Commits · fa9e60895b74f1afcb2526033597990e19df5cf1 · Roger Ferrer / llvm-epi-0.8

Jun 29, 2006
- Add shift and rotate by 1 instructions / patterns. · fa9e6089
  Evan Cheng authored Jun 29, 2006
```
llvm-svn: 28980
```
  fa9e6089
- Always use xorps to clear XMM registers. · fc8cdda0
  Evan Cheng authored Jun 29, 2006
```
llvm-svn: 28979
```
  fc8cdda0
- Move .literal4 and .literal8 support into AsmPrinter.cpp · 56737d4f
  Evan Cheng authored Jun 29, 2006
```
llvm-svn: 28978
```
  56737d4f
- Hide x86 symbols · 0cc59077
  Chris Lattner authored Jun 28, 2006
```
llvm-svn: 28976
```
  0cc59077
- Use hidden visibility to make symbols in an anonymous namespace get · 996795b0
  Chris Lattner authored Jun 28, 2006
```
dropped.  This shrinks libllvmgcc.dylib another 67K

llvm-svn: 28975
```
  996795b0
- shrink libllvmgcc.dylib another 25K · 2f8c2d8e
  Chris Lattner authored Jun 28, 2006
```
llvm-svn: 28971
```
  2f8c2d8e
Jun 28, 2006
- Doh. · 87813744
  Evan Cheng authored Jun 28, 2006
```
llvm-svn: 28963
```
  87813744
- Oops. Need to keep CP index. · 0687b044
  Evan Cheng authored Jun 28, 2006
```
llvm-svn: 28958
```
  0687b044
- Darwin puts float and double literal constants into literal4 and literal8 sections. · 7f88856d
  Evan Cheng authored Jun 28, 2006
```
llvm-svn: 28957
```
  7f88856d
- this case isn't handled · a53a22e5
  Andrew Lenharth authored Jun 27, 2006
```
llvm-svn: 28948
```
  a53a22e5
Jun 27, 2006
- handle the "mov reg1, reg2" case in isMoveInstr · f6f5aff0
  Rafael Espindola authored Jun 27, 2006
```
llvm-svn: 28945
```
  f6f5aff0
- Don't match 64-bit bitfield inserts into rlwimi's. todo add rldimi. :) · ca9c4885
  Chris Lattner authored Jun 27, 2006
```
llvm-svn: 28944
```
  ca9c4885
- Fix ppc64 jump tables · f882c545
  Chris Lattner authored Jun 27, 2006
```
llvm-svn: 28941
```
  f882c545
- Remove dead code. · 2aed9ebd
  Evan Cheng authored Jun 27, 2006
```
llvm-svn: 28938
```
  2aed9ebd
- Print stubs for external globals right. · 82ab3e21
  Chris Lattner authored Jun 27, 2006
```
llvm-svn: 28936
```
  82ab3e21
- Implement 64-bit select, bswap, etc. · 8aed3cc4
  Chris Lattner authored Jun 27, 2006
```
llvm-svn: 28935
```
  8aed3cc4
- Add a pattern for i64 sra. Print 8-byte units with a space between the .quad · a2af3f47
  Chris Lattner authored Jun 27, 2006
```
and the data

llvm-svn: 28934
```
  a2af3f47
- Fix rewriting frame offsets with ixaddr instructions, which implicitly shift · db9a95b7
  Chris Lattner authored Jun 27, 2006
```
the offset two bits to the left.

llvm-svn: 28933
```
  db9a95b7
- PPC doesn't have bit converts to/from i64 · a07410c9
  Chris Lattner authored Jun 27, 2006
```
llvm-svn: 28932
```
  a07410c9
- Add 64-bit MTCTR so that indirect calls work. · 3b587345
  Chris Lattner authored Jun 27, 2006
```
llvm-svn: 28931
```
  3b587345
- Fix an incorrect store pattern. This fixes em3d. · e27d51e0
  Chris Lattner authored Jun 27, 2006
```
llvm-svn: 28930
```
  e27d51e0
- Implement 64-bit undef, sub, shl/shr, srem/urem · d48ce275
  Chris Lattner authored Jun 27, 2006
```
llvm-svn: 28929
```
  d48ce275
- Use i32 for shift amounts instead of i64. This gets bisort working. · cb5a84f4
  Chris Lattner authored Jun 27, 2006
```
llvm-svn: 28927
```
  cb5a84f4
- Add zextload from i32 -> i64, with this, perimeter works. · f7fd8835
  Chris Lattner authored Jun 27, 2006
```
llvm-svn: 28926
```
  f7fd8835
- Print darwin stub stuff correctly in 64-bit mode. With this, treeadd works in · 1df08390
  Chris Lattner authored Jun 27, 2006
```
ppc64 mode!

llvm-svn: 28923
```
  1df08390
- Fix variable shadowing issue · 9a40cca4
  Chris Lattner authored Jun 27, 2006
```
llvm-svn: 28922
```
  9a40cca4
- Implement a bunch of 64-bit cleanliness work. With this, treeadd builds (but · 97b3da15
  Chris Lattner authored Jun 27, 2006
```
doesn't work right).

llvm-svn: 28921
```
  97b3da15
- Rearrange compares, add ADDI8, add sext from 32-to-64 bit register · 7ecbd301
  Chris Lattner authored Jun 26, 2006
```
llvm-svn: 28920
```
  7ecbd301
- Improve PPC64 calling convention support · ec78cade
  Chris Lattner authored Jun 26, 2006
```
llvm-svn: 28919
```
  ec78cade
- Remove two more definitions · b6a65f46
  Chris Lattner authored Jun 26, 2006
```
llvm-svn: 28918
```
  b6a65f46
- remove two unused instructions. · 86e60465
  Chris Lattner authored Jun 26, 2006
```
llvm-svn: 28917
```
  86e60465
Jun 24, 2006
- Simplify X86CompilationCallback: always align to 16-byte boundary; don't save... · 38c5aee9
  Evan Cheng authored Jun 24, 2006
```
Simplify X86CompilationCallback: always align to 16-byte boundary; don't save EAX/EDX if unnecessary.

llvm-svn: 28910
```
  38c5aee9
Jun 23, 2006
- Add and sort "sections" in debug lines. This always stepping through · a7b2bd59
  Jim Laskey authored Jun 23, 2006
```
code in sections other than ".text", including weak sections like ctors and
dtors.

llvm-svn: 28909
```
  a7b2bd59
Jun 22, 2006
- Eliminate unneeded parameter. · 0c9b90ab
  Evan Cheng authored Jun 22, 2006
```
llvm-svn: 28907
```
  0c9b90ab
- variable_ops instructions such as call can have any number of operands. · fc1b27da
  Evan Cheng authored Jun 21, 2006
```
llvm-svn: 28906
```
  fc1b27da
Jun 21, 2006
- Add memory operand and int regs · 680ac12e
  Andrew Lenharth authored Jun 21, 2006
```
llvm-svn: 28896
```
  680ac12e
- inline asm, at least for floats · b0316ead
  Andrew Lenharth authored Jun 21, 2006
```
llvm-svn: 28895
```
  b0316ead
- fix argument problem · 336313ce
  Andrew Lenharth authored Jun 21, 2006
```
llvm-svn: 28893
```
  336313ce
- Correct returns of 64-bit values, though they seemed to work before... · dc38e6f3
  Chris Lattner authored Jun 21, 2006
```
llvm-svn: 28892
```
  dc38e6f3
- Make these predicates correct in 64-bit mode too. · 1f1b0961
  Chris Lattner authored Jun 20, 2006
```
llvm-svn: 28890
```
  1f1b0961