[Webkit-unassigned] [Bug 146440] Crash on xLarge memory allocation using bmalloc on 32bit systems
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Wed Jul 1 18:45:19 PDT 2015
https://bugs.webkit.org/show_bug.cgi?id=146440
--- Comment #4 from Carlos Alberto Lopez Perez <clopez at igalia.com> ---
(In reply to comment #3)
> (In reply to comment #2)
> > [...]
> > Also, should disabling this optimization could make sense as a reasonable
> > workaround for 2.8.3 (similar to what it's done in bug 127777 with
> > -fno-omit-frame-pointer and -fno-tree-dce), would it be ok to propose a
> > patch for the CMake files for WebKitGTK+? (Adding Martin to CC)
>
> To be more precise, I was thinking perhaps of something like this:
>
> diff --git a/Source/cmake/OptionsCommon.cmake
> b/Source/cmake/OptionsCommon.cmake
> index 6691526..355d475 100644
> --- a/Source/cmake/OptionsCommon.cmake
> +++ b/Source/cmake/OptionsCommon.cmake
> @@ -99,6 +99,12 @@ endif ()
>
> string(TOLOWER ${CMAKE_HOST_SYSTEM_PROCESSOR}
> LOWERCASE_CMAKE_HOST_SYSTEM_PROCESSOR)
> if ("${CMAKE_CXX_COMPILER_ID}" STREQUAL "GNU" AND
> "${LOWERCASE_CMAKE_HOST_SYSTEM_PROCESSOR}" MATCHES "(i[3-6]86|x86)")
> + # The -ftree-sra optimization (implicit with -O2) causes crashes when
> + # allocating large chunks of memory using bmalloc on Intel 32bit.
> + # See https://bugs.webkit.org/show_bug.cgi?id=146440
> + set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -fno-tree-sra")
> + set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fno-tree-sra")
> +
> # To avoid out of memory when building with debug option in 32bit
> system.
> # See https://bugs.webkit.org/show_bug.cgi?id=77327
> set(CMAKE_SHARED_LINKER_FLAGS_DEBUG "-Wl,--no-keep-memory
> ${CMAKE_SHARED_LINKER_FLAGS_DEBUG}")
What about setting this flag only for the bmalloc source files at Source/bmalloc/CMakeLists.txt . Would it be enough?
--
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20150702/b7d69ade/attachment.html>
More information about the webkit-unassigned
mailing list