linux-dev/tools/testing/selftests/powerpc/tm/Makefile, branch master

selftests/powerpc: Always build the tm-poison test 64-bit

2020-04-04T10:41:40Z

The tm-poison test includes inline asm which is 64-bit only, so the test must be built 64-bit in order to work. Otherwise it fails, eg: # file tm-poison tm-poison: ELF 32-bit MSB executable, PowerPC or cisco 4500, version 1 (SYSV) ... # ./tm-poison test: tm_poison_test Unknown value 0x1fff71150 leaked into f31! Unknown value 0x1fff710c0 leaked into vr31! failure: tm_poison_test Fixes: a003365cab64 ("powerpc/tm: Add tm-poison test") Signed-off-by: Michael Ellerman Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20200403095656.3772005-1-mpe@ellerman.id.au

selftests/powerpc: Turn off timeout setting for benchmarks, dscr, signal, tm

2020-03-25T01:09:30Z

Some specific tests in powerpc can take longer than the default 45 seconds that added in commit 852c8cbf34d3 ("selftests/kselftest/runner.sh: Add 45 second timeout per test") to run, the following test result was collected across 2 Power8 nodes and 1 Power9 node in our pool: powerpc/benchmarks/futex_bench - 52s powerpc/dscr/dscr_sysfs_test - 116s powerpc/signal/signal_fuzzer - 88s powerpc/tm/tm_unavailable_test - 168s powerpc/tm/tm-poison - 240s Thus they will fail with TIMEOUT error. Disable the timeout setting for these sub-tests to allow them finish properly. https://bugs.launchpad.net/bugs/1864642 Fixes: 852c8cbf34d3 ("selftests/kselftest/runner.sh: Add 45 second timeout per test") Signed-off-by: Po-Hsu Lin Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20200318060004.10685-1-po-hsu.lin@canonical.com

selftests/powerpc: Add tm-signal-pagefault test

2020-03-20T02:10:23Z

This test triggers a TM Bad Thing by raising a signal in transactional state and forcing a pagefault to happen in kernelspace when the kernel signal handling code first touches the user signal stack. This is inspired by the test tm-signal-context-force-tm but uses userfaultfd to make the test deterministic. While this test always triggers the bug in one run, I had to execute tm-signal-context-force-tm several times (the test runs 5000 times each execution) to trigger the same bug. tm-signal-context-force-tm is kept instead of replaced because, while this test is more reliable and triggers the same bug, tm-signal-context-force-tm has a better coverage, in the sense that by running the test several times it might trigger the pagefault and/or be preempted at different places. v3: skip test if userfaultfd is unavailable. Signed-off-by: Gustavo Luiz Duarte Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20200211033831.11165-2-gustavold@linux.ibm.com

powerpc/tm: Add tm-poison test

2019-09-20T22:36:53Z

Add TM selftest to check if FP or VEC register values from one process can leak into another process when both run on the same CPU. Signed-off-by: Gustavo Romero Signed-off-by: Michael Neuling Signed-off-by: Michael Ellerman Link: https://lore.kernel.org/r/20190904045529.23002-3-gromero@linux.vnet.ibm.com

selftests/powerpc: New TM signal self test

2019-01-15T00:17:09Z

A new self test that forces MSR[TS] to be set without calling any TM instruction. This test also tries to cause a page fault at a signal handler, exactly between MSR[TS] set and tm_recheckpoint(), forcing thread->texasr to be rewritten with TEXASR[FS] = 0, which will cause a BUG when tm_recheckpoint() is called. This test is not deterministic, since it is hard to guarantee that the page access will cause a page fault. In order to force more page faults at signal context, the signal handler and the ucontext are being mapped into a MADV_DONTNEED memory chunks. Tests have shown that the bug could be exposed with few interactions in a buggy kernel. This test is configured to loop 5000x, having a good chance to hit the kernel issue in just one run. This self test takes less than two seconds to run. This test uses set/getcontext because the kernel will recheckpoint zeroed structures, causing the test to segfault, which is undesired because the test needs to rerun, so, there is a signal handler for SIGSEGV which will restart the test. v2: Uses the MADV_DONTNEED memory advice v3: Fix memcpy and 32-bits compilation v4: Does not define unused macros Signed-off-by: Breno Leitao Signed-off-by: Michael Ellerman

selftests/powerpc: Add checks for transactional sigreturn

2018-12-21T03:46:50Z

This is a new test case that creates a signal and starts a suspended transaction inside the signal handler. It returns from the signal handler with the CPU at suspended state, but without setting user context MSR Transaction State (TS) field. The kernel signal handler code should be able to handle this discrepancy instead of crashing. This code could be compiled and used to test 32 and 64-bits signal handlers. Signed-off-by: Breno Leitao Signed-off-by: Gustavo Romero Signed-off-by: Michael Ellerman

selftests/powerpc: Fix Makefiles for headers_install change

2018-09-28T05:07:45Z

Commit b2d35fa5fc80 ("selftests: add headers_install to lib.mk") introduced a requirement that Makefiles more than one level below the selftests directory need to define top_srcdir, but it didn't update any of the powerpc Makefiles. This broke building all the powerpc selftests with eg: make[1]: Entering directory '/src/linux/tools/testing/selftests/powerpc' BUILD_TARGET=/src/linux/tools/testing/selftests/powerpc/alignment; mkdir -p $BUILD_TARGET; make OUTPUT=$BUILD_TARGET -k -C alignment all make[2]: Entering directory '/src/linux/tools/testing/selftests/powerpc/alignment' ../../lib.mk:20: ../../../../scripts/subarch.include: No such file or directory make[2]: *** No rule to make target '../../../../scripts/subarch.include'. make[2]: Failed to remake makefile '../../../../scripts/subarch.include'. Makefile:38: recipe for target 'alignment' failed Fix it by setting top_srcdir in the affected Makefiles. Fixes: b2d35fa5fc80 ("selftests: add headers_install to lib.mk") Signed-off-by: Michael Ellerman

Merge branch 'fixes' into next

2018-03-28T11:59:50Z

Merge our fixes branch from the 4.16 cycle. There were a number of important fixes merged, in particular some Power9 workarounds that we want in next for testing purposes. There's also been some conflicting changes in the CPU features code which are best merged and tested before going upstream.

selftest/powerpc: Add test for sigreturn in transaction

2018-03-13T04:10:14Z

Ensure that kernel is throwing away the suspended transaction when sigreturn() is called otherwise it if fails to restore the signal frame's TM SPRS. Signed-off-by: Laurent Dufour Reviewed-by: Cyril Bur [mpe: Add have_htm() check, minor formatting, add SPDX tag] Signed-off-by: Michael Ellerman

selftests/powerpc: Fix missing clean of pmu/lib.o

2018-02-28T11:28:35Z

The tm-resched-dscr test links against pmu/lib.o, but we don't have a rule to clean pmu/lib.o. This can lead to a build break if you build for big endian and then little, or vice versa. Fix it by making tm-resched-dscr depend on pmu/lib.c, causing the code to be built directly in, meaning no .o is generated. Signed-off-by: Michael Ellerman