RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2025-02-10 06:25:01 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
27,007 Commits 6 Branches 6 Tags
Commit Graph

27007 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Evan Cheng
2113bdf34d Reflect MemOpAddrMode change; added a helper to create pre-indexed load. 
llvm-svn: 31016
 2006-10-17 21:14:32 +00:00
Evan Cheng
582d22a821 Split PRE_INDEXED to PRE_INC / PRE_DEC and similarly for POST_INDEXED. 
llvm-svn: 31015
 2006-10-17 21:12:56 +00:00
Rafael Espindola
58233f9db2 expand ISD::SDIV, ISD::UDIV, ISD::SREM and ISD::UREM 
llvm-svn: 31014
 2006-10-17 21:05:33 +00:00
Rafael Espindola
4e8824608e add the FPUnaryOp and DFPUnaryOp classes 
llvm-svn: 31013
 2006-10-17 20:45:22 +00:00
Rafael Espindola
fe0a9a6fe2 add FABSS and FABSD 
llvm-svn: 31012
 2006-10-17 20:33:13 +00:00
Anton Korobeynikov
c86485a397 Adding linkonce linkage codegeneration support for mingw32\cygwin 
targets.

llvm-svn: 31011
 2006-10-17 20:29:49 +00:00
Rafael Espindola
a8f8a3ef26 cleanup some tests 
llvm-svn: 31010
 2006-10-17 20:20:07 +00:00
Jim Laskey
7f16ca4872 Make it simplier to dump DAGs while in DAGCombiner. Remove a nasty optimization. 
llvm-svn: 31009
 2006-10-17 19:33:52 +00:00
Rafael Espindola
64f93033bc remove extra [] in stores 
llvm-svn: 31008
 2006-10-17 18:29:14 +00:00
Chris Lattner
ef7c2e1dc8 Enable deleting branches to successor blocks. With the previous patches, 
branch folding can now compile stuff like this:

void foo(int W, int X, int Y, int Z) {
  if (W & 1) {
    for (; X;--X) bar();
  } else if (W & 2) {
    for (; Y;--Y) bar();
  } else if (W & 4) {
    for (; Z;--Z) bar();
  } else if (W & 8) {
    for (; W;--W) bar();
  }

  if (W) {
    bar();
  }
}

contrived testcase where loops exits all end up merging together.  To have
the loop merges be:

...
        cmplw cr0, r30, r27
        bne cr0, LBB1_14        ;bb38
LBB1_16:        ;cond_next48.loopexit
        mr r27, r29
LBB1_20:        ;cond_next48
        cmplwi cr0, r27, 0
        beq cr0, LBB1_22        ;UnifiedReturnBlock
...

instead of:


...
        cmplw cr0, r30, r27
        bne cr0, LBB1_14        ;bb38
LBB1_16:        ;cond_next48.loopexit
        mr r27, r29
        b LBB1_20       ;cond_next48
LBB1_17:        ;cond_next48.loopexit1
        b LBB1_20       ;cond_next48
LBB1_18:        ;cond_next48.loopexit2
        b LBB1_20       ;cond_next48
LBB1_19:        ;cond_next48.loopexit3
LBB1_20:        ;cond_next48
        cmplwi cr0, r27, 0
        beq cr0, LBB1_22        ;UnifiedReturnBlock
...


This is CodeGen/PowerPC/branch-opt.ll

llvm-svn: 31006
 2006-10-17 18:16:40 +00:00
Chris Lattner
d5554a07c0 new testcase 
llvm-svn: 31005
 2006-10-17 18:14:39 +00:00
Chris Lattner
bf15d2f503 add support for inserting an uncond branch 
llvm-svn: 31003
 2006-10-17 18:06:55 +00:00
Rafael Espindola
47970f96ac initial implementation of addressing mode 5 
llvm-svn: 31002
 2006-10-17 18:04:53 +00:00
Jim Laskey
37679253ae Clean up interface to getGlobalLinkName. 
llvm-svn: 31001
 2006-10-17 17:17:24 +00:00
Chris Lattner
8464c7e3e1 Reenable this pass, fixing the bugs in it. 
It now correctly deletes unreachable blocks and blocks that are empty.

llvm-svn: 31000
 2006-10-17 17:13:52 +00:00
Evan Cheng
d9bec725a2 Make sure operand does have size and element type operands. 
llvm-svn: 30999
 2006-10-17 17:06:35 +00:00
Rafael Espindola
d5a6eaec14 add the immediate to the Offset in eliminateFrameIndex 
llvm-svn: 30998
 2006-10-17 14:34:02 +00:00
Jim Laskey
620df3ea5f Basic support for getGlobalLinkName. 
llvm-svn: 30997
 2006-10-17 13:41:07 +00:00
Rafael Espindola
31f59f5b94 add FSTD and FSTS 
llvm-svn: 30996
 2006-10-17 13:36:07 +00:00
Rafael Espindola
01400015fc add FCPYS and FCPYD 
llvm-svn: 30995
 2006-10-17 13:13:23 +00:00
Jim Laskey
d266d999cb Use S_debug for dwarf info. 
llvm-svn: 30994
 2006-10-17 11:30:57 +00:00
Evan Cheng
bf3e5a1d7a Proper fix. 
llvm-svn: 30993
 2006-10-17 00:24:49 +00:00
Evan Cheng
36ab7976a0 One more try. 
llvm-svn: 30992
 2006-10-16 23:44:08 +00:00
Reid Spencer
c6aa794a41 Undo Chris' last patch, it caused a regression. 
llvm-svn: 30991
 2006-10-16 23:08:08 +00:00
Evan Cheng
1cf362ec18 Unbreak x86-64 build. 
llvm-svn: 30990
 2006-10-16 22:53:28 +00:00
Evan Cheng
2d9318cff1 Be careful when looking through a vbit_convert. Optimizing this: 
(vector_shuffle
  (vbitconvert (vbuildvector (copyfromreg v4f32), 1, v4f32), 4, f32),
  (undef, undef, undef, undef), (0, 0, 0, 0), 4, f32)
to the
  vbitconvert
is a very bad idea.

llvm-svn: 30989
 2006-10-16 22:49:37 +00:00
Rafael Espindola
a156538e34 add fdivs e fdivd 
llvm-svn: 30988
 2006-10-16 21:50:04 +00:00
Rafael Espindola
4f61431679 expand ISD::SHL_PARTS, ISD::SRA_PARTS and ISD::SRL_PARTS 
llvm-svn: 30987
 2006-10-16 21:10:32 +00:00
Evan Cheng
04bec886a9 Added a X86CompilationCallback variant which saves XMM argument registers for targets with SSE. 
llvm-svn: 30986
 2006-10-16 21:01:55 +00:00
Evan Cheng
b4d8323e54 Proper fix for rdar://problem/4770604 Thanks to Stuart Hastings! 
llvm-svn: 30985
 2006-10-16 21:00:37 +00:00
Jim Laskey
06f4428abc Pass AliasAnalysis thru to DAGCombiner. 
llvm-svn: 30984
 2006-10-16 20:52:31 +00:00
Chris Lattner
3bfbb5eabd add a ReplaceMBBInJumpTables method 
llvm-svn: 30983
 2006-10-16 20:41:42 +00:00
Jim Laskey
5fe261bc52 Global name regression. 
llvm-svn: 30982
 2006-10-16 19:38:41 +00:00
Rafael Espindola
25fbeb5ec2 define the DFPBinOp class 
llvm-svn: 30981
 2006-10-16 18:39:22 +00:00
Rafael Espindola
57dc40a3a8 add the FPBinOp class 
llvm-svn: 30980
 2006-10-16 18:32:36 +00:00
Rafael Espindola
8a40de94f2 define the Addr1BinOp class 
llvm-svn: 30979
 2006-10-16 18:18:14 +00:00
Rafael Espindola
6554b1014d define the IntBinOp class and use it to implement the multiply instructions 
llvm-svn: 30978
 2006-10-16 17:57:20 +00:00
Rafael Espindola
71ae8c3d4a fix assembly syntax 
llvm-svn: 30977
 2006-10-16 17:38:12 +00:00
Rafael Espindola
d255bfb09b implement LDRB, LDRSB, LDRH and LDRSH 
llvm-svn: 30976
 2006-10-16 17:17:22 +00:00
Rafael Espindola
93359fa883 implement smull and umull 
llvm-svn: 30975
 2006-10-16 16:33:29 +00:00
Evan Cheng
ca5eaf4020 Avoid getting into an infinite loop when -disable-x86-shuffle-opti is specified. 
llvm-svn: 30974
 2006-10-16 06:36:00 +00:00
Evan Cheng
c73547a71d SelectScalarSSELoad should call CanBeFoldedBy as well. 
llvm-svn: 30973
 2006-10-16 06:34:55 +00:00
Evan Cheng
f872538d4b Passing isel root and use operands to ComplexPattern functions, these should do the usual load folding checks as well. 
llvm-svn: 30972
 2006-10-16 06:33:44 +00:00
Chris Lattner
9ecd9bcde4 regenerate 
llvm-svn: 30971
 2006-10-15 23:27:25 +00:00
Chris Lattner
2ac6ca0f82 Produce a useful error message for Regression/Verifier/2006-10-15-AddrLabel.ll 
llvm-svn: 30970
 2006-10-15 23:26:46 +00:00
Chris Lattner
a5383a5b2d testcase that causes the asmparser to assert 
llvm-svn: 30969
 2006-10-15 23:26:28 +00:00
Chris Lattner
5a4d067b26 label* is not a valid type, add an assertion. 
llvm-svn: 30968
 2006-10-15 23:21:12 +00:00
Chris Lattner
fd983f91e7 fix a buggy check that accidentally disabled this xform 
llvm-svn: 30967
 2006-10-15 22:42:15 +00:00
Chris Lattner
99ce7667fb clarify 
llvm-svn: 30966
 2006-10-15 20:05:59 +00:00
Reid Spencer
88ad34206e Document the VAArg, and vector instruction opcodes. 
llvm-svn: 30965
 2006-10-15 00:11:05 +00:00