February 2023 - Linux-kselftest-mirror

[PATCH] kunit: Add printf attribute to fail_current_test_impl

by David Gow

Add the gnu_printf (__printf()) attribute to the kunit_fail_current_test() implementation in __kunit_fail_current_test_impl(). While it's not actually useful here, as this function is never called directly, it nevertheless was triggering -Wsuggest-attribute=format warnings, so we should add it to reduce the noise. Fixes: cc3ed2fe5c93 ("kunit: Add "hooks" to call into KUnit when it's built as a module") Reported-by: kernel test robot <lkp(a)intel.com> Signed-off-by: David Gow <davidgow(a)google.com> --- lib/kunit/hooks-impl.h | 4 +++- lib/kunit/test.c | 2 +- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/lib/kunit/hooks-impl.h b/lib/kunit/hooks-impl.h index ec745a39832c..4e71b2d0143b 100644 --- a/lib/kunit/hooks-impl.h +++ b/lib/kunit/hooks-impl.h @@ -15,7 +15,9 @@ #include <kunit/test-bug.h> /* List of declarations. */ -void __kunit_fail_current_test_impl(const char *file, int line, const char *fmt, ...); +void __printf(3, 4) __kunit_fail_current_test_impl(const char *file, + int line, + const char *fmt, ...); void *__kunit_get_static_stub_address_impl(struct kunit *test, void *real_fn_addr); /* Code to set all of the function pointers. */ diff --git a/lib/kunit/test.c b/lib/kunit/test.c index 51cae59d8aae..c9e15bb60058 100644 --- a/lib/kunit/test.c +++ b/lib/kunit/test.c @@ -24,7 +24,7 @@ /* * Hook to fail the current test and print an error message to the log. */ -void __kunit_fail_current_test_impl(const char *file, int line, const char *fmt, ...) +void __printf(3, 4) __kunit_fail_current_test_impl(const char *file, int line, const char *fmt, ...) { va_list args; int len; -- 2.39.1.519.gcb327c4b5f-goog

1 year, 2 months

2
1
0 0

[PATCH v2] selftests: use printf instead of echo -ne

by Guillaume Tucker

On some systems, the default echo command doesn't handle the -e option and the output looks like this (arm64 build): -ne Emit Tests for alsa -ne Emit Tests for amd-pstate -ne Emit Tests for arm64 This is for example the case with the KernelCI Docker images e.g. kernelci/gcc-10:x86-kselftest-kernelci. To avoid this issue, use printf which handles escape characters as a standard feature and is more widespread among modern shells. The output is now formatted as expected (x86 build this time): Emit Tests for alsa Emit Tests for amd-pstate Skipping non-existent dir: arm64 Reported-by: "kernelci.org bot" <bot(a)kernelci.org> Suggested-by: David Laight <David.Laight(a)ACULAB.COM> Fixes: 3297a4df805d ("kselftests: Enable the echo command to print newlines in Makefile") Signed-off-by: Guillaume Tucker <guillaume.tucker(a)collabora.com> --- Notes: v2: use printf insead of $(which echo) tools/testing/selftests/Makefile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/Makefile b/tools/testing/selftests/Makefile index 41b649452560..06578963f4f1 100644 --- a/tools/testing/selftests/Makefile +++ b/tools/testing/selftests/Makefile @@ -236,8 +236,8 @@ ifdef INSTALL_PATH @# included in the generated runlist. for TARGET in $(TARGETS); do \ BUILD_TARGET=$$BUILD/$$TARGET; \ - [ ! -d $(INSTALL_PATH)/$$TARGET ] && echo "Skipping non-existent dir: $$TARGET" && continue; \ - echo -ne "Emit Tests for $$TARGET\n"; \ + [ ! -d $(INSTALL_PATH)/$$TARGET ] && printf "Skipping non-existent dir: $$TARGET\n" && continue; \ + printf "Emit Tests for $$TARGET\n"; \ $(MAKE) -s --no-print-directory OUTPUT=$$BUILD_TARGET COLLECTION=$$TARGET \ -C $$TARGET emit_tests >> $(TEST_LIST); \ done; -- 2.30.2

1 year, 2 months

3
3
0 0

[PATCH] kselftest/arm64: Don't require FA64 for streaming SVE+ZA tests

by Mark Brown

During early development a dependedncy was added on having FA64 available so we could use the full FPSIMD register set in the signal handler which got copied over into the SSVE+ZA registers test case. Subsequently the ABI was finialised so the handler is run with streaming mode disabled meaning this is redundant but the dependency was never removed, do so now. Signed-off-by: Mark Brown <broonie(a)kernel.org> --- tools/testing/selftests/arm64/signal/testcases/ssve_za_regs.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) diff --git a/tools/testing/selftests/arm64/signal/testcases/ssve_za_regs.c b/tools/testing/selftests/arm64/signal/testcases/ssve_za_regs.c index 1f62621794d5..9dc5f128bbc0 100644 --- a/tools/testing/selftests/arm64/signal/testcases/ssve_za_regs.c +++ b/tools/testing/selftests/arm64/signal/testcases/ssve_za_regs.c @@ -154,12 +154,7 @@ static int sme_regs(struct tdescr *td, siginfo_t *si, ucontext_t *uc) struct tdescr tde = { .name = "Streaming SVE registers", .descr = "Check that we get the right Streaming SVE registers reported", - /* - * We shouldn't require FA64 but things like memset() used in the - * helpers might use unsupported instructions so for now disable - * the test unless we've got the full instruction set. - */ - .feats_required = FEAT_SME | FEAT_SME_FA64, + .feats_required = FEAT_SME, .timeout = 3, .init = sme_get_vls, .run = sme_regs, --- base-commit: 7294f24db4fa5ebb5a6bde104f08d3345ecee053 change-id: 20230202-arm64-kselftest-sve-za-fa64-9a04f0c49052 Best regards, -- Mark Brown <broonie(a)kernel.org>

1 year, 2 months

2
1
0 0

[PATCH] kselftest/arm64: Copy whole EXTRA context

by Mark Brown

When copying the EXTRA context our calculation of the amount of data we need to copy is incorrect, we only calculate the amount of data needed within uc_mcontext.__reserved, not taking account of the fixed portion of the context. Add in the offset of the reserved data so that we copy everything we should. This will only cause test failures in cases where the last context in the EXTRA context is smaller than the missing data since we don't currently validate any of the register data and all the buffers we copy into are statically allocated so default to zero meaning that if we walk beyond the end of what we copied we'll encounter what looks like a context with magic and length both 0 which is a valid terminator record. Signed-off-by: Mark Brown <broonie(a)kernel.org> --- tools/testing/selftests/arm64/signal/test_signals_utils.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/arm64/signal/test_signals_utils.c b/tools/testing/selftests/arm64/signal/test_signals_utils.c index 308e229e58ab..746a4f70f082 100644 --- a/tools/testing/selftests/arm64/signal/test_signals_utils.c +++ b/tools/testing/selftests/arm64/signal/test_signals_utils.c @@ -192,8 +192,10 @@ static bool handle_signal_copyctx(struct tdescr *td, * in the copy, this was previously validated in * ASSERT_GOOD_CONTEXT(). */ - to_copy = offset + sizeof(struct extra_context) + 16 + - extra->size; + to_copy = __builtin_offsetof(ucontext_t, + uc_mcontext.__reserved); + to_copy += offset + sizeof(struct extra_context) + 16; + to_copy += extra->size; copied_extra = (struct extra_context *)&(td->live_uc->uc_mcontext.__reserved[offset]); } else { copied_extra = NULL; --- base-commit: b7bfaa761d760e72a969d116517eaa12e404c262 change-id: 20230201-arm64-kselftest-full-extra-164baae78412 Best regards, -- Mark Brown <broonie(a)kernel.org>

1 year, 2 months

2
1
0 0

[PATCH v7 00/14] KVM: s390: Extend MEM_OP ioctl by storage key checked cmpxchg

by Janis Schoetterl-Glausch

User space can use the MEM_OP ioctl to make storage key checked reads and writes to the guest, however, it has no way of performing atomic, key checked, accesses to the guest. Extend the MEM_OP ioctl in order to allow for this, by adding a cmpxchg operation. For now, support this operation for absolute accesses only. This operation can be use, for example, to set the device-state-change indicator and the adapter-local-summary indicator atomically. Also contains some fixes/changes for the memop selftest independent of the cmpxchg changes. v6 -> v7 * cast to __user pointers in cmpxchg_guest_abs_with_key * dropped function for allocating buffer for mem_op * use goto if CHECK_ONLY in mem_op refactoring * changed assert message in bad address test * picked up R-b's (thanks Thomas & Janosch) * fix nits, fix typos, improve commit descriptions v5 -> v6 * move memop selftest fixes/refactoring to front of series so they can be picked independently from the rest * use op instead of flag to indicate cmpxchg * no longer indicate success of cmpxchg to user space, which can infer it by observing a change in the old value instead * refactor functions implementing the ioctl * adjust documentation (drop R-b) * adjust selftest * rebase v4 -> v5 * refuse cmpxchg if not write (thanks Thomas) * minor doc changes (thanks Claudio) * picked up R-b's (thanks Thomas & Claudio) * memop selftest fixes * rebased v3 -> v4 * no functional change intended * rework documentation a bit * name extension cap cmpxchg bit * picked up R-b (thanks Thomas) * various changes (rename variable, comments, ...) see range-diff below v2 -> v3 * rebase onto the wip/cmpxchg_user_key branch in the s390 kernel repo * use __uint128_t instead of unsigned __int128 * put moving of testlist into main into separate patch * pick up R-b's (thanks Nico) v1 -> v2 * get rid of xrk instruction for cmpxchg byte and short implementation * pass old parameter via pointer instead of in mem_op struct * indicate failure of cmpxchg due to wrong old value by special return code * picked up R-b's (thanks Thomas) Janis Schoetterl-Glausch (14): KVM: s390: selftest: memop: Pass mop_desc via pointer KVM: s390: selftest: memop: Replace macros by functions KVM: s390: selftest: memop: Move testlist into main KVM: s390: selftest: memop: Add bad address test KVM: s390: selftest: memop: Fix typo KVM: s390: selftest: memop: Fix wrong address being used in test KVM: s390: selftest: memop: Fix integer literal KVM: s390: Move common code of mem_op functions into function KVM: s390: Dispatch to implementing function at top level of vm mem_op KVM: s390: Refactor absolute vm mem_op function KVM: s390: Refactor vcpu mem_op function KVM: s390: Extend MEM_OP ioctl by storage key checked cmpxchg Documentation: KVM: s390: Describe KVM_S390_MEMOP_F_CMPXCHG KVM: s390: selftest: memop: Add cmpxchg tests Documentation/virt/kvm/api.rst | 29 +- include/uapi/linux/kvm.h | 8 + arch/s390/kvm/gaccess.h | 3 + arch/s390/kvm/gaccess.c | 109 ++++ arch/s390/kvm/kvm-s390.c | 221 ++++--- tools/testing/selftests/kvm/s390x/memop.c | 675 +++++++++++++++++----- 6 files changed, 809 insertions(+), 236 deletions(-) Range-diff against v6: 1: 512e1a3e0ae5 ! 1: fb51df0930d8 KVM: s390: selftest: memop: Pass mop_desc via pointer @@ Commit message Signed-off-by: Janis Schoetterl-Glausch <scgl(a)linux.ibm.com> Reviewed-by: Thomas Huth <thuth(a)redhat.com> + Reviewed-by: Janosch Frank <frankja(a)linux.ibm.com> ## tools/testing/selftests/kvm/s390x/memop.c ## @@ tools/testing/selftests/kvm/s390x/memop.c: struct mop_desc { 2: 47328ea64f80 ! 2: 6f09b06574f0 KVM: s390: selftest: memop: Replace macros by functions @@ Commit message KVM: s390: selftest: memop: Replace macros by functions Replace the DEFAULT_* test helpers by functions, as they don't - need the exta flexibility. + need the extra flexibility. Signed-off-by: Janis Schoetterl-Glausch <scgl(a)linux.ibm.com> Reviewed-by: Thomas Huth <thuth(a)redhat.com> + Acked-by: Janosch Frank <frankja(a)linux.ibm.com> ## tools/testing/selftests/kvm/s390x/memop.c ## @@ tools/testing/selftests/kvm/s390x/memop.c: struct mop_desc { 3: 224fe37eeec7 ! 3: 108c1af396fe KVM: s390: selftest: memop: Move testlist into main @@ Commit message Signed-off-by: Janis Schoetterl-Glausch <scgl(a)linux.ibm.com> Reviewed-by: Thomas Huth <thuth(a)redhat.com> + Reviewed-by: Janosch Frank <frankja(a)linux.ibm.com> ## tools/testing/selftests/kvm/s390x/memop.c ## @@ tools/testing/selftests/kvm/s390x/memop.c: static void test_errors(void) 4: f622d3413cf0 ! 4: 4bcad9b4bf9d KVM: s390: selftest: memop: Add bad address test @@ Metadata ## Commit message ## KVM: s390: selftest: memop: Add bad address test - Add test that tries to access, instead of CHECK_ONLY. + Add a test that tries a real write to a bad address. + The existing CHECK_ONLY test doesn't cover all paths. Signed-off-by: Janis Schoetterl-Glausch <scgl(a)linux.ibm.com> Reviewed-by: Nico Boehr <nrb(a)linux.ibm.com> @@ tools/testing/selftests/kvm/s390x/memop.c: static void _test_errors_common(struc /* Bad guest address: */ rv = ERR_MOP(info, target, WRITE, mem1, size, GADDR((void *)~0xfffUL), CHECK_ONLY); - TEST_ASSERT(rv > 0, "ioctl does not report bad guest memory access"); -+ TEST_ASSERT(rv > 0, "ioctl does not report bad guest memory address"); ++ TEST_ASSERT(rv > 0, "ioctl does not report bad guest memory address with CHECK_ONLY"); + rv = ERR_MOP(info, target, WRITE, mem1, size, GADDR((void *)~0xfffUL)); -+ TEST_ASSERT(rv > 0, "ioctl does not report bad guest memory address"); ++ TEST_ASSERT(rv > 0, "ioctl does not report bad guest memory address on write"); /* Bad host address: */ rv = ERR_MOP(info, target, WRITE, 0, size, GADDR_V(mem1)); 5: 431f191a8a57 = 5: 7af4c710045c KVM: s390: selftest: memop: Fix typo 6: 3122187435fb = 6: 2ab1854f3959 KVM: s390: selftest: memop: Fix wrong address being used in test 7: 401f51f3ef55 ! 7: 980644e1ce74 KVM: s390: selftest: memop: Fix integer literal @@ Commit message Signed-off-by: Janis Schoetterl-Glausch <scgl(a)linux.ibm.com> Fixes: 1bb873495a9e ("KVM: s390: selftests: Add more copy memop tests") + Reviewed-by: Thomas Huth <thuth(a)redhat.com> ## tools/testing/selftests/kvm/s390x/memop.c ## @@ tools/testing/selftests/kvm/s390x/memop.c: static void guest_copy_key_fetch_prot_override(void) 8: 185e5cd33df6 ! 8: a9930caadf5e KVM: s390: Move common code of mem_op functions into functions @@ Metadata Author: Janis Schoetterl-Glausch <scgl(a)linux.ibm.com> ## Commit message ## - KVM: s390: Move common code of mem_op functions into functions + KVM: s390: Move common code of mem_op functions into function The vcpu and vm mem_op ioctl implementations share some functionality. - Move argument checking and buffer allocation into functions and call - them from both implementations. - This allows code reuse in case of additional future mem_op operations. + Move argument checking into a function and call it from both + implementations. This allows code reuse in case of additional future + mem_op operations. Suggested-by: Janosch Frank <frankja(a)linux.ibm.com> Signed-off-by: Janis Schoetterl-Glausch <scgl(a)linux.ibm.com> @@ arch/s390/kvm/kvm-s390.c: static int kvm_s390_handle_pv(struct kvm *kvm, struct + mop->key = 0; + } + return 0; -+} -+ -+static void *mem_op_alloc_buf(struct kvm_s390_mem_op *mop) -+{ -+ void *buf; -+ -+ if (mop->flags & KVM_S390_MEMOP_F_CHECK_ONLY) -+ return NULL; -+ buf = vmalloc(mop->size); -+ if (!buf) -+ return ERR_PTR(-ENOMEM); -+ return buf; } static int kvm_s390_vm_mem_op(struct kvm *kvm, struct kvm_s390_mem_op *mop) @@ arch/s390/kvm/kvm-s390.c: static int kvm_s390_vm_mem_op(struct kvm *kvm, struct - } else { - mop->key = 0; - } -- if (!(mop->flags & KVM_S390_MEMOP_F_CHECK_ONLY)) { -- tmpbuf = vmalloc(mop->size); -- if (!tmpbuf) -- return -ENOMEM; -- } -+ tmpbuf = mem_op_alloc_buf(mop); -+ if (IS_ERR(tmpbuf)) -+ return PTR_ERR(tmpbuf); - - srcu_idx = srcu_read_lock(&kvm->srcu); - + if (!(mop->flags & KVM_S390_MEMOP_F_CHECK_ONLY)) { + tmpbuf = vmalloc(mop->size); + if (!tmpbuf) @@ arch/s390/kvm/kvm-s390.c: static long kvm_s390_vcpu_mem_op(struct kvm_vcpu *vcpu, { void __user *uaddr = (void __user *)mop->buf; @@ arch/s390/kvm/kvm-s390.c: static long kvm_s390_vcpu_mem_op(struct kvm_vcpu *vcpu - } else { - mop->key = 0; - } -- if (!(mop->flags & KVM_S390_MEMOP_F_CHECK_ONLY)) { -- tmpbuf = vmalloc(mop->size); -- if (!tmpbuf) -- return -ENOMEM; -- } -+ tmpbuf = mem_op_alloc_buf(mop); -+ if (IS_ERR(tmpbuf)) -+ return PTR_ERR(tmpbuf); - - switch (mop->op) { - case KVM_S390_MEMOP_LOGICAL_READ: + if (!(mop->flags & KVM_S390_MEMOP_F_CHECK_ONLY)) { + tmpbuf = vmalloc(mop->size); + if (!tmpbuf) 9: d7c8b94cb351 ! 9: 2715c1ceabdb KVM: s390: Dispatch to implementing function at top level of vm mem_op @@ Commit message Suggested-by: Janosch Frank <frankja(a)linux.ibm.com> Signed-off-by: Janis Schoetterl-Glausch <scgl(a)linux.ibm.com> + Reviewed-by: Thomas Huth <thuth(a)redhat.com> ## arch/s390/kvm/kvm-s390.c ## -@@ arch/s390/kvm/kvm-s390.c: static void *mem_op_alloc_buf(struct kvm_s390_mem_op *mop) - return buf; +@@ arch/s390/kvm/kvm-s390.c: static int mem_op_validate_common(struct kvm_s390_mem_op *mop, u64 supported_fla + return 0; } -static int kvm_s390_vm_mem_op(struct kvm *kvm, struct kvm_s390_mem_op *mop) @@ arch/s390/kvm/kvm-s390.c: static int kvm_s390_vm_mem_op(struct kvm *kvm, struct - */ - if (kvm_s390_pv_get_handle(kvm)) - return -EINVAL; - tmpbuf = mem_op_alloc_buf(mop); - if (IS_ERR(tmpbuf)) - return PTR_ERR(tmpbuf); + if (!(mop->flags & KVM_S390_MEMOP_F_CHECK_ONLY)) { + tmpbuf = vmalloc(mop->size); + if (!tmpbuf) @@ arch/s390/kvm/kvm-s390.c: static int kvm_s390_vm_mem_op(struct kvm *kvm, struct kvm_s390_mem_op *mop) } break; 10: 8acb425a9e93 ! 10: 91f4e887f311 KVM: s390: Refactor absolute vm mem_op function @@ Commit message Suggested-by: Janosch Frank <frankja(a)linux.ibm.com> Signed-off-by: Janis Schoetterl-Glausch <scgl(a)linux.ibm.com> + Reviewed-by: Thomas Huth <thuth(a)redhat.com> ## arch/s390/kvm/kvm-s390.c ## -@@ arch/s390/kvm/kvm-s390.c: static void *mem_op_alloc_buf(struct kvm_s390_mem_op *mop) +@@ arch/s390/kvm/kvm-s390.c: static int mem_op_validate_common(struct kvm_s390_mem_op *mop, u64 supported_fla static int kvm_s390_vm_mem_op_abs(struct kvm *kvm, struct kvm_s390_mem_op *mop) { void __user *uaddr = (void __user *)mop->buf; @@ arch/s390/kvm/kvm-s390.c: static int kvm_s390_vm_mem_op_abs(struct kvm *kvm, str - } - } - break; -- } ++ acc_mode = mop->op == KVM_S390_MEMOP_ABSOLUTE_READ ? GACC_FETCH : GACC_STORE; ++ if (mop->flags & KVM_S390_MEMOP_F_CHECK_ONLY) { ++ r = check_gpa_range(kvm, mop->gaddr, mop->size, acc_mode, mop->key); ++ goto out_unlock; + } - case KVM_S390_MEMOP_ABSOLUTE_WRITE: { - if (mop->flags & KVM_S390_MEMOP_F_CHECK_ONLY) { - r = check_gpa_range(kvm, mop->gaddr, mop->size, GACC_STORE, mop->key); @@ arch/s390/kvm/kvm-s390.c: static int kvm_s390_vm_mem_op_abs(struct kvm *kvm, str - } - r = access_guest_abs_with_key(kvm, mop->gaddr, tmpbuf, - mop->size, GACC_STORE, mop->key); -+ acc_mode = mop->op == KVM_S390_MEMOP_ABSOLUTE_READ ? GACC_FETCH : GACC_STORE; -+ if (mop->flags & KVM_S390_MEMOP_F_CHECK_ONLY) { -+ r = check_gpa_range(kvm, mop->gaddr, mop->size, acc_mode, mop->key); -+ } else if (acc_mode == GACC_FETCH) { ++ if (acc_mode == GACC_FETCH) { + r = access_guest_abs_with_key(kvm, mop->gaddr, tmpbuf, + mop->size, GACC_FETCH, mop->key); + if (r) 11: f62dfb06ea55 ! 11: d16ac0db04c7 KVM: s390: Refactor absolute vcpu mem_op function @@ Metadata Author: Janis Schoetterl-Glausch <scgl(a)linux.ibm.com> ## Commit message ## - KVM: s390: Refactor absolute vcpu mem_op function + KVM: s390: Refactor vcpu mem_op function Remove code duplication with regards to the CHECK_ONLY flag. Decrease the number of indents. @@ arch/s390/kvm/kvm-s390.c: static long kvm_s390_vcpu_mem_op(struct kvm_vcpu *vcpu int r; @@ arch/s390/kvm/kvm-s390.c: static long kvm_s390_vcpu_mem_op(struct kvm_vcpu *vcpu, - if (IS_ERR(tmpbuf)) - return PTR_ERR(tmpbuf); + return -ENOMEM; + } - switch (mop->op) { - case KVM_S390_MEMOP_LOGICAL_READ: @@ arch/s390/kvm/kvm-s390.c: static long kvm_s390_vcpu_mem_op(struct kvm_vcpu *vcpu + if (mop->flags & KVM_S390_MEMOP_F_CHECK_ONLY) { + r = check_gva_range(vcpu, mop->gaddr, mop->ar, mop->size, + acc_mode, mop->key); -+ } else if (acc_mode == GACC_FETCH) { ++ goto out_inject; ++ } ++ if (acc_mode == GACC_FETCH) { r = read_guest_with_key(vcpu, mop->gaddr, mop->ar, tmpbuf, mop->size, mop->key); - if (r == 0) { 12: 2199327cb484 ! 12: afec9544eec4 KVM: s390: Extend MEM_OP ioctl by storage key checked cmpxchg @@ Commit message Extend the MEM_OP ioctl in order to allow for this, by adding a cmpxchg op. For now, support this op for absolute accesses only. - This op can be use, for example, to set the device-state-change + This op can be used, for example, to set the device-state-change indicator and the adapter-local-summary indicator atomically. Signed-off-by: Janis Schoetterl-Glausch <scgl(a)linux.ibm.com> @@ arch/s390/kvm/gaccess.c: int access_guest_real(struct kvm_vcpu *vcpu, unsigned l + * Honors storage keys. + * + * Return: * 0: successful exchange -+ * * a program interruption code indicating the reason cmpxchg could -+ * not be attempted ++ * * >0: a program interruption code indicating the reason cmpxchg could ++ * not be attempted + * * -EINVAL: address misaligned or len not power of two + * * -EAGAIN: transient failure (len 1 or 2) + * * -EOPNOTSUPP: read-only memslot (should never occur) @@ arch/s390/kvm/gaccess.c: int access_guest_real(struct kvm_vcpu *vcpu, unsigned l + __uint128_t *old_addr, __uint128_t new, + u8 access_key, bool *success) +{ -+ gfn_t gfn = gpa >> PAGE_SHIFT; ++ gfn_t gfn = gpa_to_gfn(gpa); + struct kvm_memory_slot *slot = gfn_to_memslot(kvm, gfn); + bool writable; + hva_t hva; @@ arch/s390/kvm/gaccess.c: int access_guest_real(struct kvm_vcpu *vcpu, unsigned l + return -EOPNOTSUPP; + + hva += offset_in_page(gpa); ++ /* ++ * The cmpxchg_user_key macro depends on the type of "old", so we need ++ * a case for each valid length and get some code duplication as long ++ * as we don't introduce a new macro. ++ */ + switch (len) { + case 1: { + u8 old; + -+ ret = cmpxchg_user_key((u8 *)hva, &old, *old_addr, new, access_key); ++ ret = cmpxchg_user_key((u8 __user *)hva, &old, *old_addr, new, access_key); + *success = !ret && old == *old_addr; + *old_addr = old; + break; @@ arch/s390/kvm/gaccess.c: int access_guest_real(struct kvm_vcpu *vcpu, unsigned l + case 2: { + u16 old; + -+ ret = cmpxchg_user_key((u16 *)hva, &old, *old_addr, new, access_key); ++ ret = cmpxchg_user_key((u16 __user *)hva, &old, *old_addr, new, access_key); + *success = !ret && old == *old_addr; + *old_addr = old; + break; @@ arch/s390/kvm/gaccess.c: int access_guest_real(struct kvm_vcpu *vcpu, unsigned l + case 4: { + u32 old; + -+ ret = cmpxchg_user_key((u32 *)hva, &old, *old_addr, new, access_key); ++ ret = cmpxchg_user_key((u32 __user *)hva, &old, *old_addr, new, access_key); + *success = !ret && old == *old_addr; + *old_addr = old; + break; @@ arch/s390/kvm/gaccess.c: int access_guest_real(struct kvm_vcpu *vcpu, unsigned l + case 8: { + u64 old; + -+ ret = cmpxchg_user_key((u64 *)hva, &old, *old_addr, new, access_key); ++ ret = cmpxchg_user_key((u64 __user *)hva, &old, *old_addr, new, access_key); + *success = !ret && old == *old_addr; + *old_addr = old; + break; @@ arch/s390/kvm/gaccess.c: int access_guest_real(struct kvm_vcpu *vcpu, unsigned l + case 16: { + __uint128_t old; + -+ ret = cmpxchg_user_key((__uint128_t *)hva, &old, *old_addr, new, access_key); ++ ret = cmpxchg_user_key((__uint128_t __user *)hva, &old, *old_addr, new, access_key); + *success = !ret && old == *old_addr; + *old_addr = old; + break; @@ arch/s390/kvm/gaccess.c: int access_guest_real(struct kvm_vcpu *vcpu, unsigned l + default: + return -EINVAL; + } -+ mark_page_dirty_in_slot(kvm, slot, gfn); ++ if (*success) ++ mark_page_dirty_in_slot(kvm, slot, gfn); + /* + * Assume that the fault is caused by protection, either key protection + * or user page write protection. 13: 05d7682a90ba = 13: 2c006a4015c0 Documentation: KVM: s390: Describe KVM_S390_MEMOP_F_CMPXCHG 14: 993293dcdf13 = 14: ac16991fbe27 KVM: s390: selftest: memop: Add cmpxchg tests base-commit: 4ec5183ec48656cec489c49f989c508b68b518e3 prerequisite-patch-id: 02346a556c2af3340683322ef71816f813599c21 prerequisite-patch-id: 770ae375220957f1db8eb2657bf2fb7868a501af prerequisite-patch-id: fba46548bd1b60742eed8669a15cadad3c5ec19c prerequisite-patch-id: 8df365bc587d8768618759d77e3a6806bc36130d prerequisite-patch-id: 0fc74ea6b8342b1dd69b034be308647d47d31746 prerequisite-patch-id: c5cdc3ce7cdffc18c5e56abfb657c84141fb623a prerequisite-patch-id: 837715a7a75d9175027343d727c98a398b927c9b prerequisite-patch-id: 6ccea6ff976ae44347f266d7a9e19e09a2ae9063 -- 2.37.2

1 year, 2 months

2
24
0 0

[PATCH][next] selftests/bpf: Fix spelling mistake "detecion" -> "detection"

by Colin Ian King

There is a spelling mistake in a literal string. Fix it. Signed-off-by: Colin Ian King <colin.i.king(a)gmail.com> --- tools/testing/selftests/bpf/xdp_features.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/bpf/xdp_features.c b/tools/testing/selftests/bpf/xdp_features.c index 10fad1243573..fce12165213b 100644 --- a/tools/testing/selftests/bpf/xdp_features.c +++ b/tools/testing/selftests/bpf/xdp_features.c @@ -57,7 +57,7 @@ static void sig_handler(int sig) const char *argp_program_version = "xdp-features 0.0"; const char argp_program_doc[] = -"XDP features detecion application.\n" +"XDP features detection application.\n" "\n" "XDP features application checks the XDP advertised features match detected ones.\n" "\n" -- 2.30.2

1 year, 2 months

2
1
0 0

[PATCH] KVM: selftests: Remove redundant setbuf()

by shahuang＠redhat.com

From: Shaoqin Huang <shahuang(a)redhat.com> Since setbuf(stdout, NULL) has been called in kvm_util.c with __attribute((constructor)). Selftests no need to setup it in their own code. Signed-off-by: Shaoqin Huang <shahuang(a)redhat.com> --- tools/testing/selftests/kvm/aarch64/page_fault_test.c | 2 -- .../selftests/kvm/x86_64/exit_on_emulation_failure_test.c | 3 --- 2 files changed, 5 deletions(-) diff --git a/tools/testing/selftests/kvm/aarch64/page_fault_test.c b/tools/testing/selftests/kvm/aarch64/page_fault_test.c index beb944fa6fd4..513b20bec3c2 100644 --- a/tools/testing/selftests/kvm/aarch64/page_fault_test.c +++ b/tools/testing/selftests/kvm/aarch64/page_fault_test.c @@ -1093,8 +1093,6 @@ int main(int argc, char *argv[]) enum vm_mem_backing_src_type src_type; int opt; - setbuf(stdout, NULL); - src_type = DEFAULT_VM_MEM_SRC; while ((opt = getopt(argc, argv, "hm:s:")) != -1) { diff --git a/tools/testing/selftests/kvm/x86_64/exit_on_emulation_failure_test.c b/tools/testing/selftests/kvm/x86_64/exit_on_emulation_failure_test.c index 37c61f712fd5..e334844d6e1d 100644 --- a/tools/testing/selftests/kvm/x86_64/exit_on_emulation_failure_test.c +++ b/tools/testing/selftests/kvm/x86_64/exit_on_emulation_failure_test.c @@ -26,9 +26,6 @@ int main(int argc, char *argv[]) struct kvm_vcpu *vcpu; struct kvm_vm *vm; - /* Tell stdout not to buffer its content */ - setbuf(stdout, NULL); - TEST_REQUIRE(kvm_has_cap(KVM_CAP_EXIT_ON_EMULATION_FAILURE)); vm = vm_create_with_one_vcpu(&vcpu, guest_code); -- 2.39.0

1 year, 2 months

3
4
0 0

[PATCH v4 00/21] arm64/sme: Support SME 2 and SME 2.1

by Mark Brown

Arm have recently released versions 2 and 2.1 of the SME extension. Among the features introduced by SME 2 is some new architectural state, the ZT0 register. This series adds support for this and all the other features of the new SME versions. Since the architecture has been designed with the possibility of adding further ZTn registers in mind the interfaces added for ZT0 are done with this possibility in mind. As ZT0 is a simple fixed size register these interfaces are all fairly simple, the main complication is that ZT0 is only accessible when PSTATE.ZA is enabled. The memory allocation that we already do for PSTATE.ZA is extended to include space for ZT0. Due to textual collisions especially around the addition of hwcaps this is based on the recently merged series "arm64: Support for 2022 data processing instructions" but there is no meaningful interaction. There will be collisions with "arm64/signal: Signal handling cleanups" if that is applied but again not super substantial. v4: - Rebase onto v6.2-rc3. - Add SME2 value to ID_AA64PFR1_EL1.SME and move cpufeature to key off it. - Fix cut'n'paste errors and missing capability in hwcap table. - Fix bitrot in za-test program. - Typo and cut'n'paste fixes. v3: - Rebase onto merged series for the 2022 architectur extensions. - Clarifications and typo fixes in the ABI documentation. v2: - Add missing initialisation of user->zt in signal context parsing. - Change the magic for ZT signal frames to 0x5a544e01 (ZTN0). To: Catalin Marinas <catalin.marinas(a)arm.com> To: Will Deacon <will(a)kernel.org> To: Oleg Nesterov <oleg(a)redhat.com> To: Marc Zyngier <maz(a)kernel.org> To: James Morse <james.morse(a)arm.com> To: Alexandru Elisei <alexandru.elisei(a)arm.com> To: Suzuki K Poulose <suzuki.poulose(a)arm.com> To: Oliver Upton <oliver.upton(a)linux.dev> Cc: Alan Hayward <alan.hayward(a)arm.com> Cc: Luis Machado <luis.machado(a)arm.com>, Cc: Szabolcs Nagy <szabolcs.nagy(a)arm.com> Cc: linux-arm-kernel(a)lists.infradead.org Cc: linux-kernel(a)vger.kernel.org Cc: kvmarm(a)lists.linux.dev To: Shuah Khan <shuah(a)kernel.org> Cc: linux-kselftest(a)vger.kernel.org --- Mark Brown (21): arm64/sme: Rename za_state to sme_state arm64: Document boot requirements for SME 2 arm64/sysreg: Update system registers for SME 2 and 2.1 arm64/sme: Document SME 2 and SME 2.1 ABI arm64/esr: Document ISS for ZT0 being disabled arm64/sme: Manually encode ZT0 load and store instructions arm64/sme: Enable host kernel to access ZT0 arm64/sme: Add basic enumeration for SME2 arm64/sme: Provide storage for ZT0 arm64/sme: Implement context switching for ZT0 arm64/sme: Implement signal handling for ZT arm64/sme: Implement ZT0 ptrace support arm64/sme: Add hwcaps for SME 2 and 2.1 features kselftest/arm64: Add a stress test program for ZT0 kselftest/arm64: Cover ZT in the FP stress test kselftest/arm64: Enumerate SME2 in the signal test utility code kselftest/arm64: Teach the generic signal context validation about ZT kselftest/arm64: Add test coverage for ZT register signal frames kselftest/arm64: Add SME2 coverage to syscall-abi kselftest/arm64: Add coverage of the ZT ptrace regset kselftest/arm64: Add coverage of SME 2 and 2.1 hwcaps Documentation/arm64/booting.rst | 10 + Documentation/arm64/elf_hwcaps.rst | 18 + Documentation/arm64/sme.rst | 52 ++- arch/arm64/include/asm/cpufeature.h | 6 + arch/arm64/include/asm/esr.h | 1 + arch/arm64/include/asm/fpsimd.h | 30 +- arch/arm64/include/asm/fpsimdmacros.h | 22 ++ arch/arm64/include/asm/hwcap.h | 6 + arch/arm64/include/asm/processor.h | 2 +- arch/arm64/include/uapi/asm/hwcap.h | 6 + arch/arm64/include/uapi/asm/sigcontext.h | 19 ++ arch/arm64/kernel/cpufeature.c | 28 ++ arch/arm64/kernel/cpuinfo.c | 6 + arch/arm64/kernel/entry-fpsimd.S | 30 +- arch/arm64/kernel/fpsimd.c | 47 ++- arch/arm64/kernel/hyp-stub.S | 6 + arch/arm64/kernel/idreg-override.c | 1 + arch/arm64/kernel/process.c | 21 +- arch/arm64/kernel/ptrace.c | 60 +++- arch/arm64/kernel/signal.c | 113 ++++++- arch/arm64/kvm/fpsimd.c | 2 +- arch/arm64/tools/cpucaps | 1 + arch/arm64/tools/sysreg | 27 +- include/uapi/linux/elf.h | 1 + tools/testing/selftests/arm64/abi/hwcap.c | 115 +++++++ .../testing/selftests/arm64/abi/syscall-abi-asm.S | 43 ++- tools/testing/selftests/arm64/abi/syscall-abi.c | 40 ++- tools/testing/selftests/arm64/fp/.gitignore | 2 + tools/testing/selftests/arm64/fp/Makefile | 5 + tools/testing/selftests/arm64/fp/fp-stress.c | 29 +- tools/testing/selftests/arm64/fp/sme-inst.h | 20 ++ tools/testing/selftests/arm64/fp/zt-ptrace.c | 365 +++++++++++++++++++++ tools/testing/selftests/arm64/fp/zt-test.S | 317 ++++++++++++++++++ tools/testing/selftests/arm64/signal/.gitignore | 1 + .../testing/selftests/arm64/signal/test_signals.h | 2 + .../selftests/arm64/signal/test_signals_utils.c | 3 + .../selftests/arm64/signal/testcases/testcases.c | 36 ++ .../selftests/arm64/signal/testcases/testcases.h | 1 + .../selftests/arm64/signal/testcases/zt_no_regs.c | 51 +++ .../selftests/arm64/signal/testcases/zt_regs.c | 85 +++++ 40 files changed, 1558 insertions(+), 72 deletions(-) --- base-commit: b7bfaa761d760e72a969d116517eaa12e404c262 change-id: 20221208-arm64-sme2-363c27227a32 Best regards, -- Mark Brown <broonie(a)kernel.org>

1 year, 2 months

3
25
0 0

[PATCH 19/19] selftests/vm: Add pins-cgroup selftest for mlock/mmap

by Alistair Popple

Add some basic tests of mlock/mmap cgroup accounting for pinned memory. Signed-off-by: Alistair Popple <apopple(a)nvidia.com> Cc: Shuah Khan <shuah(a)kernel.org> Cc: linux-kernel(a)vger.kernel.org Cc: linux-mm(a)kvack.org Cc: linux-kselftest(a)vger.kernel.org Cc: cgroups(a)vger.kernel.org --- MAINTAINERS | 1 +- tools/testing/selftests/vm/Makefile | 1 +- tools/testing/selftests/vm/pins-cgroup.c | 271 ++++++++++++++++++++++++- 3 files changed, 273 insertions(+) create mode 100644 tools/testing/selftests/vm/pins-cgroup.c diff --git a/MAINTAINERS b/MAINTAINERS index f8526e2..4c4eed9 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5387,6 +5387,7 @@ L: cgroups(a)vger.kernel.org L: linux-mm(a)kvack.org S: Maintained F: mm/pins_cgroup.c +F: tools/testing/selftests/vm/pins-cgroup.c CORETEMP HARDWARE MONITORING DRIVER M: Fenghua Yu <fenghua.yu(a)intel.com> diff --git a/tools/testing/selftests/vm/Makefile b/tools/testing/selftests/vm/Makefile index 89c14e4..0653720 100644 --- a/tools/testing/selftests/vm/Makefile +++ b/tools/testing/selftests/vm/Makefile @@ -56,6 +56,7 @@ TEST_GEN_PROGS += soft-dirty TEST_GEN_PROGS += split_huge_page_test TEST_GEN_FILES += ksm_tests TEST_GEN_PROGS += ksm_functional_tests +TEST_GEN_FILES += pins-cgroup ifeq ($(MACHINE),x86_64) CAN_BUILD_I386 := $(shell ./../x86/check_cc.sh "$(CC)" ../x86/trivial_32bit_program.c -m32) diff --git a/tools/testing/selftests/vm/pins-cgroup.c b/tools/testing/selftests/vm/pins-cgroup.c new file mode 100644 index 0000000..c2eabc2 --- /dev/null +++ b/tools/testing/selftests/vm/pins-cgroup.c @@ -0,0 +1,271 @@ +// SPDX-License-Identifier: GPL-2.0 +#include "../kselftest_harness.h" + +#define _GNU_SOURCE +#include <fcntl.h> +#include <assert.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/mman.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <sys/ioctl.h> +#include <sys/prctl.h> +#include <sys/resource.h> +#include <sys/capability.h> +#include <unistd.h> + +#define CGROUP_TEMP "/sys/fs/cgroup/pins_XXXXXX" +#define PINS_MAX (-1UL) + +FIXTURE(pins_cg) +{ + char *cg_path; + long page_size; +}; + +static char *cgroup_new(void) +{ + char *cg; + + cg = malloc(sizeof(CGROUP_TEMP)); + strcpy(cg, CGROUP_TEMP); + if (!mkdtemp(cg)) { + perror("Failed to create cgroup"); + return NULL; + } + + return cg; +} + +static int cgroup_add_proc(char *cg, pid_t pid) +{ + char *cg_proc; + FILE *f; + int ret = 0; + + if (asprintf(&cg_proc, "%s/cgroup.procs", cg) < 0) + return -1; + + f = fopen(cg_proc, "w"); + free(cg_proc); + if (!f) + return -1; + + if (fprintf(f, "%ld\n", (long) pid) < 0) + ret = -1; + + fclose(f); + return ret; +} + +static int cgroup_set_limit(char *cg, unsigned long limit) +{ + char *cg_pins_max; + FILE *f; + int ret = 0; + + if (asprintf(&cg_pins_max, "%s/pins.max", cg) < 0) + return -1; + + f = fopen(cg_pins_max, "w"); + free(cg_pins_max); + if (!f) + return -1; + + if (limit != PINS_MAX) { + if (fprintf(f, "%ld\n", limit) < 0) + ret = -1; + } else { + if (fprintf(f, "max\n") < 0) + ret = -1; + } + + fclose(f); + return ret; +} + +FIXTURE_SETUP(pins_cg) +{ + char *cg_subtree_control; + FILE *f; + + if (asprintf(&cg_subtree_control, + "/sys/fs/cgroup/cgroup.subtree_control") < 0) + return; + + f = fopen(cg_subtree_control, "w"); + free(cg_subtree_control); + if (!f) + return; + + fprintf(f, "+pins\n"); + fclose(f); + + self->cg_path = cgroup_new(); + self->page_size = sysconf(_SC_PAGE_SIZE); +} + +FIXTURE_TEARDOWN(pins_cg) +{ + cgroup_add_proc("/sys/fs/cgroup", getpid()); + + rmdir(self->cg_path); + free(self->cg_path); +} + +static long cgroup_pins(char *cg) +{ + long pin_count; + char *cg_pins_current; + FILE *f; + int ret; + + if (asprintf(&cg_pins_current, "%s/pins.current", cg) < 0) + return -1; + + f = fopen(cg_pins_current, "r"); + if (!f) { + printf("Can't open %s\n", cg_pins_current); + getchar(); + free(cg_pins_current); + return -2; + } + + free(cg_pins_current); + + if (fscanf(f, "%ld", &pin_count) == EOF) + ret = -3; + else + ret = pin_count; + + fclose(f); + return ret; +} + +static int set_rlim_memlock(unsigned long size) +{ + struct rlimit rlim_memlock = { + .rlim_cur = size, + .rlim_max = size, + }; + cap_t cap; + cap_value_t capability[1] = { CAP_IPC_LOCK }; + + /* + * Many of the rlimit checks are skipped if a process has + * CAP_IP_LOCK. As this test should be run as root we need to + * explicitly drop it. + */ + cap = cap_get_proc(); + if (!cap) + return -1; + if (cap_set_flag(cap, CAP_EFFECTIVE, 1, capability, CAP_CLEAR)) + return -1; + if (cap_set_proc(cap)) + return -1; + return setrlimit(RLIMIT_MEMLOCK, &rlim_memlock); +} + +TEST_F(pins_cg, basic) +{ + pid_t child_pid; + long page_size = self->page_size; + char *p; + + ASSERT_EQ(cgroup_add_proc(self->cg_path, getpid()), 0); + p = mmap(NULL, 32*page_size, PROT_READ | PROT_WRITE, + MAP_ANONYMOUS | MAP_PRIVATE, -1, 0); + ASSERT_NE(p, MAP_FAILED); + + ASSERT_EQ(cgroup_pins(self->cg_path), 0); + memset(p, 0, 16*page_size); + ASSERT_EQ(mlock(p, page_size), 0); + ASSERT_EQ(cgroup_pins(self->cg_path), 1); + ASSERT_EQ(mlock(p + page_size, page_size), 0); + ASSERT_EQ(cgroup_pins(self->cg_path), 2); + ASSERT_EQ(mlock(p, page_size), 0); + ASSERT_EQ(cgroup_pins(self->cg_path), 2); + ASSERT_EQ(mlock(p, 4*page_size), 0); + ASSERT_EQ(cgroup_pins(self->cg_path), 4); + ASSERT_EQ(munlock(p + 2*page_size, 2*page_size), 0); + ASSERT_EQ(cgroup_pins(self->cg_path), 2); + ASSERT_EQ(cgroup_set_limit(self->cg_path, 8), 0); + ASSERT_EQ(mlock(p, 16*page_size), -1); + ASSERT_EQ(errno, ENOMEM); + ASSERT_EQ(cgroup_pins(self->cg_path), 2); + ASSERT_EQ(cgroup_set_limit(self->cg_path, PINS_MAX), 0); + + /* check mremap() a locked region correctly accounts locked pages */ + ASSERT_EQ(mlock(p, 32*page_size), 0); + ASSERT_EQ(cgroup_pins(self->cg_path), 32); + p = mremap(p, 32*page_size, 64*page_size, MREMAP_MAYMOVE); + ASSERT_NE(p, MAP_FAILED); + ASSERT_EQ(cgroup_pins(self->cg_path), 64); + ASSERT_EQ(munmap(p + 32*page_size, 32*page_size), 0) + ASSERT_EQ(cgroup_pins(self->cg_path), 32); + p = mremap(p, 32*page_size, 32*page_size, MREMAP_MAYMOVE | MREMAP_DONTUNMAP); + ASSERT_NE(p, MAP_FAILED); + ASSERT_EQ(cgroup_pins(self->cg_path), 32); + ASSERT_EQ(munlock(p, 32*page_size), 0); + + /* mremap() a locked region should fail if limit exceeded */ + ASSERT_EQ(set_rlim_memlock(32*page_size), 0); + ASSERT_EQ(mlock(p, 32*page_size), 0); + ASSERT_EQ(mremap(p, 32*page_size, 64*page_size, 0), MAP_FAILED); + ASSERT_EQ(munlock(p, 32*page_size), 0); + + /* Exceeds rlimit, expected to fail */ + ASSERT_EQ(set_rlim_memlock(16*page_size), 0); + ASSERT_EQ(mlock(p, 32*page_size), -1); + ASSERT_EQ(errno, ENOMEM); + + /* memory in the child isn't locked so shouldn't increase pin_cg count */ + ASSERT_EQ(mlock(p, 16*page_size), 0); + child_pid = fork(); + if (!child_pid) { + ASSERT_EQ(cgroup_pins(self->cg_path), 16); + ASSERT_EQ(mlock(p, 16*page_size), 0); + ASSERT_EQ(cgroup_pins(self->cg_path), 32); + return; + + } + waitpid(child_pid, NULL, 0); + + /* check that child exit uncharged the pins */ + ASSERT_EQ(cgroup_pins(self->cg_path), 16); +} + +TEST_F(pins_cg, mmap) +{ + char *p; + + ASSERT_EQ(cgroup_add_proc(self->cg_path, getpid()), 0); + p = mmap(NULL, 4*self->page_size, PROT_READ | PROT_WRITE, + MAP_ANONYMOUS | MAP_PRIVATE | MAP_LOCKED, -1, 0); + ASSERT_NE(p, MAP_FAILED); + ASSERT_EQ(cgroup_pins(self->cg_path), 4); +} + +/* + * Test moving to a different cgroup. + */ +TEST_F(pins_cg, move_cg) +{ + char *p, *new_cg; + + ASSERT_EQ(cgroup_add_proc(self->cg_path, getpid()), 0); + p = mmap(NULL, 16*self->page_size, PROT_READ | PROT_WRITE, + MAP_ANONYMOUS | MAP_PRIVATE, -1, 0); + ASSERT_NE(p, MAP_FAILED); + memset(p, 0, 16*self->page_size); + ASSERT_EQ(mlock(p, 16*self->page_size), 0); + ASSERT_EQ(cgroup_pins(self->cg_path), 16); + ASSERT_NE(new_cg = cgroup_new(), NULL); + ASSERT_EQ(cgroup_add_proc(new_cg, getpid()), 0); + ASSERT_EQ(cgroup_pins(new_cg), 16); + ASSERT_EQ(cgroup_add_proc(self->cg_path, getpid()), 0); + rmdir(new_cg); +} +TEST_HARNESS_MAIN -- git-series 0.9.1

1 year, 2 months

1
0
0 0

[PATCH 01/19] mm: Introduce vm_account

by Alistair Popple

Kernel drivers that pin pages should account these pages against either user->locked_vm and/or mm->pinned_vm and fail the pinning if RLIMIT_MEMLOCK is exceeded and CAP_IPC_LOCK isn't held. Currently drivers open-code this accounting and use various methods to update the atomic variables and check against the limits leading to various bugs and inconsistencies. To fix this introduce a standard interface for charging pinned and locked memory. As this involves taking references on kernel objects such as mm_struct or user_struct we introduce a new vm_account struct to hold these references. Several helper functions are then introduced to grab references and check limits. As the way these limits are charged and enforced is visible to userspace we need to be careful not to break existing applications by charging to different counters. As a result the vm_account functions support accounting to different counters as required. A future change will extend this to also account against a cgroup for pinned pages. Signed-off-by: Alistair Popple <apopple(a)nvidia.com> Cc: linux-kernel(a)vger.kernel.org Cc: linuxppc-dev(a)lists.ozlabs.org Cc: linux-fpga(a)vger.kernel.org Cc: linux-rdma(a)vger.kernel.org Cc: virtualization(a)lists.linux-foundation.org Cc: kvm(a)vger.kernel.org Cc: netdev(a)vger.kernel.org Cc: cgroups(a)vger.kernel.org Cc: io-uring(a)vger.kernel.org Cc: linux-mm(a)kvack.org Cc: bpf(a)vger.kernel.org Cc: rds-devel(a)oss.oracle.com Cc: linux-kselftest(a)vger.kernel.org --- include/linux/vm_account.h | 56 +++++++++++++++++- mm/util.c | 127 ++++++++++++++++++++++++++++++++++++++- 2 files changed, 183 insertions(+) create mode 100644 include/linux/vm_account.h diff --git a/include/linux/vm_account.h b/include/linux/vm_account.h new file mode 100644 index 0000000..b4b2e90 --- /dev/null +++ b/include/linux/vm_account.h @@ -0,0 +1,56 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _LINUX_VM_ACCOUNT_H +#define _LINUX_VM_ACCOUNT_H + +/** + * enum vm_account_flags - Determine how pinned/locked memory is accounted. + * @VM_ACCOUNT_TASK: Account pinned memory to mm->pinned_vm. + * @VM_ACCOUNT_BYPASS: Don't enforce rlimit on any charges. + * @VM_ACCOUNT_USER: Account locked memory to user->locked_vm. + * + * Determines which statistic pinned/locked memory is accounted + * against. All limits will be enforced against RLIMIT_MEMLOCK and the + * pins cgroup if CONFIG_CGROUP_PINS is enabled. + * + * New drivers should use VM_ACCOUNT_USER. VM_ACCOUNT_TASK is used by + * pre-existing drivers to maintain existing accounting against + * mm->pinned_mm rather than user->locked_mm. + * + * VM_ACCOUNT_BYPASS may also be specified to bypass rlimit + * checks. Typically this is used to cache CAP_IPC_LOCK from when a + * driver is first initialised. Note that this does not bypass cgroup + * limit checks. + */ +enum vm_account_flags { + VM_ACCOUNT_USER = 0, + VM_ACCOUNT_BYPASS = 1, + VM_ACCOUNT_TASK = 1, +}; + +struct vm_account { + struct task_struct *task; + struct mm_struct *mm; + struct user_struct *user; + enum vm_account_flags flags; +}; + +void vm_account_init(struct vm_account *vm_account, struct task_struct *task, + struct user_struct *user, enum vm_account_flags flags); + +/** + * vm_account_init_current - Initialise a new struct vm_account. + * @vm_account: pointer to uninitialised vm_account. + * + * Helper to initialise a vm_account for the common case of charging + * with VM_ACCOUNT_TASK against current. + */ +static inline void vm_account_init_current(struct vm_account *vm_account) +{ + vm_account_init(vm_account, current, NULL, VM_ACCOUNT_TASK); +} + +void vm_account_release(struct vm_account *vm_account); +int vm_account_pinned(struct vm_account *vm_account, unsigned long npages); +void vm_unaccount_pinned(struct vm_account *vm_account, unsigned long npages); + +#endif /* _LINUX_VM_ACCOUNT_H */ diff --git a/mm/util.c b/mm/util.c index b56c92f..d8c19f8 100644 --- a/mm/util.c +++ b/mm/util.c @@ -23,6 +23,7 @@ #include <linux/processor.h> #include <linux/sizes.h> #include <linux/compat.h> +#include <linux/vm_account.h> #include <linux/uaccess.h> @@ -431,6 +432,132 @@ void arch_pick_mmap_layout(struct mm_struct *mm, struct rlimit *rlim_stack) #endif /** + * vm_account_init - Initialise a new struct vm_account. + * @vm_account: pointer to uninitialised vm_account. + * @task: task to charge against. + * @user: user to charge against. Must be non-NULL for VM_ACCOUNT_USER. + * @flags: flags to use when charging to vm_account. + * + * Initialise a new uninitialised struct vm_account. Takes references + * on the task/mm/user/cgroup as required although callers must ensure + * any references passed in remain valid for the duration of this + * call. + */ +void vm_account_init(struct vm_account *vm_account, struct task_struct *task, + struct user_struct *user, enum vm_account_flags flags) +{ + vm_account->task = get_task_struct(task); + + if (flags & VM_ACCOUNT_USER) + vm_account->user = get_uid(user); + + mmgrab(task->mm); + vm_account->mm = task->mm; + vm_account->flags = flags; +} +EXPORT_SYMBOL_GPL(vm_account_init); + +/** + * vm_account_release - Initialise a new struct vm_account. + * @vm_account: pointer to initialised vm_account. + * + * Drop any object references obtained by vm_account_init(). The + * vm_account must not be used after calling this unless reinitialised + * with vm_account_init(). + */ +void vm_account_release(struct vm_account *vm_account) +{ + put_task_struct(vm_account->task); + if (vm_account->flags & VM_ACCOUNT_USER) + free_uid(vm_account->user); + + mmdrop(vm_account->mm); +} +EXPORT_SYMBOL_GPL(vm_account_release); + +/* + * Charge pages with an atomic compare and swap. Returns -ENOMEM on + * failure, 1 on success and 0 for retry. + */ +static int vm_account_cmpxchg(struct vm_account *vm_account, + unsigned long npages, unsigned long lock_limit) +{ + u64 cur_pages, new_pages; + + if (vm_account->flags & VM_ACCOUNT_USER) + cur_pages = atomic_long_read(&vm_account->user->locked_vm); + else + cur_pages = atomic64_read(&vm_account->mm->pinned_vm); + + new_pages = cur_pages + npages; + if (lock_limit != RLIM_INFINITY && new_pages > lock_limit) + return -ENOMEM; + + if (vm_account->flags & VM_ACCOUNT_USER) { + return atomic_long_cmpxchg(&vm_account->user->locked_vm, + cur_pages, new_pages) == cur_pages; + } else { + return atomic64_cmpxchg(&vm_account->mm->pinned_vm, + cur_pages, new_pages) == cur_pages; + } +} + +/** + * vm_account_pinned - Charge pinned or locked memory to the vm_account. + * @vm_account: pointer to an initialised vm_account. + * @npages: number of pages to charge. + * + * Return: 0 on success, -ENOMEM if a limit would be exceeded. + * + * Note: All pages must be explicitly uncharged with + * vm_unaccount_pinned() prior to releasing the vm_account with + * vm_account_release(). + */ +int vm_account_pinned(struct vm_account *vm_account, unsigned long npages) +{ + unsigned long lock_limit = RLIM_INFINITY; + int ret; + + if (!(vm_account->flags & VM_ACCOUNT_BYPASS) && !capable(CAP_IPC_LOCK)) + lock_limit = task_rlimit(vm_account->task, + RLIMIT_MEMLOCK) >> PAGE_SHIFT; + + while (true) { + ret = vm_account_cmpxchg(vm_account, npages, lock_limit); + if (ret > 0) + break; + else if (ret < 0) + return ret; + } + + /* + * Always add pinned pages to mm->pinned_vm even when we're + * not enforcing the limit against that. + */ + if (vm_account->flags & VM_ACCOUNT_USER) + atomic64_add(npages, &vm_account->mm->pinned_vm); + + return 0; +} +EXPORT_SYMBOL_GPL(vm_account_pinned); + +/** + * vm_unaccount_pinned - Uncharge pinned or locked memory to the vm_account. + * @vm_account: pointer to an initialised vm_account. + * @npages: number of pages to uncharge. + */ +void vm_unaccount_pinned(struct vm_account *vm_account, unsigned long npages) +{ + if (vm_account->flags & VM_ACCOUNT_USER) { + atomic_long_sub(npages, &vm_account->user->locked_vm); + atomic64_sub(npages, &vm_account->mm->pinned_vm); + } else { + atomic64_sub(npages, &vm_account->mm->pinned_vm); + } +} +EXPORT_SYMBOL_GPL(vm_unaccount_pinned); + +/** * __account_locked_vm - account locked pages to an mm's locked_vm * @mm: mm to account against * @pages: number of pages to account -- git-series 0.9.1

1 year, 2 months

1
0
0 0

2024

2023

2022

2021

2020

2019

2018

2017

Linux-kselftest-mirror February 2023