Commits · d1dda5339d43ddc5b5522b059bf66b1280ab6db4 · Roger Ferrer / llvm-epi · GitLab

Nov 27, 2008
- Add -march=x86. · d1dda533
  Evan Cheng authored Nov 27, 2008
  d1dda533
Nov 26, 2008
- Add x86-specific test for add-with-overflow intrinsics. · a69ced6b
  Bill Wendling authored Nov 26, 2008
  a69ced6b
- Turn on my codegen prepare heuristic by default. It doesn't affect · 397a11cc
  Chris Lattner authored Nov 26, 2008
  397a11cc
- Check that running the DAG combiner between type · d1ba7908
  Duncan Sands authored Nov 26, 2008
  d1ba7908
- Add test for rdar://6394879 . · 3d14916b
  Bill Wendling authored Nov 26, 2008
  3d14916b
- This adds in some code (currently disabled unless you pass · eb3e4fb6
  Chris Lattner authored Nov 26, 2008
  eb3e4fb6
- Teach CodeGenPrepare to look through Bitcast instructions when attempting to · a9ab165b
  Chris Lattner authored Nov 26, 2008
  a9ab165b
- fix an over-reduced test. · f0e01def
  Chris Lattner authored Nov 26, 2008
  f0e01def
- this doesn't need EH · 0f98f74c
  Chris Lattner authored Nov 26, 2008
  0f98f74c
Nov 25, 2008
- Since the old llvmc was removed, rename llvmc2 to llvmc. · 98d5ed5c
  Mikhail Glushenkov authored Nov 25, 2008
  98d5ed5c
- convertToSignExtendedInteger should return opInvalidOp instead of asserting if... · 2e5aeff6
  Evan Cheng authored Nov 25, 2008
  2e5aeff6
- CellSPU: · 910046d1
  Scott Michel authored Nov 25, 2008
  910046d1
- CellSPU: test should use shlqby, not shlqbyi · 5149430c
  Scott Michel authored Nov 25, 2008
  5149430c
- XFAIL this test. A recent CellSPU check-in broke it. · aec5a564
  Bill Wendling authored Nov 25, 2008
  aec5a564
- Initial support for anti-dependence breaking. Currently this code does not · ad2134d4
  Dan Gohman authored Nov 25, 2008
  ad2134d4
- Testcase for constant CFStrings. · a3070208
  Bill Wendling authored Nov 24, 2008
  a3070208
Nov 24, 2008
- reenable test · 18065ce9
  Chris Lattner authored Nov 24, 2008
  18065ce9
- Temporarily XFAIL this test. r59976 and r59972 broke it. · e6fe59df
  Bill Wendling authored Nov 24, 2008
  e6fe59df
- Fix 3113: If we have a dead cyclic PHI, replace the whole thing · 53d6a078
  Chris Lattner authored Nov 24, 2008
  53d6a078
- CellSPU: · 2e5df906
  Scott Michel authored Nov 24, 2008
  2e5df906
- CellSPU: · efc8c7a2
  Scott Michel authored Nov 24, 2008
  efc8c7a2
- Test add-with-overflow with fast ISel. · 411eaa5c
  Bill Wendling authored Nov 24, 2008
  411eaa5c
- Extend the 'noalias' attribute to function return values. This is intended to · f5ffcbcd
  Nick Lewycky authored Nov 24, 2008
  f5ffcbcd
- Add support for llvm.uadd.with.overflow. · 2278f8f5
  Bill Wendling authored Nov 24, 2008
  2278f8f5
Nov 23, 2008
- CellSPU: Fix bug 3056. Varadic extract_element was not implemented (nor was it · 0be0339b
  Scott Michel authored Nov 22, 2008
  0be0339b
Nov 21, 2008
- Optimize (x/y)*y into x-(x%y) in general. Div and rem are about the same, and · 07d726ec
  Nick Lewycky authored Nov 21, 2008
  07d726ec
- CellSPU: · c6918c1f
  Scott Michel authored Nov 21, 2008
  c6918c1f
- Add generic test for add with overflow. · d10202bf
  Bill Wendling authored Nov 21, 2008
  d10202bf
Nov 20, 2008
- Test -pre-RA-sched=fast too, for completeness. · 52aece7e
  Dan Gohman authored Nov 20, 2008
  52aece7e
- - Register scavenger should use MachineRegisterInfo and internal map to find... · c22215a0
  Evan Cheng authored Nov 20, 2008
  c22215a0
Nov 19, 2008
- · 7ed6c531
  Devang Patel authored Nov 19, 2008
  7ed6c531
- Revert r59640. It broke this test for builds that aren't · 7f480e04
  Dan Gohman authored Nov 19, 2008
  7f480e04
- Use %llvmgcc -xassembler instead of invoking as directly. This avoids · f09aabfe
  Dan Gohman authored Nov 19, 2008
  f09aabfe
- Add support for rematerialization in pre-alloc-splitting. · 1af37c2f
  Owen Anderson authored Nov 19, 2008
  1af37c2f
- LLVMC2: -emit-llvm stops compilation. · 45535f3f
  Daniel Dunbar authored Nov 19, 2008
  45535f3f
- LLVMC2: Teach llvm_gcc_c tool about -include and -fsyntax-only. · 2dd5aa4f
  Daniel Dunbar authored Nov 19, 2008
  2dd5aa4f
Nov 18, 2008
- Register scavenger should process early clobber defs first. A dead early... · 3095c32c
  Evan Cheng authored Nov 18, 2008
  3095c32c
- Add a utility function that detects whether a loop is guaranteed to be finite. · 1c451ae4
  Nick Lewycky authored Nov 18, 2008
  1c451ae4
- Reapply r59464, this time using the correct type · 1315f80e
  Duncan Sands authored Nov 18, 2008
  1315f80e
- A simple test for stack protectors. This should be valid on all platforms. · c6c6dc5b
  Bill Wendling authored Nov 18, 2008
  c6c6dc5b