summaryrefslogtreecommitdiffstats
path: root/tools/clang
Commit message (Collapse)AuthorAgeFilesLines
* clang/blink_gc_plugin: Replace 'WebCore' with 'blink'mkwst@chromium.org2014-07-1773-96/+96
| | | | | | | | TBR=abarth@chromium.org Review URL: https://codereview.chromium.org/401613002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@283795 0039d316-1c4b-4281-b951-d872f2087c98
* Roll ASan/Win to 213197 to pick up LLVM r213151timurrrr@chromium.org2014-07-171-1/+1
| | | | | | | | | | | This should fix ASan CHECK failures when a report happens after Chrome used DbgHelp once BUG=394220,345874 TBR=rnk@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/393323005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@283718 0039d316-1c4b-4281-b951-d872f2087c98
* Roll ASan/Win to r212815timurrrr@chromium.org2014-07-111-1/+1
| | | | | | | | | | | | This should enable NULL deref and wild memory access detection! TBR=rnk@chromium.org BUG=345874 NOTRY=true Review URL: https://codereview.chromium.org/387013002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@282632 0039d316-1c4b-4281-b951-d872f2087c98
* Revert of Revert of Turn on clang by default on linux, 3rd try. ↵yutak@chromium.org2014-07-101-7/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (https://codereview.chromium.org/378363006/) Reason for revert: This change should not have been reverted, since it was easy to just disable these tests. thakis explicitly requested "please ping thakis before reverting it". Original issue's description: > Revert of Turn on clang by default on linux, 3rd try. (https://codereview.chromium.org/379413002/) > > Reason for revert: > NaClBrowserTestNonSfiMode.Messaging and NaClBrowserTestNonSfiMode.Irt failed. > > http://build.chromium.org/p/chromium.linux/builders/Linux%20Tests/builds/12124/steps/browser_tests/logs/Irt > http://build.chromium.org/p/chromium.linux/builders/Linux%20Tests/builds/12124/steps/browser_tests/logs/Messaging > > Original issue's description: > > Turn on clang by default on linux, 3rd try. > > > > This is mostly to collect performance and size data for now, and to find out > > if more bots need work. Unless things look really good, I'm going to revert > > this by thursday morning (PDT). > > > > BUG=360311 > > TBR=ajwong@chromium.org > > > > Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=282246 > > TBR=awong@chromium.org,thakis@chromium.org > NOTREECHECKS=true > NOTRY=true > BUG=360311 > > Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=282265 TBR=awong@chromium.org,thakis@chromium.org,horo@chromium.org NOTREECHECKS=true NOTRY=true BUG=360311 Review URL: https://codereview.chromium.org/382743002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@282266 0039d316-1c4b-4281-b951-d872f2087c98
* Revert of Turn on clang by default on linux, 3rd try. ↵horo@chromium.org2014-07-101-6/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (https://codereview.chromium.org/379413002/) Reason for revert: NaClBrowserTestNonSfiMode.Messaging and NaClBrowserTestNonSfiMode.Irt failed. http://build.chromium.org/p/chromium.linux/builders/Linux%20Tests/builds/12124/steps/browser_tests/logs/Irt http://build.chromium.org/p/chromium.linux/builders/Linux%20Tests/builds/12124/steps/browser_tests/logs/Messaging Original issue's description: > Turn on clang by default on linux, 3rd try. > > This is mostly to collect performance and size data for now, and to find out > if more bots need work. Unless things look really good, I'm going to revert > this by thursday morning (PDT). > > BUG=360311 > TBR=ajwong@chromium.org > > Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=282246 TBR=awong@chromium.org,thakis@chromium.org NOTREECHECKS=true NOTRY=true BUG=360311 Review URL: https://codereview.chromium.org/378363006 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@282265 0039d316-1c4b-4281-b951-d872f2087c98
* Turn on clang by default on linux, 3rd try.thakis@chromium.org2014-07-101-7/+6
| | | | | | | | | | | | | This is mostly to collect performance and size data for now, and to find out if more bots need work. Unless things look really good, I'm going to revert this by thursday morning (PDT). BUG=360311 TBR=ajwong@chromium.org Review URL: https://codereview.chromium.org/379413002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@282246 0039d316-1c4b-4281-b951-d872f2087c98
* Bump ASan/Win revision by ~2000timurrrr@chromium.org2014-07-091-1/+1
| | | | | | | | | | | | By the way, we should do this more often than once in 2000 revisions :) BUG=345874 TBR=rnk@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/374373002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281996 0039d316-1c4b-4281-b951-d872f2087c98
* Revert r281914thakis@chromium.org2014-07-091-6/+7
| | | | | | | | | | | | | | | Many 32bit bots fail with "bits/predef" not found. Examples: http://build.chromium.org/p/chromium.webkit/builders/WebKit%20Linux%2032/builds/25814/steps/compile/logs/stdio http://build.chromium.org/p/chromium/builders/Linux/builds/51225 BUG=360311 NOTRY=true NOTREECHECKS=true TBR=ajwong@chromium.org Review URL: https://codereview.chromium.org/378183002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281928 0039d316-1c4b-4281-b951-d872f2087c98
* Turn on clang by default on linux, 2nd try.thakis@chromium.org2014-07-091-6/+6
| | | | | | | | | | | | This is mostly to collect performance and size data for now. Unless things look really good, I'm going to revert this by wednesday morning (PDT). BUG=360311 TBR=awong@chromium.org Review URL: https://codereview.chromium.org/380573002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281914 0039d316-1c4b-4281-b951-d872f2087c98
* Support ignorance of base class finalizers.sigbjornf@opera.com2014-07-0812-17/+137
| | | | | | | | | | | | | | | | | | | | When checking if a class needs a finalizer, allow some of its base classes to be ignored, even if they do have a declared destructor. This is done so as to handle WebCore::ScriptWrappable use better while in transition. i.e., WebCore::ScriptWrappable's destructor does not have to be invoked when sweeping, and when Oilpan will be on by default, ScriptWrappable will not define one. However, until that time, it still provides a user-defined destructor, and to allow the class to be used as base in conjunction with GarbageCollected<>, we will now look past it when checking the finalizer requirement. R=zerny@chromium.org BUG= Review URL: https://codereview.chromium.org/374593002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@281758 0039d316-1c4b-4281-b951-d872f2087c98
* Clang update.py: check for LLVM_FORCE_HEAD_REVISION in the environmenthans@chromium.org2014-06-191-2/+2
| | | | | | | | | | | The FYI builders will set this to enforce using the latest revision. BUG=82385 NOTRY=true Review URL: https://codereview.chromium.org/336163004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278500 0039d316-1c4b-4281-b951-d872f2087c98
* Chrome clang plugin cleanup part 1 of ??dcheng@chromium.org2014-06-196-779/+899
| | | | | | | | | | | | | The clang plugin has grown organically over its lifetime. This patch just splits apart the options struct, the PluginASTAction (which most developers don't care about), and the actual ASTConsumer into separate units. R=erg@chromium.org Review URL: https://codereview.chromium.org/348543002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278487 0039d316-1c4b-4281-b951-d872f2087c98
* Use a faster rolling fixed Clang revision for ASan on Windowsrnk@chromium.org2014-06-191-0/+6
| | | | | | | | | | | | | | | | | | | We want a fixed Clang revision for ASan on Windows because ClusterFuzz produces archived builds. We don't want Clang to change underfoot and produce a bad Chromium build that gets archived. Long term, when ASan for Windows stabilizes, we should remove this extra fixed revision and use the mainline Chromium revision. For now, we don't build with -Werror with Clang on Windows, so it isn't as important to be in sync with other platforms. R=thakis@chromium.org BUG=82385, 345874 NOTRY=true Review URL: https://codereview.chromium.org/341963002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@278271 0039d316-1c4b-4281-b951-d872f2087c98
* Revert of Pick a fixed Clang revision for Windows now that we have an LKGR ↵thakis@chromium.org2014-06-171-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | ASan builder (https://codereview.chromium.org/320383004/) Reason for revert: This was landed despite an explicit "not lgtm" Original issue's description: > Pick a fixed Clang revision for Windows now that we have an LKGR ASan builder > This way, the LKGR builder and the developers who want to try ASan (or just Clang) on Windows won't get a broken Clang revision. > > The FYI bots got LLVM_WIN_REVISION=HEAD in r276301 so won't be affected. > > BUG=82385,345874 > NOTRY=true > > Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=276366 TBR=hans@chromium.org,rnk@chromium.org,timurrrr@chromium.org NOTREECHECKS=true NOTRY=true BUG=82385,345874 Review URL: https://codereview.chromium.org/342533002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277837 0039d316-1c4b-4281-b951-d872f2087c98
* Pick a fixed Clang revision for Windows now that we have an LKGR ASan buildertimurrrr@chromium.org2014-06-111-4/+4
| | | | | | | | | | | | | This way, the LKGR builder and the developers who want to try ASan (or just Clang) on Windows won't get a broken Clang revision. The FYI bots got LLVM_WIN_REVISION=HEAD in r276301 so won't be affected. BUG=82385,345874 NOTRY=true Review URL: https://codereview.chromium.org/320383004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@276366 0039d316-1c4b-4281-b951-d872f2087c98
* Blink GC plugin: disallow own pointers to GC allocated collections.zerny@chromium.org2014-06-043-2/+11
| | | | | | | | | | BUG=334149 R=haraken@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/318623002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@274816 0039d316-1c4b-4281-b951-d872f2087c98
* Package libclang_rt.san-{i386,x86_64}.a for Chromium Clang.glider@chromium.org2014-05-281-3/+7
| | | | | | | | | | BUG=174801 R=thakis@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/303473011 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@273223 0039d316-1c4b-4281-b951-d872f2087c98
* Blink GC plugin: support checking structures in anonymous namespaces.zerny@chromium.org2014-05-281-11/+9
| | | | | | | | | | BUG=334149 R=wibling@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/306473005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@273166 0039d316-1c4b-4281-b951-d872f2087c98
* Roll Clang 206824:209387hans@chromium.org2014-05-233-4/+6
| | | | | | | | | BUG=370448 NOTRY=true Review URL: https://codereview.chromium.org/296983008 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@272541 0039d316-1c4b-4281-b951-d872f2087c98
* Remove no longer needed "svn revert" lines.thakis@chromium.org2014-05-221-16/+0
| | | | | | | | | | | | They were added after we had to add some local patches for a tricky clang roll two months ago. We've done several rolls without local patches since then. BUG=341352 NOTRY=true Review URL: https://codereview.chromium.org/292333005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@272342 0039d316-1c4b-4281-b951-d872f2087c98
* Blink GC plugin: disallow GC derived types as part-object fields.zerny@chromium.org2014-05-214-24/+112
| | | | | | | | | | BUG=334149 R=ager@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/292743006 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@271880 0039d316-1c4b-4281-b951-d872f2087c98
* Blink GC plugin: look for checked namespaces within the list of namespaces.zerny@chromium.org2014-05-201-4/+6
| | | | | | | | | | | | Previously only the innermost namespace was matched against causing structures in, eg, namespace WebCore::Foo to not be checked. BUG=334149 R=haraken@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/294903002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@271608 0039d316-1c4b-4281-b951-d872f2087c98
* Win-clang update script: make an auxiliary copy of the sanitizer headershans@chromium.org2014-05-161-0/+14
| | | | | | | | | | | | This dir will be put on the include path for MSVC to pick up when invoked as fallback compiler. BUG=345874 NOTRY=true Review URL: https://codereview.chromium.org/281413006 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@271039 0039d316-1c4b-4281-b951-d872f2087c98
* Win-clang update script: copy asan_blacklist.txt into the right place.hans@chromium.org2014-05-161-2/+8
| | | | | | | | | BUG=371303 NOTRY=true Review URL: https://codereview.chromium.org/286193002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@270905 0039d316-1c4b-4281-b951-d872f2087c98
* Blink GC plugin: remove temporary exceptions for the TreeShared/Node hierarchy.zerny@chromium.org2014-05-134-28/+2
| | | | | | | | | | BUG=334149 R=ager@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/270623009 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@270096 0039d316-1c4b-4281-b951-d872f2087c98
* Bundle libc++ headers with clang.thakis@chromium.org2014-05-131-0/+4
| | | | | | | | | | | | | That way, -stdlib=libc++ works better, at least on OS X. BUG=371034 TEST=`echo '#include <vector>" > test.cc ; bin/clang -c test.cc -stdlib=libc++ -isysroot $(xcrun -show-sdk-path)` no longer complains about not finding the vector header on OS X. NOTRY=true Review URL: https://codereview.chromium.org/281753002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@270095 0039d316-1c4b-4281-b951-d872f2087c98
* Blink GC plugin: require that trace call base::trace and traceAfterDispatch ↵zerny@chromium.org2014-05-135-12/+17
| | | | | | | | | | | | call base::traceAfterDispatch. BUG=334149 R=wibling@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/278523004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@270057 0039d316-1c4b-4281-b951-d872f2087c98
* Blink GC plugin: require a trace method if a class derives multiple bases ↵zerny@chromium.org2014-05-129-9/+91
| | | | | | | | | | | | that need tracing. BUG=334149 R=haraken@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/273873002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@269764 0039d316-1c4b-4281-b951-d872f2087c98
* Blink GC plugin: disallow pure-virtual trace methods.zerny@chromium.org2014-05-106-30/+35
| | | | | | | | | | BUG=334149 R=ager@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/278513005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@269547 0039d316-1c4b-4281-b951-d872f2087c98
* clang: Instead of having update.sh clobber out/, add a define with the ↵thakis@chromium.org2014-05-091-13/+5
| | | | | | | | | | | | | | | | | | | | | | current revision. When clang is turned on or off, or when clang is updated, all .o files and all precompiled headers need to be rebuilt. This is currently done by having the update script remove the out/ directory. This has issues: * It fails to catch build directories with different names (e.g. out_android) * It removes other build artifacts (like resources) that don't need rebuilding * It doesn't happen when turning clang off (i.e. moving clang -> gcc). Instead, let common.gypi add a define with the current clang revision to each source file. This way, the clang revision is on each compile's command line and the build system's commandline tracking can take care of the rebuilding. BUG=nativeclient:3840 NOTRY=true Review URL: https://codereview.chromium.org/268363020 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@269427 0039d316-1c4b-4281-b951-d872f2087c98
* Blink GC plugin: require that GC derived classes do not override their ↵zerny@chromium.org2014-05-057-0/+71
| | | | | | | | | | | new(size_t) operator. BUG=334149 R=wibling@chromium.org Review URL: https://codereview.chromium.org/261973002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@268164 0039d316-1c4b-4281-b951-d872f2087c98
* Clang update script (Windows): fix asan runtime copyinghans@chromium.org2014-05-021-2/+5
| | | | | | | | | | | | Fix the copying of asan runtime. I'll try to come up with a more permanent solution, but this should get the ASan buildbot going in the meantime. BUG=82385 NOTRY=true Review URL: https://codereview.chromium.org/263003002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@267932 0039d316-1c4b-4281-b951-d872f2087c98
* Blink GC plugin: fix construction of collection edges containing primitive ↵zerny@chromium.org2014-05-029-18/+276
| | | | | | | | | | | types. R=ager@chromium.org BUG= Review URL: https://codereview.chromium.org/262883004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@267836 0039d316-1c4b-4281-b951-d872f2087c98
* Changed clang update script so it clobbers the entire chromium out dir.dyen@chromium.org2014-04-302-30/+9
| | | | | | | | | | | | | | | | When updating Clang, we must delete the all object files to force a rebuild of all source files using the new compiler. It is now safe to simply delete the chromium output directory since the update script is run from DEPS before the gyp files are generated. This is much safer as deleting object files may introduce unintended consequences with dependencies. BUG=https://code.google.com/p/nativeclient/issues/detail?id=3840 TEST=None Review URL: https://codereview.chromium.org/263443003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@267370 0039d316-1c4b-4281-b951-d872f2087c98
* Fix IsGCMixin predicate to correctly identify classes with multiple mixin bases.zerny@chromium.org2014-04-291-10/+13
| | | | | | | | | R=ager@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/260073010 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@266851 0039d316-1c4b-4281-b951-d872f2087c98
* Roll Clang 206155:206824hans@chromium.org2014-04-232-2/+2
| | | | | | | | BUG=365727 Review URL: https://codereview.chromium.org/247813003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@265694 0039d316-1c4b-4281-b951-d872f2087c98
* Roll Clang 204777:206155hans@chromium.org2014-04-162-2/+5
| | | | | | | | BUG=363667 Review URL: https://codereview.chromium.org/239243004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@264340 0039d316-1c4b-4281-b951-d872f2087c98
* clang plugin: Fix tests when running on OS X 10.9thakis@chromium.org2014-04-161-0/+5
| | | | | | | | | | | | | | | | | On 10.9, trunk clang defaults to targeting libc++. Clang expects to bundle the libc++ headers to be able to use libc++, but we don't currently do this. Instead, tell the plugin tests to use libstdc++ and pass a sysroot flag. (The blink oilpan plugin doesn't use and standard library includes, so it doesn't need this change at the moment.) BUG=none TEST=tools/clang/plugins/tests/test.sh third_party/llvm-build/Release+Asserts NOTRY=true Review URL: https://codereview.chromium.org/239753005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@264089 0039d316-1c4b-4281-b951-d872f2087c98
* Clang update.py: tiny fix to the regexeshans@chromium.org2014-04-141-2/+2
| | | | | | | | | | | | | | Remove erroneous word boundary marker. It seems "\b(make_clang_dir)=\b" wouldn't match "make_clang_dir=../foo/bar". Also, much less importantly, "\b(clang|asan)=1\b" wouldn't match "clang=10". BUG=none NOTRY=true Review URL: https://codereview.chromium.org/237323003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@263687 0039d316-1c4b-4281-b951-d872f2087c98
* Check that GC base classes are left-most derived and that RawPtr<T> are not ↵zerny@chromium.org2014-04-086-13/+89
| | | | | | | | | | | | used as fields. BUG=334149 R=ager@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/228503002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@262387 0039d316-1c4b-4281-b951-d872f2087c98
* Revert "Turn on clang by default on linux."thakis@chromium.org2014-04-051-6/+7
| | | | | | | | | | | | | We don't have 32bit clang binaries, but several 32bit builders. (And since clang now needs a c++11 compiler newer than gcc4.6 to compile, we can't tell these bots to just build clang themselves either.) BUG=360311 Review URL: https://codereview.chromium.org/227133002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@262030 0039d316-1c4b-4281-b951-d872f2087c98
* Turn on clang by default on linux.thakis@chromium.org2014-04-051-0/+6
| | | | | | | | | | | | This is mostly to collect performance and size data for now. Unless things look really good, I'm going to revert this by end-of-day sunday. BUG=360311 TBR=awong@chromium.org Review URL: https://codereview.chromium.org/226253008 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@262025 0039d316-1c4b-4281-b951-d872f2087c98
* Upgrade blink plugin to pick up r261132.thakis@chromium.org2014-04-031-1/+1
| | | | | | | | | | BUG=334149 TBR=zerny@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/223043003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@261274 0039d316-1c4b-4281-b951-d872f2087c98
* clang update script cleanups: Remove Xcode 3.2 step, replace "cd -" with "popd".thakis@chromium.org2014-04-022-33/+8
| | | | | | | | | | | No intended behavior change. BUG=none NOTRY=true Review URL: https://codereview.chromium.org/221273002 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@261133 0039d316-1c4b-4281-b951-d872f2087c98
* Add temporary exceptions for the TreeShared hierarchy to the Blink GC plugin.zerny@chromium.org2014-04-025-5/+33
| | | | | | | | | | | | | | While transitioning the Node hierarchy we allow Persistent fields from within the hierarchy and allow raw pointers to the hierarchy from stack-allocated objects. Also, fixes a typo introduced in r260833. BUG=334149 R=ager@chromium.org NOTRY=true Review URL: https://codereview.chromium.org/220403003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@261132 0039d316-1c4b-4281-b951-d872f2087c98
* roll clang 202555:204777thakis@chromium.org2014-04-021-644/+19
| | | | | | | | | BUG=357890 TBR=hans@chromium.org Review URL: https://codereview.chromium.org/221623003 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@261042 0039d316-1c4b-4281-b951-d872f2087c98
* roll clang 202554:202555thakis@chromium.org2014-04-012-6/+10
| | | | | | | | | | | | | | | | | | | | | | | | This rolls in just clang r202555, which is a formatting change -- so this change doesn't modify the compiler itself. It does deploy the following things: 1.) On Mac, clang now packages libc++. This will hopefully allow clang to continue running on OS X 10.6, even when clang is built with a dependency on a standard library that provides c++11 features. (http://crbug.com/202555) 2.) It picks up r260249 for the plugin, fixing a crash in the plugin (http://crbug.com/202555) and turning the warning off by default. 3.) It picks up oilplan plugin changes up to r260911. This also changes the bundle build script to use the bootstrap compiler to compile libc++ in bootstrap builds, and tweaks the script to keep clang asan/android buildable with the r9d ndk (http://crbug.com/357890) BUG=358500,356816,357890 TBR=hans@chromium.org Review URL: https://codereview.chromium.org/217323005 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@260962 0039d316-1c4b-4281-b951-d872f2087c98
* clang/mac: Bundle libc++.dylib with clang.thakis@chromium.org2014-04-014-13/+63
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | clang now uses c++11, which on OS X means it depends on libc++. OS X 10.6 doesn't have libc++ yet, but some bots still are on 10.6 and upgrading them will take a bit longer. Bundle libc++ with the clang package, and set rpaths so that both clang and plugins refer to the bundled libc++ instead of /usr/lib/libc++.1.dylib. (System libraries will still cause libstdc++ to be loaded as well. Since libc++ keeps its symbols in std::__1:: and aliases everything to that, loading libc++ and libstdc++ in the same process is safe.) Also don't set MACOSX_DEPLOYMENT_TARGET while building the bootstrap compiler (Xcode's clang complains when using -stdlib=libc++ when targeting 10.6 or older; open-source luckily doesn't for some reason). Also pass -isysroot; when building on OS X 10.9, /usr/include doesn't exist unless the command line tools are installed. BUG=358500 TEST=Run package.sh (with a tweaked clang revision, to not clobber an existing package). Upload produced tgz file somewhere, download it on a 10.6 box. Clang should run. R=hans@chromium.org Review URL: https://codereview.chromium.org/218613009 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@260908 0039d316-1c4b-4281-b951-d872f2087c98
* Add flag to gather statistics from the Blink points-to graph.zerny@chromium.org2014-04-012-26/+150
| | | | | | | | | | | This also fixes an error where the edge key of subclass edges was non-unique causing subclass edge entries to be incorrectly overwritten. BUG=334149 NOTRY=true Review URL: https://codereview.chromium.org/219213004 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@260833 0039d316-1c4b-4281-b951-d872f2087c98
* Re-enable the clang check for enum constants.tsepez@chromium.org2014-03-2810-1/+172
| | | | | | | | | | | | | | | | | | | The patch adds a flag to disable the feature, so as to avoid future reverts. There is also a check to avoid mixed signed comparisons, as can arise with enums from C files. We need not bother checking these. BUG=356816 R=mark@chromium.org R=thakis@chromium.org Re-enable flag. patch from issue 150943005 Review URL: https://codereview.chromium.org/213403010 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@260249 0039d316-1c4b-4281-b951-d872f2087c98
generated by cgit v1.1 at 2025-03-30 14:38:34 +0000