Commits · 35944dda103ed588b68970c24bcd79b26d1b2dfd · Roger Ferrer / llvm-epi-0.8

Mar 17, 2011
- Reapply: Add type output to llvm-dis annotations. Patch by Yuri! · ec54bd75
  Stuart Hastings authored Mar 17, 2011
```
llvm-svn: 127824
```
  ec54bd75
- Add XCore intrinsic for setpsc. · 6120962d
  Richard Osborne authored Mar 17, 2011
```
llvm-svn: 127821
```
  6120962d
- Revert 127813 while fixing broken test. · ead01c40
  Stuart Hastings authored Mar 17, 2011
```
llvm-svn: 127814
```
  ead01c40
- Add type output to llvm-dis. Patch by Yuri! · 8e08a786
  Stuart Hastings authored Mar 17, 2011
```
llvm-svn: 127813
```
  8e08a786
- MC/Mach-O: Fix regression introduced in r126127, this assignment shouldn't have · f1d62cfc
  Daniel Dunbar authored Mar 17, 2011
```
been removed.

llvm-svn: 127812
```
  f1d62cfc
- Move more logic into getTypeForExtArgOrReturn. · 2ef0c69d
  Cameron Zwarich authored Mar 17, 2011
```
llvm-svn: 127809
```
  2ef0c69d
- Change the signext language in LangRef to closer match zeroext. · 341c36dc
  Cameron Zwarich authored Mar 17, 2011
```
llvm-svn: 127808
```
  341c36dc
- Rename getTypeForExtendedInteger() to getTypeForExtArgOrReturn(). · 34e7b3f7
  Cameron Zwarich authored Mar 17, 2011
```
llvm-svn: 127807
```
  34e7b3f7
- Add comments for the demanglings. Correct mangled form of operator delete! · b4d763b3
  Nick Lewycky authored Mar 17, 2011
```
llvm-svn: 127801
```
  b4d763b3
- test/CodeGen/X86/h-registers-1.ll: Add explicit -mtriple=x86_64-linux. It does... · bf9ff6f6
  NAKAMURA Takumi authored Mar 17, 2011
```
test/CodeGen/X86/h-registers-1.ll: Add explicit -mtriple=x86_64-linux. It does not need to be checked on x86_64-win32 (aka Win64).

llvm-svn: 127800
```
  bf9ff6f6
- Add "swi" which is an obsolete mnemonic for "svc". · 881e1871
  Nick Lewycky authored Mar 17, 2011
```
llvm-svn: 127788
```
  881e1871
- A couple new README entries. · e8f2be0c
  Eli Friedman authored Mar 17, 2011
```
llvm-svn: 127786
```
  e8f2be0c
- Add support in the LTO library for loading an object from the middle · b39c7c7b
  Rafael Espindola authored Mar 17, 2011
```
of an file.

llvm-svn: 127781
```
  b39c7c7b
- Fix handling of @IDNTPOFF relocations, they need to get STT_TLS. · 07de07ee
  Joerg Sonnenberger authored Mar 17, 2011
```
While here, add VK_ARM_TPOFF and VK_ARM_GOTTPOFF, too.

llvm-svn: 127780
```
  07de07ee
- Rewrite instructions as part of ConnectedVNInfoEqClasses::Distribute. · 315b42c3
  Jakob Stoklund Olesen authored Mar 17, 2011
```
llvm-svn: 127779
```
  315b42c3
- test/CodeGen/X86/constant-pool-remat-0.ll: FileCheck-ize and add explicit -mtriple=x86_64-linux. · 5b6198df
  NAKAMURA Takumi authored Mar 16, 2011
```
llvm-svn: 127775
```
  5b6198df
Mar 16, 2011
- Add a LiveRangeEdit delegate callback before shrinking a live range. · e14b2b22
  Jakob Stoklund Olesen authored Mar 16, 2011
```
The register allocator needs to adjust its live interval unions when that happens.

llvm-svn: 127774
```
  e14b2b22
- Erase virtual registers that are unused after DCE. · c738c965
  Jakob Stoklund Olesen authored Mar 16, 2011
```
llvm-svn: 127773
```
  c738c965
- Tag cached interference with a user-provided tag instead of the virtual register number. · e29d63e9
  Jakob Stoklund Olesen authored Mar 16, 2011
```
The live range of a virtual register may change which invalidates the cached
interference information.

llvm-svn: 127772
```
  e29d63e9
- Clarify debugging output. · 557a82c0
  Jakob Stoklund Olesen authored Mar 16, 2011
```
llvm-svn: 127771
```
  557a82c0
- The x86-64 ABI says that a bool is only guaranteed to be sign-extended to a byte · ac106273
  Cameron Zwarich authored Mar 16, 2011
```
rather than an int. Thankfully, this only causes LLVM to miss optimizations, not
generate incorrect code.

This just fixes the zext at the return. We still insert an i32 ZextAssert when
reading a function's arguments, but it is followed by a truncate and another i8
ZextAssert so it is not optimized.

llvm-svn: 127766
```
  ac106273
- Rename a test to be more inclusive. · 40a92003
  Cameron Zwarich authored Mar 16, 2011
```
llvm-svn: 127765
```
  40a92003
- Don't recompute something that we already have in a local variable. · d1ad9bc2
  Cameron Zwarich authored Mar 16, 2011
```
llvm-svn: 127764
```
  d1ad9bc2
- Revert r127757, "Patch to a fix dwarf relocation problem on ARM. One-line fix · fd95b016
  Daniel Dunbar authored Mar 16, 2011
```
plus the test where it used to break.", which broke Clang self-host of a
Debug+Asserts compiler, on OS X.

llvm-svn: 127763
```
  fd95b016
- Add XCore intrinsics for setclk, setrdy. · c871eff3
  Richard Osborne authored Mar 16, 2011
```
llvm-svn: 127761
```
  c871eff3
- Patch to a fix dwarf relocation problem on ARM. One-line fix plus the test where it used to break. · a3aeafeb
  Renato Golin authored Mar 16, 2011
```
llvm-svn: 127757
```
  a3aeafeb
- Add a test for i1 zeroext arguments on x86-64. We currently generate code that · 49e354bc
  Cameron Zwarich authored Mar 16, 2011
```
conforms to the ABI, but DAGCombine could in theory recognize the sequence of
zext asserts and truncates and generate incorrect code.

llvm-svn: 127754
```
  49e354bc
- Add checkevent intrinsic to check if any resources owned by the current thread · d4346f23
  Richard Osborne authored Mar 16, 2011
```
can event.

llvm-svn: 127741
```
  d4346f23
- test/CodeGen/X86: FileCheck-ize and add actions for x86_64-linux and x86_64-win32. · d60e4101
  NAKAMURA Takumi authored Mar 16, 2011
```
llvm-svn: 127734
```
  d60e4101
- test/CodeGen/X86: Add a pattern for Win64. · 0b9e2b02
  NAKAMURA Takumi authored Mar 16, 2011
```
llvm-svn: 127733
```
  0b9e2b02
- test/CodeGen/X86: FileCheck-ize and add explicit -mtriple=x86_64-linux. They... · c10801e8
  NAKAMURA Takumi authored Mar 16, 2011
```
test/CodeGen/X86: FileCheck-ize and add explicit -mtriple=x86_64-linux. They are useless to Win64 target.

llvm-svn: 127732
```
  c10801e8
- test/CodeGen/X86/byval*.ll: Win64 has not supported byval yet. · 662892df
  NAKAMURA Takumi authored Mar 16, 2011
```
llvm-svn: 127731
```
  662892df
- test/CodeGen/X86/dyn-stackalloc.ll: FileCheck-ize. · 406f02c9
  NAKAMURA Takumi authored Mar 16, 2011
```
llvm-svn: 127730
```
  406f02c9
- Fix a comment. · 7599b106
  Cameron Zwarich authored Mar 16, 2011
```
llvm-svn: 127728
```
  7599b106
- lib/Support/raw_ostream.cpp: On mingw, report_fatal_error() should not be called at dtor context. · 76e68ea9
  NAKAMURA Takumi authored Mar 16, 2011
```
report_fatal_error() invokes exit(). We know report_fatal_error() might not write messages to stderr when any errors were detected on FD == 2.

llvm-svn: 127726
```
  76e68ea9
- Windows/PathV2.inc: [PR8520] Recognize "NUL" as special (character) file. · a3d4749e
  NAKAMURA Takumi authored Mar 16, 2011
```
FIXME: It is a temporal hack. We should detect as many "special file name" as possible.
llvm-svn: 127724
```
  a3d4749e
- Windows/Path.inc: [PR6270] PathV1::makeUnique(): Give arbitrary initial seed for workaround. · aaa9b4f8
  NAKAMURA Takumi authored Mar 16, 2011
```
FIXME: We should use sys::fs::unique_file() in future.
llvm-svn: 127723
```
  aaa9b4f8
- Tidy up. Whitespace and 80 column. · c91fa6d1
  Jim Grosbach authored Mar 16, 2011
```
llvm-svn: 127721
```
  c91fa6d1
- Do not accidently initialize NumDbgValueLost and NumDbgLineLost counts. · 39c041ca
  Devang Patel authored Mar 16, 2011
```
llvm-svn: 127720
```
  39c041ca
- Only convert allocas to scalars if it is profitable. The profitability metric I · 0454253d
  Cameron Zwarich authored Mar 16, 2011
```
chose is having a non-memcpy/memset use and being larger than any native integer
type. Originally I chose having an access of a size smaller than the total size
of the alloca, but this caused some minor issues on the spirit benchmark where
SRoA runs again after some inlining.

This fixes <rdar://problem/8613163>.

llvm-svn: 127718
```
  0454253d