similar to: [LLVMdev] [PATCH] Removing -fsanitize-address-zero-base-shadow

I can. I've removed every other compilation flags from clang and even used GCC, with the exact same behaviour. cheers, --renato On 29 July 2014 15:15, Evgeniy Stepanov <eugeni.stepanov at gmail.com> wrote: > OK, we can switch to SIGHUP. Could you please verify that this SIGUSR1 > behavior is not caused by MSan? > > On Tue, Jul 29, 2014 at 6:09 PM, Renato Golin

Yes, it would be great to get that fixed. On Wed, Feb 5, 2014 at 4:09 PM, Evgeniy Stepanov <eugeni.stepanov at gmail.com>wrote: > On Thu, Feb 6, 2014 at 12:21 AM, Keno Fischer > <kfischer at college.harvard.edu> wrote: > > Looks like when you materialize the stores, you should check the size of > the > > the store and emit an appropriate amount of stores to the

Yup, using SIGHUP works. On 29 July 2014 13:14, Evgeniy Stepanov <eugeni.stepanov at gmail.com> wrote: > Could it be that I'm misunderstanding signal semantics, and SIGUSR1 is > not guaranteed to be delivered (to the same process!) before kill() > returns? Could you check if that's what happens by adding a sleep() > somewhere? > > On Tue, Jul 29, 2014 at 2:17 AM,

Hi Evgeniy, Yes, it is. The problem here is that the program doesn't fail on my box (release and debug builds), so I have no idea how to debug the problem. According to the test, it's ran as "not chained_origin_with_signals.cc.tmp", expecting it to fail. Also, the FileCheck expects to find warnings on the output, for which there is none Maybe I'm missing some CMake flag or

On 29 July 2014 15:02, Evgeniy Stepanov <eugeni.stepanov at gmail.com> wrote: > You mean replacing SIGUSR1 with SIGHUP in the test case? Weird, I > don't see how they are different. So, AFAIK, they should be identical. But I put some printfs and sleeps around and it wasn't a synchronization issue. My man page says that SIGUSR1 should terminate if there isn't a handler for

Looks like when you materialize the stores, you should check the size of the the store and emit an appropriate amount of stores to the origin shadow (or just a memset intrinsic?). On Wed, Feb 5, 2014 at 2:13 PM, Keno Fischer <kfischer at college.harvard.edu>wrote: > The @entry stuff is just a gdb artifact. I've been tracking this back a > little further, and it seems there's

The code for ccall looks right. Sounds like you have a very small range of instructions where an uninitialized value appear. You could try debugging at asm level. Shadow for b should be passed at offset 0 in __msan_param_tls. MSan could propagate shadow through arithmetic and even some logic operations (like select). It could be that b is clean on function entry, but then something uninitialized

On Mon, May 13, 2013 at 11:03 AM, İsmail Dönmez <ismail at donmez.ws> wrote: > Hi, > > > On Mon, May 13, 2013 at 10:49 AM, Evgeniy Stepanov > <eugeni.stepanov at gmail.com> wrote: >> >> A recent change added defined(__linux__) condition to the code below. >> Now it says that on linux with --std=c++0x (or --std=c++11) the system >> stdlib.h header

It does sound like Android is better suited for "honest" cross-compilation, rather than "build compiler-rt for all targets we can find" model. I'm still not convinced that we must require the "ninja install" step. Could we just "ninja clang" and then build the second stage against the first stage build directory? Will this "find_package" thing

How is ccall() implemented? If it manually sets up a stack frame, then it also needs to store argument shadow values in paramtls. I don't think there is an overflow, unless you have a _lot_ of arguments in a function call. On Sun, Feb 2, 2014 at 9:26 AM, Keno Fischer <kfischer at college.harvard.edu> wrote: > Also, I was looking at the instrumented LLVM code and I noticed that the

This buildbot seems to have been failing continuously for a couple of weeks now ( http://lab.llvm.org:8011/builders/llvm-mips-linux/builds/14658 ) - is anyone watching it/caring about it? ---------- Forwarded message ---------- From: <llvm.buildmaster at lab.llvm.org> Date: Wed, Sep 30, 2015 at 11:34 PM Subject: buildbot failure in LLVM on llvm-mips-linux To: Ahmed Bougacha

Thanks! I'll try that. In order to avoid starting a new thread, let me ask you the next question. One of the shared libraries I load calls strtol and msan fails to intercept it. Why would this be? The library seems to be otherwise implemented. One of the potential culprits I saw is that strtol is marked as strong in libc. Is there any workaround? Keno On Thu, Apr 9, 2015 at 3:00 PM, Evgeniy

Thanks, it works. 2.15 has quick_exit and at_quick_exit. The attached patch also works. On Mon, May 13, 2013 at 11:22 AM, İsmail Dönmez <ismail at donmez.ws> wrote: > > On Mon, May 13, 2013 at 11:16 AM, Kostya Serebryany <kcc at google.com> wrote: >> >> __GLIBC_PREREQ(2, 17) > > > Attached patch should work. Please test. > > Regards. >

On 27 November 2013 08:43, Evgeniy Stepanov <eugeni.stepanov at gmail.com>wrote: > Also note that tail merging of calls happens in CodeGen, not in > SimplifyCFG. > Hi Evgenly, What we need is the general information that we want to avoid too much code motion, from choosing the passes, to choosing steps on the passes, to lowering code differently. On the front-end layer, It's

Does it build with libstdc++? I've got this with fresh clang, -std=c++11: In file included from ../projects/lld/lib/ReaderWriter/ELF/./SectionChunks.h:19: In file included from ../projects/lld/include/lld/Core/Parallel.h:28: In file included from /usr/lib/gcc/x86_64-linux-gnu/4.6/../../../../include/c++/4.6/condition_variable:38:

Hi Greg, On Wed, Apr 2, 2014 at 2:50 AM, Greg Fitzgerald <garious at gmail.com> wrote: > Alexey, Evgeniy, > > I propose the following steps to unify multi-arch support in compiler-rt: > > 1) The compiler-rt test suite adds "-L${CMAKE_CURRENT_BINARY_DIR}/lib" > to its 'clang' variables. This way we can test the sanitizers without > installing any libs

Could we move this setting to function attributes? We already have OptimizeForSize / MinSize there, but not the other opt levels. We also have OptimizeNone, which seems to be completely unused. This would let us support __attribute__((optimize())) in the future, which is currently ignored. Another example would be an LTO link of objects compiled with different optimization settings. I'm not

I have verified that both TLS implementations indeed find the same area of memory. Anything else I could look for? On Tue, Jan 28, 2014 at 4:28 PM, Keno Fischer <kfischer at college.harvard.edu>wrote: > Yes, both JIT code and the native runtime are instrumented. I am under the > impressions that the the C library should guarantee that from the way the > relocations are

I assume there are transitions between JITted code and native helper functions. How are you handling them? Are native functions MSan-instrumented? MSan is passing shadow across function calls in TLS slots. Does your TLS implementation guarantee that accesses to __msan_param_tls from JITted and from native code map to the same memory? On Mon, Jan 27, 2014 at 11:36 PM, Evgeniy Stepanov

> Note that ASan tests on Android require llvm-symbolizer binary. That's a really good point. And I see that llvm-symbolizer can't just be pulled into compiler-rt because it has dependencies on DebugInfo, Object, and Support libraries. This throws a big wrench in Alexey's plan to have the native compiler-rt build generate the cross-compiled binaries for all supported targets. We