alekseyshl (Aleksey Shlyapnikov)
User

Projects

User does not belong to any projects.

User Details

User Since
Oct 31 2016, 4:57 PM (58 w, 1 d)

Recent Activity

Today

alekseyshl accepted D41121: [sanitizer] Introduce a vDSO aware timing function.
Tue, Dec 12, 2:06 PM
alekseyshl added a comment to D41128: [scudo] Adding a public Scudo interface.

To address your security concern, maybe have a define to compile Scudo without public API?

Tue, Dec 12, 2:02 PM
alekseyshl added inline comments to D40903: [Sanitizers] Basic Solaris sanitizer support (PR 33274).
Tue, Dec 12, 1:55 PM · Restricted Project
alekseyshl accepted D40899: [Sanitizers, CMake] Basic sanitizer Solaris support (PR 33274).
Tue, Dec 12, 1:39 PM · Restricted Project
alekseyshl accepted D40898: [Sanitizers] Basic sanitizer Solaris support (PR 33274).
Tue, Dec 12, 11:45 AM · Restricted Project
alekseyshl added inline comments to D41121: [sanitizer] Introduce a vDSO aware timing function.
Tue, Dec 12, 11:32 AM

Yesterday

alekseyshl added a comment to D40898: [Sanitizers] Basic sanitizer Solaris support (PR 33274).

Mostly looks good, just a few minor changes.

Mon, Dec 11, 6:11 PM · Restricted Project
alekseyshl accepted D41088: [sanitizer] Revert rL320409.
Mon, Dec 11, 1:01 PM

Fri, Dec 8

alekseyshl accepted D40935: Hardware-assisted AddressSanitizer (compiler-rt).
Fri, Dec 8, 3:45 PM

Thu, Dec 7

alekseyshl accepted D40932: Hardware-assisted AddressSanitizer (llvm part)..
Thu, Dec 7, 4:22 PM
alekseyshl added inline comments to D40951: [ASan] Add interceptor for printf_chk.
Thu, Dec 7, 12:59 PM · Restricted Project
alekseyshl accepted D40936: Hardware-assisted AddressSanitizer (clang part)..
Thu, Dec 7, 11:16 AM

Wed, Dec 6

alekseyshl committed rCRT319929: [TSan] Make more TSan interceptors symbolizer-aware..
[TSan] Make more TSan interceptors symbolizer-aware.
Wed, Dec 6, 11:13 AM
alekseyshl committed rL319929: [TSan] Make more TSan interceptors symbolizer-aware..
[TSan] Make more TSan interceptors symbolizer-aware.
Wed, Dec 6, 11:13 AM
alekseyshl closed D40877: [TSan] Make more TSan interceptors symbolizer-aware. by committing rL319929: [TSan] Make more TSan interceptors symbolizer-aware..
Wed, Dec 6, 11:13 AM

Tue, Dec 5

alekseyshl created D40877: [TSan] Make more TSan interceptors symbolizer-aware..
Tue, Dec 5, 7:12 PM
alekseyshl accepted D40862: [scudo] Correct performance regression in Secondary.
Tue, Dec 5, 3:26 PM
alekseyshl removed a reviewer for D38971: Enhance libsanitizer support for invalid-pointer-pair.: alekseyshl.
Tue, Dec 5, 12:00 PM
alekseyshl accepted D40679: [sanitizer] Introduce a vDSO aware time function, and use it in the allocator [redo].
Tue, Dec 5, 11:59 AM

Mon, Dec 4

alekseyshl accepted D40796: [scudo] Get rid of the thread local PRNG & header salt.
Mon, Dec 4, 6:08 PM
alekseyshl committed rL319681: [ASan] Disable invalid-pointer-pairs-threads.cc on Darwin..
[ASan] Disable invalid-pointer-pairs-threads.cc on Darwin.
Mon, Dec 4, 11:41 AM
alekseyshl committed rCRT319681: [ASan] Disable invalid-pointer-pairs-threads.cc on Darwin..
[ASan] Disable invalid-pointer-pairs-threads.cc on Darwin.
Mon, Dec 4, 11:41 AM
alekseyshl accepted D40754: [sanitizer] 64-bit allocator's PopulateFreeArray partial refactor.

Great, thanks!

Mon, Dec 4, 10:45 AM
alekseyshl committed rCRT319671: [ASan] Fix test/asan/TestCases/Posix/invalid-pointer-pairs-threads.cc.
[ASan] Fix test/asan/TestCases/Posix/invalid-pointer-pairs-threads.cc
Mon, Dec 4, 10:36 AM
alekseyshl committed rL319671: [ASan] Fix test/asan/TestCases/Posix/invalid-pointer-pairs-threads.cc.
[ASan] Fix test/asan/TestCases/Posix/invalid-pointer-pairs-threads.cc
Mon, Dec 4, 10:35 AM
alekseyshl added a comment to D38971: Enhance libsanitizer support for invalid-pointer-pair..

D40600 is committed, I guess, we can abandon this one now.

Mon, Dec 4, 10:03 AM
alekseyshl committed rCRT319668: [ASan] Enhance libsanitizer support for invalid-pointer-pair..
[ASan] Enhance libsanitizer support for invalid-pointer-pair.
Mon, Dec 4, 10:01 AM
alekseyshl committed rL319668: [ASan] Enhance libsanitizer support for invalid-pointer-pair..
[ASan] Enhance libsanitizer support for invalid-pointer-pair.
Mon, Dec 4, 10:01 AM
alekseyshl closed D40600: Enhance libsanitizer support for invalid-pointer-pair. by committing rL319668: [ASan] Enhance libsanitizer support for invalid-pointer-pair..
Mon, Dec 4, 10:01 AM · Restricted Project
alekseyshl accepted D40754: [sanitizer] 64-bit allocator's PopulateFreeArray partial refactor.
Mon, Dec 4, 9:18 AM

Fri, Dec 1

alekseyshl added inline comments to D40754: [sanitizer] 64-bit allocator's PopulateFreeArray partial refactor.
Fri, Dec 1, 3:33 PM
alekseyshl added inline comments to D40754: [sanitizer] 64-bit allocator's PopulateFreeArray partial refactor.
Fri, Dec 1, 2:56 PM
alekseyshl added inline comments to D40679: [sanitizer] Introduce a vDSO aware time function, and use it in the allocator [redo].
Fri, Dec 1, 11:26 AM
alekseyshl added inline comments to D40679: [sanitizer] Introduce a vDSO aware time function, and use it in the allocator [redo].
Fri, Dec 1, 10:24 AM

Thu, Nov 30

alekseyshl added a comment to D40665: [sanitizer] Implement NanoTime() on Darwin.

FYI, D40679 adds MonotonicNanoTime() for that NanoTime() is too expensive for the allocator's purposes.

Thu, Nov 30, 3:40 PM · Restricted Project
alekseyshl added a comment to D40349: [LSan] New experimental flag for background leak checking before exit..

Can it be implemented with timer_create() instead of the thread?

Thu, Nov 30, 3:18 PM
alekseyshl accepted D40627: [asan] Use linker initialization for the allocator.

Ah, sorry. linker_initialized should of course be "true".

Thu, Nov 30, 12:29 PM · Restricted Project
alekseyshl added inline comments to D40657: [sanitizer] Introduce a vDSO aware time function, and use it in the allocator.
Thu, Nov 30, 11:41 AM
alekseyshl added inline comments to D40349: [LSan] New experimental flag for background leak checking before exit..
Thu, Nov 30, 11:33 AM
alekseyshl added inline comments to D40627: [asan] Use linker initialization for the allocator.
Thu, Nov 30, 11:08 AM · Restricted Project

Wed, Nov 29

alekseyshl added a comment to D40600: Enhance libsanitizer support for invalid-pointer-pair..

I'll do it

Wed, Nov 29, 4:40 PM · Restricted Project
alekseyshl added inline comments to D40627: [asan] Use linker initialization for the allocator.
Wed, Nov 29, 4:35 PM · Restricted Project
alekseyshl accepted D40600: Enhance libsanitizer support for invalid-pointer-pair..
Wed, Nov 29, 2:36 PM · Restricted Project
alekseyshl accepted D40521: [scudo] Allow for compile-time choice of the SizeClassMap.
Wed, Nov 29, 11:39 AM

Tue, Nov 28

alekseyshl committed rL319237: [LSan] Fix one source of stale segments in the process memory mapping..
[LSan] Fix one source of stale segments in the process memory mapping.
Tue, Nov 28, 2:15 PM
alekseyshl closed D40529: [LSan] Fix one source of stale segments in the process memory mapping. by committing rL319237: [LSan] Fix one source of stale segments in the process memory mapping..
Tue, Nov 28, 2:15 PM
alekseyshl added a comment to D40521: [scudo] Allow for compile-time choice of the SizeClassMap.

I guess another question that I had for you @alekseyshl : would you rather see #if or #ifdef?
I am not sure SCUDO_SCM_DEFAULT=1 makes sense as opposed to it just being defined.

Tue, Nov 28, 1:32 PM

Mon, Nov 27

alekseyshl accepted D40521: [scudo] Allow for compile-time choice of the SizeClassMap.

What's wrong with SCUDO_SCM_{DEFAULT|COMPACT|VERY_COMPACT}? If someone wants to tinker with it, they probably aware what SCM stands for and what it affects.

Mon, Nov 27, 4:19 PM
alekseyshl created D40529: [LSan] Fix one source of stale segments in the process memory mapping..
Mon, Nov 27, 4:00 PM
alekseyshl accepted D40504: [scudo] Workaround for uninitialized Bionic globals.
Mon, Nov 27, 11:22 AM

Tue, Nov 21

alekseyshl added inline comments to D38595: [fuchsia] Update Fuchsia with a new mmap implementation..
Tue, Nov 21, 3:16 PM
alekseyshl accepted D40322: [scudo] Overhaul hardware CRC32 feature detection.
Tue, Nov 21, 2:43 PM
alekseyshl added inline comments to D38971: Enhance libsanitizer support for invalid-pointer-pair..
Tue, Nov 21, 2:02 PM

Mon, Nov 20

alekseyshl added a comment to D38971: Enhance libsanitizer support for invalid-pointer-pair..

Very sorry for the delay!

Mon, Nov 20, 5:15 PM
alekseyshl accepted D40260: [sanitizer] Define SANITIZER_USE_GETAUXVAL for Android.
Mon, Nov 20, 3:34 PM
alekseyshl accepted D39473: [sanitizers] Add init function to set alignment of low level allocator.
Mon, Nov 20, 3:32 PM
alekseyshl added inline comments to D39473: [sanitizers] Add init function to set alignment of low level allocator.
Mon, Nov 20, 10:53 AM

Fri, Nov 17

alekseyshl added a comment to D39473: [sanitizers] Add init function to set alignment of low level allocator.

Why do we need this at all for internal allocator? It's internal to sanitizer and sanitizer code is not instrumented, that means no shadow.

Here is a typical backtrace. It seems that asan installs a callback to the internal allocator that poisons those memory.

#0 0xf7fd7c89 in kernel_vsyscall ()
#1 0xf7e3c627 in syscall () from /lib/i386-linux-gnu/libc.so.6
#2 0x0811c3cc in
sanitizer::internal_sched_yield() () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_linux.cc:401
#3 0x08062c75 in LockSlow () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/asan/../sanitizer_common/sanitizer_mutex.h:55
#4 0x08129696 in Lock () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_mutex.h:32
#5 GenericScopedLock () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_mutex.h:187
#6 GetOrInit () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_symbolizer_libcdep.cc:21
#7 0x08128147 in Print () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_stacktrace_libcdep.cc:35
#8 0x0810f31f in AsanCheckFailed () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/asan/asan_rtl.cc:69
#9 0x08124884 in CheckFailed () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_termination.cc:79
#10 0x08108b26 in PoisonShadow () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/asan/asan_poisoning.cc:38
#11 0x0810f355 in __asan::OnLowLevelAllocate(unsigned long, unsigned long) () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/asan/asan_rtl.cc:89
#12 0x0811328b in Allocate () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_allocator.cc:193
#13 0x0812b49a in operator new () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_common.h:937
#14 ChooseExternalSymbolizer () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_symbolizer_posix_libcdep.cc:486
#15 ChooseSymbolizerTools () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_symbolizer_posix_libcdep.cc:516
#16 PlatformInit () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_symbolizer_posix_libcdep.cc:529
#17 0x081296aa in GetOrInit () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_symbolizer_libcdep.cc:24
#18 0x0812b635 in LateInitialize () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/sanitizer_common/sanitizer_symbolizer_posix_libcdep.cc:534
#19 0x0810ee62 in AsanInitInternal () at /usr/local/google/home/waltl/Work/llvm-scale5/scale5/compiler-rt/lib/asan/asan_rtl.cc:479
#20 0xf7fe8de4 in ?? () from /lib/ld-linux.so.2
#21 0xf7fd9a5f in ?? () from /lib/ld-linux.so.2

Fri, Nov 17, 1:11 PM

Thu, Nov 16

alekseyshl added a comment to D39473: [sanitizers] Add init function to set alignment of low level allocator.

Why do we need this at all for internal allocator? It's internal to sanitizer and sanitizer code is not instrumented, that means no shadow.

Thu, Nov 16, 3:26 PM

Wed, Nov 15

alekseyshl committed rL318349: [LSan] Export __lsan_init.
[LSan] Export __lsan_init
Wed, Nov 15, 3:43 PM
alekseyshl closed D39889: Export __lsan_init by committing rL318349: [LSan] Export __lsan_init.
Wed, Nov 15, 3:43 PM

Tue, Nov 14

alekseyshl accepted D40038: [scudo] Soft and hard RSS limit checks.

Ok, add a TODO with the required improvements and let's get it in.

Tue, Nov 14, 3:11 PM
alekseyshl added inline comments to D40038: [scudo] Soft and hard RSS limit checks.
Tue, Nov 14, 2:15 PM
alekseyshl added inline comments to D40038: [scudo] Soft and hard RSS limit checks.
Tue, Nov 14, 1:35 PM
alekseyshl added a comment to D39970: AMDGPU: Fix producing saveexec when the copy is spilled.

At least this bot is unhappy about this revision: http://lab.llvm.org:8011/builders/sanitizer-x86_64-linux-bootstrap/builds/3020/steps/check-llvm%20check-clang%20stage3%2Fasan/logs/stdio, please fix.

Tue, Nov 14, 7:51 AM

Mon, Nov 13

alekseyshl accepted D39974: [scudo] Simplify initialization and flags.
Mon, Nov 13, 5:38 PM

Nov 10 2017

alekseyshl accepted D39905: [scudo] Bump the Android API level requirement to 21 for getauxval.
Nov 10 2017, 10:00 AM
alekseyshl accepted D39889: Export __lsan_init.
Nov 10 2017, 9:59 AM

Nov 7 2017

alekseyshl accepted D39753: [scudo] Support syslog logging on Android and in dieWithMessage.
Nov 7 2017, 12:50 PM

Nov 6 2017

alekseyshl committed rL317512: [LSan] Detect dynamic loader by its base address..
[LSan] Detect dynamic loader by its base address.
Nov 6 2017, 1:27 PM
alekseyshl closed D39275: [LSan] Detect dynamic loader by its base address. by committing rL317512: [LSan] Detect dynamic loader by its base address..
Nov 6 2017, 1:27 PM
alekseyshl committed rL317494: [Sanitizers] Check pthread_setcancel{state|type} interceptor arguments for !=….
[Sanitizers] Check pthread_setcancel{state|type} interceptor arguments for !=…
Nov 6 2017, 9:44 AM
alekseyshl closed D39626: [Sanitizers] Check pthread_setcancel{state|type} interceptor arguments for != nullptr. by committing rL317494: [Sanitizers] Check pthread_setcancel{state|type} interceptor arguments for !=….
Nov 6 2017, 9:44 AM

Nov 3 2017

alekseyshl accepted D38592: Update sanitizer_allocator to use new API..

The root bug is that, if you have decorate_proc_maps enabled, *and* actually use a named mapping, *and* are using a sanitizer intercepting open (just TSAN and ESAN), then we'll get an explosion during initialization.

I've updated this CL with what seems like a reasonable fix to me: change ReservedAddressRange::Init to not use name on POSIX, since that's where the problem happens (Fuchsia is fine with names).

I can file a bug for allowing named mappings can be used everywhere more easily (the fix may involve e.g. porting shm_open s.t. it's not invoking open, which is what's causing the problem), if that's desired?

Please let me know if this minor change is satisfactory so I can go ahead with it, thanks~

Nov 3 2017, 5:07 PM
alekseyshl created D39626: [Sanitizers] Check pthread_setcancel{state|type} interceptor arguments for != nullptr..
Nov 3 2017, 4:52 PM
alekseyshl committed rL317386: [Sanitizers] Call NanoTime() conditionally..
[Sanitizers] Call NanoTime() conditionally.
Nov 3 2017, 4:31 PM
alekseyshl closed D39624: [Sanitizers] Call NanoTime() conditionally. by committing rL317386: [Sanitizers] Call NanoTime() conditionally..
Nov 3 2017, 4:31 PM
alekseyshl accepted D39623: [scudo] Rearrange #include order.
Nov 3 2017, 4:27 PM
alekseyshl created D39624: [Sanitizers] Call NanoTime() conditionally..
Nov 3 2017, 4:21 PM

Nov 1 2017

alekseyshl accepted D39334: [Driver] Add Scudo as a possible -fsanitize= option.
Nov 1 2017, 11:16 AM
alekseyshl accepted D39497: [scudo] Fix standlone build -lrt requirement.
Nov 1 2017, 9:55 AM

Oct 31 2017

alekseyshl accepted D38592: Update sanitizer_allocator to use new API..
Oct 31 2017, 3:12 PM
alekseyshl accepted D39461: [scudo] Implement stricter separation of C vs C++.
Oct 31 2017, 1:46 PM
alekseyshl accepted D39461: [scudo] Implement stricter separation of C vs C++.
Oct 31 2017, 10:56 AM

Oct 30 2017

alekseyshl accepted D39072: Introduce ReservedAddressRange to sanitizer_common..
Oct 30 2017, 10:46 AM

Oct 27 2017

alekseyshl removed a reviewer for D33716: Provide path to Solaris.h include (Bug 28370): alekseyshl.
Oct 27 2017, 11:06 AM
alekseyshl accepted D39334: [Driver] Add Scudo as a possible -fsanitize= option.
Oct 27 2017, 10:50 AM
alekseyshl accepted D39338: [scudo] Allow to specify the maximum number of TSDs at compile time.
Oct 27 2017, 10:37 AM

Oct 26 2017

alekseyshl committed rL316720: [LSan] Disable a couple of failing tests on PPC64 (pending investigation)..
[LSan] Disable a couple of failing tests on PPC64 (pending investigation).
Oct 26 2017, 5:30 PM
alekseyshl added inline comments to D39338: [scudo] Allow to specify the maximum number of TSDs at compile time.
Oct 26 2017, 3:35 PM
alekseyshl committed rL316698: [LSan] Enable LSan tests on PPC64 Linux..
[LSan] Enable LSan tests on PPC64 Linux.
Oct 26 2017, 1:59 PM
alekseyshl closed D39316: [LSan] Enable LSan tests on PPC64 Linux. by committing rL316698: [LSan] Enable LSan tests on PPC64 Linux..
Oct 26 2017, 1:59 PM
alekseyshl added inline comments to D39334: [Driver] Add Scudo as a possible -fsanitize= option.
Oct 26 2017, 1:08 PM
alekseyshl added inline comments to D39338: [scudo] Allow to specify the maximum number of TSDs at compile time.
Oct 26 2017, 11:55 AM
alekseyshl added inline comments to D39334: [Driver] Add Scudo as a possible -fsanitize= option.
Oct 26 2017, 11:22 AM
alekseyshl accepted D38592: Update sanitizer_allocator to use new API..
Oct 26 2017, 11:02 AM
alekseyshl committed rL316683: [Sanitizers] Set default allocator_release_to_os_interval_ms to 5 seconds.
[Sanitizers] Set default allocator_release_to_os_interval_ms to 5 seconds
Oct 26 2017, 10:59 AM
alekseyshl closed D39318: [Sanitizers] Set default allocator_release_to_os_interval_ms to 5 seconds by committing rL316683: [Sanitizers] Set default allocator_release_to_os_interval_ms to 5 seconds.
Oct 26 2017, 10:59 AM
alekseyshl updated the diff for D39318: [Sanitizers] Set default allocator_release_to_os_interval_ms to 5 seconds.
  • Comment edit.
Oct 26 2017, 10:58 AM

Oct 25 2017

alekseyshl created D39318: [Sanitizers] Set default allocator_release_to_os_interval_ms to 5 seconds.
Oct 25 2017, 9:48 PM