[llvm-branch-commits] [NFC][msan] Extract `handleSelectLikeInst` (PR #94881)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/94881 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][msan] Prepare function to extract main logic (PR #94880)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/94880 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [nfc][tsan] Clang format includes (PR #96599)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/96599 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][hwasan] Use %p to print addresses (PR #98576)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/98576 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][sanitizer_common] Use %p to print addresses (PR #98578)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/98578 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [tsan] Replace ALIGNED with alignas (PR #98959)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/98959 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][hwasan] Remove duplicated checks (PR #100678)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/100678 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][ubsan] Add mixed filter test case (PR #100679)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/100679 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [InstCombine][asan] Don't speculate loads before `select ptr` (PR #100773)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/100773 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [docs] Fix code-block formating (PR #100772)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/100772 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][asan] Report DynInitUnPoison (PR #101586)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/101586 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [ModuleUtils] Add updateGlobalCtors/updateGlobalDtors (PR #101757)
@@ -44,6 +45,14 @@ void appendToGlobalCtors(Module &M, Function *F, int Priority, void appendToGlobalDtors(Module &M, Function *F, int Priority, Constant *Data = nullptr); +/// Apply 'Fn' to the list of global ctors of module M and replace contructor +/// record with the one returned by `Fn`. If `nullptr` was returned, the +/// corresponding constructor will be removed from the array. For details see +/// https://llvm.org/docs/LangRef.html#the-llvm-global-ctors-global-variable +using GlobalCtorUpdateFn = llvm::function_ref; +void updateGlobalCtors(Module &M, const GlobalCtorUpdateFn &Fn); thurstond wrote: 'updateGlobalCtors' is a very generic name - is there a name that could be descriptive? (e.g., it's somewhat similar to 'map' in Haskell/Python or std::transform.) https://github.com/llvm/llvm-project/pull/101757 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [ModuleUtils] Add updateGlobalCtors/updateGlobalDtors (PR #101757)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/101757 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][asan] Track current dynamic init module (PR #101597)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/101597 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [asan] Reduce priority of "contiguous_container:" VPrintf (PR #104402)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/104402 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][sanitizer] Use `UNLIKELY` in VReport/VPrintf (PR #104403)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/104403 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [asan] Remove debug tracing from `report_globals` (PR #104404)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/104404 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [asan] Remove debug tracing from `report_globals` (PR #104404)
@@ -4,7 +4,7 @@ // RUN: %clangxx_asan -g -O0 -DSHARED_LIB -DSIZE=1 %s -fPIC -shared -o %t-so-1.so // RUN: %clangxx_asan -g -O0 -DSHARED_LIB -DSIZE=2 %s -fPIC -shared -o %t-so-2.so // RUN: %clangxx_asan -g -O0 %s %libdl -Wl,--export-dynamic -o %t -// RUN: %env_asan_opts=report_globals=2:detect_odr_violation=1 %run %t 2>&1 | FileCheck %s +// RUN: %env_asan_opts=report_globals=1:detect_odr_violation=1:verbosity=4 %run %t 2>&1 | FileCheck %s thurstond wrote: Is there a particular reason why this uses verbosity=4 instead of 3? (it's the same output since verbosity >= 3, but a bit perplexing since nothing is gated on verbosity >= 4) https://github.com/llvm/llvm-project/pull/104404 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][asan] Better `___asan_gen_` names (PR #104728)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/104728 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][asan] Create `ModuleName` lazily (PR #104729)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/104729 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][asan] Don't `cd` after `split-file` (PR #104727)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/104727 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] Reland "[asan] Catch `initialization-order-fiasco` in modules without…" (PR #104730)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/104730 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] release/18.x: [msan] Re-exec with no ASLR if memory layout is incompatible on Linux (#85142) (PR #86201)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/86201 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [tsan] Set `thr->is_inited` after SlotAttachAndLock (PR #86342)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/86342 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [tsan] Process SIGPROF as sync signal only if thread is alive (PR #86343)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/86343 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [HWASAN] Don't instrument loads from global if globals are not tagged (PR #86774)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/86774 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] release/18.x: [msan] Re-exec with no ASLR if memory layout is incompatible on Linux (#85142) (PR #86201)
thurstond wrote: > Hi @thurstond (or anyone else). If you would like to add a note about this > fix in the release notes (completely optional). Please reply to this comment > with a one or two sentence description of the fix. "MemorySanitizer on Linux can now run even when maximum-entropy address-space layout randomization is configured globally (as is becoming increasingly common); MSan can automatically re-exec ASLR-off for the process if needed, instead of aborting." Thanks @tstellar! https://github.com/llvm/llvm-project/pull/86201 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [clang] [clang][builtin] Implement __builtin_allow_runtime_check (PR #87568)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/87568 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [llvm] [NFC][msan] Remove unused parameter from getOriginPtrForVAArgument (PR #72687)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/72687 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [llvm] [compiler-rt] [msan][x86] Fix shadow if vararg overflow beyond kParamTLSSize (PR #72707)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/72707 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] [test][sanitizer] Allow fork_threaded test on Msan, Tsan (PR #75260)
@@ -1,6 +1,6 @@ // RUN: %clangxx -O0 %s -o %t && %env_tool_opts=die_after_fork=0 %run %t -// UNSUPPORTED: * +// UNSUPPORTED: asan, lsan, hwasan thurstond wrote: This would also enable the test for ubsan, which is not mentioned in the title https://github.com/llvm/llvm-project/pull/75260 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] [test][sanitizer] Allow fork_threaded test on Msan, Tsan (PR #75260)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/75260 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] [test][sanitizer] Allow fork_threaded test on Msan, Tsan (PR #75260)
https://github.com/thurstond edited https://github.com/llvm/llvm-project/pull/75260 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [clang-tools-extra] [lld] [libcxx] [compiler-rt] [llvm] [test][hwasan] Implement sanitizer_specific for HWASAN (PR #75280)
@@ -31,6 +31,19 @@ static void make_mem_good(void *p, size_t s) {
static void make_mem_bad(void *p, size_t s) {
__asan_poison_memory_region(p, s);
}
+#elif __has_feature(hwaddress_sanitizer)
+# include
+# include
+static void check_mem_is_good(void *p, size_t s) {
+ if (__hwasan_test_shadow(p, s) != -1)
+abort();
+}
+static void make_mem_good(void *p, size_t s) {
+ __hwasan_tag_memory(p, __hwasan_get_tag_from_pointer(p), s);
+}
+static void make_mem_bad(void *p, size_t s) {
+ __hwasan_tag_memory(p, ~__hwasan_get_tag_from_pointer(p), s);
thurstond wrote:
Is this guaranteed to make memory bad if it's a short granule?
Also, in the rare case where the tag was 255, bitwise NOT results in the
confusing tag value of zero.
https://github.com/llvm/llvm-project/pull/75280
___
llvm-branch-commits mailing list
llvm-branch-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [clang-tools-extra] [lld] [libcxx] [compiler-rt] [llvm] [lsan] Install `pthread_atfork` (PR #75281)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/75281 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] [asan] Install `pthread_atfork` (PR #75290)
https://github.com/thurstond edited https://github.com/llvm/llvm-project/pull/75290 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] [asan] Install `pthread_atfork` (PR #75290)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/75290 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] [asan] Install `pthread_atfork` (PR #75290)
@@ -148,6 +148,30 @@ void PlatformTSDDtor(void *tsd) {
}
#endif
+void InstallAtForkHandler() {
+ auto before = []() {
+if (CAN_SANITIZE_LEAKS) {
+ __lsan::LockGlobal();
+}
+// `_lsan` functions defined regardless of `CAN_SANITIZE_LEAKS` and do the
+// job.
thurstond wrote:
Not sure what "and do the job" means.
https://github.com/llvm/llvm-project/pull/75290
___
llvm-branch-commits mailing list
llvm-branch-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [lld] [clang] [compiler-rt] [clang-tools-extra] [llvm] [libcxx] [test][hwasan] Implement sanitizer_specific for HWASAN (PR #75280)
@@ -31,6 +31,30 @@ static void make_mem_good(void *p, size_t s) {
static void make_mem_bad(void *p, size_t s) {
__asan_poison_memory_region(p, s);
}
+#elif __has_feature(hwaddress_sanitizer)
+# include
+# include
+static void check_mem_is_good(void *p, size_t s) {
+ if (__hwasan_test_shadow(p, s) != -1)
+abort();
+}
+static void make_mem_good(void *p, size_t s) {
+ __hwasan_tag_memory(p, __hwasan_get_tag_from_pointer(p), s);
+}
+static void make_mem_bad(void *p, size_t s) {
+ uint8_t tag = ~__hwasan_get_tag_from_pointer(p);
+ if (!tag) {
+// Nothing wrong with tag zero, but non-zero tags help to detect never
+// tagged memory.
+tag = 1;
+ }
+ __hwasan_tag_memory(p, , s);
+ // With missaligned `p` or short granules we can't guaranlee tag mismatch.
thurstond wrote:
Nits: "misaligned", "guaranlee"
https://github.com/llvm/llvm-project/pull/75280
___
llvm-branch-commits mailing list
llvm-branch-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [lld] [libcxx] [llvm] [compiler-rt] [clang-tools-extra] [clang] [test][hwasan] Implement sanitizer_specific for HWASAN (PR #75280)
@@ -42,7 +42,18 @@ static void make_mem_good(void *p, size_t s) {
__hwasan_tag_memory(p, __hwasan_get_tag_from_pointer(p), s);
}
static void make_mem_bad(void *p, size_t s) {
- __hwasan_tag_memory(p, ~__hwasan_get_tag_from_pointer(p), s);
+ uint8_t tag = ~__hwasan_get_tag_from_pointer(p);
+ if (!tag) {
+// Nothing wrong with tag zero, but non-zero tags help to detect never
+// tagged memory.
+tag = 1;
+ }
+ __hwasan_tag_memory(p, , s);
thurstond wrote:
Is this missing the tag parameter?
https://github.com/llvm/llvm-project/pull/75280
___
llvm-branch-commits mailing list
llvm-branch-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] [clang] [hwasan] Respect strip_path_prefix printing locals (PR #76132)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/76132 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] [NFC][tsan] `ptr != 0` to implicit check (PR #77144)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/77144 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] [tsan] Fallback to top frame (PR #77145)
thurstond wrote: > Probably it's not important, as it should not happen, but if all frames are > internal we should pick top, not the bottom frame. If we encounter a case that "should not happen", wouldn't it be better to fail/assert? https://github.com/llvm/llvm-project/pull/77145 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [libc] [llvm] [libcxx] [libunwind] [clang] [compiler-rt] [tsan] Fallback to top frame (PR #77145)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/77145 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [sanitizer] Allow to override GetDTLSRange (PR #108348)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/108348 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [sanitizer] Test for #108348 (PR #108349)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/108349 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][sanitizer] Switch to `gnu_get_libc_version` (PR #108724)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/108724 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [sanitizer] Add CHECK that static TLS info is ready (PR #108684)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/108684 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][sanitizer] Remove `else if` from ThreadDescriptorSizeFallback (PR #108909)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/108909 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][sanitizer] Move ThreadDescriptorSize into GLIBC/FREEBSD block (PR #108913)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/108913 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][sanitizer] Move `InitTlsSize` into `InitializePlatformEarly` (PR #108921)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/108921 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][sanitizer] Move InitTlsSize (PR #108922)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/108922 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [sanitizer] Switch from lazy `ThreadDescriptorSize` (PR #108923)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/108923 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] release/19.x: Reapply "[sanitizer_common] AND signals in BlockSignals instead of deleting (#113443)" for non-Android Linux only (#115790) (PR #116670)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/116670 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] release/19.x: Reapply "[sanitizer_common] AND signals in BlockSignals instead of deleting (#113443)" for non-Android Linux only (#115790) (PR #116670)
thurstond wrote: > @thurstond (or anyone else). If you would like to add a note about this fix > in the release notes (completely optional). Please reply to this comment with > a one or two sentence description of the fix. When you are done, please add > the release:note label to this PR. Fixes a regression whereby the sanitizers may incorrectly unblock signals that had been blocked by the user. https://github.com/llvm/llvm-project/pull/116670 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [llvm] [NFC][ModuleUtils] Test appendToGlobalCtors/appendToGlobalDtors (PR #101753)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/101753 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [clang] [ubsan] Remove -fsanitizer=vptr from -fsanitizer=undefined (PR #121115)
https://github.com/thurstond approved this pull request. LGTM if no objections pop up on discourse in the next few days https://github.com/llvm/llvm-project/pull/121115 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [compiler-rt] [ubsan] Runtime and driver support for local-bounds (PR #120515)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/120515 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][BoundsChecking] Add TrapBB local variable (PR #119983)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/119983 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [ubsan] Add runtime test for -fsanitize=local-bounds (PR #120038)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/120038 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [ubsan] Pass fsanitize-skip-hot-cutoff into -fsanitize=bounds (PR #122576)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/122576 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFCI][BoundsChecking] Apply nosanitize on local-bounds instrumentation (PR #122416)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/122416 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [nfc][ubsan] Add local-bounds test (PR #122415)
https://github.com/thurstond commented: Would it be useful to keep -fsanitize=...null and add local-bounds instead of replacing null with local-bounds? It would make the test more rigorous and also make the diff smaller. https://github.com/llvm/llvm-project/pull/122415 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [nfc][BoundsChecking] Rename BoundsCheckingOptions into Options (PR #122359)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/122359 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [nfc][ubsan] Add local-bounds test (PR #122415)
https://github.com/thurstond edited https://github.com/llvm/llvm-project/pull/122415 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [nfc][ubsan] Add local-bounds test (PR #122415)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/122415 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [ubsan] Pass fsanitize-skip-hot-cutoff into -fsanitize=bounds (PR #122576)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/122576 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [NFC][CFI] Avoid failing CFI tests (PR #135981)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/135981 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [llvm] [MSAN] handle assorted AVX permutations (PR #143462)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/143462 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [llvm] [MSAN] handle AVX vpermi2var (PR #143463)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/143463 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [llvm] [MSAN] handle AVX vpermi2var (PR #143463)
@@ -4191,6 +4191,15 @@ struct MemorySanitizerVisitor : public
InstVisitor {
// We apply the same permutation (argument index 1) to the shadows.
void handleAVXVpermil2var(IntrinsicInst &I) {
assert(I.arg_size() == 3);
+assert(isa(I.getArgOperand(0)->getType()));
+assert(isa(I.getArgOperand(1)->getType()));
+assert(isa(I.getArgOperand(2)->getType()));
+[[maybe_unused]] auto ArgVectorSize =
+cast(I.getArgOperand(0)->getType())->getNumElements();
+assert(cast(I.getArgOperand(1)->getType())
+ ->getNumElements() == ArgVectorSize);
+assert(cast(I.getArgOperand(2)->getType())
+ ->getNumElements() == ArgVectorSize);
thurstond wrote:
Some of the assertions are redundant (e.g., if operand 0 is a vector, and
operand 0's type is the same as operand 2's type, then operand 2 must be a
vector with the same number of elements as operand 0), but that's fine.
https://github.com/llvm/llvm-project/pull/143463
___
llvm-branch-commits mailing list
llvm-branch-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [MSAN] handle AVX vpermi2var (PR #143463)
https://github.com/thurstond edited https://github.com/llvm/llvm-project/pull/143463 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [MSAN] handle AVX vpermi2var (PR #143463)
https://github.com/thurstond requested changes to this pull request. Please add assertions and more tests (comments inline). https://github.com/llvm/llvm-project/pull/143463 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [MSAN] handle AVX vpermi2var (PR #143463)
@@ -5130,6 +5149,26 @@ struct MemorySanitizerVisitor : public
InstVisitor {
handleAVXVpermilvar(I);
break;
}
+case Intrinsic::x86_avx512_vpermi2var_d_128:
+case Intrinsic::x86_avx512_vpermi2var_d_256:
+case Intrinsic::x86_avx512_vpermi2var_d_512:
+case Intrinsic::x86_avx512_vpermi2var_hi_128:
+case Intrinsic::x86_avx512_vpermi2var_hi_256:
+case Intrinsic::x86_avx512_vpermi2var_hi_512:
+case Intrinsic::x86_avx512_vpermi2var_pd_128:
+case Intrinsic::x86_avx512_vpermi2var_pd_256:
+case Intrinsic::x86_avx512_vpermi2var_pd_512:
+case Intrinsic::x86_avx512_vpermi2var_ps_128:
+case Intrinsic::x86_avx512_vpermi2var_ps_256:
+case Intrinsic::x86_avx512_vpermi2var_ps_512:
+case Intrinsic::x86_avx512_vpermi2var_q_128:
+case Intrinsic::x86_avx512_vpermi2var_q_256:
+case Intrinsic::x86_avx512_vpermi2var_q_512:
+case Intrinsic::x86_avx512_vpermi2var_qi_128:
+case Intrinsic::x86_avx512_vpermi2var_qi_256:
+case Intrinsic::x86_avx512_vpermi2var_qi_512:
+ handleAVXVpermil2var(I);
thurstond wrote:
Please precommit tests to cover all these instructions.
(AFAICS the tests only cover llvm.x86.avx512.vpermi2var.d.512,
llvm.x86.avx512.vpermi2var.pd.512, llvm.x86.avx512.vpermi2var.ps.512,
llvm.x86.avx512.vpermi2var.q.512.)
https://github.com/llvm/llvm-project/pull/143463
___
llvm-branch-commits mailing list
llvm-branch-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [MSAN] handle AVX vpermi2var (PR #143463)
@@ -4175,6 +4175,25 @@ struct MemorySanitizerVisitor : public
InstVisitor {
setShadow(&I, IRB.CreateBitCast(CI, getShadowTy(&I)));
setOriginForNaryOp(I);
}
+ // Instrument AVX permutation intrinsic.
+ // We apply the same permutation (argument index 1) to the shadows.
+ void handleAVXVpermil2var(IntrinsicInst &I) {
+IRBuilder<> IRB(&I);
thurstond wrote:
Please add some assertions that check whether the instruction is suitable for
this handler:
- number of operands
- type and size of each operand
- relation of input and output operands
https://github.com/llvm/llvm-project/pull/143463
___
llvm-branch-commits mailing list
llvm-branch-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [MSAN] handle assorted AVX permutations (PR #143462)
https://github.com/thurstond requested changes to this pull request. Please add assertions and more tests (comments inline). https://github.com/llvm/llvm-project/pull/143462 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [MSAN] handle assorted AVX permutations (PR #143462)
https://github.com/thurstond edited https://github.com/llvm/llvm-project/pull/143462 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [MSAN] handle assorted AVX permutations (PR #143462)
@@ -4161,7 +4161,7 @@ struct MemorySanitizerVisitor : public
InstVisitor {
// Instrument AVX permutation intrinsic.
// We apply the same permutation (argument index 1) to the shadow.
- void handleAVXVpermilvar(IntrinsicInst &I) {
+ void handleAVXPermutation(IntrinsicInst &I) {
thurstond wrote:
Since this function is being used for many more cases, please augment it with
assertions that check whether the instruction is suitable for this handler:
- number of operands
- type and size of each operand
- relation of input and output operands
https://github.com/llvm/llvm-project/pull/143462
___
llvm-branch-commits mailing list
llvm-branch-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [MSAN] handle assorted AVX permutations (PR #143462)
@@ -5139,14 +5139,30 @@ struct MemorySanitizerVisitor : public
InstVisitor {
assert(Success);
break;
}
-
+case Intrinsic::x86_avx2_permd:
+case Intrinsic::x86_avx2_permps:
+case Intrinsic::x86_ssse3_pshuf_b_128:
+case Intrinsic::x86_avx2_pshuf_b:
+case Intrinsic::x86_avx512_pshuf_b_512:
+case Intrinsic::x86_avx512_permvar_df_256:
+case Intrinsic::x86_avx512_permvar_df_512:
+case Intrinsic::x86_avx512_permvar_di_256:
+case Intrinsic::x86_avx512_permvar_di_512:
+case Intrinsic::x86_avx512_permvar_hi_128:
+case Intrinsic::x86_avx512_permvar_hi_256:
+case Intrinsic::x86_avx512_permvar_hi_512:
+case Intrinsic::x86_avx512_permvar_qi_128:
+case Intrinsic::x86_avx512_permvar_qi_256:
+case Intrinsic::x86_avx512_permvar_qi_512:
+case Intrinsic::x86_avx512_permvar_sf_512:
+case Intrinsic::x86_avx512_permvar_si_512:
thurstond wrote:
Please precommit tests to cover all these instructions.
(AFAICS the tests only cover llvm.x86.avx2.permd, llvm.x86.avx2.permps,
llvm.x86.avx2.pshuf.b, llvm.x86.avx512.permvar.df.512,
llvm.x86.avx512.permvar.di.512, llvm.x86.avx512.permvar.sf.512,
llvm.x86.avx512.permvar.si.512.)
https://github.com/llvm/llvm-project/pull/143462
___
llvm-branch-commits mailing list
llvm-branch-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [clang] [clang] Add flag fallow-runtime-check-skip-hot-cutoff (PR #145999)
@@ -2668,6 +2668,16 @@ def fsanitize_skip_hot_cutoff_EQ } // end -f[no-]sanitize* flags +def fallow_runtime_check_skip_hot_cutoff_EQ +: Joined<["-"], "fallow-runtime-check-skip-hot-cutoff=">, + Group, + Visibility<[ClangOption, CC1Option]>, + HelpText<"Exclude __allow_runtime_check for the top hottest code " + "responsible for " + "the given fraction of PGO counters " + "(0.0 [default] = skip none; 1/.0 = skip all). " thurstond wrote: Typo: "1/.0" https://github.com/llvm/llvm-project/pull/145999 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [clang] [clang] Add flag fallow-runtime-check-skip-hot-cutoff (PR #145999)
https://github.com/thurstond commented: W https://github.com/llvm/llvm-project/pull/145999 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [clang] [clang] Add flag fallow-runtime-check-skip-hot-cutoff (PR #145999)
https://github.com/thurstond edited https://github.com/llvm/llvm-project/pull/145999 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [clang] [clang] Add flag fallow-runtime-check-skip-hot-cutoff (PR #145999)
@@ -2668,6 +2668,16 @@ def fsanitize_skip_hot_cutoff_EQ } // end -f[no-]sanitize* flags +def fallow_runtime_check_skip_hot_cutoff_EQ thurstond wrote: "fallow" is either unfortunate or brilliant naming, similar to fun safe math. (I'm looking forward to the "fallow_fields" flag.) https://github.com/llvm/llvm-project/pull/145999 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [clang] [clang] Add flag fallow-runtime-check-skip-hot-cutoff (PR #145999)
https://github.com/thurstond edited https://github.com/llvm/llvm-project/pull/145999 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [clang] [clang] Add flag fallow-runtime-check-skip-hot-cutoff (PR #145999)
@@ -2668,6 +2668,16 @@ def fsanitize_skip_hot_cutoff_EQ } // end -f[no-]sanitize* flags +def fallow_runtime_check_skip_hot_cutoff_EQ +: Joined<["-"], "fallow-runtime-check-skip-hot-cutoff=">, thurstond wrote: Would it be more user-friendly (and possibly less code duplication) to add "allow_runtime_check" as one of the options for the existing `-fsanitize-skip-hot-cutoff` (possibly renaming the flag since it would no longer be entirely sanitizers)? e.g., allow `-fsanitize-skip-hot-cutoff=undefined=0.99,allow_runtime_check=1.0` https://github.com/llvm/llvm-project/pull/145999 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [LSan] skip leaks from dlerror (PR #142876)
thurstond wrote: There's a flag: ``` LSAN_FLAG(bool, use_ld_allocations, true, "Root set: mark as reachable all allocations made from dynamic " "linker. This was the old way to handle dynamic TLS, and will " "be removed soon. Do not use this flag.") ``` Obviously we shouldn't use this flag, but I wonder if something like that would be better than specifically suppressing `*dlerror*` (are other `dl*` functions going to break at some point?). https://github.com/llvm/llvm-project/pull/142876 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [LSan] skip leaks from dlerror (PR #142876)
https://github.com/thurstond approved this pull request. https://github.com/llvm/llvm-project/pull/142876 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [LSan] skip leaks from dlerror (PR #142876)
thurstond wrote: > Yeah, it would be better to fix the root cause. But until then, this > exclusion makes sense IMO. WDYT? SGTM. https://github.com/llvm/llvm-project/pull/142876 ___ llvm-branch-commits mailing list llvm-branch-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
[llvm-branch-commits] [llvm] [MSAN] handle AVX vpermi2var (PR #143463)
@@ -4187,6 +4187,29 @@ struct MemorySanitizerVisitor : public
InstVisitor {
setShadow(&I, IRB.CreateBitCast(CI, getShadowTy(&I)));
setOriginForNaryOp(I);
}
+ // Instrument AVX permutation intrinsic.
+ // We apply the same permutation (argument index 1) to the shadows.
+ void handleAVXVpermil2var(IntrinsicInst &I) {
+assert(I.arg_size() == 3);
+assert(I.getArgOperand(0)->getType() == I.getArgOperand(2)->getType());
+assert(I.getType() == I.getArgOperand(0)->getType());
thurstond wrote:
Please also check that I.getType() is a vector type, and that
`I.getArgOperand(0)` has the same number of elements as `I.getArgOperand(1)`
https://github.com/llvm/llvm-project/pull/143463
___
llvm-branch-commits mailing list
llvm-branch-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
