The original order of cases in kunit_module_notify() is confusing and
misleading.
And the best practice is return the err code from
MODULE_STATE_COMING func.
And the test suits should be executed when notify MODULE_STATE_LIVE.
Jinjie Ruan (3):
kunit: Make the cases sequence more reasonable for
kunit_module_notify()
kunit: Return error from kunit_module_init()
kunit: Init and run test suites in the right state
lib/kunit/test.c | 37 +++++++++++++++++++++++--------------
1 file changed, 23 insertions(+), 14 deletions(-)
--
2.34.1
Attn: linux-kselftest(a)vger.kernel.org
Date: 03-10-2023
Subject: Letter of Intent (LOI) (03-10-2023)
This is a Letter of Intent concerning my Board of investors intent to fund some of your available investment projects.
If this is of any interest to you then kindly advise.
Yours Faithfully,
Ahmad R. Deeb
Head of Investment Team
This patch series introduces UFFDIO_REMAP feature to userfaultfd, which
has long been implemented and maintained by Andrea in his local tree [1],
but was not upstreamed due to lack of use cases where this approach would
be better than allocating a new page and copying the contents.
UFFDIO_COPY performs ~20% better than UFFDIO_REMAP when the application
needs pages to be allocated [2]. However, with UFFDIO_REMAP, if pages are
available (in userspace) for recycling, as is usually the case in heap
compaction algorithms, then we can avoid the page allocation and memcpy
(done by UFFDIO_COPY). Also, since the pages are recycled in the
userspace, we avoid the need to release (via madvise) the pages back to
the kernel [3].
We see over 40% reduction (on a Google pixel 6 device) in the compacting
thread’s completion time by using UFFDIO_REMAP vs. UFFDIO_COPY. This was
measured using a benchmark that emulates a heap compaction implementation
using userfaultfd (to allow concurrent accesses by application threads).
More details of the usecase are explained in [3].
Furthermore, UFFDIO_REMAP enables remapping swapped-out pages without
touching them within the same vma. Today, it can only be done by mremap,
however it forces splitting the vma.
Main changes since Andrea's last version [1]:
- Trivial translations from page to folio, mmap_sem to mmap_lock
- Replace pmd_trans_unstable() with pte_offset_map_nolock() and handle its
possible failure
- Move pte mapping into remap_pages_pte to allow for retries when source
page or anon_vma is contended. Since pte_offset_map_nolock() start RCU
read section, we can't block anymore after mapping a pte, so have to unmap
the ptesm do the locking and retry.
- Add and use anon_vma_trylock_write() to avoid blocking while in RCU
read section.
- Accommodate changes in mmu_notifier_range_init() API, switch to
mmu_notifier_invalidate_range_start_nonblock() to avoid blocking while in
RCU read section.
- Open-code now removed __swp_swapcount()
- Replace pmd_read_atomic() with pmdp_get_lockless()
- Add new selftest for UFFDIO_REMAP
Changes since v1 [4]:
- add mmget_not_zero in userfaultfd_remap, per Jann Horn
- removed extern from function definitions, per Matthew Wilcox
- converted to folios in remap_pages_huge_pmd, per Matthew Wilcox
- use PageAnonExclusive in remap_pages_huge_pmd, per David Hildenbrand
- handle pgtable transfers between MMs, per Jann Horn
- ignore concurrent A/D pte bit changes, per Jann Horn
- split functions into smaller units, per David Hildenbrand
- test for folio_test_large in remap_anon_pte, per Matthew Wilcox
- use pte_swp_exclusive for swapcount check, per David Hildenbrand
- eliminated use of mmu_notifier_invalidate_range_start_nonblock,
per Jann Horn
- simplified THP alignment checks, per Jann Horn
- refactored the loop inside remap_pages, per Jann Horn
- additional clarifying comments, per Jann Horn
[1] https://gitlab.com/aarcange/aa/-/commit/2aec7aea56b10438a3881a20a411aa4b1fc…
[2] https://lore.kernel.org/all/1425575884-2574-1-git-send-email-aarcange@redha…
[3] https://lore.kernel.org/linux-mm/CA+EESO4uO84SSnBhArH4HvLNhaUQ5nZKNKXqxRCyj…
[4] https://lore.kernel.org/all/20230914152620.2743033-1-surenb@google.com/
Andrea Arcangeli (2):
userfaultfd: UFFDIO_REMAP: rmap preparation
userfaultfd: UFFDIO_REMAP uABI
Suren Baghdasaryan (1):
selftests/mm: add UFFDIO_REMAP ioctl test
fs/userfaultfd.c | 63 ++
include/linux/rmap.h | 5 +
include/linux/userfaultfd_k.h | 12 +
include/uapi/linux/userfaultfd.h | 22 +
mm/huge_memory.c | 130 ++++
mm/khugepaged.c | 3 +
mm/rmap.c | 13 +
mm/userfaultfd.c | 590 +++++++++++++++++++
tools/testing/selftests/mm/uffd-common.c | 41 +-
tools/testing/selftests/mm/uffd-common.h | 1 +
tools/testing/selftests/mm/uffd-unit-tests.c | 62 ++
11 files changed, 940 insertions(+), 2 deletions(-)
--
2.42.0.515.g380fc7ccd1-goog
The abi_test currently uses a long sized test value for enablement
checks. On LE this works fine, however, on BE this results in inaccurate
assert checks due to a bit being used and assuming it's value is the
same on both LE and BE.
Use int type for 32-bit values and long type for 64-bit values to ensure
appropriate behavior on both LE and BE.
Fixes: 60b1af8de8c1 ("tracing/user_events: Add ABI self-test")
Signed-off-by: Beau Belgrave <beaub(a)linux.microsoft.com>
---
V2 Changes:
Rebase to linux-kselftest/fixes.
tools/testing/selftests/user_events/abi_test.c | 14 ++++++++------
1 file changed, 8 insertions(+), 6 deletions(-)
diff --git a/tools/testing/selftests/user_events/abi_test.c b/tools/testing/selftests/user_events/abi_test.c
index 8202f1327c39..aa297d3ad95e 100644
--- a/tools/testing/selftests/user_events/abi_test.c
+++ b/tools/testing/selftests/user_events/abi_test.c
@@ -47,7 +47,7 @@ static int change_event(bool enable)
return ret;
}
-static int reg_enable(long *enable, int size, int bit)
+static int reg_enable(void *enable, int size, int bit)
{
struct user_reg reg = {0};
int fd = open(data_file, O_RDWR);
@@ -69,7 +69,7 @@ static int reg_enable(long *enable, int size, int bit)
return ret;
}
-static int reg_disable(long *enable, int bit)
+static int reg_disable(void *enable, int bit)
{
struct user_unreg reg = {0};
int fd = open(data_file, O_RDWR);
@@ -90,7 +90,8 @@ static int reg_disable(long *enable, int bit)
}
FIXTURE(user) {
- long check;
+ int check;
+ long check_long;
bool umount;
};
@@ -99,6 +100,7 @@ FIXTURE_SETUP(user) {
change_event(false);
self->check = 0;
+ self->check_long = 0;
}
FIXTURE_TEARDOWN(user) {
@@ -136,9 +138,9 @@ TEST_F(user, bit_sizes) {
#if BITS_PER_LONG == 8
/* Allow 0-64 bits for 64-bit */
- ASSERT_EQ(0, reg_enable(&self->check, sizeof(long), 63));
- ASSERT_NE(0, reg_enable(&self->check, sizeof(long), 64));
- ASSERT_EQ(0, reg_disable(&self->check, 63));
+ ASSERT_EQ(0, reg_enable(&self->check_long, sizeof(long), 63));
+ ASSERT_NE(0, reg_enable(&self->check_long, sizeof(long), 64));
+ ASSERT_EQ(0, reg_disable(&self->check_long, 63));
#endif
/* Disallowed sizes (everything beside 4 and 8) */
base-commit: 6f874fa021dfc7bf37f4f37da3a5aaa41fe9c39c
--
2.34.1
All architectures should use a long aligned address passed to set_bit().
User processes can pass either a 32-bit or 64-bit sized value to be
updated when tracing is enabled when on a 64-bit kernel. Both cases are
ensured to be naturally aligned, however, that is not enough. The
address must be long aligned without affecting checks on the value
within the user process which require different adjustments for the bit
for little and big endian CPUs.
32 bit on 64 bit, even when properly long aligned, still require a 32 bit
offset to be done for BE. Due to this, it cannot be easily put into a
generic method.
The abi_test also used a long, which broke the test on 64-bit BE machines.
The change simply uses an int for 32-bit value checks and a long when on
64-bit kernels for 64-bit specific checks.
I've run these changes and self tests for user_events on ppc64 BE, x86_64
LE, and aarch64 LE. It'd be great to test this also on RISC-V, but I do
not have one.
Clément Léger originally put a patch together for the alignment issue, but
we uncovered more issues as we went further into the problem. Clément felt
my version was better [1] so I am sending this series out that addresses
the selftest, BE bit offset, and the alignment issue.
1. https://lore.kernel.org/linux-trace-kernel/713f4916-00ff-4a24-82d1-72884500…
Beau Belgrave (2):
tracing/user_events: Align set_bit() address for all archs
selftests/user_events: Fix abi_test for BE archs
kernel/trace/trace_events_user.c | 58 ++++++++++++++++---
.../testing/selftests/user_events/abi_test.c | 16 ++---
2 files changed, 60 insertions(+), 14 deletions(-)
base-commit: fc1653abba0d554aad80224e51bcad42b09895ed
--
2.34.1