Changes in lsan_common.cpp

b15cf06f - [lsan][nfc] Add missing newline to VReport

Fri, 15 Jul 2022 19:44:23 +0000

[lsan][nfc] Add missing newline to VReport List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

e06a81d8 - [LSAN] Fix up LSAN weak symbols for Windows

Tue, 31 May 2022 14:51:13 +0000

[LSAN] Fix up LSAN weak symbols for WindowsDifferential Revision: https://reviews.llvm.org/D126703 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

8246b2e1 - [Sanitizers][Darwin] Replace SANITIZER_MAC with SANITIZER_APPLE in source files

Mon, 23 May 2022 21:35:42 +0000

[Sanitizers][Darwin] Replace SANITIZER_MAC with SANITIZER_APPLE in source filesThis is a follow up to [Sanitizers][Darwin] Rename Apple macro SANITIZER_MAC -> SANITIZER_APPLE (D125816)Performed a global search/replace as in title against LLVM sourcesDifferential Revision: https://reviews.llvm.org/D126263 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

c8644ea8 - [compiler-rt][lsan] Update CanBeAHeapPointer for AArch64

Tue, 24 May 2022 17:22:46 +0000

[compiler-rt][lsan] Update CanBeAHeapPointer for AArch64While attempting to get the 64-bit lsan allocator working for Fuchsia, Inoticed this function would incorrectly return false for pointers returnedby the 64-bit allocator. On AArch64, this function attempts to get the VMAsize dynamically by counting the number of leading zeros from the functionframe address. This will fail if the frame address is significantly below anallocated pointer (that is, the frame address has more leading zeros than anallocated pointer). This is possible on Fuchsia and linux (when not calledfrom the initial thread stack).It seems the intended use of this function is to speed up pointer scanning byfiltering out addresses that user code might not be able to access. Otherplatforms this check is done on seem to hardcode the VMA size/shift, so itseems appropriate to do this for aarch64 as well. This implies pointers onaarch64 where the VMA size is <64 will pass through, but bad pointers willstill be caught by subsequent scan checks.This patch also renames the function to something more fitting of what it'strying to do.Differential Revision: https://reviews.llvm.org/D123814 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

a63932a8 - [lsan] Allow suppression of "unknown module"

Tue, 15 Mar 2022 07:30:14 +0000

[lsan] Allow suppression of "unknown module"If sanitizer cannot determine name of the module itwill use "". Then it can be suppressedif needed.Reviewed By: kdaDifferential Revision: https://reviews.llvm.org/D121674 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

f86deb18 - [lsan] Move out suppression of invalid PCs from StopTheWorld

Thu, 09 Dec 2021 19:21:16 +0000

[lsan] Move out suppression of invalid PCs from StopTheWorldThis removes the last use of StackDepot from StopTheWorld.Depends on D115284.Reviewed By: morehouseDifferential Revision: https://reviews.llvm.org/D115319 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

f72e5094 - [lsan] Reduce StopTheWorld access to StackDepot

Wed, 08 Dec 2021 05:40:20 +0000

[lsan] Reduce StopTheWorld access to StackDepotStackDepot locks some stuff. As is there is small probability todeadlock if we stop thread which locked the Depot.We need either Lock/Unlock StackDepot for StopTheWorld, or don'tinteract with StackDepot from there.This patch does not run LeakReport under StopTheWorld. LeakReportcontains most of StackDepot access.As a bonus this patch will help to resolve kMaxLeaksConsidered FIXME.Depends on D114498.Reviewed By: morehouse, kstoimenovDifferential Revision: https://reviews.llvm.org/D115284 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

a9a14990 - [NFC][lsan] Change LeakSuppressionContext interface

Wed, 08 Dec 2021 05:37:02 +0000

[NFC][lsan] Change LeakSuppressionContext interfaceReviewed By: morehouseDifferential Revision: https://reviews.llvm.org/D115318 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

b1c369e0 - [lsan] Replace atomic_store/atomic_load with atomic_fetch_add

Wed, 08 Dec 2021 06:39:12 +0000

[lsan] Replace atomic_store/atomic_load with atomic_fetch_add List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

dd1abb11 - [NFC][lsan] Move method implementation up

Wed, 08 Dec 2021 05:34:57 +0000

[NFC][lsan] Move method implementation up List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

b79ea567 - [NFC][lsan] Clang-format lsan_common.cpp

Wed, 08 Dec 2021 05:34:06 +0000

[NFC][lsan] Clang-format lsan_common.cpp List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

8ed4d281 - [NFC][lsan] Use kIgnored in MarkInvalidPCCb

Wed, 08 Dec 2021 03:49:44 +0000

[NFC][lsan] Use kIgnored in MarkInvalidPCCbkIgnored didn't exist when the code was added, but it should beequivalent to kReachable.The goal is to refactor MarkInvalidPCCb to avoid StackDepotGetin StopTheWorld. List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

64d44208 - [NFC][lsan] Simplify root_regions initialization

Fri, 12 Nov 2021 07:32:58 +0000

[NFC][lsan] Simplify root_regions initialization List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

629b40da - Fix -Wformat warnings reported by GCC.

Wed, 03 Nov 2021 12:49:21 +0000

Fix -Wformat warnings reported by GCC.Differential Revision: https://reviews.llvm.org/D113099 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

ca0036df - [sanitizer] Remove StackDepotReverseMap

Tue, 12 Oct 2021 05:18:00 +0000

[sanitizer] Remove StackDepotReverseMapNow StackDepotGet can retrive the stack in O(1).Depends on D111612.Reviewed By: dvyukovDifferential Revision: https://reviews.llvm.org/D111613 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

051d766b - [NFC][lsan] Expand use StackDepotReverseMap

Fri, 01 Oct 2021 02:24:53 +0000

[NFC][lsan] Expand use StackDepotReverseMapBefore StackDepotReverseMap was used only by ProcessPC. List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

56debbf5 - sanitizers: switch BlockingMutex(LINKER_INITIALIZED) to Mutex

Wed, 28 Jul 2021 11:12:24 +0000

sanitizers: switch BlockingMutex(LINKER_INITIALIZED) to MutexMutex does not support LINKER_INITIALIZED support.As preparation to switching BlockingMutex to Mutex,proactively replace all BlockingMutex(LINKER_INITIALIZED) to Mutex.All of these are objects with static storage duration and Mutex ctoris constexpr, so it should be equivalent.Reviewed By: melverDifferential Revision: https://reviews.llvm.org/D106944 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

e0dadf3d - [sanitizer] Remove max_len parameter from InternalScopedString

Tue, 16 Mar 2021 23:33:04 +0000

[sanitizer] Remove max_len parameter from InternalScopedStringInternalScopedString uses InternalMmapVector internallyso it can be resized dynamically as needed.Reviewed By: eugenisDifferential Revision: https://reviews.llvm.org/D98751 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

dd922bc2 - [LSan] Introduce a callback mechanism to allow adding data reachable from ThreadContexts to the frontier.

Thu, 21 Jan 2021 23:48:41 +0000

[LSan] Introduce a callback mechanism to allow adding data reachable from ThreadContexts to the frontier.This mechanism is intended to provide a way to treat the `arg` pointerof a created (but not yet started) thread as reachable. In futurepatches this will be implemented in `GetAdditionalThreadContextPtrs`.A separate implementation of `GetAdditionalThreadContextPtrs` existsfor ASan and LSan runtimes because they need to be implementeddifferently in future patches.rdar://problem/63537240Differential Revision: https://reviews.llvm.org/D95183 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp

9a023701 - [lsan] Ignore inderect leaks referenced by suppressed blocks

Tue, 29 Dec 2020 03:19:38 +0000

[lsan] Ignore inderect leaks referenced by suppressed blocksThis makes suppression list to work similar to __lsan_ignore_object.Existing behavior was inconsistent and very inconvenient for complexdata structures.Example:struct B;struct A { B* ptr; };A* t = makeA();t->ptr = makeB();Before the patch: if makeA suppressed by suppression file, lsan willstill report the makeB() leak, so we need two suppressions.After the patch: a single makeA suppression is enough (the same as asingle __lsan_ignore_object(t)).Differential Revision: https://reviews.llvm.org/D93884 List of files: /llvm-project-15.0.7/compiler-rt/lib/lsan/lsan_common.cpp