Commits · 3241b74f69e845fec49bf77632865c66df4f6486 · Roger Ferrer / llvm-epi · GitLab

Mar 27, 2009
- Revert r67798: it breaks llvm-gcc bootstrap on x86-64-linux, presumably due to · 3241b74f
  Duncan Sands authored Mar 27, 2009
  3241b74f
- -write-graph now can be used with -o. · f4be420d
  Mikhail Glushenkov authored Mar 27, 2009
  f4be420d
- Shrink wrapping in PEI: initial release. Finishing development, enable with --shrink-wrap. · a1be2dcd
  John Mosby authored Mar 27, 2009
  a1be2dcd
- One more place to skip debug info. · 4026b041
  Dale Johannesen authored Mar 27, 2009
  4026b041
- While hoisting an instruction, update alias info set tracker. · fe7c0492
  Devang Patel authored Mar 26, 2009
  fe7c0492
- -no-implicit-float means explicit fp operations are legal. · d88ebc35
  Evan Cheng authored Mar 26, 2009
  d88ebc35
Mar 26, 2009
- Fix misc. small issues with debug visualization. · 87aa96dd
  Mikhail Glushenkov authored Mar 26, 2009
  87aa96dd
- tADDhirr is a thumb instruction. Do not allow this code to be reached in non-thumb mode. · 904f1466
  Evan Cheng authored Mar 26, 2009
  904f1466
- Don't assign a new stack slot if the pre-alloc splitter already assigned one. · 76a561f3
  Owen Anderson authored Mar 26, 2009
  76a561f3
- Pull transform from target-dependent code into target-independent code. · aa28be65
  Bill Wendling authored Mar 26, 2009
  aa28be65
- fix warning in -asserts mode. · 78b7cbe6
  Chris Lattner authored Mar 26, 2009
  78b7cbe6
- fix some warnings in release-asserts mode. · f38fd11a
  Chris Lattner authored Mar 26, 2009
  f38fd11a
- fix an apparently real bug exposed by a warning in -asserts mode. · 3c04970f
  Chris Lattner authored Mar 26, 2009
  3c04970f
- fix warning in -asserts build. · 3871781c
  Chris Lattner authored Mar 26, 2009
  3871781c
- Match this pattern so that we can generate simpler code: · 94f299f2
  Bill Wendling authored Mar 26, 2009
  94f299f2
- Doxygen-ify comments. · 189d6718
  Bill Wendling authored Mar 26, 2009
  189d6718
- Skip debug info one more place. (This one gets · db90560c
  Dale Johannesen authored Mar 26, 2009
  db90560c
- Before deleting a basic block, give other loop passes a chance cleanup... · 45556188
  Devang Patel authored Mar 25, 2009
  45556188
Mar 25, 2009
- Revert 67132. This is breaking some objective-c apps. · 2e9f42be
  Evan Cheng authored Mar 25, 2009
  2e9f42be
- When optimizing with debug info, don't keep the · eb1646d2
  Dale Johannesen authored Mar 25, 2009
  eb1646d2
- Fix a bug in our autoupgrade support: in an argument list to a function · 56a95f6c
  Chris Lattner authored Mar 25, 2009
  56a95f6c
- do not rely on callee being operand 0 · 2755edf4
  Gabor Greif authored Mar 25, 2009
  2755edf4
- · 68b0881d
  Devang Patel authored Mar 25, 2009
  68b0881d
- CodeGen still defaults to non-verbose asm, but llc now overrides it and default to verbose. · 5e5a63cf
  Evan Cheng authored Mar 25, 2009
  5e5a63cf
- Don't print global names twice with -asm-verbose. · 9966403e
  Evan Cheng authored Mar 25, 2009
  9966403e
- Fix PR3874 by restoring a condition I removed, but making it more · c3b2111d
  Chris Lattner authored Mar 25, 2009
  c3b2111d
- Do not ignore DW_TAG_class_type! · a6e85a19
  Devang Patel authored Mar 25, 2009
  a6e85a19
- oops, I intended to remove this, not comment it out. Thanks Duncan! · 9e945380
  Chris Lattner authored Mar 24, 2009
  9e945380
Mar 24, 2009
- Fix PR3845: Avoid stale MachineInstruction pointer reference. · f413a888
  Evan Cheng authored Mar 24, 2009
  f413a888
- simplify logic and get rid of the assumption that operand 0 is the callee · e7e5becd
  Gabor Greif authored Mar 24, 2009
  e7e5becd
- canonicalize inttoptr and ptrtoint instructions which cast pointers · 306813cb
  Chris Lattner authored Mar 24, 2009
  306813cb
- fix one more fp80 case (used only by Interpreter) · 4d7e4ee7
  Dale Johannesen authored Mar 24, 2009
  4d7e4ee7
- two changes: · d9eb4117
  Chris Lattner authored Mar 24, 2009
  d9eb4117
- Change JIT for different layout of fp80. · f259f3f3
  Dale Johannesen authored Mar 24, 2009
  f259f3f3
- more tidying: name the components of PhysReg in the case when · c35847e1
  Chris Lattner authored Mar 24, 2009
  c35847e1
- Tidy a bit more. · 42eceb34
  Chris Lattner authored Mar 24, 2009
  42eceb34
- simplify this code a bit now that "allocation to a vreg class" can never · 246eda43
  Chris Lattner authored Mar 24, 2009
  246eda43
- I was convinced that it's ok to allow a second i8 return value · efd2d44a
  Dan Gohman authored Mar 24, 2009
  efd2d44a
- Minor compile-time optimization; don't bother checking · f3746cbc
  Dan Gohman authored Mar 24, 2009
  f3746cbc
- Add a pre-pass to the burr-list scheduler which makes adjustments to · 9a658d72
  Dan Gohman authored Mar 24, 2009
  9a658d72