Mark Rutland's recent SME fixes updated the SME ABI to reject any attempt to write FPSIMD register data via the streaming mode SVE register set but did not update the sve-ptrace test to take account of this, resulting in spurious failures. Update the test for this, and also fix another preexisting issue I noticed while looking at this.
Signed-off-by: Mark Brown broonie@kernel.org --- Changes in v2: - Rebase onto v6.16-rc1. - Update fixes tag for patch 1. - Link to v1: https://lore.kernel.org/r/20250523-kselftest-arm64-ssve-fixups-v1-0-65069a26...
--- Mark Brown (3): kselftest/arm64: Fix check for setting new VLs in sve-ptrace kselftest/arm64: Fix test for streaming FPSIMD write in sve-ptrace kselftest/arm64: Specify SVE data when testing VL set in sve-ptrace
tools/testing/selftests/arm64/fp/sve-ptrace.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) --- base-commit: 19272b37aa4f83ca52bdf9c16d5d81bdd1354494 change-id: 20250523-kselftest-arm64-ssve-fixups-b68ae61c1ebf
Best regards,
The check that the new vector length we set was the expected one was typoed to an assignment statement which for some reason the compilers didn't spot, most likely due to the macros involved.
Fixes: a1d7111257cd ("selftests: arm64: More comprehensively test the SVE ptrace interface") Acked-by: Mark Rutland mark.rutland@arm.com Acked-by: Dev Jain dev.jain@arm.com Signed-off-by: Mark Brown broonie@kernel.org --- tools/testing/selftests/arm64/fp/sve-ptrace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tools/testing/selftests/arm64/fp/sve-ptrace.c b/tools/testing/selftests/arm64/fp/sve-ptrace.c index 577b6e05e860..c499d5789dd5 100644 --- a/tools/testing/selftests/arm64/fp/sve-ptrace.c +++ b/tools/testing/selftests/arm64/fp/sve-ptrace.c @@ -253,7 +253,7 @@ static void ptrace_set_get_vl(pid_t child, const struct vec_type *type, return; }
- ksft_test_result(new_sve->vl = prctl_vl, "Set %s VL %u\n", + ksft_test_result(new_sve->vl == prctl_vl, "Set %s VL %u\n", type->name, vl);
free(new_sve);
Since f916dd32a943 ("arm64/fpsimd: ptrace: Mandate SVE payload for streaming-mode state") we do not support writing FPSIMD payload data when writing NT_ARM_SSVE but the sve-ptrace test has an explicit test for this being supported which was not updated to reflect the new behaviour. Fix the test to expect a failure when writing FPSIMD data to the streaming mode register set.
Acked-by: Mark Rutland mark.rutland@arm.com Signed-off-by: Mark Brown broonie@kernel.org --- tools/testing/selftests/arm64/fp/sve-ptrace.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/tools/testing/selftests/arm64/fp/sve-ptrace.c b/tools/testing/selftests/arm64/fp/sve-ptrace.c index c499d5789dd5..7e259907805b 100644 --- a/tools/testing/selftests/arm64/fp/sve-ptrace.c +++ b/tools/testing/selftests/arm64/fp/sve-ptrace.c @@ -301,8 +301,10 @@ static void ptrace_sve_fpsimd(pid_t child, const struct vec_type *type) p[j] = j; }
+ /* This should only succeed for SVE */ ret = set_sve(child, type, sve); - ksft_test_result(ret == 0, "%s FPSIMD set via SVE: %d\n", + ksft_test_result((type->regset == NT_ARM_SVE) == (ret == 0), + "%s FPSIMD set via SVE: %d\n", type->name, ret); if (ret) goto out;
Since f916dd32a943 ("arm64/fpsimd: ptrace: Mandate SVE payload for streaming-mode state") we reject attempts to write to the streaming mode regset even if there is no register data supplied, causing the tests for setting vector lengths and setting SVE_VL_INHERIT in sve-ptrace to spuriously fail. Set the flag to avoid the issue, we still support not supplying register data.
Acked-by: Mark Rutland mark.rutland@arm.com Signed-off-by: Mark Brown broonie@kernel.org --- tools/testing/selftests/arm64/fp/sve-ptrace.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/tools/testing/selftests/arm64/fp/sve-ptrace.c b/tools/testing/selftests/arm64/fp/sve-ptrace.c index 7e259907805b..7f9b6a61d369 100644 --- a/tools/testing/selftests/arm64/fp/sve-ptrace.c +++ b/tools/testing/selftests/arm64/fp/sve-ptrace.c @@ -170,7 +170,7 @@ static void ptrace_set_get_inherit(pid_t child, const struct vec_type *type) memset(&sve, 0, sizeof(sve)); sve.size = sizeof(sve); sve.vl = sve_vl_from_vq(SVE_VQ_MIN); - sve.flags = SVE_PT_VL_INHERIT; + sve.flags = SVE_PT_VL_INHERIT | SVE_PT_REGS_SVE; ret = set_sve(child, type, &sve); if (ret != 0) { ksft_test_result_fail("Failed to set %s SVE_PT_VL_INHERIT\n", @@ -235,6 +235,7 @@ static void ptrace_set_get_vl(pid_t child, const struct vec_type *type, /* Set the VL by doing a set with no register payload */ memset(&sve, 0, sizeof(sve)); sve.size = sizeof(sve); + sve.flags = SVE_PT_REGS_SVE; sve.vl = vl; ret = set_sve(child, type, &sve); if (ret != 0) {
On Mon, 09 Jun 2025 16:25:30 +0100, Mark Brown wrote:
Mark Rutland's recent SME fixes updated the SME ABI to reject any attempt to write FPSIMD register data via the streaming mode SVE register set but did not update the sve-ptrace test to take account of this, resulting in spurious failures. Update the test for this, and also fix another preexisting issue I noticed while looking at this.
[...]
Applied to arm64 (for-next/kselftest), thanks!
[1/3] kselftest/arm64: Fix check for setting new VLs in sve-ptrace https://git.kernel.org/arm64/c/867446f09058 [2/3] kselftest/arm64: Fix test for streaming FPSIMD write in sve-ptrace https://git.kernel.org/arm64/c/94ab150010f4 [3/3] kselftest/arm64: Specify SVE data when testing VL set in sve-ptrace https://git.kernel.org/arm64/c/9e8ebfe677f9
linux-kselftest-mirror@lists.linaro.org
-
Catalin Marinas -
Mark Brown