==================
1. Pulled in the memory-provider dependency from Jakub's RFC[1] to make the
series reviewable and mergeable.
2. Implemented multi-rx-queue binding which was a todo in v2.
3. Fix to cmsg handling.
The sticking point in RFC v2[2] was the device reset required to refill
the device rx-queues after the dmabuf bind/unbind. The solution
suggested as I understand is a subset of the per-queue management ops
Jakub suggested or similar:
https://lore.kernel.org/netdev/20230815171638.4c057dcd@kernel.org/
This is not addressed in this revision, because:
1. This point was discussed at netconf & netdev and there is openness to
using the current approach of requiring a device reset.
2. Implementing individual queue resetting seems to be difficult for my
test bed with GVE. My prototype to test this ran into issues with the
rx-queues not coming back up properly if reset individually. At the
moment I'm unsure if it's a mistake in the POC or a genuine issue in
the virtualization stack behind GVE, which currently doesn't test
individual rx-queue restart.
3. Our usecases are not bothered by requiring a device reset to refill
the buffer queues, and we'd like to support NICs that run into this
limitation with resetting individual queues.
My thought is that drivers that have trouble with per-queue configs can
use the support in this series, while drivers that support new netdev
ops to reset individual queues can automatically reset the queue as
part of the dma-buf bind/unbind.
The same approach with device resets is presented again for consideration
with other sticking points addressed.
This proposal includes the rx devmem path only proposed for merge. For a
snapshot of my entire tree which includes the GVE POC page pool support &
device memory support:
https://github.com/torvalds/linux/compare/master...mina:linux:tcpdevmem-v3
[1] https://lore.kernel.org/netdev/f8270765-a27b-6ccf-33ea-cda097168d79@redhat.…
[2] https://lore.kernel.org/netdev/CAHS8izOVJGJH5WF68OsRWFKJid1_huzzUK+hpKbLcL4…
Changes in RFC v2:
==================
The sticking point in RFC v1[1] was the dma-buf pages approach we used to
deliver the device memory to the TCP stack. RFC v2 is a proof-of-concept
that attempts to resolve this by implementing scatterlist support in the
networking stack, such that we can import the dma-buf scatterlist
directly. This is the approach proposed at a high level here[2].
Detailed changes:
1. Replaced dma-buf pages approach with importing scatterlist into the
page pool.
2. Replace the dma-buf pages centric API with a netlink API.
3. Removed the TX path implementation - there is no issue with
implementing the TX path with scatterlist approach, but leaving
out the TX path makes it easier to review.
4. Functionality is tested with this proposal, but I have not conducted
perf testing yet. I'm not sure there are regressions, but I removed
perf claims from the cover letter until they can be re-confirmed.
5. Added Signed-off-by: contributors to the implementation.
6. Fixed some bugs with the RX path since RFC v1.
Any feedback welcome, but specifically the biggest pending questions
needing feedback IMO are:
1. Feedback on the scatterlist-based approach in general.
2. Netlink API (Patch 1 & 2).
3. Approach to handle all the drivers that expect to receive pages from
the page pool (Patch 6).
[1] https://lore.kernel.org/netdev/dfe4bae7-13a0-3c5d-d671-f61b375cb0b4@gmail.c…
[2] https://lore.kernel.org/netdev/CAHS8izPm6XRS54LdCDZVd0C75tA1zHSu6jLVO8nzTLX…
==================
* TL;DR:
Device memory TCP (devmem TCP) is a proposal for transferring data to and/or
from device memory efficiently, without bouncing the data to a host memory
buffer.
* Problem:
A large amount of data transfers have device memory as the source and/or
destination. Accelerators drastically increased the volume of such transfers.
Some examples include:
- ML accelerators transferring large amounts of training data from storage into
GPU/TPU memory. In some cases ML training setup time can be as long as 50% of
TPU compute time, improving data transfer throughput & efficiency can help
improving GPU/TPU utilization.
- Distributed training, where ML accelerators, such as GPUs on different hosts,
exchange data among them.
- Distributed raw block storage applications transfer large amounts of data with
remote SSDs, much of this data does not require host processing.
Today, the majority of the Device-to-Device data transfers the network are
implemented as the following low level operations: Device-to-Host copy,
Host-to-Host network transfer, and Host-to-Device copy.
The implementation is suboptimal, especially for bulk data transfers, and can
put significant strains on system resources, such as host memory bandwidth,
PCIe bandwidth, etc. One important reason behind the current state is the
kernel’s lack of semantics to express device to network transfers.
* Proposal:
In this patch series we attempt to optimize this use case by implementing
socket APIs that enable the user to:
1. send device memory across the network directly, and
2. receive incoming network packets directly into device memory.
Packet _payloads_ go directly from the NIC to device memory for receive and from
device memory to NIC for transmit.
Packet _headers_ go to/from host memory and are processed by the TCP/IP stack
normally. The NIC _must_ support header split to achieve this.
Advantages:
- Alleviate host memory bandwidth pressure, compared to existing
network-transfer + device-copy semantics.
- Alleviate PCIe BW pressure, by limiting data transfer to the lowest level
of the PCIe tree, compared to traditional path which sends data through the
root complex.
* Patch overview:
** Part 1: netlink API
Gives user ability to bind dma-buf to an RX queue.
** Part 2: scatterlist support
Currently the standard for device memory sharing is DMABUF, which doesn't
generate struct pages. On the other hand, networking stack (skbs, drivers, and
page pool) operate on pages. We have 2 options:
1. Generate struct pages for dmabuf device memory, or,
2. Modify the networking stack to process scatterlist.
Approach #1 was attempted in RFC v1. RFC v2 implements approach #2.
** part 3: page pool support
We piggy back on page pool memory providers proposal:
https://github.com/kuba-moo/linux/tree/pp-providers
It allows the page pool to define a memory provider that provides the
page allocation and freeing. It helps abstract most of the device memory
TCP changes from the driver.
** part 4: support for unreadable skb frags
Page pool iovs are not accessible by the host; we implement changes
throughput the networking stack to correctly handle skbs with unreadable
frags.
** Part 5: recvmsg() APIs
We define user APIs for the user to send and receive device memory.
Not included with this series is the GVE devmem TCP support, just to
simplify the review. Code available here if desired:
https://github.com/mina/linux/tree/tcpdevmem
This series is built on top of net-next with Jakub's pp-providers changes
cherry-picked.
* NIC dependencies:
1. (strict) Devmem TCP require the NIC to support header split, i.e. the
capability to split incoming packets into a header + payload and to put
each into a separate buffer. Devmem TCP works by using device memory
for the packet payload, and host memory for the packet headers.
2. (optional) Devmem TCP works better with flow steering support & RSS support,
i.e. the NIC's ability to steer flows into certain rx queues. This allows the
sysadmin to enable devmem TCP on a subset of the rx queues, and steer
devmem TCP traffic onto these queues and non devmem TCP elsewhere.
The NIC I have access to with these properties is the GVE with DQO support
running in Google Cloud, but any NIC that supports these features would suffice.
I may be able to help reviewers bring up devmem TCP on their NICs.
* Testing:
The series includes a udmabuf kselftest that show a simple use case of
devmem TCP and validates the entire data path end to end without
a dependency on a specific dmabuf provider.
** Test Setup
Kernel: net-next with this series and memory provider API cherry-picked
locally.
Hardware: Google Cloud A3 VMs.
NIC: GVE with header split & RSS & flow steering support.
Cc: Pavel Begunkov <asml.silence(a)gmail.com>
Cc: David Wei <dw(a)davidwei.uk>
Cc: Jason Gunthorpe <jgg(a)ziepe.ca>
Cc: Yunsheng Lin <linyunsheng(a)huawei.com>
Cc: Shailend Chand <shailend(a)google.com>
Cc: Harshitha Ramamurthy <hramamurthy(a)google.com>
Cc: Shakeel Butt <shakeel.butt(a)linux.dev>
Cc: Jeroen de Borst <jeroendb(a)google.com>
Cc: Praveen Kaligineedi <pkaligineedi(a)google.com>
Mina Almasry (14):
netdev: add netdev_rx_queue_restart()
net: netdev netlink api to bind dma-buf to a net device
netdev: support binding dma-buf to netdevice
netdev: netdevice devmem allocator
page_pool: move dmaddr helpers to .c file
page_pool: devmem support
memory-provider: dmabuf devmem memory provider
net: support non paged skb frags
net: add support for skbs with unreadable frags
tcp: RX path for devmem TCP
net: add SO_DEVMEM_DONTNEED setsockopt to release RX frags
net: add devmem TCP documentation
selftests: add ncdevmem, netcat for devmem TCP
netdev: add dmabuf introspection
Documentation/netlink/specs/netdev.yaml | 61 +++
Documentation/networking/devmem.rst | 269 ++++++++++++
Documentation/networking/index.rst | 1 +
arch/alpha/include/uapi/asm/socket.h | 6 +
arch/mips/include/uapi/asm/socket.h | 6 +
arch/parisc/include/uapi/asm/socket.h | 6 +
arch/sparc/include/uapi/asm/socket.h | 6 +
include/linux/skbuff.h | 61 ++-
include/linux/skbuff_ref.h | 9 +-
include/linux/socket.h | 1 +
include/net/devmem.h | 128 ++++++
include/net/mp_dmabuf_devmem.h | 44 ++
include/net/netdev_rx_queue.h | 5 +
include/net/netmem.h | 164 +++++++-
include/net/page_pool/helpers.h | 42 +-
include/net/page_pool/types.h | 8 +
include/net/sock.h | 2 +
include/net/tcp.h | 5 +-
include/trace/events/page_pool.h | 12 +-
include/uapi/asm-generic/socket.h | 6 +
include/uapi/linux/netdev.h | 13 +
include/uapi/linux/uio.h | 17 +
net/core/Makefile | 3 +-
net/core/datagram.c | 6 +
net/core/dev.c | 7 +-
net/core/devmem.c | 378 +++++++++++++++++
net/core/gro.c | 3 +-
net/core/netdev-genl-gen.c | 23 +
net/core/netdev-genl-gen.h | 6 +
net/core/netdev-genl.c | 111 +++++
net/core/netdev_rx_queue.c | 74 ++++
net/core/netmem_priv.h | 36 ++
net/core/page_pool.c | 147 +++++--
net/core/page_pool_priv.h | 3 +
net/core/page_pool_user.c | 4 +
net/core/skbuff.c | 77 +++-
net/core/sock.c | 68 +++
net/ipv4/esp4.c | 3 +-
net/ipv4/tcp.c | 261 +++++++++++-
net/ipv4/tcp_input.c | 13 +-
net/ipv4/tcp_ipv4.c | 16 +
net/ipv4/tcp_minisocks.c | 2 +
net/ipv4/tcp_output.c | 5 +-
net/ipv6/esp6.c | 3 +-
net/packet/af_packet.c | 4 +-
tools/include/uapi/linux/netdev.h | 13 +
tools/testing/selftests/net/.gitignore | 1 +
tools/testing/selftests/net/Makefile | 9 +
tools/testing/selftests/net/ncdevmem.c | 536 ++++++++++++++++++++++++
49 files changed, 2563 insertions(+), 121 deletions(-)
create mode 100644 Documentation/networking/devmem.rst
create mode 100644 include/net/devmem.h
create mode 100644 include/net/mp_dmabuf_devmem.h
create mode 100644 net/core/devmem.c
create mode 100644 net/core/netdev_rx_queue.c
create mode 100644 net/core/netmem_priv.h
create mode 100644 tools/testing/selftests/net/ncdevmem.c
--
2.46.0.rc1.232.g9752f9e123-goog
The relative RPATH ("./") supplied to linker options in CFLAGS is resolved
relative to current working directory and not the executable directory,
which will lead in incorrect resolution when the test executable is run
from elsewhere. Changing it to $ORIGIN makes it resolve relative
to the directory in which the executable resides, which is supposedly
the desired behaviour.
Discovered by the check-rpaths script[1][2] that checks for insecure
RPATH/RUNPATH[3], such as relative directories, during an attempt
to package BPF selftests for later use in CI:
ERROR 0004: file '/usr/libexec/kselftests/bpf/urandom_read' contains an insecure runpath '.' in [.]
[1] https://github.com/rpm-software-management/rpm/blob/master/scripts/check-rp…
[2] https://github.com/rpm-software-management/rpm/blob/master/scripts/check-rp…
[3] https://cwe.mitre.org/data/definitions/426.html
Signed-off-by: Eugene Syromiatnikov <esyr(a)redhat.com>
---
tools/testing/selftests/bpf/Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile
index dd49c1d23a60..6a3dc9b99159 100644
--- a/tools/testing/selftests/bpf/Makefile
+++ b/tools/testing/selftests/bpf/Makefile
@@ -241,7 +241,7 @@ $(OUTPUT)/urandom_read: urandom_read.c urandom_read_aux.c $(OUTPUT)/liburandom_r
$(filter-out -static,$(CFLAGS) $(LDFLAGS)) $(filter %.c,$^) \
-lurandom_read $(filter-out -static,$(LDLIBS)) -L$(OUTPUT) \
-fuse-ld=$(LLD) -Wl,-znoseparate-code -Wl,--build-id=sha1 \
- -Wl,-rpath=. -o $@
+ -Wl,-rpath=\$$ORIGIN/ -o $@
$(OUTPUT)/sign-file: ../../../../scripts/sign-file.c
$(call msg,SIGN-FILE,,$@)
--
2.28.0
The relative RPATH ("./") supplied to linker options in CFLAGS is resolved
relative to current working directory and not the executable directory,
which will lead in incorrect resolution when the test executables are run
from elsewhere. However, the sole sched test (cs_prctl_test)
does not require any locally-built libraries to run, so the RPATH
directive can be removed.
Discovered by the /usr/lib/rpm/check-rpaths script[1][2] that checks
for insecure RPATH/RUNPATH[3], such as containing relative directories,
during an attempt to package BPF selftests for later use in CI:
ERROR 0004: file '/usr/libexec/kselftests/bpf/urandom_read' contains an insecure runpath '.' in [.]
[1] https://github.com/rpm-software-management/rpm/blob/master/scripts/check-rp…
[2] https://github.com/rpm-software-management/rpm/blob/master/scripts/check-rp…
[3] https://cwe.mitre.org/data/definitions/426.html
Signed-off-by: Eugene Syromiatnikov <esyr(a)redhat.com>
---
tools/testing/selftests/sched/Makefile | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/tools/testing/selftests/sched/Makefile b/tools/testing/selftests/sched/Makefile
index 099ee9213557..0e4581ded9d6 100644
--- a/tools/testing/selftests/sched/Makefile
+++ b/tools/testing/selftests/sched/Makefile
@@ -4,8 +4,7 @@ ifneq ($(shell $(CC) --version 2>&1 | head -n 1 | grep clang),)
CLANG_FLAGS += -no-integrated-as
endif
-CFLAGS += -O2 -Wall -g -I./ $(KHDR_INCLUDES) -Wl,-rpath=./ \
- $(CLANG_FLAGS)
+CFLAGS += -O2 -Wall -g -I./ $(KHDR_INCLUDES) $(CLANG_FLAGS)
LDLIBS += -lpthread
TEST_GEN_FILES := cs_prctl_test
--
2.28.0
The relative RPATH ("./") supplied to linker options in CFLAGS is resolved
relative to current working directory and not the executable directory,
which will lead in incorrect resolution when the test executables are run
from elsewhere. Changing it to $ORIGIN makes it resolve relative
to the directory in which the executables reside, which is supposedly
the desired behaviour.
Discovered by the /usr/lib/rpm/check-rpaths script[1][2] that checks
for insecure RPATH/RUNPATH[3], such as containing relative directories,
during an attempt to package BPF selftests for later use in CI:
ERROR 0004: file '/usr/libexec/kselftests/bpf/urandom_read' contains an insecure runpath '.' in [.]
[1] https://github.com/rpm-software-management/rpm/blob/master/scripts/check-rp…
[2] https://github.com/rpm-software-management/rpm/blob/master/scripts/check-rp…
[3] https://cwe.mitre.org/data/definitions/426.html
Signed-off-by: Eugene Syromiatnikov <esyr(a)redhat.com>
---
tools/testing/selftests/rseq/Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/testing/selftests/rseq/Makefile b/tools/testing/selftests/rseq/Makefile
index 5a3432fceb58..27544a67d6f0 100644
--- a/tools/testing/selftests/rseq/Makefile
+++ b/tools/testing/selftests/rseq/Makefile
@@ -6,7 +6,7 @@ endif
top_srcdir = ../../../..
-CFLAGS += -O2 -Wall -g -I./ $(KHDR_INCLUDES) -L$(OUTPUT) -Wl,-rpath=./ \
+CFLAGS += -O2 -Wall -g -I./ $(KHDR_INCLUDES) -L$(OUTPUT) -Wl,-rpath=\$$ORIGIN/ \
$(CLANG_FLAGS) -I$(top_srcdir)/tools/include
LDLIBS += -lpthread -ldl
--
2.28.0
From: Geliang Tang <tanggeliang(a)kylinos.cn>
So many "Address not found" messages occur at the end of forwarding tests
when using "ip address del" command for an invalid address:
TEST: FDB limits interacting with FDB type local [ OK ]
Error: ipv4: Address not found.
... ...
TEST: IGMPv3 S,G port entry automatic add to a *,G port [ OK ]
Error: ipv4: Address not found.
Error: ipv6: address not found.
... ...
TEST: Isolated port flooding [ OK ]
Error: ipv4: Address not found.
Error: ipv6: address not found.
... ...
TEST: Externally learned FDB entry - ageing & roaming [ OK ]
Error: ipv4: Address not found.
Error: ipv6: address not found.
This patch gnores these messages and redirects them to /dev/null in
__addr_add_del().
Signed-off-by: Geliang Tang <tanggeliang(a)kylinos.cn>
---
tools/testing/selftests/net/forwarding/lib.sh | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/testing/selftests/net/forwarding/lib.sh b/tools/testing/selftests/net/forwarding/lib.sh
index ff96bb7535ff..8670b6053cde 100644
--- a/tools/testing/selftests/net/forwarding/lib.sh
+++ b/tools/testing/selftests/net/forwarding/lib.sh
@@ -839,7 +839,7 @@ __addr_add_del()
array=("${@}")
for addrstr in "${array[@]}"; do
- ip address $add_del $addrstr dev $if_name
+ ip address $add_del $addrstr dev $if_name &> /dev/null
done
}
--
2.43.0
This small series includes fixes for creation of veth pairs for
networkless kernels & adds tests for turning the different network
interface features on and off in selftests/net/netdevice.sh script.
Changes in v4:
Move veth creation/removal to the main shell script.
Tested using vng on a networkless kernel and the script works, sample
output below the changes.
Changes in v3:
https://lore.kernel.org/all/20240614113240.41550-1-jain.abhinav177@gmail.co…
Add a check for netdev, create veth pair for testing.
Restore feature to its initial state.
Changes in v2:
https://lore.kernel.org/all/20240609132124.51683-1-jain.abhinav177@gmail.co…
Remove tail usage; use read to parse the features from temp file.
v1:
https://lore.kernel.org/all/20240606212714.27472-1-jain.abhinav177@gmail.co…
```
# selftests: net: netdevice.sh
# No valid network device found, creating veth pair
# PASS: veth0: set interface up
# PASS: veth0: set MAC address
# SKIP: veth0: set IP address
# PASS: veth0: ethtool list features
# PASS: veth0: Turned off feature: rx-checksumming
# PASS: veth0: Turned on feature: rx-checksumming
# PASS: veth0: Restore feature rx-checksumming to initial state on
# Actual changes:
# tx-checksum-ip-generic: off
# tx-tcp-segmentation: off [not requested]
....
....
....
# PASS: veth1: Restore feature tx-nocache-copy to initial state off
# PASS: veth1: Turned off feature: tx-vlan-stag-hw-insert
# PASS: veth1: Turned on feature: tx-vlan-stag-hw-insert
# PASS: veth1: Restore feature tx-vlan-stag-hw-insert to initial state on
# PASS: veth1: Turned off feature: rx-vlan-stag-hw-parse
# PASS: veth1: Turned on feature: rx-vlan-stag-hw-parse
# PASS: veth1: Restore feature rx-vlan-stag-hw-parse to initial state on
# PASS: veth1: Turned off feature: rx-gro-list
# PASS: veth1: Turned on feature: rx-gro-list
# PASS: veth1: Restore feature rx-gro-list to initial state off
# PASS: veth1: Turned off feature: rx-udp-gro-forwarding
# PASS: veth1: Turned on feature: rx-udp-gro-forwarding
# PASS: veth1: Restore feature rx-udp-gro-forwarding to initial state off
# Cannot get register dump: Operation not supported
# SKIP: veth1: ethtool dump not supported
# PASS: veth1: ethtool stats
# PASS: veth1: stop interface
# Removed veth pair
ok 12 selftests: net: netdevice.sh
```
Abhinav Jain (2):
selftests: net: Create veth pair for testing in networkless kernel
selftests: net: Add on/off checks for non-fixed features of interface
tools/testing/selftests/net/netdevice.sh | 55 +++++++++++++++++++++++-
1 file changed, 54 insertions(+), 1 deletion(-)
--
2.34.1
Add a new kselftest to detect and report slowdowns in key boot events. The
test uses ftrace to track timings for specific boot events and compares
these timestamps against reference values provided in YAML format.
The test includes the following files:
- `bootconfig` file: configures ftrace and lists reference key boot
events.
- `config` fragment: enables boot time tracing and attaches the
bootconfig file to the kernel image.
- `kprobe_timestamps_to_yaml.py` script: parses the current trace file to
extract event names and timestamps and writes them to a YAML file. The
script is intended to be run once to generate initial reference values;
the generated file is not meant to be stored in the kernel sources but
should be provided as input to the test itself. YAML format was chosen
to allow easy integration with per-platform data used in other tests,
such as the discoverable devices probe test in
tools/testing/selftests/devices. Another option is to use JSON, as the
file is not intended for manual editing and JSON is already supported
by the Python standard library.
- `test_boot_time.py` script: parses the current trace file and compares
timestamps against the values in the YAML file provided as input.
Reports a failure if any timestamp differs from the reference value by
more than the specified delta.
- `trace_utils.py` file: utility functions to mount debugfs and parse the
trace file to extract relevant information.
The bootconfig file provided is an initial draft with some reference kprobe
events to showcase how the test works. I would appreciate feedback from
those interested in running this test on which boot events should be added.
Different key events might be relevant depending on the platform and its
boot time requirements. This file should serve as a common ground and be
populated with critical events and functions common to different platforms.
Feedback on the overall approach of this test and suggestions for
additional boot events to trace would be greatly appreciated.
Example output with a deliberately small delta of 0.01 to demonstrate failures:
TAP version 13
1..4
ok 1 populate_rootfs_begin
# 'run_init_process_begin' differs by 0.033990 seconds.
not ok 2 run_init_process_begin
# 'run_init_process_end' differs by 0.033796 seconds.
not ok 3 run_init_process_end
ok 4 unpack_to_rootfs_begin
# Totals: pass:2 fail:2 xfail:0 xpass:0 skip:0 error:0
This patch depends on "kselftest: Move ksft helper module to common
directory":
https://lore.kernel.org/all/20240705-dev-err-log-selftest-v2-2-163b9cd7b3c1…
which was picked through the usb tree and is queued for 6.11-rc1.
Best,
Laura
Laura Nao (1):
kselftests: Add test to detect boot event slowdowns
tools/testing/selftests/Makefile | 1 +
tools/testing/selftests/boot-time/Makefile | 17 ++++
tools/testing/selftests/boot-time/bootconfig | 8 ++
tools/testing/selftests/boot-time/config | 4 +
.../boot-time/kprobe_timestamps_to_yaml.py | 55 +++++++++++
.../selftests/boot-time/test_boot_time.py | 94 +++++++++++++++++++
.../selftests/boot-time/trace_utils.py | 63 +++++++++++++
7 files changed, 242 insertions(+)
create mode 100644 tools/testing/selftests/boot-time/Makefile
create mode 100644 tools/testing/selftests/boot-time/bootconfig
create mode 100644 tools/testing/selftests/boot-time/config
create mode 100755 tools/testing/selftests/boot-time/kprobe_timestamps_to_yaml.py
create mode 100755 tools/testing/selftests/boot-time/test_boot_time.py
create mode 100644 tools/testing/selftests/boot-time/trace_utils.py
--
2.30.2
There are multiple possible timer sources which could be useful for
the sound stream synchronization: hrtimers, hardware clocks (e.g. PTP),
timer wheels (jiffies). Currently, using one of them to synchronize
the audio stream of snd-aloop module would require writing a
kernel-space driver which exports an ALSA timer through the
snd_timer interface.
However, it is not really convenient for application developers, who may
want to define their custom timer sources for audio synchronization.
For instance, we could have a network application which receives frames
and sends them to snd-aloop pcm device, and another application
listening on the other end of snd-aloop. It makes sense to transfer a
new period of data only when certain amount of frames is received
through the network, but definitely not when a certain amount of jiffies
on a local system elapses. Since all of the devices are purely virtual
it won't introduce any glitches and will help the application developers
to avoid using sample-rate conversion.
This patch series introduces userspace-driven ALSA timers: virtual
timers which are created and controlled from userspace. The timer can
be created from the userspace using the new ioctl SNDRV_TIMER_IOCTL_CREATE.
After creating a timer, it becomes available for use system-wide, so it
can be passed to snd-aloop as a timer source (timer_source parameter
would be "-1.SNDRV_TIMER_GLOBAL_UDRIVEN.{timer_id}"). When the userspace
app decides to trigger a timer, it calls another ioctl
SNDRV_TIMER_IOCTL_TRIGGER on the file descriptor of a timer. It
initiates a transfer of a new period of data.
Userspace-driven timers are associated with file descriptors. If the
application wishes to destroy the timer, it can simply release the file
descriptor of a virtual timer.
I believe introducing new ioctl calls is quite inconvenient (as we have
a limited amount of them), but other possible ways of app <-> kernel
communication (like virtual FS) seem completely inappropriate for this
task (but I'd love to discuss alternative solutions).
This patch series also updates the snd-aloop module so the global timers
can be used as a timer_source for it (it allows using userspace-driven
timers as timer source).
V1 -> V2:
- Fix some problems found by Christophe Jaillet
<christophe.jaillet(a)wanadoo.fr>
V2 -> V3:
- Add improvements suggested by Takashi Iwai <tiwai(a)suse.de>
Please, find the patch-specific changelog in the following patches.
Ivan Orlov (4):
ALSA: aloop: Allow using global timers
Docs/sound: Add documentation for userspace-driven ALSA timers
ALSA: timer: Introduce virtual userspace-driven timers
selftests: ALSA: Cover userspace-driven timers with test
Documentation/sound/index.rst | 1 +
Documentation/sound/utimers.rst | 120 +++++++++++
include/uapi/sound/asound.h | 20 +-
sound/core/Kconfig | 10 +
sound/core/timer.c | 221 ++++++++++++++++++++
sound/drivers/aloop.c | 2 +
tools/testing/selftests/alsa/Makefile | 2 +-
tools/testing/selftests/alsa/global-timer.c | 87 ++++++++
tools/testing/selftests/alsa/utimer-test.c | 170 +++++++++++++++
9 files changed, 631 insertions(+), 2 deletions(-)
create mode 100644 Documentation/sound/utimers.rst
create mode 100644 tools/testing/selftests/alsa/global-timer.c
create mode 100644 tools/testing/selftests/alsa/utimer-test.c
--
2.34.1
Hi Kees and All,
There are several tests in kselftest subsystem which load modules to tests
the internals of the kernel. Most of these test modules are just loaded by
the kselftest, their status isn't read and reported to the user logs. Hence
they don't provide benefit of executing those tests.
I've found patches from Kees where he has been converting such kselftests
to kunit tests [1]. The probable motivation is to move tests output of
kselftest subsystem which only triggers tests without correctly reporting
the results. On the other hand, kunit is there to test the kernel's
internal functions which can't be done by userspace.
Kselftest: Test user facing APIs from userspace
Kunit: Test kernel's internal functions from kernelspace
This brings me to conclusion that kselftest which are loading modules to
test kernelspace should be converted to kunit tests. I've noted several
such kselftests.
This is just my understanding. Please mention if I'm correct above or more
reasons to support kselftest test modules transformation into kunit test.
[1] https://lore.kernel.org/all/20221018082824.never.845-kees@kernel.org/
--
BR,
Muhammad Usama Anjum
The kernel has recently added support for shadow stacks, currently
x86 only using their CET feature but both arm64 and RISC-V have
equivalent features (GCS and Zicfiss respectively), I am actively
working on GCS[1]. With shadow stacks the hardware maintains an
additional stack containing only the return addresses for branch
instructions which is not generally writeable by userspace and ensures
that any returns are to the recorded addresses. This provides some
protection against ROP attacks and making it easier to collect call
stacks. These shadow stacks are allocated in the address space of the
userspace process.
Our API for shadow stacks does not currently offer userspace any
flexiblity for managing the allocation of shadow stacks for newly
created threads, instead the kernel allocates a new shadow stack with
the same size as the normal stack whenever a thread is created with the
feature enabled. The stacks allocated in this way are freed by the
kernel when the thread exits or shadow stacks are disabled for the
thread. This lack of flexibility and control isn't ideal, in the vast
majority of cases the shadow stack will be over allocated and the
implicit allocation and deallocation is not consistent with other
interfaces. As far as I can tell the interface is done in this manner
mainly because the shadow stack patches were in development since before
clone3() was implemented.
Since clone3() is readily extensible let's add support for specifying a
shadow stack when creating a new thread or process in a similar manner
to how the normal stack is specified, keeping the current implicit
allocation behaviour if one is not specified either with clone3() or
through the use of clone(). The user must provide a shadow stack
address and size, this must point to memory mapped for use as a shadow
stackby map_shadow_stack() with a shadow stack token at the top of the
stack.
Please note that the x86 portions of this code are build tested only, I
don't appear to have a system that can run CET avaible to me, I have
done testing with an integration into my pending work for GCS. There is
some possibility that the arm64 implementation may require the use of
clone3() and explicit userspace allocation of shadow stacks, this is
still under discussion.
Please further note that the token consumption done by clone3() is not
currently implemented in an atomic fashion, Rick indicated that he would
look into fixing this if people are OK with the implementation.
A new architecture feature Kconfig option for shadow stacks is added as
here, this was suggested as part of the review comments for the arm64
GCS series and since we need to detect if shadow stacks are supported it
seemed sensible to roll it in here.
[1] https://lore.kernel.org/r/20231009-arm64-gcs-v6-0-78e55deaa4dd@kernel.org/
Signed-off-by: Mark Brown <broonie(a)kernel.org>
---
Changes in v7:
- Rebase onto v6.11-rc1.
- Typo fixes.
- Link to v6: https://lore.kernel.org/r/20240623-clone3-shadow-stack-v6-0-9ee7783b1fb9@ke…
Changes in v6:
- Rebase onto v6.10-rc3.
- Ensure we don't try to free the parent shadow stack in error paths of
x86 arch code.
- Spelling fixes in userspace API document.
- Additional cleanups and improvements to the clone3() tests to support
the shadow stack tests.
- Link to v5: https://lore.kernel.org/r/20240203-clone3-shadow-stack-v5-0-322c69598e4b@ke…
Changes in v5:
- Rebase onto v6.8-rc2.
- Rework ABI to have the user allocate the shadow stack memory with
map_shadow_stack() and a token.
- Force inlining of the x86 shadow stack enablement.
- Move shadow stack enablement out into a shared header for reuse by
other tests.
- Link to v4: https://lore.kernel.org/r/20231128-clone3-shadow-stack-v4-0-8b28ffe4f676@ke…
Changes in v4:
- Formatting changes.
- Use a define for minimum shadow stack size and move some basic
validation to fork.c.
- Link to v3: https://lore.kernel.org/r/20231120-clone3-shadow-stack-v3-0-a7b8ed3e2acc@ke…
Changes in v3:
- Rebase onto v6.7-rc2.
- Remove stale shadow_stack in internal kargs.
- If a shadow stack is specified unconditionally use it regardless of
CLONE_ parameters.
- Force enable shadow stacks in the selftest.
- Update changelogs for RISC-V feature rename.
- Link to v2: https://lore.kernel.org/r/20231114-clone3-shadow-stack-v2-0-b613f8681155@ke…
Changes in v2:
- Rebase onto v6.7-rc1.
- Remove ability to provide preallocated shadow stack, just specify the
desired size.
- Link to v1: https://lore.kernel.org/r/20231023-clone3-shadow-stack-v1-0-d867d0b5d4d0@ke…
---
Mark Brown (9):
Documentation: userspace-api: Add shadow stack API documentation
selftests: Provide helper header for shadow stack testing
mm: Introduce ARCH_HAS_USER_SHADOW_STACK
fork: Add shadow stack support to clone3()
selftests/clone3: Remove redundant flushes of output streams
selftests/clone3: Factor more of main loop into test_clone3()
selftests/clone3: Explicitly handle child exits due to signals
selftests/clone3: Allow tests to flag if -E2BIG is a valid error code
selftests/clone3: Test shadow stack support
Documentation/userspace-api/index.rst | 1 +
Documentation/userspace-api/shadow_stack.rst | 41 ++++
arch/x86/Kconfig | 1 +
arch/x86/include/asm/shstk.h | 11 +-
arch/x86/kernel/process.c | 2 +-
arch/x86/kernel/shstk.c | 104 +++++++---
fs/proc/task_mmu.c | 2 +-
include/linux/mm.h | 2 +-
include/linux/sched/task.h | 13 ++
include/uapi/linux/sched.h | 13 +-
kernel/fork.c | 76 ++++++--
mm/Kconfig | 6 +
tools/testing/selftests/clone3/clone3.c | 224 ++++++++++++++++++----
tools/testing/selftests/clone3/clone3_selftests.h | 40 +++-
tools/testing/selftests/ksft_shstk.h | 63 ++++++
15 files changed, 511 insertions(+), 88 deletions(-)
---
base-commit: 8400291e289ee6b2bf9779ff1c83a291501f017b
change-id: 20231019-clone3-shadow-stack-15d40d2bf536
Best regards,
--
Mark Brown <broonie(a)kernel.org>
From: "Steven Rostedt (Google)" <rostedt(a)goodmis.org>
A regression happened where running the ownership test passes on the first
iteration but fails running it a second time. This was caught and fixed,
but a later change brought it back. The regression was missed because the
automated tests only run the tests once per boot.
Change the ownership test to iterate through the tests twice, as this will
catch the regression with a single run.
Signed-off-by: Steven Rostedt (Google) <rostedt(a)goodmis.org>
---
.../ftrace/test.d/00basic/test_ownership.tc | 34 +++++++++++--------
1 file changed, 20 insertions(+), 14 deletions(-)
diff --git a/tools/testing/selftests/ftrace/test.d/00basic/test_ownership.tc b/tools/testing/selftests/ftrace/test.d/00basic/test_ownership.tc
index c45094d1e1d2..71e43a92352a 100644
--- a/tools/testing/selftests/ftrace/test.d/00basic/test_ownership.tc
+++ b/tools/testing/selftests/ftrace/test.d/00basic/test_ownership.tc
@@ -83,32 +83,38 @@ run_tests() {
done
}
-mount -o remount,"$new_options" .
+# Run the tests twice as leftovers can cause issues
+for loop in 1 2 ; do
-run_tests
+ echo "Running iteration $loop"
-mount -o remount,"$mount_options" .
+ mount -o remount,"$new_options" .
-for d in "." "events" "events/sched" "events/sched/sched_switch" "events/sched/sched_switch/enable" $canary; do
- test "$d" $original_group
-done
+ run_tests
+
+ mount -o remount,"$mount_options" .
+
+ for d in "." "events" "events/sched" "events/sched/sched_switch" "events/sched/sched_switch/enable" $canary; do
+ test "$d" $original_group
+ done
# check instances as well
-chgrp $other_group instances
+ chgrp $other_group instances
-instance="$(mktemp -u test-XXXXXX)"
+ instance="$(mktemp -u test-XXXXXX)"
-mkdir instances/$instance
+ mkdir instances/$instance
-cd instances/$instance
+ cd instances/$instance
-run_tests
+ run_tests
-cd ../..
+ cd ../..
-rmdir instances/$instance
+ rmdir instances/$instance
-chgrp $original_group instances
+ chgrp $original_group instances
+done
exit 0
--
2.43.0
This small series fixes is_madv_discard() and adds a small sanity check
test to selftests/mm/mseal_test. Without this patch, is_madv_discard()
erroneously thinks innocent ops like MADV_RANDOM are discard operations
(which they are not, and are supposed to be allowed, per the overall
design).
Based on Linus's tree and taken from my mseal depessimization series[1].
[1]: https://lore.kernel.org/all/20240806212808.1885309-1-pedro.falcato@gmail.co…
Pedro Falcato (2):
mseal: Fix is_madv_discard()
selftests/mm: Add mseal test for no-discard madvise
mm/mseal.c | 14 +++++++---
tools/testing/selftests/mm/mseal_test.c | 34 +++++++++++++++++++++++++
2 files changed, 45 insertions(+), 3 deletions(-)
--
2.46.0
The dma-iommu needs to find the correct domain for MSI mapping. With an
IOMMU_DOMAIN_NESTED, the mapping resides in its parent paging domain.
Add a get_msi_mapping_domain op for drivers to return paging domains.
Add an iommufd selftest coverage for that, by doing a loopback test.
Add arm_smmu_get_msi_mapping_domain in the SMMUv3 driver so its nesting
feature could work with MSI correctly.
This is based on top of the reserved-IOVA change:
https://lore.kernel.org/all/20240802053458.2754673-1-nicolinc@nvidia.com/
And Jason's SMMUv3 nesting series:
https://lore.kernel.org/all/0-v1-54e734311a7f+14f72-smmuv3_nesting_jgg@nvid…
This series is on Github:
https://github.com/nicolinc/iommufd/commits/iommufd_nesting_sw_msi/
[changelog]
v3:
* Refined PATCH-2 commit message
* Added domain->ops check in PATCH-2
* Added PATCH-4 to implement in SMMUv3 driver
v2:
https://lore.kernel.org/all/cover.1722644866.git.nicolinc@nvidia.com/
* Resent with a proper bug fix.
Thanks
Nicolin
Nicolin Chen (3):
iommufd: Reorder include files
iommufd/selftest: Add coverage for IOMMU_RESV_SW_MSI
iommu/arm-smmu-v3: Implement arm_smmu_get_msi_mapping_domain
Robin Murphy (1):
iommu/dma: Support MSIs through nested domains
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 10 ++
drivers/iommu/dma-iommu.c | 18 +++-
drivers/iommu/iommufd/device.c | 4 +-
drivers/iommu/iommufd/fault.c | 4 +-
drivers/iommu/iommufd/io_pagetable.c | 8 +-
drivers/iommu/iommufd/io_pagetable.h | 2 +-
drivers/iommu/iommufd/ioas.c | 2 +-
drivers/iommu/iommufd/iommufd_private.h | 9 +-
drivers/iommu/iommufd/iommufd_test.h | 6 +-
drivers/iommu/iommufd/iova_bitmap.c | 2 +-
drivers/iommu/iommufd/main.c | 8 +-
drivers/iommu/iommufd/pages.c | 10 +-
drivers/iommu/iommufd/selftest.c | 92 ++++++++++++++++++-
include/linux/iommu.h | 4 +
include/linux/iommufd.h | 4 +-
include/uapi/linux/iommufd.h | 2 +-
tools/testing/selftests/iommu/iommufd_utils.h | 9 ++
17 files changed, 160 insertions(+), 34 deletions(-)
--
2.43.0
This patch series adds a selftest suite to validate the s390x
architecture specific ucontrol KVM interface.
When creating a VM on s390x it is possible to create it as userspace
controlled VM or in short ucontrol VM.
These VMs delegates the management of the VM to userspace instead
of handling most events within the kernel. Consequently the userspace
has to manage interrupts, memory allocation etc.
Before this patch set this functionality lacks any public test cases.
It is desirable to add test cases for this interface to be able to
reduce the risk of breaking changes in the future.
In order to provision a ucontrol VM the kernel needs to be compiled with
the CONFIG_KVM_S390_UCONTROL enabled. The users with sys_admin capability
can then create a new ucontrol VM providing the KVM_VM_S390_UCONTROL
parameter to the KVM_CREATE_VM ioctl.
The kernels existing selftest helper functions can only be partially be
reused for these tests.
The test cases cover existing special handling of ucontrol VMs within the
implementation and basic VM creation and handling cases:
* Reject setting HPAGE when VM is ucontrol
* Assert KVM_GET_DIRTY_LOG is rejected
* Assert KVM_S390_VM_MEM_LIMIT_SIZE is rejected
* Assert state of initial SIE flags setup by the kernel
* Run simple program in VM with and without DAT
* Assert KVM_EXIT_S390_UCONTROL exit on not mapped memory access
* Assert functionality of storage keys in ucontrol VM
Running the test cases requires sys_admin capabilities to start the
ucontrol VM.
This can be achieved by running as root or with a command like:
sudo setpriv --reuid nobody --inh-caps -all,+sys_admin \
--ambient-caps -all,+sys_admin --bounding-set -all,+sys_admin \
./ucontrol_test
The patch set does also contain some code cleanup / consolidation of
architecture specific defines that are now used in multiple test cases.
---
v4:
- PATCH 5: Remove not yet used include for debug print functions
- PATCH 6: Add include for debug print functions (removed from patch 5)
Remove no longer needed code since stopped but is reset
before starting since v3 (thanks Janosch)
Adjust test output to use leading zeros instead of spaces in sieic
- PATCH 7: Rename constant to PGM_SEGMENT_TRANSLATION (thanks Janosch)
Put comments on their own lines
v3:
- Remove stopped bit before starting the VM (no initial stop in multiple
test cases) (thanks Janosch)
- PATCH 2: Clarified SIE control block vs SIE instruction (thanks
Janosch)
- PATCH 3: Make use of CAP_TO_MASK(CAP_SYS_ADMIN) instead of custom
define (thanks Janosch)
Removed Reviewed-By: Claudio
- PATCH 4: Remove erroneous 1MB offset from self->base_hva (thanks
Janosch)
- PATCH 6-8: Change name of test program _pgm to _asm to prevent confusion
- PATCH 10: Move KVM_S390_UCONTROL default option to actual debug config
(thanks Christian)
v2:
- add ucontrol to s390 debug config (new patch)
- PATCH 2: changed atomic_t to __u32 (thanks Claudio)
- PATCH 4: reformatted comment in FIXTURE_SETUP(uc_kvm)
- PATCH 5: refactored to display 8 byte blocks + more internal reuse
(thanks Claudio)
- PATCH 7: make use of more declarative defines instead of magic values
- PATCH 8: make use of more declarative defines instead of magic values
(thanks Claudio)
- PATCH 9: add reference to fix verified by the test case
Christoph Schlameuss (10):
selftests: kvm: s390: Define page sizes in shared header
selftests: kvm: s390: Add kvm_s390_sie_block definition for userspace
tests
selftests: kvm: s390: Add s390x ucontrol test suite with hpage test
selftests: kvm: s390: Add test fixture and simple VM setup tests
selftests: kvm: s390: Add debug print functions
selftests: kvm: s390: Add VM run test case
selftests: kvm: s390: Add uc_map_unmap VM test case
selftests: kvm: s390: Add uc_skey VM test case
selftests: kvm: s390: Verify reject memory region operations for
ucontrol VMs
s390: Enable KVM_S390_UCONTROL config in debug_defconfig
arch/s390/configs/debug_defconfig | 1 +
tools/testing/selftests/kvm/.gitignore | 1 +
tools/testing/selftests/kvm/Makefile | 1 +
.../selftests/kvm/include/s390x/debug_print.h | 69 ++
.../selftests/kvm/include/s390x/processor.h | 5 +
.../testing/selftests/kvm/include/s390x/sie.h | 240 +++++++
.../selftests/kvm/lib/s390x/processor.c | 10 +-
tools/testing/selftests/kvm/s390x/cmma_test.c | 7 +-
tools/testing/selftests/kvm/s390x/config | 2 +
.../testing/selftests/kvm/s390x/debug_test.c | 4 +-
tools/testing/selftests/kvm/s390x/memop.c | 4 +-
tools/testing/selftests/kvm/s390x/tprot.c | 5 +-
.../selftests/kvm/s390x/ucontrol_test.c | 596 ++++++++++++++++++
13 files changed, 929 insertions(+), 16 deletions(-)
create mode 100644 tools/testing/selftests/kvm/include/s390x/debug_print.h
create mode 100644 tools/testing/selftests/kvm/include/s390x/sie.h
create mode 100644 tools/testing/selftests/kvm/s390x/config
create mode 100644 tools/testing/selftests/kvm/s390x/ucontrol_test.c
base-commit: c0ecd6388360d930440cc5554026818895199923
--
2.45.2
First 4 patches are more-or-less cleanups/preparations.
Patch 5 was sent to me/contributed off-list by Mohammad, who wants 32-bit
kernels to run TCP-AO.
Patch 6 is a workaround/fix for slow VMs. Albeit, I can't reproduce
the issue, but I hope it will fix netdev flakes for connect-deny-*
tests.
And the biggest change is adding TCP-AO tracepoints to selftests.
I think it's a good addition by the following reasons:
- The related tracepoints are now tested;
- It allows tcp-ao selftests to raise expectations on the kernel
behavior - up from the syscalls exit statuses + net counters.
- Provides tracepoints usage samples.
As tracepoints are not a stable ABI, any kernel changes done to them
will be reflected to the selftests, which also will allow users
to see how to change their code. It's quite better than parsing dmesg
(what BGP was doing pre-tracepoints, ugh).
Somewhat arguably, the code parses trace_pipe, rather than uses
libtraceevent (which any sane user should do). The reason behind that is
the same as for rt-netlink macros instead of libmnl: I'm trying
to minimize the library dependencies of the selftests. And the
performance of formatting text in kernel and parsing it again in a test
is not critical.
Current output sample:
> ok 73 Trace events matched expectations: 13 tcp_hash_md5_required[2] tcp_hash_md5_unexpected[4] tcp_hash_ao_required[3] tcp_ao_key_not_found[4]
Previously, tracepoints selftests were part of kernel tcp tracepoints
submission [1], but since then the code was quite changed:
- Now generic tracing setup is in lib/ftrace.c, separate from
lib/ftrace-tcp.c which utilizes TCP trace points. This separation
allows future selftests to trace non-TCP events, i.e. to find out
an skb's drop reason, which was useful in the creation of TCP-CLOSE
stress-test (not in this patch set, but used in attempt to reproduce
the issue from [2]).
- Another change is that in the previous submission the trace events
where used only to detect unexpected TCP-AO/TCP-MD5 events. In this
version the selftests will fail if an expected trace event didn't
appear.
Let's see how reliable this is on the netdev bot - it obviously passes
on my testing, but potentially may require a temporary XFAIL patch
if it misbehaves on a slow VM.
[1] https://lore.kernel.org/lkml/20240224-tcp-ao-tracepoints-v1-0-15f31b7f30a7@…
[2] https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net.git/commit/?id=3…
Signed-off-by: Dmitry Safonov <0x7f454c46(a)gmail.com>
---
Changes in v2:
- Fixed two issues with parsing TCP-AO events: the socket state and TCP
segment flags. Hopefully, won't fail on netdev.
- Reword patch 1 & 2 messages to be more informative and at some degree
formal (Paolo)
- Since commit e33a02ed6a4f ("selftests: Add printf attribute to
kselftest prints") it's possible to use __printf instead of "raw" gcc
attribute - switch using that, as checkpatch suggests.
- Link to v1: https://lore.kernel.org/r/20240730-tcp-ao-selftests-upd-6-12-v1-0-ffd4bf15d…
---
Dmitry Safonov (6):
selftests/net: Clean-up double assignment
selftests/net: Provide test_snprintf() helper
selftests/net: Be consistent in kconfig checks
selftests/net: Don't forget to close nsfd after switch_save_ns()
selftests/net: Synchronize client/server before counters checks
selftests/net: Add trace events matching to tcp_ao
Mohammad Nassiri (1):
selftests/tcp_ao: Fix printing format for uint64_t
tools/testing/selftests/net/tcp_ao/Makefile | 3 +-
tools/testing/selftests/net/tcp_ao/bench-lookups.c | 2 +-
tools/testing/selftests/net/tcp_ao/config | 1 +
tools/testing/selftests/net/tcp_ao/connect-deny.c | 25 +-
tools/testing/selftests/net/tcp_ao/connect.c | 6 +-
tools/testing/selftests/net/tcp_ao/icmps-discard.c | 2 +-
.../testing/selftests/net/tcp_ao/key-management.c | 18 +-
tools/testing/selftests/net/tcp_ao/lib/aolib.h | 176 ++++++-
.../testing/selftests/net/tcp_ao/lib/ftrace-tcp.c | 549 +++++++++++++++++++++
tools/testing/selftests/net/tcp_ao/lib/ftrace.c | 466 +++++++++++++++++
tools/testing/selftests/net/tcp_ao/lib/kconfig.c | 31 +-
tools/testing/selftests/net/tcp_ao/lib/setup.c | 15 +-
tools/testing/selftests/net/tcp_ao/lib/sock.c | 1 -
tools/testing/selftests/net/tcp_ao/lib/utils.c | 26 +
tools/testing/selftests/net/tcp_ao/restore.c | 30 +-
tools/testing/selftests/net/tcp_ao/rst.c | 2 +-
tools/testing/selftests/net/tcp_ao/self-connect.c | 19 +-
tools/testing/selftests/net/tcp_ao/seq-ext.c | 28 +-
.../selftests/net/tcp_ao/setsockopt-closed.c | 6 +-
tools/testing/selftests/net/tcp_ao/unsigned-md5.c | 35 +-
20 files changed, 1375 insertions(+), 66 deletions(-)
---
base-commit: 3361a6eae59664ffae640ff7a838f5bd89c24461
change-id: 20240730-tcp-ao-selftests-upd-6-12-4d3e53a74f3f
Best regards,
--
Dmitry Safonov <0x7f454c46(a)gmail.com>
This revision only updates the tests from the previous revision[1], and
integrates an Acked-by[2] and a Reviewed-By[3] into the first commit
message.
Documentation/admin-guide/cgroup-v2.rst | 22 ++-
include/linux/cgroup-defs.h | 5 +
include/linux/cgroup.h | 3 +
include/linux/memcontrol.h | 5 +
include/linux/page_counter.h | 11 +-
kernel/cgroup/cgroup-internal.h | 2 +
kernel/cgroup/cgroup.c | 7 +
mm/memcontrol.c | 116 +++++++++++++--
mm/page_counter.c | 30 +++-
tools/testing/selftests/cgroup/cgroup_util.c | 22 +++
tools/testing/selftests/cgroup/cgroup_util.h | 2 +
tools/testing/selftests/cgroup/test_memcontrol.c | 264 ++++++++++++++++++++++++++++++++-
12 files changed, 454 insertions(+), 35 deletions(-)
[1]: https://lore.kernel.org/cgroups/20240729143743.34236-1-davidf@vimeo.com/T/
[2]: https://lore.kernel.org/cgroups/20240729143743.34236-1-davidf@vimeo.com/T/#…
[3]: https://lore.kernel.org/cgroups/20240729143743.34236-1-davidf@vimeo.com/T/#…
Thank you all for the support and reviews so far!
David Finkel
Senior Principal Software Engineer
Vimeo Inc.
Hello,
this series brings a new set of test converted to the test_progs framework.
Since the tests are quite small, I chose to group three tests conversion in
the same series, but feel free to let me know if I should keep one series
per test. The series focuses on cgroup testing and converts the following
tests:
- get_cgroup_id_user
- cgroup_storage
- test_skb_cgroup_id_user
Signed-off-by: Alexis Lothoré (eBPF Foundation) <alexis.lothore(a)bootlin.com>
---
Alexis Lothoré (eBPF Foundation) (4):
selftests/bpf: convert get_current_cgroup_id_user to test_progs
selftests/bpf: convert test_cgroup_storage to test_progs
selftests/bpf: add proper section name to bpf prog and rename it
selftests/bpf: convert test_skb_cgroup_id_user to test_progs
tools/testing/selftests/bpf/.gitignore | 3 -
tools/testing/selftests/bpf/Makefile | 8 +-
tools/testing/selftests/bpf/get_cgroup_id_user.c | 151 -----------------
.../selftests/bpf/prog_tests/cgroup_ancestor.c | 159 ++++++++++++++++++
.../bpf/prog_tests/cgroup_get_current_cgroup_id.c | 58 +++++++
.../selftests/bpf/prog_tests/cgroup_storage.c | 65 ++++++++
...test_skb_cgroup_id_kern.c => cgroup_ancestor.c} | 2 +-
tools/testing/selftests/bpf/progs/cgroup_storage.c | 24 +++
tools/testing/selftests/bpf/test_cgroup_storage.c | 174 --------------------
tools/testing/selftests/bpf/test_skb_cgroup_id.sh | 63 -------
.../selftests/bpf/test_skb_cgroup_id_user.c | 183 ---------------------
11 files changed, 309 insertions(+), 581 deletions(-)
---
base-commit: 0e2eaf4b33f65e904b69bae6b956f3f610dbba9a
change-id: 20240725-convert_cgroup_tests-d07c66053225
Best regards,
--
Alexis Lothoré, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
The system register definitions in the arm64 get-reg-list are all done
with directly specified magic numbers rather than using the definitions
we import from the main kernel. This is error prone, and requires us to
audit the additions to get-reg-list separately to what we do when
specifying the registers for the main kernel. Since Marc has indicated
that this isn't a deliberate or desired choice let's start using the
constants we have defined.
We first manually update the data used to filter registers based on ID
register fields to use a simplified macro that specifies the register
and ID field in a muc more compact fashion. This is done first since
there is an error in the ID register field for the S1PIE registers. We
then replace all the remaining named system register specifications with
use of the existing KVM_ARM64_SYS_REG() macro.
This is just a first step, there's a bunch more work we could be doing
here, the main thing being making use of the encodings in
arch/arm64/tools/sysreg to convert more of the registers (including
updating as more registers are converted to use the generator).
Signed-off-by: Mark Brown <broonie(a)kernel.org>
---
Changes in v2:
- Add use of designated initalisers when converting filtering macros.
- Manual handling of CNTV_CTL_EL0 and CNTV_CVAL_EL0.
- Commit message tweaks.
- Link to v1: https://lore.kernel.org/r/20240802-kvm-arm64-get-reg-list-v1-0-3a5bf8f80765…
---
Mark Brown (3):
KVM: selftests: arm64: Simplify specification of filtered registers
KVM: selftests: arm64: Use symbolic definitions for incorrect encodings
KVM: selftests: arm64: Use generated defines for named system registers
tools/testing/selftests/kvm/aarch64/get-reg-list.c | 244 ++++++++++-----------
1 file changed, 122 insertions(+), 122 deletions(-)
---
base-commit: 8400291e289ee6b2bf9779ff1c83a291501f017b
change-id: 20240802-kvm-arm64-get-reg-list-a86a37460bdd
Best regards,
--
Mark Brown <broonie(a)kernel.org>
The system register definitions in the arm64 get-reg-list are all done
with directly specified magic numbers rather than using the definitions
we import from the main kernel. This is error prone, and requires us to
audit the additions to get-reg-list separately to what we do when
specifying the registers for the main kernel. Since Marc has indicated
that this isn't a deliberate or desired choice let's start using the
constants we have defined.
We first manually update the data used to filter registers based on ID
register fields to use a simplified macro that specifies the register
and ID field in a muc more compact fashion. This is done first since
there is an error in the ID register field for the S1PIE registers. We
then replace all the remaining named system register specifications with
use of the existing KVM_ARM64_SYS_REG() macro.
Signed-off-by: Mark Brown <broonie(a)kernel.org>
---
Mark Brown (2):
KVM: selftests: arm64: Simplify specification of filtered registers
KVM: selftests: arm64: Use generated defines for named system registers
tools/testing/selftests/kvm/aarch64/get-reg-list.c | 237 ++++++++++-----------
1 file changed, 115 insertions(+), 122 deletions(-)
---
base-commit: 8400291e289ee6b2bf9779ff1c83a291501f017b
change-id: 20240802-kvm-arm64-get-reg-list-a86a37460bdd
Best regards,
--
Mark Brown <broonie(a)kernel.org>
In commit 7d3c33b290b1 ("kunit: Device wrappers should also manage driver name"),
the kunit_kstrdup_const() and kunit_kfree_const() were introduced as an
optimisation of kunit_kstrdup(), which only copy/free strings from the
kernel rodata.
However, these are inline functions, and is_kernel_rodata() only works
for built-in code. This causes problems in two cases:
- If kunit is built as a module, __{start,end}_rodata is not defined.
- If a kunit test using these functions is built as a module, it will
suffer the same fate.
Restrict the is_kernel_rodata() case to when KUnit is built as a module,
which fixes the first case, at the cost of losing the optimisation.
Also, make kunit_{kstrdup,kfree}_const non-inline, so that other modules
using them will not accidentally depend on is_kernel_rodata(). If KUnit
is built-in, they'll benefit from the optimisation, if KUnit is not,
they won't, but the string will be properly duplicated.
(And fix a couple of typos in the doc comment, too.)
Reported-by: Nico Pache <npache(a)redhat.com>
Closes: https://lore.kernel.org/all/CAA1CXcDKht4vOL-acxrARbm6JhGna8_k8wjYJ-vHONink8…
Fixes: 7d3c33b290b1 ("kunit: Device wrappers should also manage driver name")
Signed-off-by: David Gow <davidgow(a)google.com>
---
include/kunit/test.h | 16 +++-------------
lib/kunit/test.c | 19 +++++++++++++++++++
2 files changed, 22 insertions(+), 13 deletions(-)
diff --git a/include/kunit/test.h b/include/kunit/test.h
index da9e84de14c0..5ac237c949a0 100644
--- a/include/kunit/test.h
+++ b/include/kunit/test.h
@@ -489,11 +489,7 @@ static inline void *kunit_kcalloc(struct kunit *test, size_t n, size_t size, gfp
* Calls kunit_kfree() only if @x is not in .rodata section.
* See kunit_kstrdup_const() for more information.
*/
-static inline void kunit_kfree_const(struct kunit *test, const void *x)
-{
- if (!is_kernel_rodata((unsigned long)x))
- kunit_kfree(test, x);
-}
+void kunit_kfree_const(struct kunit *test, const void *x);
/**
* kunit_kstrdup() - Duplicates a string into a test managed allocation.
@@ -527,16 +523,10 @@ static inline char *kunit_kstrdup(struct kunit *test, const char *str, gfp_t gfp
* @gfp: flags passed to underlying kmalloc().
*
* Calls kunit_kstrdup() only if @str is not in the rodata section. Must be freed with
- * kunit_free_const() -- not kunit_free().
+ * kunit_kfree_const() -- not kunit_kfree().
* See kstrdup_const() and kunit_kmalloc_array() for more information.
*/
-static inline const char *kunit_kstrdup_const(struct kunit *test, const char *str, gfp_t gfp)
-{
- if (is_kernel_rodata((unsigned long)str))
- return str;
-
- return kunit_kstrdup(test, str, gfp);
-}
+const char *kunit_kstrdup_const(struct kunit *test, const char *str, gfp_t gfp);
/**
* kunit_vm_mmap() - Allocate KUnit-tracked vm_mmap() area
diff --git a/lib/kunit/test.c b/lib/kunit/test.c
index e8b1b52a19ab..089c832e3cdb 100644
--- a/lib/kunit/test.c
+++ b/lib/kunit/test.c
@@ -874,6 +874,25 @@ void kunit_kfree(struct kunit *test, const void *ptr)
}
EXPORT_SYMBOL_GPL(kunit_kfree);
+void kunit_kfree_const(struct kunit *test, const void *x)
+{
+#if !IS_MODULE(CONFIG_KUNIT)
+ if (!is_kernel_rodata((unsigned long)x))
+#endif
+ kunit_kfree(test, x);
+}
+EXPORT_SYMBOL_GPL(kunit_kfree_const);
+
+const char *kunit_kstrdup_const(struct kunit *test, const char *str, gfp_t gfp)
+{
+#if !IS_MODULE(CONFIG_KUNIT)
+ if (is_kernel_rodata((unsigned long)str))
+ return str;
+#endif
+ return kunit_kstrdup(test, str, gfp);
+}
+EXPORT_SYMBOL_GPL(kunit_kstrdup_const);
+
void kunit_cleanup(struct kunit *test)
{
struct kunit_resource *res;
--
2.46.0.rc2.264.g509ed76dc8-goog
kunit_driver_create() accepts a name for the driver, but does not copy
it, so if that name is either on the stack, or otherwise freed, we end
up with a use-after-free when the driver is cleaned up.
Instead, strdup() the name, and manage it as another KUnit allocation.
As there was no existing kunit_kstrdup(), we add one. Further, add a
kunit_ variant of strdup_const() and kfree_const(), so we don't need to
allocate and manage the string in the majority of cases where it's a
constant.
This fixes a KASAN splat with overflow.overflow_allocation_test, when
built as a module.
Fixes: d03c720e03bd ("kunit: Add APIs for managing devices")
Reported-by: Nico Pache <npache(a)redhat.com>
Closes: https://groups.google.com/g/kunit-dev/c/81V9b9QYON0
Signed-off-by: David Gow <davidgow(a)google.com>
Reviewed-by: Kees Cook <kees(a)kernel.org>
---
There's some more serious changes since the RFC I sent, so please take a
closer look.
Thanks,
-- David
Changes since RFC:
https://groups.google.com/g/kunit-dev/c/81V9b9QYON0/m/PFKNKDKAAAAJ
- Add and use the kunit_kstrdup_const() and kunit_free_const()
functions.
- Fix a typo in the doc comments.
---
include/kunit/test.h | 58 ++++++++++++++++++++++++++++++++++++++++++++
lib/kunit/device.c | 7 ++++--
2 files changed, 63 insertions(+), 2 deletions(-)
diff --git a/include/kunit/test.h b/include/kunit/test.h
index e2a1f0928e8b..da9e84de14c0 100644
--- a/include/kunit/test.h
+++ b/include/kunit/test.h
@@ -28,6 +28,7 @@
#include <linux/types.h>
#include <asm/rwonce.h>
+#include <asm/sections.h>
/* Static key: true if any KUnit tests are currently running */
DECLARE_STATIC_KEY_FALSE(kunit_running);
@@ -480,6 +481,63 @@ static inline void *kunit_kcalloc(struct kunit *test, size_t n, size_t size, gfp
return kunit_kmalloc_array(test, n, size, gfp | __GFP_ZERO);
}
+
+/**
+ * kunit_kfree_const() - conditionally free test managed memory
+ * @x: pointer to the memory
+ *
+ * Calls kunit_kfree() only if @x is not in .rodata section.
+ * See kunit_kstrdup_const() for more information.
+ */
+static inline void kunit_kfree_const(struct kunit *test, const void *x)
+{
+ if (!is_kernel_rodata((unsigned long)x))
+ kunit_kfree(test, x);
+}
+
+/**
+ * kunit_kstrdup() - Duplicates a string into a test managed allocation.
+ *
+ * @test: The test context object.
+ * @str: The NULL-terminated string to duplicate.
+ * @gfp: flags passed to underlying kmalloc().
+ *
+ * See kstrdup() and kunit_kmalloc_array() for more information.
+ */
+static inline char *kunit_kstrdup(struct kunit *test, const char *str, gfp_t gfp)
+{
+ size_t len;
+ char *buf;
+
+ if (!str)
+ return NULL;
+
+ len = strlen(str) + 1;
+ buf = kunit_kmalloc(test, len, gfp);
+ if (buf)
+ memcpy(buf, str, len);
+ return buf;
+}
+
+/**
+ * kunit_kstrdup_const() - Conditionally duplicates a string into a test managed allocation.
+ *
+ * @test: The test context object.
+ * @str: The NULL-terminated string to duplicate.
+ * @gfp: flags passed to underlying kmalloc().
+ *
+ * Calls kunit_kstrdup() only if @str is not in the rodata section. Must be freed with
+ * kunit_free_const() -- not kunit_free().
+ * See kstrdup_const() and kunit_kmalloc_array() for more information.
+ */
+static inline const char *kunit_kstrdup_const(struct kunit *test, const char *str, gfp_t gfp)
+{
+ if (is_kernel_rodata((unsigned long)str))
+ return str;
+
+ return kunit_kstrdup(test, str, gfp);
+}
+
/**
* kunit_vm_mmap() - Allocate KUnit-tracked vm_mmap() area
* @test: The test context object.
diff --git a/lib/kunit/device.c b/lib/kunit/device.c
index 25c81ed465fb..520c1fccee8a 100644
--- a/lib/kunit/device.c
+++ b/lib/kunit/device.c
@@ -89,7 +89,7 @@ struct device_driver *kunit_driver_create(struct kunit *test, const char *name)
if (!driver)
return ERR_PTR(err);
- driver->name = name;
+ driver->name = kunit_kstrdup_const(test, name, GFP_KERNEL);
driver->bus = &kunit_bus_type;
driver->owner = THIS_MODULE;
@@ -192,8 +192,11 @@ void kunit_device_unregister(struct kunit *test, struct device *dev)
const struct device_driver *driver = to_kunit_device(dev)->driver;
kunit_release_action(test, device_unregister_wrapper, dev);
- if (driver)
+ if (driver) {
+ const char *driver_name = driver->name;
kunit_release_action(test, driver_unregister_wrapper, (void *)driver);
+ kunit_kfree_const(test, driver_name);
+ }
}
EXPORT_SYMBOL_GPL(kunit_device_unregister);
--
2.46.0.rc1.232.g9752f9e123-goog
In arm64 pKVM and QuIC's Gunyah protected VM model, we want to support
grabbing shmem user pages instead of using KVM's guestmemfd. These
hypervisors provide a different isolation model than the CoCo
implementations from x86. KVM's guest_memfd is focused on providing
memory that is more isolated than AVF requires. Some specific examples
include ability to pre-load data onto guest-private pages, dynamically
sharing/isolating guest pages without copy, and (future) migrating
guest-private pages. In sum of those differences after a discussion in
[1] and at PUCK, we want to try to stick with existing shmem and extend
GUP to support the isolation needs for arm64 pKVM and Gunyah. To that
end, we introduce the concept of "exclusive GUP pinning", which enforces
that only one pin of any kind is allowed when using the FOLL_EXCLUSIVE
flag is set. This behavior doesn't affect FOLL_GET or any other folio
refcount operations that don't go through the FOLL_PIN path.
[1]: https://lore.kernel.org/all/20240319143119.GA2736@willie-the-truck/
Tree with patches at:
https://git.codelinaro.org/clo/linux-kernel/gunyah-linux/-/tree/sent/exclus…
anup(a)brainfault.org, paul.walmsley(a)sifive.com,
palmer(a)dabbelt.com, aou(a)eecs.berkeley.edu, seanjc(a)google.com,
viro(a)zeniv.linux.org.uk, brauner(a)kernel.org,
willy(a)infradead.org, akpm(a)linux-foundation.org,
xiaoyao.li(a)intel.com, yilun.xu(a)intel.com,
chao.p.peng(a)linux.intel.com, jarkko(a)kernel.org,
amoorthy(a)google.com, dmatlack(a)google.com,
yu.c.zhang(a)linux.intel.com, isaku.yamahata(a)intel.com,
mic(a)digikod.net, vbabka(a)suse.cz, vannapurve(a)google.com,
ackerleytng(a)google.com, mail(a)maciej.szmigiero.name,
david(a)redhat.com, michael.roth(a)amd.com, wei.w.wang(a)intel.com,
liam.merwick(a)oracle.com, isaku.yamahata(a)gmail.com,
kirill.shutemov(a)linux.intel.com, suzuki.poulose(a)arm.com,
steven.price(a)arm.com, quic_eberman(a)quicinc.com,
quic_mnalajal(a)quicinc.com, quic_tsoni(a)quicinc.com,
quic_svaddagi(a)quicinc.com, quic_cvanscha(a)quicinc.com,
quic_pderrin(a)quicinc.com, quic_pheragu(a)quicinc.com,
catalin.marinas(a)arm.com, james.morse(a)arm.com,
yuzenghui(a)huawei.com, oliver.upton(a)linux.dev, maz(a)kernel.org,
will(a)kernel.org, qperret(a)google.com, keirf(a)google.com,
tabba(a)google.com
Signed-off-by: Elliot Berman <quic_eberman(a)quicinc.com>
---
Elliot Berman (2):
mm/gup-test: Verify exclusive pinned
mm/gup_test: Verify GUP grabs same pages twice
Fuad Tabba (3):
mm/gup: Move GUP_PIN_COUNTING_BIAS to page_ref.h
mm/gup: Add an option for obtaining an exclusive pin
mm/gup: Add support for re-pinning a normal pinned page as exclusive
include/linux/mm.h | 57 ++++----
include/linux/mm_types.h | 2 +
include/linux/page_ref.h | 74 ++++++++++
mm/Kconfig | 5 +
mm/gup.c | 265 ++++++++++++++++++++++++++++++----
mm/gup_test.c | 108 ++++++++++++++
mm/gup_test.h | 1 +
tools/testing/selftests/mm/gup_test.c | 5 +-
8 files changed, 457 insertions(+), 60 deletions(-)
---
base-commit: 6ba59ff4227927d3a8530fc2973b80e94b54d58f
change-id: 20240509-exclusive-gup-66259138bbff
Best regards,
--
Elliot Berman <quic_eberman(a)quicinc.com>
Hi Linus,
Please pull the kselftest fixes update for Linux 6.11-rc3.
This kselftest fixes update consists of a single fix to the conditional
in ksft.py script which incorrectly flags a test suite failed when there
are skipped tests in the mix. The logic is fixed to take skipped tests
into account and report the test as passed.
diff is attached.
thanks,
-- Shuah
----------------------------------------------------------------
The following changes since commit 8400291e289ee6b2bf9779ff1c83a291501f017b:
Linux 6.11-rc1 (2024-07-28 14:19:55 -0700)
are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest tags/linux_kselftest-fixes-6.11-rc3
for you to fetch changes up to 170c966cbe274e664288cfc12ee919d5e706dc50:
selftests: ksft: Fix finished() helper exit code on skipped tests (2024-07-31 11:38:56 -0600)
----------------------------------------------------------------
linux_kselftest-fixes-6.11-rc3
This kselftest fixes update consists of a single fix to the conditional
in ksft.py script which incorrectly flags a test suite failed when there
are skipped tests in the mix. The logic is fixed to take skipped tests
into account and report the test as passed.
----------------------------------------------------------------
Laura Nao (1):
selftests: ksft: Fix finished() helper exit code on skipped tests
tools/testing/selftests/kselftest/ksft.py | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
----------------------------------------------------------------
The first 2 patches in this series fix cpuset bugs found by Chen Ridong.
Patch 3 streamlines the sched domain rebuild process for hotplug
operation and eliminates the use of intermediate cpuset states for
sched domain generation. Patch 4 modifies generate_sched_domains()
to check the correctness of partition roots with non-overlapping CPUs.
Patch 5 adds new test cases to cover the bugs fixed in patches 1 and 2.
Chen Ridong (1):
cgroup/cpuset: fix panic caused by partcmd_update
Waiman Long (4):
cgroup/cpuset: Clear effective_xcpus on cpus_allowed clearing only if
cpus.exclusive not set
cgroup/cpuset: Eliminate unncessary sched domains rebuilds in hotplug
cgroup/cpuset: Check for partition roots with overlapping CPUs
selftest/cgroup: Add new test cases to test_cpuset_prs.sh
kernel/cgroup/cpuset.c | 70 ++++++++++---------
.../selftests/cgroup/test_cpuset_prs.sh | 12 +++-
2 files changed, 49 insertions(+), 33 deletions(-)
--
2.43.5
The current support for LLVM and clang in nolibc and its testsuite is
very limited.
* Various architectures plain do not compile
* The user *has* to specify "-Os" otherwise the program crashes
* Cross-compilation of the tests does not work
* Using clang is not wired up in run-tests.sh
This series extends this support.
Signed-off-by: Thomas Weißschuh <linux(a)weissschuh.net>
---
Thomas Weißschuh (12):
tools/nolibc: use clang-compatible asm syntax in arch-arm.h
tools/nolibc: limit powerpc stack-protector workaround to GCC
tools/nolibc: move entrypoint specifics to compiler.h
tools/nolibc: use attribute((naked)) if available
selftests/nolibc: report failure if no testcase passed
selftests/nolibc: avoid passing NULL to printf("%s")
selftests/nolibc: determine $(srctree) first
selftests/nolibc: setup objtree without Makefile.include
selftests/nolibc: add support for LLVM= parameter
selftests/nolibc: add cc-option compatible with clang cross builds
selftests/nolibc: run-tests.sh: avoid overwriting CFLAGS_EXTRA
selftests/nolibc: run-tests.sh: allow building through LLVM
tools/include/nolibc/arch-aarch64.h | 4 ++--
tools/include/nolibc/arch-arm.h | 8 ++++----
tools/include/nolibc/arch-i386.h | 4 ++--
tools/include/nolibc/arch-loongarch.h | 4 ++--
tools/include/nolibc/arch-mips.h | 4 ++--
tools/include/nolibc/arch-powerpc.h | 6 +++---
tools/include/nolibc/arch-riscv.h | 4 ++--
tools/include/nolibc/arch-s390.h | 4 ++--
tools/include/nolibc/arch-x86_64.h | 4 ++--
tools/include/nolibc/compiler.h | 12 ++++++++++++
tools/testing/selftests/nolibc/Makefile | 27 ++++++++++++++++-----------
tools/testing/selftests/nolibc/nolibc-test.c | 4 ++--
tools/testing/selftests/nolibc/run-tests.sh | 20 ++++++++++++++++----
13 files changed, 67 insertions(+), 38 deletions(-)
---
base-commit: 0db287736bc586fcd5a2925518ef09eec6924803
change-id: 20240727-nolibc-llvm-3fad68590d4c
Best regards,
--
Thomas Weißschuh <linux(a)weissschuh.net>
Don't print that 88 sub-tests are going to be executed, but then skip.
This is against TAP compliance. Instead check pre-requisites first
before printing total number of tests.
Old non-tap compliant output:
TAP version 13
1..88
ok 2 # SKIP all tests require euid == 0
# Planned tests != run tests (88 != 1)
# Totals: pass:0 fail:0 xfail:0 xpass:0 skip:1 error:0
New and correct output:
TAP version 13
1..0 # SKIP all tests require euid == 0
Signed-off-by: Muhammad Usama Anjum <usama.anjum(a)collabora.com>
---
Changes since v1:
- Remove simplifying if condition lines
- Update the patch message
---
tools/testing/selftests/openat2/resolve_test.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/tools/testing/selftests/openat2/resolve_test.c b/tools/testing/selftests/openat2/resolve_test.c
index bbafad440893c..85a4c64ee950d 100644
--- a/tools/testing/selftests/openat2/resolve_test.c
+++ b/tools/testing/selftests/openat2/resolve_test.c
@@ -508,12 +508,13 @@ void test_openat2_opath_tests(void)
int main(int argc, char **argv)
{
ksft_print_header();
- ksft_set_plan(NUM_TESTS);
/* NOTE: We should be checking for CAP_SYS_ADMIN here... */
if (geteuid() != 0)
ksft_exit_skip("all tests require euid == 0\n");
+ ksft_set_plan(NUM_TESTS);
+
test_openat2_opath_tests();
if (ksft_get_fail_cnt() + ksft_get_error_cnt() > 0)
--
2.39.2
A small bugfix for "run-user XARCH=ppc64le" and run-user support for
run-tests.sh.
Signed-off-by: Thomas Weißschuh <linux(a)weissschuh.net>
---
Thomas Weißschuh (2):
selftests/nolibc: introduce QEMU_ARCH_USER
selftests/nolibc: run-tests.sh: enable testing via qemu-user
tools/testing/selftests/nolibc/Makefile | 5 ++++-
tools/testing/selftests/nolibc/run-tests.sh | 22 +++++++++++++++++++---
2 files changed, 23 insertions(+), 4 deletions(-)
---
base-commit: ba335752620565c25c3028fff9496bb8ef373602
change-id: 20770915-nolibc-run-user-845375a3ec4f
Best regards,
--
Thomas Weißschuh <linux(a)weissschuh.net>
Extend pmu_counters_test to AMD CPUs.
As the AMD PMU is quite different from Intel with different events and
feature sets, this series introduces a new code path to test it,
specifically focusing on the core counters including the
PerfCtrExtCore and PerfMonV2 features. Northbridge counters and cache
counters exist, but are not as important and can be deferred to a
later series.
The first patch is a bug fix that could be submitted separately.
The series has been tested on both Intel and AMD machines, but I have
not found an AMD machine old enough to lack PerfCtrExtCore. I have
made efforts that no part of the code has any dependency on its
presence.
I am aware of similar work in this direction done by Jinrong Liang
[1]. He told me he is not working on it currently and I am not
intruding by making my own submission.
[1] https://lore.kernel.org/kvm/20231121115457.76269-1-cloudliang@tencent.com/
Colton Lewis (6):
KVM: x86: selftests: Fix typos in macro variable use
KVM: x86: selftests: Define AMD PMU CPUID leaves
KVM: x86: selftests: Set up AMD VM in pmu_counters_test
KVM: x86: selftests: Test read/write core counters
KVM: x86: selftests: Test core events
KVM: x86: selftests: Test PerfMonV2
.../selftests/kvm/include/x86_64/processor.h | 7 +
.../selftests/kvm/x86_64/pmu_counters_test.c | 267 ++++++++++++++++--
2 files changed, 249 insertions(+), 25 deletions(-)
--
2.46.0.rc2.264.g509ed76dc8-goog
Introduce a new test to identify regressions causing devices to go
missing on the system.
For each bus and class on the system the test checks the number of
devices present against a reference file, which needs to have been
generated by the program at a previous point on a known-good kernel, and
if there are missing devices they are reported.
Signed-off-by: Nícolas F. R. A. Prado <nfraprado(a)collabora.com>
---
Hi,
Key points about this test:
* Goal: Identify regressions causing devices to go missing on the system
* Focus:
* Ease of maintenance: the reference file is generated programatically
* Minimum of false-positives: the script makes as few assumptions as possible
about the stability of device identifiers to ensure renames/refactors don't
trigger false-positives
* How it works: For each bus and class on the system the test checks the number
of devices present against a reference file, which needs to have been
generated by the program at a previous point on a known-good kernel, and if
there are missing devices they are reported.
* Comparison to other tests: It might be possible(*) to replace the discoverable
devices test [1] with this. The benefits of this test is that it's easier
to setup and maintain and has wider coverage of devices.
Additional detail:
* Having more devices on the running system than the reference does not cause a
failure, but a warning is printed in that case to suggest that the reference
be updated.
* Missing devices are detected per bus/class based on the number of devices.
When the test fails, the known metadata for each of the expected and detected
devices is printed and some simple similitarity comparison is done to suggest
the devices that are the most likely to be missing.
* The proposed place to store the generated reference files is the
'platform-test-parameters' repository in KernelCI [2].
Example output: This is an example of a failing test case when one of the two
devices in the nvmem bus went missing:
# Missing devices for subsystem 'nvmem': 1 (Expected 2, found 1)
# =================
# Devices expected:
#
# uevent:
# OF_NAME=efuse
# OF_FULLNAME=/soc/efuse@11c10000
# OF_COMPATIBLE_0=mediatek,mt8195-efuse
# OF_COMPATIBLE_1=mediatek,efuse
# OF_COMPATIBLE_N=2
#
# uevent:
# OF_NAME=flash
# OF_FULLNAME=/soc/spi@1132c000/flash@0
# OF_COMPATIBLE_0=jedec,spi-nor
# OF_COMPATIBLE_N=1
#
# -----------------
# Devices found:
#
# uevent:
# OF_NAME=efuse
# OF_FULLNAME=/soc/efuse@11c10000
# OF_COMPATIBLE_0=mediatek,mt8195-efuse
# OF_COMPATIBLE_1=mediatek,efuse
# OF_COMPATIBLE_N=2
#
# -----------------
# Devices missing (best guess):
#
# uevent:
# OF_NAME=flash
# OF_FULLNAME=/soc/spi@1132c000/flash@0
# OF_COMPATIBLE_0=jedec,spi-nor
# OF_COMPATIBLE_N=1
#
# =================
not ok 19 bus.nvmem
Example of how the data for these devices is encoded in the reference file:
bus:
...
nvmem:
count: 2
devices:
- info:
uevent: 'OF_NAME=efuse
OF_FULLNAME=/soc/efuse@11c10000
OF_COMPATIBLE_0=mediatek,mt8195-efuse
OF_COMPATIBLE_1=mediatek,efuse
OF_COMPATIBLE_N=2
'
- info:
uevent: 'OF_NAME=flash
OF_FULLNAME=/soc/spi@1132c000/flash@0
OF_COMPATIBLE_0=jedec,spi-nor
OF_COMPATIBLE_N=1
'
(Full reference file: http://0x0.st/Xp60.yaml;)
Caveat: Relying only on the count of devices in a subsystem makes the test
susceptible to false-negatives eg. if a device goes missing and another in the
same subsystem is added the count will be the same so this regression won't be
reported. In order to avoid this we may include properties that must match
individual devices, but we must be very careful (and it's why I haven't done it)
since matching against properties that aren't guaranteed to be stable will
introduce false-positives (ie. detecting false regressions) due to eventual
renames.
Some things to improve in the near future / gather feedback on:
* (*): Currently this test only checks for the existence of devices. We could
extend it to also encode into the reference which devices are bound to drivers
to be able to completely replace the discoverable devices probe kselftest [1].
* Expanding identifying properties: Currently the properties that are stored
(when present) in the reference for each device to be used for identification
in the result output are uevent, device/uevent, firmware_node/uevent and name.
Suggestions of others properties to add are welcome.
* Adding more filtering to reduce noise:
* Ignoring buses/classes: Currently the devlink class is ignored by the test
since it seems like a kernel internal detail that userspace doesn't actually
care about. We should add others that are similar.
* Ignoring non-devices: There can be entries in /sys/class/ that aren't
devices. For now we're filtering down to only symlinks, but there might be a
better way.
* As mentioned in the caveat section above we may want to add actual matching
of devices based on properties to avoid false-negatives if we identify
suitable properties.
* It would be nice to have an option in the program to compare a newer reference
to an older one to make it easier for the user to see the differences and
decide if the new reference is ok.
* Since the reference file is not supposed to be manually edited, JSON might be
a better choice than YAML since it is included in the python standard library.
Let me know your thoughts.
Thanks,
Nícolas
[1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/too…
[2] https://github.com/kernelci/platform-test-parameters
---
tools/testing/selftests/Makefile | 1 +
tools/testing/selftests/devices/exist/Makefile | 3 +
tools/testing/selftests/devices/exist/exist.py | 268 +++++++++++++++++++++++++
3 files changed, 272 insertions(+)
diff --git a/tools/testing/selftests/Makefile b/tools/testing/selftests/Makefile
index bc8fe9e8f7f2..9c49b5ec5bef 100644
--- a/tools/testing/selftests/Makefile
+++ b/tools/testing/selftests/Makefile
@@ -14,6 +14,7 @@ TARGETS += cpufreq
TARGETS += cpu-hotplug
TARGETS += damon
TARGETS += devices/error_logs
+TARGETS += devices/exist
TARGETS += devices/probe
TARGETS += dmabuf-heaps
TARGETS += drivers/dma-buf
diff --git a/tools/testing/selftests/devices/exist/Makefile b/tools/testing/selftests/devices/exist/Makefile
new file mode 100644
index 000000000000..3075cac32092
--- /dev/null
+++ b/tools/testing/selftests/devices/exist/Makefile
@@ -0,0 +1,3 @@
+TEST_PROGS := exist.py
+
+include ../../lib.mk
diff --git a/tools/testing/selftests/devices/exist/exist.py b/tools/testing/selftests/devices/exist/exist.py
new file mode 100755
index 000000000000..8241b2fabc8e
--- /dev/null
+++ b/tools/testing/selftests/devices/exist/exist.py
@@ -0,0 +1,268 @@
+#!/usr/bin/env python3
+# SPDX-License-Identifier: GPL-2.0
+# Copyright (c) 2024 Collabora Ltd
+
+# * Goal: Identify regressions causing devices to go missing on the system
+# * Focus:
+# * Ease of maintenance: the reference file is generated programatically
+# * Minimum of false-positives: the script makes as few assumptions as
+# possible about the stability of device identifiers to ensure
+# renames/refactors don't trigger false-positives
+# * How it works: For each bus and class on the system the test checks the
+# number of devices present against a reference file, which needs to have been
+# generated by the program at a previous point on a known-good kernel, and if
+# there are missing devices they are reported.
+
+import os
+import sys
+import argparse
+
+import yaml
+
+# Allow ksft module to be imported from different directory
+this_dir = os.path.dirname(os.path.realpath(__file__))
+sys.path.append(os.path.join(this_dir, "../../kselftest/"))
+
+import ksft
+
+
+def generate_devs_obj():
+ obj = {}
+
+ device_sources = [
+ {
+ "base_dir": "/sys/class",
+ "add_path": "",
+ "key_name": "class",
+ "ignored": ["devlink"],
+ },
+ {
+ "base_dir": "/sys/bus",
+ "add_path": "devices",
+ "key_name": "bus",
+ "ignored": [],
+ },
+ ]
+
+ properties = sorted(["uevent", "device/uevent", "firmware_node/uevent", "name"])
+
+ for source in device_sources:
+ source_subsystems = {}
+ for subsystem in sorted(os.listdir(source["base_dir"])):
+ if subsystem in source["ignored"]:
+ continue
+
+ devs_path = os.path.join(source["base_dir"], subsystem, source["add_path"])
+ dev_dirs = [dev for dev in os.scandir(devs_path) if dev.is_symlink()]
+ devs_data = []
+ for dev_dir in dev_dirs:
+ dev_path = os.path.join(devs_path, dev_dir)
+ dev_data = {"info": {}}
+ for prop in properties:
+ if os.path.isfile(os.path.join(dev_path, prop)):
+ with open(os.path.join(dev_path, prop)) as f:
+ dev_data["info"][prop] = f.read()
+ devs_data.append(dev_data)
+ if len(dev_dirs):
+ source_subsystems[subsystem] = {
+ "count": len(dev_dirs),
+ "devices": devs_data,
+ }
+ obj[source["key_name"]] = source_subsystems
+
+ return obj
+
+
+def commented(s):
+ return s.replace("\n", "\n# ")
+
+
+def indented(s, n):
+ return " " * n + s.replace("\n", "\n" + " " * n)
+
+
+def stripped(s):
+ return s.strip("\n")
+
+
+def devices_difference(dev1, dev2):
+ difference = 0
+
+ for prop in dev1["info"].keys():
+ for l1, l2 in zip(
+ dev1["info"].get(prop, "").split("\n"),
+ dev2["info"].get(prop, "").split("\n"),
+ ):
+ if l1 != l2:
+ difference += 1
+ return difference
+
+
+def guess_missing_devices(cur_devs_subsystem, ref_devs_subsystem):
+ # Detect what devices on the current system are the most similar to devices
+ # on the reference one by one until the leftovers are the most dissimilar
+ # devices and therefore most likely the missing ones.
+ found_count = cur_devs_subsystem["count"]
+ expected_count = ref_devs_subsystem["count"]
+ missing_count = found_count - expected_count
+
+ diffs = []
+ for cur_d in cur_devs_subsystem["devices"]:
+ for ref_d in ref_devs_subsystem["devices"]:
+ diffs.append((devices_difference(cur_d, ref_d), cur_d, ref_d))
+
+ diffs.sort(key=lambda x: x[0])
+
+ assigned_ref_devs = []
+ assigned_cur_devs = []
+ for diff in diffs:
+ if len(assigned_ref_devs) >= expected_count - missing_count:
+ break
+ if diff[1] in assigned_cur_devs or diff[2] in assigned_ref_devs:
+ continue
+ assigned_cur_devs.append(diff[1])
+ assigned_ref_devs.append(diff[2])
+
+ missing_devices = []
+ for d in ref_devs_subsystem["devices"]:
+ if d not in assigned_ref_devs:
+ missing_devices.append(d)
+
+ return missing_devices
+
+
+def dump_devices_info(cur_devs_subsystem, ref_devs_subsystem):
+ def dump_device_info(dev):
+ for name, val in dev["info"].items():
+ ksft.print_msg(indented(name + ":", 2))
+ val = stripped(val)
+ if val:
+ ksft.print_msg(commented(indented(val, 4)))
+ ksft.print_msg("")
+
+ ksft.print_msg("=================")
+ ksft.print_msg("Devices expected:")
+ ksft.print_msg("")
+ for d in ref_devs_subsystem["devices"]:
+ dump_device_info(d)
+ ksft.print_msg("-----------------")
+ ksft.print_msg("Devices found:")
+ ksft.print_msg("")
+ for d in cur_devs_subsystem["devices"]:
+ dump_device_info(d)
+ ksft.print_msg("-----------------")
+ ksft.print_msg("Devices missing (best guess):")
+ ksft.print_msg("")
+ missing_devices = guess_missing_devices(cur_devs_subsystem, ref_devs_subsystem)
+ for d in missing_devices:
+ dump_device_info(d)
+ ksft.print_msg("=================")
+
+
+def run_test(ref_filename):
+ ksft.print_msg(f"Using reference file: {ref_filename}")
+
+ with open(ref_filename) as f:
+ ref_devs_obj = yaml.safe_load(f)
+
+ num_tests = 0
+ for dev_source in ref_devs_obj.values():
+ num_tests += len(dev_source)
+ ksft.set_plan(num_tests)
+
+ cur_devs_obj = generate_devs_obj()
+
+ reference_outdated = False
+
+ for source, ref_devs_source_obj in ref_devs_obj.items():
+ for subsystem, ref_devs_subsystem_obj in ref_devs_source_obj.items():
+ test_name = f"{source}.{subsystem}"
+ if not (
+ cur_devs_obj.get(source) and cur_devs_obj.get(source).get(subsystem)
+ ):
+ ksft.print_msg(f"Device subsystem '{subsystem}' missing")
+ ksft.test_result_fail(test_name)
+ continue
+ cur_devs_subsystem_obj = cur_devs_obj[source][subsystem]
+
+ found_count = cur_devs_subsystem_obj["count"]
+ expected_count = ref_devs_subsystem_obj["count"]
+ if found_count < expected_count:
+ ksft.print_msg(
+ f"Missing devices for subsystem '{subsystem}': {expected_count - found_count} (Expected {expected_count}, found {found_count})"
+ )
+ dump_devices_info(cur_devs_subsystem_obj, ref_devs_subsystem_obj)
+ ksft.test_result_fail(test_name)
+ else:
+ ksft.test_result_pass(test_name)
+ if found_count > expected_count:
+ reference_outdated = True
+
+ if len(cur_devs_obj[source]) > len(ref_devs_source_obj):
+ reference_outdated = True
+
+ if reference_outdated:
+ ksft.print_msg(
+ "Warning: The current system contains more devices and/or subsystems than the reference. Updating the reference is recommended."
+ )
+
+
+def get_possible_ref_filenames():
+ filenames = []
+
+ dt_board_compatible_file = "/proc/device-tree/compatible"
+ if os.path.exists(dt_board_compatible_file):
+ with open(dt_board_compatible_file) as f:
+ for line in f:
+ compatibles = [compat for compat in line.split("\0") if compat]
+ filenames.extend(compatibles)
+ else:
+ dmi_id_dir = "/sys/devices/virtual/dmi/id"
+ vendor_dmi_file = os.path.join(dmi_id_dir, "sys_vendor")
+ product_dmi_file = os.path.join(dmi_id_dir, "product_name")
+
+ with open(vendor_dmi_file) as f:
+ vendor = f.read().replace("\n", "")
+ with open(product_dmi_file) as f:
+ product = f.read().replace("\n", "")
+
+ filenames = [vendor + "," + product]
+
+ return filenames
+
+
+def get_ref_filename(ref_dir):
+ chosen_ref_filename = ""
+ full_ref_paths = [os.path.join(ref_dir, f + ".yaml") for f in get_possible_ref_filenames()]
+ for path in full_ref_paths:
+ if os.path.exists(path):
+ chosen_ref_filename = path
+ break
+
+ if not chosen_ref_filename:
+ tried_paths = ",".join(["'" + p + "'" for p in full_ref_paths])
+ ksft.print_msg(f"No matching reference file found (tried {tried_paths})")
+ ksft.exit_fail()
+
+ return chosen_ref_filename
+
+
+parser = argparse.ArgumentParser()
+parser.add_argument(
+ "--reference-dir", default=".", help="Directory containing the reference files"
+)
+parser.add_argument("--generate-reference", action="store_true", help="Generate a reference file with the devices on the running system")
+args = parser.parse_args()
+
+if args.generate_reference:
+ print(f"# Kernel version: {os.uname().release}")
+ print(yaml.dump(generate_devs_obj()))
+ sys.exit(0)
+
+ksft.print_header()
+
+ref_filename = get_ref_filename(args.reference_dir)
+
+run_test(ref_filename)
+
+ksft.finished()
---
base-commit: 73399b58e5e5a1b28a04baf42e321cfcfc663c2f
change-id: 20240724-kselftest-dev-exist-bb1bcf884654
Best regards,
--
Nícolas F. R. A. Prado <nfraprado(a)collabora.com>
Hello Davide,
In the following commit:
commit ca22da2fbd693b54dc8e3b7b54ccc9f7e9ba3640
Author: Davide Caratti <dcaratti(a)redhat.com>
Date: Fri Jan 20 18:01:40 2023 +0100
act_mirred: use the backlog for nested calls to mirred ingress
you added the mirred_egress_to_ingress_tcp_test kselftest that, I noticed, hangs with "Ncat: TIMEOUT." if openvswitch module is loaded.
Is this the right behaviour or was such configuration not intended?
Regards,
Lenar Khannanov
Adds a selftest that creates two virtual interfaces, assigns one to a
new namespace, and assigns IP addresses to both.
It listens on the destination interface using netcat and configures a
dynamic target on netconsole, pointing to the destination IP address.
The test then checks if the message was received properly on the
destination interface.
Signed-off-by: Breno Leitao <leitao(a)debian.org>
---
MAINTAINERS | 1 +
.../net/netconsole/basic_integration_test.sh | 153 ++++++++++++++++++
2 files changed, 154 insertions(+)
create mode 100755 tools/testing/selftests/net/netconsole/basic_integration_test.sh
diff --git a/MAINTAINERS b/MAINTAINERS
index c0a3d9e93689..59207365c9f5 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -15768,6 +15768,7 @@ M: Breno Leitao <leitao(a)debian.org>
S: Maintained
F: Documentation/networking/netconsole.rst
F: drivers/net/netconsole.c
+F: tools/testing/selftests/net/netconsole/
NETDEVSIM
M: Jakub Kicinski <kuba(a)kernel.org>
diff --git a/tools/testing/selftests/net/netconsole/basic_integration_test.sh b/tools/testing/selftests/net/netconsole/basic_integration_test.sh
new file mode 100755
index 000000000000..fbabbc633451
--- /dev/null
+++ b/tools/testing/selftests/net/netconsole/basic_integration_test.sh
@@ -0,0 +1,153 @@
+#!/usr/bin/env bash
+# SPDX-License-Identifier: GPL-2.0
+
+# This test creates two virtual interfaces, assigns one of them (the "destination
+# interface") to a new namespace, and assigns IP addresses to both interfaces.
+#
+# It listens on the destination interface using netcat (nc) and configures a
+# dynamic target on netconsole, pointing to the destination IP address.
+#
+# Finally, it checks whether the message was received properly on the
+# destination interface. Note that this test may pollute the kernel log buffer
+# (dmesg) and relies on dynamic configuration and namespaces being configured."
+#
+# Author: Breno Leitao <leitao(a)debian.org>
+
+SCRIPTDIR=$(dirname "$(readlink -e "${BASH_SOURCE[0]}")")
+
+# Simple script to test dynamic targets in netconsole
+SRCIF="veth0"
+SRCIP=192.168.1.1
+
+DSTIF="veth1"
+DSTIP=192.168.1.2
+
+PORT="6666"
+MSG="netconsole selftest"
+TARGET=$(mktemp -u netcons_XXXXX)
+NETCONS_CONFIGFS="/sys/kernel/config/netconsole"
+FULLPATH="${NETCONS_CONFIGFS}"/"${TARGET}"
+# This will be have some tmp values appened to it in set_network()
+NAMESPACE="netconsns"
+
+# Used to create and delete namespaces
+source "${SCRIPTDIR}"/../lib.sh
+
+function set_network() {
+ # this is coming from lib.sh
+ setup_ns "${NAMESPACE}"
+ NAMESPACE=${NS_LIST[0]}
+ ip link add "${SRCIF}" type veth peer name "${DSTIF}"
+
+ # "${DSTIF}"
+ ip link set "${DSTIF}" netns "${NAMESPACE}"
+ ip netns exec "${NAMESPACE}" ip addr add "${DSTIP}"/24 dev "${DSTIF}"
+ ip netns exec "${NAMESPACE}" ip link set "${DSTIF}" up
+
+ # later, configure "${SRCIF}"
+ ip addr add "${SRCIP}"/24 dev "${SRCIF}"
+ ip link set "${SRCIF}" up
+}
+
+function create_dynamic_target() {
+ DSTMAC=$(ip netns exec "${NAMESPACE}" ip link show "${DSTIF}" | awk '/ether/ {print $2}')
+
+ # Create a dynamic target
+ mkdir ${FULLPATH}
+
+ echo ${DSTIP} > ${FULLPATH}/remote_ip
+ echo ${SRCIP} > ${FULLPATH}/local_ip
+ echo ${DSTMAC} > ${FULLPATH}/remote_mac
+ echo "${SRCIF}" > ${FULLPATH}/dev_name
+
+ echo 1 > ${FULLPATH}/enabled
+}
+
+function cleanup() {
+ echo 0 > "${FULLPATH}"/enabled
+ rmdir "${FULLPATH}"
+ # This will delete DSTIF also
+ ip link del "${SRCIF}"
+ # this is coming from lib.sh
+ cleanup_all_ns
+}
+
+function listen_port() {
+ OUTPUT=${1}
+ echo "Saving content in ${OUTPUT}"
+ timeout 2 ip netns exec "${NAMESPACE}" nc -u -l "${PORT}" | sed '/^$/q' > ${OUTPUT}
+}
+
+function validate_result() {
+ TMPFILENAME=/tmp/"${TARGET}"
+
+ # sleep until the file isc reated
+ sleep 1
+ # Check if the file exists
+ if [ ! -f "$TMPFILENAME" ]; then
+ echo "FAIL: File was not generate." >&2
+ return ${ksft_fail}
+ fi
+
+ if ! grep -q "${MSG}" "${TMPFILENAME}"; then
+ echo "FAIL: ${MSG} not found in ${TMPFILENAME}" >&2
+ cat ${TMPFILENAME} >&2
+ return ${ksft_fail}
+ fi
+
+ rm ${TMPFILENAME}
+ return ${ksft_pass}
+}
+
+function check_for_dependencies() {
+ if [ "$(id -u)" -ne 0 ]; then
+ echo "This script must be run as root" >&2
+ exit 1
+ fi
+
+ if ! which nc > /dev/null ; then
+ echo "SKIP: nc(1) is not available" >&2
+ exit ${ksft_skip}
+ fi
+
+ if ! which ip > /dev/null ; then
+ echo "SKIP: ip(1) is not available" >&2
+ exit ${ksft_skip}
+ fi
+
+ if [ ! -d "${NETCONS_CONFIGFS}" ]; then
+ echo "SKIP: directory ${NETCONS_CONFIGFS} does not exist. Check if NETCONSOLE_DYNAMIC is enabled" >&2
+ exit ${ksft_skip}
+ fi
+
+ if ip link show veth0 2> /dev/null; then
+ echo "SKIP: interface veth0 exists in the system. Not overwriting it."
+ exit ${ksft_skip}
+ fi
+}
+
+# ========== #
+# Start here #
+# ========== #
+
+# Check for basic system dependency and exit if not found
+check_for_dependencies
+# Create one namespace and two interfaces
+set_network
+# Create a dynamic target for netconsole
+create_dynamic_target
+# Listed for netconsole port inside the namespace and destination interface
+listen_port /tmp/"${TARGET}" &
+# Wait for nc to start and listen to the port.
+sleep 1
+
+# Send the message
+echo "${MSG}: ${TARGET}" > /dev/kmsg
+
+# Make sure the message was received in the dst part
+validate_result
+ret=$?
+# Remove the namespace, interfaces and netconsole target
+cleanup
+
+exit ${ret}
--
2.43.0
When looking at improving the user experience around the MPTCP endpoints
setup, I noticed that setting an endpoint with both the 'signal' and the
'subflow' flags -- as it has been done in the past by users according to
bug reports we got -- was resulting on only announcing the endpoint, but
not using it to create subflows: the 'subflow' flag was then ignored.
My initial thought was to modify IPRoute2 to warn the user when the two
flags were set, but it doesn't sound normal to ignore one of them. I
then looked at modifying the kernel not to allow having the two flags
set, but when discussing about that with Mat, we thought it was maybe
not ideal to do that, as there might be use-cases, we might break some
configs. Then I saw it was working before v5.17. So instead, I fixed the
support on the kernel side (patch 5) using Paolo's suggestion. This also
includes a fix on the options side (patch 1: for v5.11+), an explicit
deny of some options combinations (patch 2: for v5.18+), and some
refactoring (patches 3 and 4) to ease the inclusion of the patch 5.
While at it, I added a new selftest (patch 7) to validate this case --
including a modification of the chk_add_nr helper to inverse the sides
were the counters are checked (patch 6) -- and allowed ADD_ADDR echo
just after the MP_JOIN 3WHS.
The selftests modification have the same Fixes tag as the previous
commit, but no 'Cc: Stable': if the backport can work, that's good --
but it still need to be verified by running the selftests -- if not, no
need to worry, many CIs will use the selftests from the last stable
version to validate previous stable releases.
Signed-off-by: Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
---
Matthieu Baerts (NGI0) (7):
mptcp: fully established after ADD_ADDR echo on MPJ
mptcp: pm: deny endp with signal + subflow + port
mptcp: pm: reduce indentation blocks
mptcp: pm: don't try to create sf if alloc failed
mptcp: pm: do not ignore 'subflow' if 'signal' flag is also set
selftests: mptcp: join: ability to invert ADD_ADDR check
selftests: mptcp: join: test both signal & subflow
net/mptcp/options.c | 3 +-
net/mptcp/pm_netlink.c | 47 +++++++++++++--------
tools/testing/selftests/net/mptcp/mptcp_join.sh | 55 ++++++++++++++++++-------
3 files changed, 73 insertions(+), 32 deletions(-)
---
base-commit: 0bf50cead4c4710d9f704778c32ab8af47ddf070
change-id: 20240731-upstream-net-20240731-mptcp-endp-subflow-signal-181d640cf5e8
Best regards,
--
Matthieu Baerts (NGI0) <matttbe(a)kernel.org>
Corrected the typographical of the word "different"
in the "name" field of the JSON object with ID "4319".
Signed-off-by: Karan Sanghavi <karansanghvi98(a)gmail.com>
---
tools/testing/selftests/tc-testing/tc-tests/filters/cgroup.json | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/testing/selftests/tc-testing/tc-tests/filters/cgroup.json b/tools/testing/selftests/tc-testing/tc-tests/filters/cgroup.json
index 03723cf84..6897ff5ad 100644
--- a/tools/testing/selftests/tc-testing/tc-tests/filters/cgroup.json
+++ b/tools/testing/selftests/tc-testing/tc-tests/filters/cgroup.json
@@ -1189,7 +1189,7 @@
},
{
"id": "4319",
- "name": "Replace cgroup filter with diffferent match",
+ "name": "Replace cgroup filter with different match",
"category": [
"filter",
"cgroup"
--
2.43.0
From: Zijian Zhang <zijianzhang(a)bytedance.com>
Original notification mechanism needs poll + recvmmsg which is not
easy for applcations to accommodate. And, it also incurs unignorable
overhead including extra system calls.
While making maximum reuse of the existing MSG_ZEROCOPY related code,
this patch set introduces a new zerocopy socket notification mechanism.
Users of sendmsg pass a control message as a placeholder for the incoming
notifications. Upon returning, kernel embeds notifications directly into
user arguments passed in. By doing so, we can reduce the complexity and
the overhead for managing notifications.
We also have the logic related to copying cmsg to the userspace in sendmsg
generic for any possible uses cases in the future. However, it introduces
ABI change of sendmsg.
Changelog:
v1 -> v2:
- Reuse errormsg queue in the new notification mechanism,
users can actually use these two mechanisms in hybrid way
if they want to do so.
- Update case SCM_ZC_NOTIFICATION in __sock_cmsg_send
1. Regardless of 32-bit, 64-bit program, we will always handle
u64 type user address.
2. The size of data to copy_to_user is precisely calculated
in case of kernel stack leak.
- fix (kbuild-bot)
1. Add SCM_ZC_NOTIFICATION to arch-specific header files.
2. header file types.h in include/uapi/linux/socket.h
v2 -> v3:
- 1. Users can now pass in the address of the zc_info_elem directly
with appropriate cmsg_len instead of the ugly user interface. Plus,
the handler is now compatible with MSG_CMSG_COMPAT and 32-bit
pointer.
- 2. Suggested by Willem, another strategy of getting zc info is
briefly taking the lock of sk_error_queue and move to a private
list, like net_rx_action. I thought sk_error_queue is protected by
sock_lock, so that it's impossible for the handling of zc info and
users recvmsg from the sk_error_queue at the same time.
However, sk_error_queue is protected by its own lock. I am afraid
that during the time it is handling the private list, users may
fail to get other error messages in the queue via recvmsg. Thus,
I don't implement the splice logic in this version. Any comments?
v3 -> v4:
- 1. Change SOCK_ZC_INFO_MAX to 64 to avoid large stack frame size.
- 2. Fix minor typos.
- 3. Change cfg_zerocopy from int to enum in msg_zerocopy.c
Initially, we expect users to pass the user address of the user array
as a data in cmsg, so that the kernel can copy_to_user to this address
directly.
As Willem commented,
> The main design issue with this series is this indirection, rather
> than passing the array of notifications as cmsg.
> This trick circumvents having to deal with compat issues and having to
> figure out copy_to_user in ____sys_sendmsg (as msg_control is an
> in-kernel copy).
> This is quite hacky, from an API design PoV.
> As is passing a pointer, but expecting msg_controllen to hold the
> length not of the pointer, but of the pointed to user buffer.
> I had also hoped for more significant savings. Especially with the
> higher syscall overhead due to meltdown and spectre mitigations vs
> when MSG_ZEROCOPY was introduced and I last tried this optimization.
We solve it by supporting put_cmsg to userspace in sendmsg path starting
from v5.
v4 -> v5:
- 1. Passing user address directly to kernel raises concerns about
ABI. In this version, we support put_cmsg to userspace in TX path
to solve this problem.
v5 -> v6:
- 1. Cleanly copy cmsg to user upon returning of ___sys_sendmsg
v6 -> v7:
- 1. Remove flag MSG_CMSG_COPY_TO_USER, use a member in msghdr instead
- 2. Pass msg to __sock_cmsg_send.
- 3. sendmsg_copy_cmsg_to_user should be put at the end of
____sys_sendmsg to make sure msg_sys->msg_control is a valid pointer.
- 4. Add struct zc_info to contain the array of zc_info_elem, so that
the kernel can update the zc_info->size. Another possible solution is
updating the cmsg_len directly, but it will break for_each_cmsghdr.
- 5. Update selftest to make cfg_notification_limit have the same
semantics in both methods for better comparison.
v7 -> v8:
- 1. Add a static_branch in ____sys_sendmsg to avoid overhead in the
hot path.
- 2. Add ZC_NOTIFICATION_MAX to limit the max size of zc_info->arr.
- 3. Minimize the code in SCM_ZC_NOTIFICATION handler by adding a local
sk_buff_head.
* Performance
We update selftests/net/msg_zerocopy.c to accommodate the new mechanism,
cfg_notification_limit has the same semantics for both methods. Test
results are as follows, we update skb_orphan_frags_rx to the same as
skb_orphan_frags to support zerocopy in the localhost test.
cfg_notification_limit = 1, both method get notifications after 1 calling
of sendmsg. In this case, the new method has around 17% cpu savings in TCP
and 23% cpu savings in UDP.
+----------------------+---------+---------+---------+---------+
| Test Type / Protocol | TCP v4 | TCP v6 | UDP v4 | UDP v6 |
+----------------------+---------+---------+---------+---------+
| ZCopy (MB) | 7523 | 7706 | 7489 | 7304 |
+----------------------+---------+---------+---------+---------+
| New ZCopy (MB) | 8834 | 8993 | 9053 | 9228 |
+----------------------+---------+---------+---------+---------+
| New ZCopy / ZCopy | 117.42% | 116.70% | 120.88% | 126.34% |
+----------------------+---------+---------+---------+---------+
cfg_notification_limit = 32, both get notifications after 32 calling of
sendmsg, which means more chances to coalesce notifications, and less
overhead of poll + recvmsg for the original method. In this case, the new
method has around 7% cpu savings in TCP and slightly better cpu usage in
UDP. In the env of selftest, notifications of TCP are more likely to be
out of order than UDP, it's easier to coalesce more notifications in UDP.
The original method can get one notification with range of 32 in a recvmsg
most of the time. In TCP, most notifications' range is around 2, so the
original method needs around 16 recvmsgs to get notified in one round.
That's the reason for the "New ZCopy / ZCopy" diff in TCP and UDP here.
+----------------------+---------+---------+---------+---------+
| Test Type / Protocol | TCP v4 | TCP v6 | UDP v4 | UDP v6 |
+----------------------+---------+---------+---------+---------+
| ZCopy (MB) | 8842 | 8735 | 10072 | 9380 |
+----------------------+---------+---------+---------+---------+
| New ZCopy (MB) | 9366 | 9477 | 10108 | 9385 |
+----------------------+---------+---------+---------+---------+
| New ZCopy / ZCopy | 106.00% | 108.28% | 100.31% | 100.01% |
+----------------------+---------+---------+---------+---------+
In conclusion, when notification interval is small or notifications are
hard to be coalesced, the new mechanism is highly recommended. Otherwise,
the performance gain from the new mechanism is very limited.
Zijian Zhang (3):
sock: support copying cmsgs to the user space in sendmsg
sock: add MSG_ZEROCOPY notification mechanism based on msg_control
selftests: add MSG_ZEROCOPY msg_control notification test
arch/alpha/include/uapi/asm/socket.h | 2 +
arch/mips/include/uapi/asm/socket.h | 2 +
arch/parisc/include/uapi/asm/socket.h | 2 +
arch/sparc/include/uapi/asm/socket.h | 2 +
include/linux/socket.h | 8 ++
include/net/sock.h | 2 +-
include/uapi/asm-generic/socket.h | 2 +
include/uapi/linux/socket.h | 23 +++++
net/core/sock.c | 72 +++++++++++++-
net/ipv4/ip_sockglue.c | 2 +-
net/ipv6/datagram.c | 2 +-
net/socket.c | 63 +++++++++++-
tools/testing/selftests/net/msg_zerocopy.c | 101 ++++++++++++++++++--
tools/testing/selftests/net/msg_zerocopy.sh | 1 +
14 files changed, 265 insertions(+), 19 deletions(-)
--
2.20.1
This patch series adds a selftest suite to validate the s390x
architecture specific ucontrol KVM interface.
When creating a VM on s390x it is possible to create it as userspace
controlled VM or in short ucontrol VM.
These VMs delegates the management of the VM to userspace instead
of handling most events within the kernel. Consequently the userspace
has to manage interrupts, memory allocation etc.
Before this patch set this functionality lacks any public test cases.
It is desirable to add test cases for this interface to be able to
reduce the risk of breaking changes in the future.
In order to provision a ucontrol VM the kernel needs to be compiled with
the CONFIG_KVM_S390_UCONTROL enabled. The users with sys_admin capability
can then create a new ucontrol VM providing the KVM_VM_S390_UCONTROL
parameter to the KVM_CREATE_VM ioctl.
The kernels existing selftest helper functions can only be partially be
reused for these tests.
The test cases cover existing special handling of ucontrol VMs within the
implementation and basic VM creation and handling cases:
* Reject setting HPAGE when VM is ucontrol
* Assert KVM_GET_DIRTY_LOG is rejected
* Assert KVM_S390_VM_MEM_LIMIT_SIZE is rejected
* Assert state of initial SIE flags setup by the kernel
* Run simple program in VM with and without DAT
* Assert KVM_EXIT_S390_UCONTROL exit on not mapped memory access
* Assert functionality of storage keys in ucontrol VM
Running the test cases requires sys_admin capabilities to start the
ucontrol VM.
This can be achieved by running as root or with a command like:
sudo setpriv --reuid nobody --inh-caps -all,+sys_admin \
--ambient-caps -all,+sys_admin --bounding-set -all,+sys_admin \
./ucontrol_test
The patch set does also contain some code cleanup / consolidation of
architecture specific defines that are now used in multiple test cases.
---
V1 -> V2:
- add ucontrol to s390 debug config (new patch)
- PATCH 2: changed atomic_t to __u32 (thanks Claudio)
- PATCH 4: reformatted comment in FIXTURE_SETUP(uc_kvm)
- PATCH 5: refactored to display 8 byte blocks + more internal reuse
(thanksClaudio)
- PATCH 7: make use of more declarative defines instead of magic values
- PATCH 8: make use of more declarative defines instead of magic values
(thanks Claudio)
- PATCH 9: add reference to fix verified by the test case
V2 -> V3:
- Remove stopped bit before starting the VM (no initial stop in multiple
test cases) (thanks Janosch)
- PATCH 2:
- Clarified SIE control block vs SIE instruction (thanks Janosch)
- PATCH 3:
- Make use of CAP_TO_MASK(CAP_SYS_ADMIN) instead of custom define (thanks
Janosch)
- Removed Reviewed-By: Claudio
- PATCH 4:
- Remove erroneous 1MB offset from self->base_hva (thanks Janosch)
- PATCH 6-8: Change name of test program _pgm to _asm to prevent confusion
- PATCH 10: Move KVM_S390_UCONTROL default option to actual debug config
(thanks Christian)
Christoph Schlameuss (10):
selftests: kvm: s390: Define page sizes in shared header
selftests: kvm: s390: Add kvm_s390_sie_block definition for userspace
tests
selftests: kvm: s390: Add s390x ucontrol test suite with hpage test
selftests: kvm: s390: Add test fixture and simple VM setup tests
selftests: kvm: s390: Add debug print functions
selftests: kvm: s390: Add VM run test case
selftests: kvm: s390: Add uc_map_unmap VM test case
selftests: kvm: s390: Add uc_skey VM test case
selftests: kvm: s390: Verify reject memory region operations for
ucontrol VMs
s390: Enable KVM_S390_UCONTROL config in debug_defconfig
arch/s390/configs/debug_defconfig | 1 +
tools/testing/selftests/kvm/.gitignore | 1 +
tools/testing/selftests/kvm/Makefile | 1 +
.../selftests/kvm/include/s390x/debug_print.h | 69 ++
.../selftests/kvm/include/s390x/processor.h | 5 +
.../testing/selftests/kvm/include/s390x/sie.h | 240 +++++++
.../selftests/kvm/lib/s390x/processor.c | 10 +-
tools/testing/selftests/kvm/s390x/cmma_test.c | 7 +-
tools/testing/selftests/kvm/s390x/config | 2 +
.../testing/selftests/kvm/s390x/debug_test.c | 4 +-
tools/testing/selftests/kvm/s390x/memop.c | 4 +-
tools/testing/selftests/kvm/s390x/tprot.c | 5 +-
.../selftests/kvm/s390x/ucontrol_test.c | 596 ++++++++++++++++++
13 files changed, 929 insertions(+), 16 deletions(-)
create mode 100644 tools/testing/selftests/kvm/include/s390x/debug_print.h
create mode 100644 tools/testing/selftests/kvm/include/s390x/sie.h
create mode 100644 tools/testing/selftests/kvm/s390x/config
create mode 100644 tools/testing/selftests/kvm/s390x/ucontrol_test.c
base-commit: 94ede2a3e9135764736221c080ac7c0ad993dc2d
--
2.45.2
In order to validate ITS-MSI hwirq entry in the /proc/interrupts, we
have created a shell script to check is there any duplicated ITS-MSI
hwirq entry.
Joseph Jang (1):
selftest: drivers: Add support its msi hwirq checking
tools/testing/selftests/drivers/irq/Makefile | 5 +++++
.../selftests/drivers/irq/its-msi-irq-test.sh | 20 +++++++++++++++++++
2 files changed, 25 insertions(+)
create mode 100644 tools/testing/selftests/drivers/irq/Makefile
create mode 100755 tools/testing/selftests/drivers/irq/its-msi-irq-test.sh
--
2.34.1
On 31/07/2024 15:07, Karan Sanghavi wrote:
> Corrected the typographical of the word "different"
> in the "name" field of the JSON object with ID "4319".
>
> Signed-off-by: Karan Sanghavi <karansanghvi98(a)gmail.com>
Thank you for the patch.
I'd suggest you transform this patch and the other one you sent, which
has the same subject, into a patch series.
Also please change the subject to target net-next and to specify which
test files your patches are touching.
cheers,
Victor