January 2023 - Linux-kselftest-mirror

[PATCH 00/34] selftests: Fix incorrect kernel headers search path

by Mathieu Desnoyers

Hi, This series fixes incorrect kernel header search path in kernel selftests. Near the end of the series, a few changes are not tagged as "Fixes" because the current behavior is to rely on the kernel sources uapi files rather than on the installed kernel header files. Nevertheless, those are updated for consistency. There are situations where "../../../../include/" was added to -I search path, which is bogus for userspace tests and caused issues with types.h. Those are removed. Thanks, Mathieu Mathieu Desnoyers (34): selftests: arm64: Fix incorrect kernel headers search path selftests: bpf: Fix incorrect kernel headers search path selftests: clone3: Fix incorrect kernel headers search path selftests: core: Fix incorrect kernel headers search path selftests: dma: Fix incorrect kernel headers search path selftests: dmabuf-heaps: Fix incorrect kernel headers search path selftests: drivers: Fix incorrect kernel headers search path selftests: filesystems: Fix incorrect kernel headers search path selftests: futex: Fix incorrect kernel headers search path selftests: gpio: Fix incorrect kernel headers search path selftests: ipc: Fix incorrect kernel headers search path selftests: kcmp: Fix incorrect kernel headers search path selftests: media_tests: Fix incorrect kernel headers search path selftests: membarrier: Fix incorrect kernel headers search path selftests: mount_setattr: Fix incorrect kernel headers search path selftests: move_mount_set_group: Fix incorrect kernel headers search path selftests: net: Fix incorrect kernel headers search path selftests: perf_events: Fix incorrect kernel headers search path selftests: pid_namespace: Fix incorrect kernel headers search path selftests: pidfd: Fix incorrect kernel headers search path selftests: powerpc: Fix incorrect kernel headers search path selftests: ptp: Fix incorrect kernel headers search path selftests: rseq: Fix incorrect kernel headers search path selftests: sched: Fix incorrect kernel headers search path selftests: seccomp: Fix incorrect kernel headers search path selftests: sync: Fix incorrect kernel headers search path selftests: user_events: Fix incorrect kernel headers search path selftests: vm: Fix incorrect kernel headers search path selftests: x86: Fix incorrect kernel headers search path selftests: bpf docs: Use installed kernel headers search path selftests: iommu: Use installed kernel headers search path selftests: memfd: Use installed kernel headers search path selftests: ptrace: Use installed kernel headers search path selftests: tdx: Use installed kernel headers search path tools/testing/selftests/arm64/fp/Makefile | 2 +- tools/testing/selftests/arm64/tags/Makefile | 2 +- tools/testing/selftests/bpf/Makefile | 2 +- tools/testing/selftests/bpf/Makefile.docs | 2 +- tools/testing/selftests/clone3/Makefile | 2 +- tools/testing/selftests/core/Makefile | 2 +- tools/testing/selftests/dma/Makefile | 2 +- tools/testing/selftests/dmabuf-heaps/Makefile | 2 +- tools/testing/selftests/dmabuf-heaps/dmabuf-heap.c | 3 +-- tools/testing/selftests/drivers/dma-buf/Makefile | 2 +- tools/testing/selftests/drivers/s390x/uvdevice/Makefile | 3 +-- tools/testing/selftests/filesystems/Makefile | 2 +- tools/testing/selftests/filesystems/binderfs/Makefile | 2 +- tools/testing/selftests/filesystems/epoll/Makefile | 2 +- tools/testing/selftests/futex/functional/Makefile | 2 +- tools/testing/selftests/gpio/Makefile | 2 +- tools/testing/selftests/iommu/Makefile | 3 +-- tools/testing/selftests/ipc/Makefile | 2 +- tools/testing/selftests/kcmp/Makefile | 2 +- tools/testing/selftests/media_tests/Makefile | 2 +- tools/testing/selftests/membarrier/Makefile | 2 +- tools/testing/selftests/memfd/Makefile | 4 +--- tools/testing/selftests/mount_setattr/Makefile | 2 +- tools/testing/selftests/move_mount_set_group/Makefile | 2 +- tools/testing/selftests/net/Makefile | 2 +- tools/testing/selftests/net/bpf/Makefile | 2 +- tools/testing/selftests/net/mptcp/Makefile | 2 +- tools/testing/selftests/net/openvswitch/Makefile | 2 +- tools/testing/selftests/perf_events/Makefile | 2 +- tools/testing/selftests/pid_namespace/Makefile | 2 +- tools/testing/selftests/pidfd/Makefile | 2 +- tools/testing/selftests/powerpc/ptrace/Makefile | 2 +- tools/testing/selftests/powerpc/security/Makefile | 2 +- tools/testing/selftests/powerpc/syscalls/Makefile | 2 +- tools/testing/selftests/powerpc/tm/Makefile | 2 +- tools/testing/selftests/ptp/Makefile | 2 +- tools/testing/selftests/ptrace/Makefile | 2 +- tools/testing/selftests/rseq/Makefile | 2 +- tools/testing/selftests/sched/Makefile | 2 +- tools/testing/selftests/seccomp/Makefile | 2 +- tools/testing/selftests/sync/Makefile | 2 +- tools/testing/selftests/tdx/Makefile | 2 +- tools/testing/selftests/tdx/tdx_guest_test.c | 2 +- tools/testing/selftests/user_events/Makefile | 2 +- tools/testing/selftests/vm/Makefile | 2 +- tools/testing/selftests/x86/Makefile | 2 +- 46 files changed, 46 insertions(+), 51 deletions(-) -- 2.25.1

1 year, 1 month

8
62
0 0

[RFC PATCH v2 0/2] selftests/x86: sysret_rip update for FRED system

by Ammar Faizi

From: Ammar Faizi <ammarfaizi2(a)gnuweeb.org> This is an RFC patchset v2. Xin Li reported sysret_rip test fails at: assert(ctx->uc_mcontext.gregs[REG_EFL] == ctx->uc_mcontext.gregs[REG_R11]); in a FRED system. Handle the FRED system scenario too. There are two patches in this series. Comments welcome... Note: This patchset is only tested for 'syscall' sets %rcx=%rip and %r11=%rflags case. I don't have a FRED system to test it. How to test this: $ make -C tools/testing/selftests/x86 $ tools/testing/selftests/x86/sysret_rip_64 Link: https://lore.kernel.org/lkml/5d4ad3e3-034f-c7da-d141-9c001c2343af@intel.com Signed-off-by: Ammar Faizi <ammarfaizi2(a)gnuweeb.org> --- ## Changelog v2: - Use "+r"(rsp) as the right way to avoid redzone problems per Andrew's comment (hpa). (Ref: https://lore.kernel.org/lkml/8f5c24df-514d-5d89-f58f-ec8c3eb1e049@zytor.com ) --- Ammar Faizi (2): selftests/x86: sysret_rip: Handle syscall in a FRED system selftests/x86: sysret_rip: Add more syscall tests with respect to `%rcx` and `%r11` tools/testing/selftests/x86/sysret_rip.c | 105 ++++++++++++++++++++++- 1 file changed, 104 insertions(+), 1 deletion(-) base-commit: e12ad468c22065a2826b2fc4c11d2113a7975301 -- Ammar Faizi

1 year, 2 months

4
53
0 0

[RFC PATCH v3 00/31] TDX KVM selftests

by Ackerley Tng

Hello, This is v3 of the patch series for TDX selftests. It has been updated for Intel’s V10 of the TDX host patches which was proposed in https://lkml.org/lkml/2022/8/8/877 The tree can be found at https://github.com/googleprodkernel/linux-cc/tree/tdx-selftests-rfc-v3/ Changes from RFC v2: Selftest setup now builds upon the KVM selftest framework when setting up the guest for testing. We now use the KVM selftest framework to build the guest page tables and load the ELF binary into guest memory. Inlining of the entire guest image is no longer required and that allows us to cleanly separate code into different compilation units and be able to use proper assembly instead of inline assembly (addresses Sean’s comment). To achieve this, we take a dependency on the SEV VM tests: https://lore.kernel.org/lkml/20221018205845.770121-1-pgonda@google.com/T/. Those patches provide functions for the host to allocate and track protected memory in the guest. In RFCv3, TDX selftest code is organized into: + headers in tools/testing/selftests/kvm/include/x86_64/tdx/ + common code in tools/testing/selftests/kvm/lib/x86_64/tdx/ + selftests in tools/testing/selftests/kvm/x86_64/tdx_* RFCv3 also adds additional selftests for UPM. Dependencies + Peter’s patches, which provide functions for the host to allocate and track protected memory in the guest. https://lore.kernel.org/lkml/20221018205845.770121-1-pgonda@google.com/T/ + Peter’s patches depend on Sean’s patches: + https://lore.kernel.org/linux-arm-kernel/20220825232522.3997340-1-seanjc@go… + https://lore.kernel.org/lkml/20221006004512.666529-1-seanjc@google.com/T/ + Proposed fixes for these these issues mentioned on the mailing list + https://lore.kernel.org/lkml/36cde6d6-128d-884e-1447-09b08bb5de3d@intel.com/ + https://lore.kernel.org/lkml/diqzedtubs0d.fsf@google.com/ + https://lore.kernel.org/lkml/67b782ee-c95c-d6bc-3cca-cdfe75f4bf6a@intel.com/ + https://lore.kernel.org/lkml/diqzcz7cd983.fsf@ackerleytng-cloudtop-sg.c.goo… + https://lore.kernel.org/linux-mm/20221116205025.1510291-1-ackerleytng@googl… Further work for this patch series/TODOs + Sean’s comments for the non-confidential UPM selftests patch series at https://lore.kernel.org/lkml/Y8dC8WDwEmYixJqt@google.com/T/#u apply here as well + Add ucall support for TDX selftests I would also like to acknowledge the following people, who helped review or test patches in RFCv1 and RFCv2: + Sean Christopherson <seanjc(a)google.com> + Zhenzhong Duan <zhenzhong.duan(a)intel.com> + Peter Gonda <pgonda(a)google.com> + Andrew Jones <drjones(a)redhat.com> + Maxim Levitsky <mlevitsk(a)redhat.com> + Xiaoyao Li <xiaoyao.li(a)intel.com> + David Matlack <dmatlack(a)google.com> + Marc Orr <marcorr(a)google.com> + Isaku Yamahata <isaku.yamahata(a)gmail.com> Links to earlier patch series + RFC v1: https://lore.kernel.org/lkml/20210726183816.1343022-1-erdemaktas@google.com… + RFC v2: https://lore.kernel.org/lkml/20220830222000.709028-1-sagis@google.com/T/#u Ackerley Tng (14): KVM: selftests: Add function to allow one-to-one GVA to GPA mappings KVM: selftests: Expose function that sets up sregs based on VM's mode KVM: selftests: Store initial stack address in struct kvm_vcpu KVM: selftests: Refactor steps in vCPU descriptor table initialization KVM: selftests: TDX: Use KVM_TDX_CAPABILITIES to validate TDs' attribute configuration KVM: selftests: Require GCC to realign stacks on function entry KVM: selftests: Add functions to allow mapping as shared KVM: selftests: Add support for restricted memory KVM: selftests: TDX: Update load_td_memory_region for VM memory backed by restricted memfd KVM: selftests: Expose _vm_vaddr_alloc KVM: selftests: TDX: Add support for TDG.MEM.PAGE.ACCEPT KVM: selftests: TDX: Add support for TDG.VP.VEINFO.GET KVM: selftests: TDX: Add TDX UPM selftest KVM: selftests: TDX: Add TDX UPM selftests for implicit conversion Erdem Aktas (4): KVM: selftests: Add support for creating non-default type VMs KVM: selftests: Add helper functions to create TDX VMs KVM: selftests: TDX: Add TDX lifecycle test KVM: selftests: TDX: Adding test case for TDX port IO Roger Wang (1): KVM: selftests: TDX: Add TDG.VP.INFO test Ryan Afranji (2): KVM: selftests: TDX: Verify the behavior when host consumes a TD private memory KVM: selftests: TDX: Add shared memory test Sagi Shahar (10): KVM: selftests: TDX: Add report_fatal_error test KVM: selftests: TDX: Add basic TDX CPUID test KVM: selftests: TDX: Add basic get_td_vmcall_info test KVM: selftests: TDX: Add TDX IO writes test KVM: selftests: TDX: Add TDX IO reads test KVM: selftests: TDX: Add TDX MSR read/write tests KVM: selftests: TDX: Add TDX HLT exit test KVM: selftests: TDX: Add TDX MMIO reads test KVM: selftests: TDX: Add TDX MMIO writes test KVM: selftests: TDX: Add TDX CPUID TDVMCALL test tools/testing/selftests/kvm/.gitignore | 3 + tools/testing/selftests/kvm/Makefile | 10 +- .../selftests/kvm/include/kvm_util_base.h | 43 +- .../testing/selftests/kvm/include/test_util.h | 2 + .../selftests/kvm/include/x86_64/processor.h | 4 + .../kvm/include/x86_64/tdx/td_boot.h | 82 + .../kvm/include/x86_64/tdx/td_boot_asm.h | 16 + .../selftests/kvm/include/x86_64/tdx/tdcall.h | 59 + .../selftests/kvm/include/x86_64/tdx/tdx.h | 65 + .../kvm/include/x86_64/tdx/tdx_util.h | 19 + .../kvm/include/x86_64/tdx/test_util.h | 164 ++ tools/testing/selftests/kvm/lib/kvm_util.c | 123 +- tools/testing/selftests/kvm/lib/test_util.c | 7 + .../selftests/kvm/lib/x86_64/processor.c | 77 +- tools/testing/selftests/kvm/lib/x86_64/sev.c | 2 +- .../selftests/kvm/lib/x86_64/tdx/td_boot.S | 101 ++ .../selftests/kvm/lib/x86_64/tdx/tdcall.S | 158 ++ .../selftests/kvm/lib/x86_64/tdx/tdx.c | 231 +++ .../selftests/kvm/lib/x86_64/tdx/tdx_util.c | 562 +++++++ .../selftests/kvm/lib/x86_64/tdx/test_util.c | 101 ++ .../kvm/x86_64/tdx_shared_mem_test.c | 137 ++ .../selftests/kvm/x86_64/tdx_upm_test.c | 460 ++++++ .../selftests/kvm/x86_64/tdx_vm_tests.c | 1329 +++++++++++++++++ 23 files changed, 3709 insertions(+), 46 deletions(-) create mode 100644 tools/testing/selftests/kvm/include/x86_64/tdx/td_boot.h create mode 100644 tools/testing/selftests/kvm/include/x86_64/tdx/td_boot_asm.h create mode 100644 tools/testing/selftests/kvm/include/x86_64/tdx/tdcall.h create mode 100644 tools/testing/selftests/kvm/include/x86_64/tdx/tdx.h create mode 100644 tools/testing/selftests/kvm/include/x86_64/tdx/tdx_util.h create mode 100644 tools/testing/selftests/kvm/include/x86_64/tdx/test_util.h create mode 100644 tools/testing/selftests/kvm/lib/x86_64/tdx/td_boot.S create mode 100644 tools/testing/selftests/kvm/lib/x86_64/tdx/tdcall.S create mode 100644 tools/testing/selftests/kvm/lib/x86_64/tdx/tdx.c create mode 100644 tools/testing/selftests/kvm/lib/x86_64/tdx/tdx_util.c create mode 100644 tools/testing/selftests/kvm/lib/x86_64/tdx/test_util.c create mode 100644 tools/testing/selftests/kvm/x86_64/tdx_shared_mem_test.c create mode 100644 tools/testing/selftests/kvm/x86_64/tdx_upm_test.c create mode 100644 tools/testing/selftests/kvm/x86_64/tdx_vm_tests.c -- 2.39.0.246.g2a6d74b583-goog

1 year, 2 months

4
42
0 0

[PATCH v2 net-next] selftests: forwarding: add Per-Stream Filtering and Policing test for Ocelot

by Vladimir Oltean

The Felix VSC9959 switch in NXP LS1028A supports the tc-gate action which enforced time-based access control per stream. A stream as seen by this switch is identified by {MAC DA, VID}. We use the standard forwarding selftest topology with 2 host interfaces and 2 switch interfaces. The host ports must require timestamping non-IP packets and supporting tc-etf offload, for isochron to work. The isochron program monitors network sync status (ptp4l, phc2sys) and deterministically transmits packets to the switch such that the tc-gate action either (a) always accepts them based on its schedule, or (b) always drops them. I tried to keep as much of the logic that isn't specific to the NXP LS1028A in a new tsn_lib.sh, for future reuse. This covers synchronization using ptp4l and phc2sys, and isochron. The cycle-time chosen for this selftest isn't particularly impressive (and the focus is the functionality of the switch), but I didn't really know what to do better, considering that it will mostly be run during debugging sessions, various kernel bloatware would be enabled, like lockdep, KASAN, etc, and we certainly can't run any races with those on. I tried to look through the kselftest framework for other real time applications and didn't really find any, so I'm not sure how better to prepare the environment in case we want to go for a lower cycle time. At the moment, the only thing the selftest is ensuring is that dynamic frequency scaling is disabled on the CPU that isochron runs on. It would probably be useful to have a blacklist of kernel config options (checked through zcat /proc/config.gz) and some cyclictest scripts to run beforehand, but I saw none of those. Signed-off-by: Vladimir Oltean <vladimir.oltean(a)nxp.com> --- v1->v2: - fix an off-by-one bug introduced at the last minute regarding which tc-mqprio queue was used for tc-etf and SO_TXTIME - introduce debugging for packets incorrectly received / incorrectly dropped based on "isochron report" - make the tsn_lib.sh dependency on isochron and linuxptp optional via REQUIRE_ISOCHRON and REQUIRE_LINUXPTP - avoid errors when CONFIG_CPU_FREQ is disabled - consistently use SCHED_FIFO instead of SCHED_RR for the isochron receiver .../selftests/drivers/net/ocelot/psfp.sh | 327 ++++++++++++++++++ .../selftests/net/forwarding/tsn_lib.sh | 235 +++++++++++++ 2 files changed, 562 insertions(+) create mode 100755 tools/testing/selftests/drivers/net/ocelot/psfp.sh create mode 100644 tools/testing/selftests/net/forwarding/tsn_lib.sh diff --git a/tools/testing/selftests/drivers/net/ocelot/psfp.sh b/tools/testing/selftests/drivers/net/ocelot/psfp.sh new file mode 100755 index 000000000000..5a5cee92c665 --- /dev/null +++ b/tools/testing/selftests/drivers/net/ocelot/psfp.sh @@ -0,0 +1,327 @@ +#!/bin/bash +# SPDX-License-Identifier: GPL-2.0 +# Copyright 2021-2022 NXP + +# Note: On LS1028A, in lack of enough user ports, this setup requires patching +# the device tree to use the second CPU port as a user port + +WAIT_TIME=1 +NUM_NETIFS=4 +STABLE_MAC_ADDRS=yes +NETIF_CREATE=no +lib_dir=$(dirname $0)/../../../net/forwarding +source $lib_dir/tc_common.sh +source $lib_dir/lib.sh +source $lib_dir/tsn_lib.sh + +UDS_ADDRESS_H1="/var/run/ptp4l_h1" +UDS_ADDRESS_SWP1="/var/run/ptp4l_swp1" + +# Tunables +NUM_PKTS=1000 +STREAM_VID=100 +STREAM_PRIO=6 +# Use a conservative cycle of 10 ms to allow the test to still pass when the +# kernel has some extra overhead like lockdep etc +CYCLE_TIME_NS=10000000 +# Create two Gate Control List entries, one OPEN and one CLOSE, of equal +# durations +GATE_DURATION_NS=$((${CYCLE_TIME_NS} / 2)) +# Give 2/3 of the cycle time to user space and 1/3 to the kernel +FUDGE_FACTOR=$((${CYCLE_TIME_NS} / 3)) +# Shift the isochron base time by half the gate time, so that packets are +# always received by swp1 close to the middle of the time slot, to minimize +# inaccuracies due to network sync +SHIFT_TIME_NS=$((${GATE_DURATION_NS} / 2)) + +h1=${NETIFS[p1]} +swp1=${NETIFS[p2]} +swp2=${NETIFS[p3]} +h2=${NETIFS[p4]} + +H1_IPV4="192.0.2.1" +H2_IPV4="192.0.2.2" +H1_IPV6="2001:db8:1::1" +H2_IPV6="2001:db8:1::2" + +# Chain number exported by the ocelot driver for +# Per-Stream Filtering and Policing filters +PSFP() +{ + echo 30000 +} + +psfp_chain_create() +{ + local if_name=$1 + + tc qdisc add dev $if_name clsact + + tc filter add dev $if_name ingress chain 0 pref 49152 flower \ + skip_sw action goto chain $(PSFP) +} + +psfp_chain_destroy() +{ + local if_name=$1 + + tc qdisc del dev $if_name clsact +} + +psfp_filter_check() +{ + local expected=$1 + local packets="" + local drops="" + local stats="" + + stats=$(tc -j -s filter show dev ${swp1} ingress chain $(PSFP) pref 1) + packets=$(echo ${stats} | jq ".[1].options.actions[].stats.packets") + drops=$(echo ${stats} | jq ".[1].options.actions[].stats.drops") + + if ! [ "${packets}" = "${expected}" ]; then + printf "Expected filter to match on %d packets but matched on %d instead\n" \ + "${expected}" "${packets}" + fi + + echo "Hardware filter reports ${drops} drops" +} + +h1_create() +{ + simple_if_init $h1 $H1_IPV4/24 $H1_IPV6/64 +} + +h1_destroy() +{ + simple_if_fini $h1 $H1_IPV4/24 $H1_IPV6/64 +} + +h2_create() +{ + simple_if_init $h2 $H2_IPV4/24 $H2_IPV6/64 +} + +h2_destroy() +{ + simple_if_fini $h2 $H2_IPV4/24 $H2_IPV6/64 +} + +switch_create() +{ + local h2_mac_addr=$(mac_get $h2) + + ip link set ${swp1} up + ip link set ${swp2} up + + ip link add br0 type bridge vlan_filtering 1 + ip link set ${swp1} master br0 + ip link set ${swp2} master br0 + ip link set br0 up + + bridge vlan add dev ${swp2} vid ${STREAM_VID} + bridge vlan add dev ${swp1} vid ${STREAM_VID} + # PSFP on Ocelot requires the filter to also be added to the bridge + # FDB, and not be removed + bridge fdb add dev ${swp2} \ + ${h2_mac_addr} vlan ${STREAM_VID} static master + + psfp_chain_create ${swp1} + + tc filter add dev ${swp1} ingress chain $(PSFP) pref 1 \ + protocol 802.1Q flower skip_sw \ + dst_mac ${h2_mac_addr} vlan_id ${STREAM_VID} \ + action gate base-time 0.000000000 \ + sched-entry OPEN ${GATE_DURATION_NS} -1 -1 \ + sched-entry CLOSE ${GATE_DURATION_NS} -1 -1 +} + +switch_destroy() +{ + psfp_chain_destroy ${swp1} + ip link del br0 +} + +txtime_setup() +{ + local if_name=$1 + + tc qdisc add dev ${if_name} clsact + # Classify PTP on TC 7 and isochron on TC 6 + tc filter add dev ${if_name} egress protocol 0x88f7 \ + flower action skbedit priority 7 + tc filter add dev ${if_name} egress protocol 802.1Q \ + flower vlan_ethtype 0xdead action skbedit priority 6 + tc qdisc add dev ${if_name} handle 100: parent root mqprio num_tc 8 \ + queues 1@0 1@1 1@2 1@3 1@4 1@5 1@6 1@7 \ + map 0 1 2 3 4 5 6 7 \ + hw 1 + # Set up TC 6 for SO_TXTIME. tc-mqprio queues count from 1. + tc qdisc replace dev ${if_name} parent 100:$((${STREAM_PRIO} + 1)) etf \ + clockid CLOCK_TAI offload delta ${FUDGE_FACTOR} +} + +txtime_cleanup() +{ + local if_name=$1 + + tc qdisc del dev ${if_name} root + tc qdisc del dev ${if_name} clsact +} + +setup_prepare() +{ + vrf_prepare + + h1_create + h2_create + switch_create + + txtime_setup ${h1} + + # Set up swp1 as a master PHC for h1, synchronized to the local + # CLOCK_REALTIME. + phc2sys_start ${swp1} ${UDS_ADDRESS_SWP1} + + # Assumption true for LS1028A: h1 and h2 use the same PHC. So by + # synchronizing h1 to swp1 via PTP, h2 is also implicitly synchronized + # to swp1 (and both to CLOCK_REALTIME). + ptp4l_start ${h1} true ${UDS_ADDRESS_H1} + ptp4l_start ${swp1} false ${UDS_ADDRESS_SWP1} + + # Make sure there are no filter matches at the beginning of the test + psfp_filter_check 0 +} + +cleanup() +{ + pre_cleanup + + ptp4l_stop ${swp1} + ptp4l_stop ${h1} + phc2sys_stop + isochron_recv_stop + + txtime_cleanup ${h1} + + h2_destroy + h1_destroy + switch_destroy + + vrf_cleanup +} + +debug_incorrectly_dropped_packets() +{ + local isochron_dat=$1 + local dropped_seqids + local seqid + + echo "Packets incorrectly dropped:" + + dropped_seqids=$(isochron report \ + --input-file "${isochron_dat}" \ + --printf-format "%u RX hw %T\n" \ + --printf-args "qR" | \ + grep 'RX hw 0.000000000' | \ + awk '{print $1}') + + for seqid in ${dropped_seqids}; do + isochron report \ + --input-file "${isochron_dat}" \ + --start ${seqid} --stop ${seqid} \ + --printf-format "seqid %u scheduled for %T, HW TX timestamp %T\n" \ + --printf-args "qST" + done +} + +debug_incorrectly_received_packets() +{ + local isochron_dat=$1 + + echo "Packets incorrectly received:" + + isochron report \ + --input-file "${isochron_dat}" \ + --printf-format "seqid %u scheduled for %T, HW TX timestamp %T, HW RX timestamp %T\n" \ + --printf-args "qSTR" | + grep -v 'HW RX timestamp 0.000000000' +} + +run_test() +{ + local base_time=$1 + local expected=$2 + local test_name=$3 + local debug=$4 + local isochron_dat="$(mktemp)" + local extra_args="" + local received + + isochron_do \ + "${h1}" \ + "${h2}" \ + "${UDS_ADDRESS_H1}" \ + "" \ + "${base_time}" \ + "${CYCLE_TIME_NS}" \ + "${SHIFT_TIME_NS}" \ + "${NUM_PKTS}" \ + "${STREAM_VID}" \ + "${STREAM_PRIO}" \ + "" \ + "${isochron_dat}" + + # Count all received packets by looking at the non-zero RX timestamps + received=$(isochron report \ + --input-file "${isochron_dat}" \ + --printf-format "%u\n" --printf-args "R" | \ + grep -w -v '0' | wc -l) + + if [ "${received}" = "${expected}" ]; then + RET=0 + else + RET=1 + echo "Expected isochron to receive ${expected} packets but received ${received}" + fi + + log_test "${test_name}" + + if [ "$RET" = "1" ]; then + ${debug} "${isochron_dat}" + fi + + rm ${isochron_dat} 2> /dev/null +} + +test_gate_in_band() +{ + # Send packets in-band with the OPEN gate entry + run_test 0.000000000 ${NUM_PKTS} "In band" \ + debug_incorrectly_dropped_packets + + psfp_filter_check ${NUM_PKTS} +} + +test_gate_out_of_band() +{ + # Send packets in-band with the CLOSE gate entry + run_test 0.005000000 0 "Out of band" \ + debug_incorrectly_received_packets + + psfp_filter_check $((2 * ${NUM_PKTS})) +} + +trap cleanup EXIT + +ALL_TESTS=" + test_gate_in_band + test_gate_out_of_band +" + +setup_prepare +setup_wait + +tests_run + +exit $EXIT_STATUS diff --git a/tools/testing/selftests/net/forwarding/tsn_lib.sh b/tools/testing/selftests/net/forwarding/tsn_lib.sh new file mode 100644 index 000000000000..60a1423e8116 --- /dev/null +++ b/tools/testing/selftests/net/forwarding/tsn_lib.sh @@ -0,0 +1,235 @@ +#!/bin/bash +# SPDX-License-Identifier: GPL-2.0 +# Copyright 2021-2022 NXP + +REQUIRE_ISOCHRON=${REQUIRE_ISOCHRON:=yes} +REQUIRE_LINUXPTP=${REQUIRE_LINUXPTP:=yes} + +# Tunables +UTC_TAI_OFFSET=37 +ISOCHRON_CPU=1 + +if [[ "$REQUIRE_ISOCHRON" = "yes" ]]; then + # https://github.com/vladimiroltean/tsn-scripts + # WARNING: isochron versions pre-1.0 are unstable, + # always use the latest version + require_command isochron +fi +if [[ "$REQUIRE_LINUXPTP" = "yes" ]]; then + require_command phc2sys + require_command ptp4l +fi + +phc2sys_start() +{ + local if_name=$1 + local uds_address=$2 + local extra_args="" + + if ! [ -z "${uds_address}" ]; then + extra_args="${extra_args} -z ${uds_address}" + fi + + phc2sys_log="$(mktemp)" + + chrt -f 10 phc2sys -m \ + -c ${if_name} \ + -s CLOCK_REALTIME \ + -O ${UTC_TAI_OFFSET} \ + --step_threshold 0.00002 \ + --first_step_threshold 0.00002 \ + ${extra_args} \ + > "${phc2sys_log}" 2>&1 & + phc2sys_pid=$! + + echo "phc2sys logs to ${phc2sys_log} and has pid ${phc2sys_pid}" + + sleep 1 +} + +phc2sys_stop() +{ + { kill ${phc2sys_pid} && wait ${phc2sys_pid}; } 2> /dev/null + rm "${phc2sys_log}" 2> /dev/null +} + +ptp4l_start() +{ + local if_name=$1 + local slave_only=$2 + local uds_address=$3 + local log="ptp4l_log_${if_name}" + local pid="ptp4l_pid_${if_name}" + local extra_args="" + + if [ "${slave_only}" = true ]; then + extra_args="${extra_args} -s" + fi + + # declare dynamic variables ptp4l_log_${if_name} and ptp4l_pid_${if_name} + # as global, so that they can be referenced later + declare -g "${log}=$(mktemp)" + + chrt -f 10 ptp4l -m -2 -P \ + -i ${if_name} \ + --step_threshold 0.00002 \ + --first_step_threshold 0.00002 \ + --tx_timestamp_timeout 100 \ + --uds_address="${uds_address}" \ + ${extra_args} \ + > "${!log}" 2>&1 & + declare -g "${pid}=$!" + + echo "ptp4l for interface ${if_name} logs to ${!log} and has pid ${!pid}" + + sleep 1 +} + +ptp4l_stop() +{ + local if_name=$1 + local log="ptp4l_log_${if_name}" + local pid="ptp4l_pid_${if_name}" + + { kill ${!pid} && wait ${!pid}; } 2> /dev/null + rm "${!log}" 2> /dev/null +} + +cpufreq_max() +{ + local cpu=$1 + local freq="cpu${cpu}_freq" + local governor="cpu${cpu}_governor" + + # Kernel may be compiled with CONFIG_CPU_FREQ disabled + if ! [ -d /sys/bus/cpu/devices/cpu${cpu}/cpufreq ]; then + return + fi + + # declare dynamic variables cpu${cpu}_freq and cpu${cpu}_governor as + # global, so they can be referenced later + declare -g "${freq}=$(cat /sys/bus/cpu/devices/cpu${cpu}/cpufreq/scaling_min_freq)" + declare -g "${governor}=$(cat /sys/bus/cpu/devices/cpu${cpu}/cpufreq/scaling_governor)" + + cat /sys/bus/cpu/devices/cpu${cpu}/cpufreq/scaling_max_freq > \ + /sys/bus/cpu/devices/cpu${cpu}/cpufreq/scaling_min_freq + echo -n "performance" > \ + /sys/bus/cpu/devices/cpu${cpu}/cpufreq/scaling_governor +} + +cpufreq_restore() +{ + local cpu=$1 + local freq="cpu${cpu}_freq" + local governor="cpu${cpu}_governor" + + if ! [ -d /sys/bus/cpu/devices/cpu${cpu}/cpufreq ]; then + return + fi + + echo "${!freq}" > /sys/bus/cpu/devices/cpu${cpu}/cpufreq/scaling_min_freq + echo -n "${!governor}" > \ + /sys/bus/cpu/devices/cpu${cpu}/cpufreq/scaling_governor +} + +isochron_recv_start() +{ + local if_name=$1 + local uds=$2 + local extra_args=$3 + + if ! [ -z "${uds}" ]; then + extra_args="--unix-domain-socket ${uds}" + fi + + isochron rcv \ + --interface ${if_name} \ + --sched-priority 98 \ + --sched-fifo \ + --utc-tai-offset ${UTC_TAI_OFFSET} \ + --quiet \ + ${extra_args} & \ + isochron_pid=$! + + sleep 1 +} + +isochron_recv_stop() +{ + { kill ${isochron_pid} && wait ${isochron_pid}; } 2> /dev/null +} + +isochron_do() +{ + local sender_if_name=$1; shift + local receiver_if_name=$1; shift + local sender_uds=$1; shift + local receiver_uds=$1; shift + local base_time=$1; shift + local cycle_time=$1; shift + local shift_time=$1; shift + local num_pkts=$1; shift + local vid=$1; shift + local priority=$1; shift + local dst_ip=$1; shift + local isochron_dat=$1; shift + local extra_args="" + local receiver_extra_args="" + local vrf="$(master_name_get ${sender_if_name})" + local use_l2="true" + + if ! [ -z "${dst_ip}" ]; then + use_l2="false" + fi + + if ! [ -z "${vrf}" ]; then + dst_ip="${dst_ip}%${vrf}" + fi + + if ! [ -z "${vid}" ]; then + vid="--vid=${vid}" + fi + + if [ -z "${receiver_uds}" ]; then + extra_args="${extra_args} --omit-remote-sync" + fi + + if ! [ -z "${shift_time}" ]; then + extra_args="${extra_args} --shift-time=${shift_time}" + fi + + if [ "${use_l2}" = "true" ]; then + extra_args="${extra_args} --l2 --etype=0xdead ${vid}" + receiver_extra_args="--l2 --etype=0xdead" + else + extra_args="${extra_args} --l4 --ip-destination=${dst_ip}" + receiver_extra_args="--l4" + fi + + cpufreq_max ${ISOCHRON_CPU} + + isochron_recv_start "${h2}" "${receiver_uds}" "${receiver_extra_args}" + + isochron send \ + --interface ${sender_if_name} \ + --unix-domain-socket ${sender_uds} \ + --priority ${priority} \ + --base-time ${base_time} \ + --cycle-time ${cycle_time} \ + --num-frames ${num_pkts} \ + --frame-size 64 \ + --txtime \ + --utc-tai-offset ${UTC_TAI_OFFSET} \ + --cpu-mask $((1 << ${ISOCHRON_CPU})) \ + --sched-fifo \ + --sched-priority 98 \ + --client 127.0.0.1 \ + --sync-threshold 5000 \ + --output-file ${isochron_dat} \ + ${extra_args} \ + --quiet + + isochron_recv_stop + + cpufreq_restore ${ISOCHRON_CPU} +} -- 2.25.1

1 year, 2 months

5
13
0 0

[PATCH RFC net-next v2 0/3] vsock: add support for sockmap

by Bobby Eshleman

Add support for sockmap to vsock. We're testing usage of vsock as a way to redirect guest-local UDS requests to the host and this patch series greatly improves the performance of such a setup. Compared to copying packets via userspace, this improves throughput by 121% in basic testing. Tested as follows. Setup: guest unix dgram sender -> guest vsock redirector -> host vsock server Threads: 1 Payload: 64k No sockmap: - 76.3 MB/s - The guest vsock redirector was "socat VSOCK-CONNECT:2:1234 UNIX-RECV:/path/to/sock" Using sockmap (this patch): - 168.8 MB/s (+121%) - The guest redirector was a simple sockmap echo server, redirecting unix ingress to vsock 2:1234 egress. - Same sender and server programs *Note: these numbers are from RFC v1 Only the virtio transport has been tested. The loopback transport was used in writing bpf/selftests, but not thoroughly tested otherwise. This series requires the skb patch. Changes in v2: - vsock/bpf: rename vsock_dgram_* -> vsock_* - vsock/bpf: change sk_psock_{get,put} and {lock,release}_sock() order to minimize slock hold time - vsock/bpf: use "new style" wait - vsock/bpf: fix bug in wait log - vsock/bpf: add check that recvmsg sk_type is one dgram, seqpacket, or stream. Return error if not one of the three. - virtio/vsock: comment __skb_recv_datagram() usage - virtio/vsock: do not init copied in read_skb() - vsock/bpf: add ifdef guard around struct proto in dgram_recvmsg() - selftests/bpf: add vsock loopback config for aarch64 - selftests/bpf: add vsock loopback config for s390x - selftests/bpf: remove vsock device from vmtest.sh qemu machine - selftests/bpf: remove CONFIG_VIRTIO_VSOCKETS=y from config.x86_64 - vsock/bpf: move transport-related (e.g., if (!vsk->transport)) checks out of fast path Signed-off-by: Bobby Eshleman <bobby.eshleman(a)bytedance.com> --- Bobby Eshleman (3): vsock: support sockmap selftests/bpf: add vsock to vmtest.sh selftests/bpf: Add a test case for vsock sockmap drivers/vhost/vsock.c | 1 + include/linux/virtio_vsock.h | 1 + include/net/af_vsock.h | 17 ++ net/vmw_vsock/Makefile | 1 + net/vmw_vsock/af_vsock.c | 55 ++++++- net/vmw_vsock/virtio_transport.c | 2 + net/vmw_vsock/virtio_transport_common.c | 24 +++ net/vmw_vsock/vsock_bpf.c | 175 +++++++++++++++++++++ net/vmw_vsock/vsock_loopback.c | 2 + tools/testing/selftests/bpf/config.aarch64 | 2 + tools/testing/selftests/bpf/config.s390x | 3 + tools/testing/selftests/bpf/config.x86_64 | 3 + .../selftests/bpf/prog_tests/sockmap_listen.c | 163 +++++++++++++++++++ 13 files changed, 443 insertions(+), 6 deletions(-) --- base-commit: d83115ce337a632f996e44c9f9e18cadfcf5a094 change-id: 20230118-support-vsock-sockmap-connectible-2e1297d2111a Best regards, -- Bobby Eshleman <bobby.eshleman(a)bytedance.com>

1 year, 2 months

4
10
0 0

[RFC PATCH v1 0/2] selftests/x86: sysret_rip update for FRED system

by Ammar Faizi

From: Ammar Faizi <ammarfaizi2(a)gnuweeb.org> On Mon, 23 Jan 2023 15:58:12 -0800, "H. Peter Anvin" wrote: > On 1/23/23 15:43, Ammar Faizi wrote: > > > > Align them to spot differences: > > > > 0x200893 = 0b1000000000100010010011 > > 0x200a93 = 0b1000000000101010010011 > > ^ > > > > Or just xor them to find the differences: > > > > (gdb) p/x 0x200893 ^ 0x200a93 > > $3 = 0x200 > > > > ** Checks my Intel SDM cheat sheets. ** > > > > Then, I was like "Oh, that's (1 << 9) a.k.a. IF. Of course we can't > > change rflags[IF] from userspace!!!". > > > > In short, we can't use 0x200893 as the rflags_sentinel value because it > > clears the interrupt flag. > > > > Right, my mistake. I changed it to 0x200a93. The test passed on my machine. But I don't have a FRED system to test the special case. Didn't manage to apply the feedback from Andrew about the way to handle redzone properly, though. Something like this... ---------- This is just an RFC patchset. Xin Li reported sysret_rip test fails at: assert(ctx->uc_mcontext.gregs[REG_EFL] == ctx->uc_mcontext.gregs[REG_R11]); in a FRED system. Handle the FRED system scenario too. There are two patches in this series. Comments welcome... Note: Only tested for 'syscall' sets %rcx=%rip and %r11=%rflags case. I don't have a FRED system to test it. How to test this: $ make -C tools/testing/selftests/x86 $ tools/testing/selftests/x86/sysret_rip_64 Link: https://lore.kernel.org/lkml/5d4ad3e3-034f-c7da-d141-9c001c2343af@intel.com Signed-off-by: Ammar Faizi <ammarfaizi2(a)gnuweeb.org> --- Ammar Faizi (2): selftests/x86: sysret_rip: Handle syscall in a FRED system selftests/x86: sysret_rip: Add more syscall tests with respect to `%rcx` and `%r11` tools/testing/selftests/x86/sysret_rip.c | 105 ++++++++++++++++++++++- 1 file changed, 104 insertions(+), 1 deletion(-) base-commit: e12ad468c22065a2826b2fc4c11d2113a7975301 -- Ammar Faizi

1 year, 2 months

4
10
0 0

[PATCH i-g-t v2 0/4] Add support for KUnit tests

by Isabella Basso

This patch series was first developed as part of the LKCamp hackathon that happened last year[1], mainly focusing on refactoring DRM tests to use KUnit. KUnit[2][3] is a unified test framework that provides helper tools, simplifying their development and execution. Using an x86-64 machine it's possible to run tests in the host's kernel natively using user-mode Linux[4] (aka UML), which simplifies usage in a wide variety of scenarios, including integration to CI. As the tool's adoption widens into graphics testing territory, I and LKCamp members figured it would be important to support it in IGT, as it's a core tool for GPU drivers maintainers. I have then added KUnit support into IGT mainly following the KTAP specs, and it can be tested using patch 4/4 in this series together with a DRM selftests patch series available at [5]. Changes since v1: - Major rework of parsing function structure: - It is not longer recursive - Adapt kselftests functions and structs to be used with KUnit - Switch DRM selftests to KUnit parsing as they're updated in the kernel - Replace AMD KUnit tests by DRM selftests [1]: https://groups.google.com/g/kunit-dev/c/YqFR1q2uZvk/m/IbvItSfHBAAJ [2]: https://kunit.dev [3]: https://docs.kernel.org/dev-tools/kunit/index.html [4]: http://user-mode-linux.sourceforge.net [5]: https://lore.kernel.org/all/20220708203052.236290-1-maira.canal@usp.br/ Isabella Basso (4): lib/igt_kmod: rename kselftest functions to ktest lib/igt_kmod.c: check if module is builtin before attempting to unload it lib/igt_kmod: add compatibility for KUnit tests: DRM selftests: switch to KUnit lib/igt_kmod.c | 315 +++++++++++++++++++++++++++++++++++++++++-- lib/igt_kmod.h | 14 +- tests/drm_buddy.c | 7 +- tests/drm_mm.c | 7 +- tests/kms_selftest.c | 12 +- 5 files changed, 329 insertions(+), 26 deletions(-) -- 2.37.2

1 year, 2 months

6
29
0 0

[PATCH 0/4] Add extra checkings to amx_test

by Mingwei Zhang

AMX architecture involves several entities such as xstate, XCR0, IA32_XFD. This series add several missing checks on top of the existing amx_test. In particular, the 1st commit just fixes a typo in comment. The 2nd and 4th commit focus on the properties of IA32_XFD/IA32_XFD_ERR when interating with xsavec and #NM, while the 3rd commit adds the checking of xcomp_bv in xstate. Mingwei Zhang (4): KVM: selftests: x86: Fix an error in comment of amx_test KVM: selftests: x86: Add check of IA32_XFD in amx_test KVM: selftests: x86: Enable checking on xcomp_bv in amx_test KVM: selftests: x86: Repeat the checking of xheader when IA32_XFD[18] is set in amx_test tools/testing/selftests/kvm/x86_64/amx_test.c | 43 +++++++++++++++++-- 1 file changed, 40 insertions(+), 3 deletions(-) -- 2.39.0.314.g84b9a713c41-goog

1 year, 2 months

2
9
0 0

[PATCH v1 0/3] kunit: fix bugs in debugfs logs

by Rae Moar

Currently, the KUnit debugfs logs have a few issues: 1. The results of parameterized tests don’t show up 2. The order of the lines in the logs is sometimes incorrect 3. There are extra new lines in the logs This patch series aims to fix these issues. This is an example of a debugfs log prior to these fixes: KTAP version 1 # Subtest: kunit_status 1..2 # kunit_status: pass:2 fail:0 skip:0 total:2 # Totals: pass:2 fail:0 skip:0 total:2 ok 1 kunit_status_set_failure_test ok 2 kunit_status_mark_skipped_test ok 1 kunit_status Note there is an extra line and a few of the lines are out of order. This is the same debugfs log after the fixes: KTAP version 1 # Subtest: kunit_status 1..2 ok 1 kunit_status_set_failure_test ok 2 kunit_status_mark_skipped_test # kunit_status: pass:2 fail:0 skip:0 total:2 # Totals: pass:2 fail:0 skip:0 total:2 ok 4 kunit_status This is now equivalent to the regular KTAP output for the kunit_status test. Thanks! -Rae Rae Moar (3): kunit: fix bug in debugfs logs of parameterized tests kunit: fix bug in the order of lines in debugfs logs kunit: fix bug of extra newline characters in debugfs logs include/kunit/test.h | 2 +- lib/kunit/debugfs.c | 13 ++++++++----- lib/kunit/kunit-test.c | 36 ++++++++++++++++++++++++------------ lib/kunit/test.c | 36 ++++++++++++++++++++++-------------- 4 files changed, 55 insertions(+), 32 deletions(-) base-commit: 766f4f2514d2d18bcbd60a058188fb502dea5ddf -- 2.39.1.456.gfc5497dd1b-goog

1 year, 2 months

3
11
0 0

[PATCH] KVM: selftests: Added eBPF program and selftest to collect vmx exit stat

by Kevin Cheng

Introduce a new selftest that loads an eBPF program that stores the number of vmx exit counts per vcpu per vm. A process is created per vm_create to load a separate eBPF program to collect its own stats unique to the pid. This test aims to serve as a proof-of-concept and example for using eBPF to collect stats that are not provided by the other stats interfaces such as kvm_binary_stats. Since there will be no further stats being added to kvm_binary_stats, developers can use this selftest as a reference for writing their own eBPF program + selftest to collect whatever stat they may need for debugging/monitoring. Signed-off-by: Kevin Cheng <chengkev(a)google.com> --- tools/testing/selftests/kvm/Makefile | 4 +- tools/testing/selftests/kvm/build_ebpf.sh | 5 + .../testing/selftests/kvm/kvm_vmx_exit_ebpf.c | 128 ++++++++++++++++++ .../selftests/kvm/kvm_vmx_exit_ebpf_kern.c | 74 ++++++++++ 4 files changed, 210 insertions(+), 1 deletion(-) create mode 100644 tools/testing/selftests/kvm/build_ebpf.sh create mode 100644 tools/testing/selftests/kvm/kvm_vmx_exit_ebpf.c create mode 100644 tools/testing/selftests/kvm/kvm_vmx_exit_ebpf_kern.c diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile index 1750f91dd936..d9f56ccbc7bb 100644 --- a/tools/testing/selftests/kvm/Makefile +++ b/tools/testing/selftests/kvm/Makefile @@ -129,6 +129,7 @@ TEST_GEN_PROGS_x86_64 += set_memory_region_test TEST_GEN_PROGS_x86_64 += steal_time TEST_GEN_PROGS_x86_64 += kvm_binary_stats_test TEST_GEN_PROGS_x86_64 += system_counter_offset_test +TEST_GEN_PROGS_x86_64 += kvm_vmx_exit_ebpf # Compiled outputs used by test targets TEST_GEN_PROGS_EXTENDED_x86_64 += x86_64/nx_huge_pages_test @@ -176,6 +177,7 @@ TEST_GEN_PROGS_riscv += set_memory_region_test TEST_GEN_PROGS_riscv += kvm_binary_stats_test TEST_PROGS += $(TEST_PROGS_$(ARCH_DIR)) +TEST_PROGS := build_ebpf.sh TEST_GEN_PROGS += $(TEST_GEN_PROGS_$(ARCH_DIR)) TEST_GEN_PROGS_EXTENDED += $(TEST_GEN_PROGS_EXTENDED_$(ARCH_DIR)) LIBKVM += $(LIBKVM_$(ARCH_DIR)) @@ -208,7 +210,7 @@ no-pie-option := $(call try-run, echo 'int main(void) { return 0; }' | \ pgste-option = $(call try-run, echo 'int main(void) { return 0; }' | \ $(CC) -Werror -Wl$(comma)--s390-pgste -x c - -o "$$TMP",-Wl$(comma)--s390-pgste) -LDLIBS += -ldl +LDLIBS += -ldl -L$(top_srcdir)/tools/lib/bpf -lbpf -lelf -lz LDFLAGS += -pthread $(no-pie-option) $(pgste-option) LIBKVM_C := $(filter %.c,$(LIBKVM)) diff --git a/tools/testing/selftests/kvm/build_ebpf.sh b/tools/testing/selftests/kvm/build_ebpf.sh new file mode 100644 index 000000000000..b8038b0a0da5 --- /dev/null +++ b/tools/testing/selftests/kvm/build_ebpf.sh @@ -0,0 +1,5 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 +clang -g -O2 -target bpf -D__TARGET_ARCH_x86_64 -I . -c kvm_vmx_exit_ebpf_kern.c + -o kvm_vmx_exit_ebpf_kern.o +make -C ../../../lib/bpf || exit diff --git a/tools/testing/selftests/kvm/kvm_vmx_exit_ebpf.c b/tools/testing/selftests/kvm/kvm_vmx_exit_ebpf.c new file mode 100644 index 000000000000..a4bd2c549207 --- /dev/null +++ b/tools/testing/selftests/kvm/kvm_vmx_exit_ebpf.c @@ -0,0 +1,128 @@ +// SPDX-License-Identifier: GPL-2.0-only + +#include <stdio.h> +#include <stdlib.h> +#include <signal.h> +#include <sys/types.h> +#include <sys/wait.h> +#include <unistd.h> +#include <bpf/bpf.h> +#include <../bpf/libbpf.h> +#include <linux/btf.h> + +#include "test_util.h" + +#include "kvm_util.h" +#include "linux/kvm.h" + +#define VCPU_ID 0 + +struct stats_map_key { + __u32 pid; + __u32 vcpu_id; + __u32 exit_reason; +}; + +static void guest_code(void) +{ + __asm__ __volatile__("cpuid"); +} + +int main(int argc, char **argv) +{ + if (argc < 2) { + fprintf(stderr, "Expected arguments: <number_of_vms>\n"); + return EXIT_FAILURE; + } + int n = atoi(argv[1]); + + for (int i = 0; i < n; i++) { + if (fork() == 0) { + struct kvm_vm *vm; + struct kvm_vcpu *vcpu; + + vm = vm_create_with_one_vcpu(&vcpu, guest_code); + + // BPF userspace code + struct bpf_object *obj; + struct bpf_program *prog; + struct bpf_map *map_obj; + struct bpf_link *link = NULL; + + obj = bpf_object__open_file("kvm_vmx_exit_ebpf_kern.o", NULL); + if (libbpf_get_error(obj)) { + fprintf(stderr, "ERROR: opening BPF object file failed\n"); + return 0; + } + + map_obj = bpf_object__find_map_by_name(obj, "vmx_exit_map"); + if (!map_obj) { + fprintf(stderr, "ERROR: loading of vmx BPF map failed\n"); + goto cleanup; + } + + struct bpf_map *pid_map = bpf_object__find_map_by_name(obj, "pid_map"); + + if (!pid_map) { + fprintf(stderr, "ERROR: loading of pid BPF map failed\n"); + goto cleanup; + } + + /* load BPF program */ + if (bpf_object__load(obj)) { + fprintf(stderr, "ERROR: loading BPF object file failed\n"); + goto cleanup; + } + + __u32 userspace_pid = (__u32)getpid(); + __u32 val = (__u32)getpid(); + + bpf_map_update_elem(bpf_map__fd(pid_map), &userspace_pid, &val, 0); + + prog = bpf_object__find_program_by_name(obj, "bpf_exit_prog"); + if (libbpf_get_error(prog)) { + fprintf(stderr, "ERROR: finding a prog in obj file failed\n"); + goto cleanup; + } + + link = bpf_program__attach(prog); + if (libbpf_get_error(link)) { + fprintf(stderr, "ERROR: bpf_program__attach failed\n"); + link = NULL; + goto cleanup; + } + + for (int j = 0; j < 10000; j++) + vcpu_run(vcpu); + + struct stats_map_key key = { + .pid = 0, + .vcpu_id = 0, + .exit_reason = 18, + }; + + + struct stats_map_key next_key, lookup_key; + + lookup_key = key; + while (bpf_map_get_next_key(bpf_map__fd(map_obj), &lookup_key, &next_key) + == 0) { + int count; + + bpf_map_lookup_elem(bpf_map__fd(map_obj), &next_key, &count); + fprintf(stdout, "exit reason: '%d'\ncount: %d\npid: %d\n", + next_key.exit_reason, count, next_key.pid); + lookup_key = next_key; + } + +cleanup: + bpf_link__destroy(link); + bpf_object__close(obj); + kvm_vm_free(vm); + } + } + + for (int i = 0; i < n; i++) + wait(NULL); + return 0; +} diff --git a/tools/testing/selftests/kvm/kvm_vmx_exit_ebpf_kern.c b/tools/testing/selftests/kvm/kvm_vmx_exit_ebpf_kern.c new file mode 100644 index 000000000000..b9c076f93171 --- /dev/null +++ b/tools/testing/selftests/kvm/kvm_vmx_exit_ebpf_kern.c @@ -0,0 +1,73 @@ +// SPDX-License-Identifier: GPL-2.0-only +#include <linux/bpf.h> +#include <stdint.h> +#include <bpf/bpf_helpers.h> +#include <bpf/bpf_tracing.h> +#include <bpf/bpf_core_read.h> + +struct kvm_vcpu { + int vcpu_id; +}; + +struct vmx_args { + __u64 pad; + unsigned int exit_reason; + __u32 isa; + struct kvm_vcpu *vcpu; +}; + +struct stats_map_key { + __u32 pid; + __u32 vcpu_id; + __u32 exit_reason; +}; + +struct { + __uint(type, BPF_MAP_TYPE_HASH); + __uint(max_entries, 1024); + __type(key, struct stats_map_key); + __type(value, int); +} vmx_exit_map SEC(".maps"); + +struct { + __uint(type, BPF_MAP_TYPE_HASH); + __uint(max_entries, 1); + __type(key, __u32); + __type(value, __u32); +} pid_map SEC(".maps"); + + +SEC("tracepoint/kvm/kvm_exit") +int bpf_exit_prog(struct vmx_args *ctx) +{ + __u32 curr_pid = (bpf_get_current_pid_tgid() >> 32); + + __u32 *userspace_pid = bpf_map_lookup_elem(&pid_map, &curr_pid); + + if (!userspace_pid || *userspace_pid != curr_pid) + return 0; + + struct kvm_vcpu *vcpu = ctx->vcpu; + int _vcpu_id = BPF_CORE_READ(vcpu, vcpu_id); + + struct stats_map_key key = { + .pid = (bpf_get_current_pid_tgid() >> 32), + .vcpu_id = _vcpu_id, + .exit_reason = ctx->exit_reason, + }; + + int *value = bpf_map_lookup_elem(&vmx_exit_map, &key); + + if (value) { + *value = *value + 1; + bpf_map_update_elem(&vmx_exit_map, &key, value, BPF_ANY); + } else { + int temp = 1; + + bpf_map_update_elem(&vmx_exit_map, &key, &temp, BPF_ANY); + } + + return 0; +} + +char _license[] SEC("license") = "GPL"; -- 2.39.1.456.gfc5497dd1b-goog

1 year, 2 months

3
3
0 0

2024

2023

2022

2021

2020

2019

2018

2017

Linux-kselftest-mirror January 2023