llvm/utils/TableGen
Filip Pizlo 23ffb3ea10 Stackmaps are used for OSR exits, which is a custom kind of unwinding. Hence, they
should not be marked nounwind.

Marking them nounwind caused crashes in the WebKit FTL JIT, because if we enable 
sufficient optimizations, LLVM starts eliding compact_unwind sections (or any unwind 
data for that matter), making deoptimization via stackmaps impossible.

This changes the stackmap intrinsic to be may-throw, adds a test for exactly the 
sympton that WebKit saw, and fixes TableGen to handle un-attributed intrinsics.

Thanks to atrick and philipreames for reviewing this.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@201826 91177308-0d34-0410-b5e6-96231b3b80d8
2014-02-20 23:57:31 +00:00
..
AsmMatcherEmitter.cpp Add missing #include's to cctype when using isdigit/alpha/etc. 2013-10-12 00:55:57 +00:00
AsmWriterEmitter.cpp Get rid of a vector copy by just making a pointer out of the reference returned by getInstructionsByEnumValue instead of assigning it to a new vector. 2014-02-05 07:56:49 +00:00
AsmWriterInst.cpp Remove dead code. 2013-12-02 05:10:04 +00:00
AsmWriterInst.h Remove dead code. 2013-12-02 05:10:04 +00:00
CallingConvEmitter.cpp LLVM-1163: AAPCS-VFP violation when CPRC allocated to stack 2014-02-07 11:19:53 +00:00
CMakeLists.txt Remove TGValueTypes.cpp from CMakeLists.txt which I forgot to do in r200036. 2014-01-24 20:51:32 +00:00
CodeEmitterGen.cpp Propagate MCSubtargetInfo through TableGen's getBinaryCodeForInstr() 2014-01-28 23:13:18 +00:00
CodeGenDAGPatterns.cpp Improve handling of EnforceSmallerThan. Remove all types that are smaller from the larger set not just the smallest type from the smaller set. Ensure 'smaller' vectors have the same or fewer total bits. Similar for 'larger' vectors. 2014-01-28 04:49:01 +00:00
CodeGenDAGPatterns.h Improve handling of EnforceSmallerThan. Remove all types that are smaller from the larger set not just the smallest type from the smaller set. Ensure 'smaller' vectors have the same or fewer total bits. Similar for 'larger' vectors. 2014-01-28 04:49:01 +00:00
CodeGenInstruction.cpp Shrink the size of CodeGenInstruction a little bit by using bitfields. 32 bools seemed excessive. 2014-02-05 09:10:40 +00:00
CodeGenInstruction.h Shrink the size of CodeGenInstruction a little bit by using bitfields. 32 bools seemed excessive. 2014-02-05 09:10:40 +00:00
CodeGenIntrinsics.h
CodeGenMapTable.cpp Correct word hyphenations 2013-12-05 05:44:44 +00:00
CodeGenRegisters.cpp Replace tablegen uses of EVT with MVT. Add isOverloaded() to MVT to facilitate. Remove TGValueTypes.cpp since its unused now (and may have been before). 2014-01-24 20:50:47 +00:00
CodeGenRegisters.h Fix known typos 2014-01-24 17:20:08 +00:00
CodeGenSchedule.cpp Fix known typos 2014-01-24 17:20:08 +00:00
CodeGenSchedule.h Fix known typos 2014-01-24 17:20:08 +00:00
CodeGenTarget.cpp Delete all of the CodeGenInstructions from CodeGenTarget destructor. 2014-02-06 06:27:59 +00:00
CodeGenTarget.h Support little-endian encodings in the FixedLenDecoderEmitter 2013-12-17 22:37:50 +00:00
CTagsEmitter.cpp Remove unnecessary include. 2014-02-09 07:55:19 +00:00
DAGISelEmitter.cpp Use ArrayRef to simplify some code. 2014-01-21 07:20:05 +00:00
DAGISelMatcher.cpp Add CheckChildInteger to ISelMatcher operations. Removes nearly 2000 bytes from X86 matcher table. 2014-02-05 05:44:28 +00:00
DAGISelMatcher.h Add CheckChildInteger to ISelMatcher operations. Removes nearly 2000 bytes from X86 matcher table. 2014-02-05 05:44:28 +00:00
DAGISelMatcherEmitter.cpp Add CheckChildInteger to ISelMatcher operations. Removes nearly 2000 bytes from X86 matcher table. 2014-02-05 05:44:28 +00:00
DAGISelMatcherGen.cpp Fix typo in commment tyep->type. 2014-01-25 17:34:23 +00:00
DAGISelMatcherOpt.cpp Add CheckChildInteger to ISelMatcher operations. Removes nearly 2000 bytes from X86 matcher table. 2014-02-05 05:44:28 +00:00
DFAPacketizerEmitter.cpp
DisassemblerEmitter.cpp Don't use PrintFatalError(which calls exit) for 'Primary decode conflict'. Just skip emitting the table. This way the main function will delete the output file instead of it remaining empty and confusing dependency checks if build is invoked a second time. 2014-01-05 01:34:12 +00:00
FastISelEmitter.cpp Rename a DebugLoc variable to DbgLoc and a DataLayout to DL. 2014-02-18 22:05:46 +00:00
FixedLenDecoderEmitter.cpp Remove unnecessary include. 2014-02-09 07:55:19 +00:00
InstrInfoEmitter.cpp Fix a vector that was passed by value instead of reference. 2014-02-05 07:27:49 +00:00
IntrinsicEmitter.cpp Stackmaps are used for OSR exits, which is a custom kind of unwinding. Hence, they 2014-02-20 23:57:31 +00:00
LLVMBuild.txt
Makefile
OptParserEmitter.cpp Typo. 2013-10-20 03:19:25 +00:00
PseudoLoweringEmitter.cpp Change MCStreamer EmitInstruction interface to take subtarget info 2014-01-28 23:12:42 +00:00
RegisterInfoEmitter.cpp Replace some unnecessary vector copies with references. 2013-09-15 22:04:42 +00:00
SequenceToOffsetTable.h
SetTheory.cpp [weak vtables] Remove a bunch of weak vtables 2013-11-19 00:57:56 +00:00
SetTheory.h
SubtargetEmitter.cpp Mark the x86 machine model as incomplete. PR17367. 2013-09-25 18:14:12 +00:00
TableGen.cpp replace LeakSanitizerIsTurnedOffForTheCurrentProcess with __lsan_is_turned_off, but this time hide it under __has_feature(address_sanitizer); also include <sanitizer/lsan_interface.h> 2014-01-15 07:59:37 +00:00
TableGenBackends.h
tdtags
X86DisassemblerShared.h Remove filtering concept from X86 disassembler table generation. It's no longer necessary. 2014-02-13 07:07:16 +00:00
X86DisassemblerTables.cpp Remove A6/A7 opcode maps. They can all be handled with a TB map, opcode of 0xa6/0xa7, and adding MRM_C0/MRM_E0 forms. Removes 376K from the disassembler tables. 2014-02-19 05:34:21 +00:00
X86DisassemblerTables.h Remove A6/A7 opcode maps. They can all be handled with a TB map, opcode of 0xa6/0xa7, and adding MRM_C0/MRM_E0 forms. Removes 376K from the disassembler tables. 2014-02-19 05:34:21 +00:00
X86ModRMFilters.cpp Second attempt at Removing special form of AddRegFrm used by FP instructions. These instructions can be handled by MRMXr instead. 2014-01-01 14:22:37 +00:00
X86ModRMFilters.h Second attempt at Removing special form of AddRegFrm used by FP instructions. These instructions can be handled by MRMXr instead. 2014-01-01 14:22:37 +00:00
X86RecognizableInstr.cpp [x86] Switch PAUSE instruction to use XS prefix instead of HasREPPrefix. Remove HasREPPrefix support from disassembler table generator since its now only used by CodeGenOnly instructions. 2014-02-20 07:59:43 +00:00
X86RecognizableInstr.h [x86] Switch PAUSE instruction to use XS prefix instead of HasREPPrefix. Remove HasREPPrefix support from disassembler table generator since its now only used by CodeGenOnly instructions. 2014-02-20 07:59:43 +00:00