RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2025-01-08 20:30:50 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity
27,148 Commits 4 Branches 5 Tags 634 MiB
1c299088c7
Commit Graph

27148 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Devang Patel
1c299088c7 Document removeModule(). 
llvm-svn: 31237
 2006-10-27 22:02:30 +00:00
Chris Lattner
838d0cbb0a this doesn't occur any more in mason 
llvm-svn: 31236
 2006-10-27 22:02:19 +00:00
Chris Lattner
6fe4efd730 the code in question is now: 
cmpw cr0, r7, r3
        ble cr0, LBB1_5 ;bb25
LBB1_8: ;bb17
        cmpw cr0, r8, r5
        bgt cr0, LBB1_2 ;bb

which is just as good as crnand.

llvm-svn: 31235
 2006-10-27 22:00:55 +00:00
Devang Patel
6ebc85d294 Document getAlignment() 
llvm-svn: 31234
 2006-10-27 21:58:31 +00:00
Chris Lattner
e2297cd5a6 remove debug code 
llvm-svn: 31233
 2006-10-27 21:58:03 +00:00
Chris Lattner
0fab97080c Codegen cond&cond with two branches. This compiles (f.e.) PowerPC/and-branch.ll to: 
cmpwi cr0, r4, 4
        bgt cr0, LBB1_2 ;UnifiedReturnBlock
LBB1_3: ;entry
        cmplwi cr0, r3, 0
        bne cr0, LBB1_2 ;UnifiedReturnBlock

instead of:

        cmpwi cr7, r4, 4
        mfcr r2
        addic r4, r3, -1
        subfe r3, r4, r3
        rlwinm r2, r2, 30, 31, 31
        or r2, r2, r3
        cmplwi cr0, r2, 0
        bne cr0, LBB1_2 ;UnifiedReturnBlock
LBB1_1: ;cond_true

llvm-svn: 31232
 2006-10-27 21:54:23 +00:00
Chris Lattner
2c97968c1c new testcase 
llvm-svn: 31231
 2006-10-27 21:52:39 +00:00
Chris Lattner
05159debc4 Turn conditions like x<Y|z==q into multiple blocks. 
This compiles Regression/CodeGen/X86/or-branch.ll into:

_foo:
        subl $12, %esp
        call L_bar$stub
        movl 20(%esp), %eax
        movl 16(%esp), %ecx
        cmpl $5, %eax
        jl LBB1_1       #cond_true
LBB1_3: #entry
        testl %ecx, %ecx
        jne LBB1_2      #UnifiedReturnBlock
LBB1_1: #cond_true
        call L_bar$stub
        addl $12, %esp
        ret
LBB1_2: #UnifiedReturnBlock
        addl $12, %esp
        ret

instead of:

_foo:
        subl $12, %esp
        call L_bar$stub
        movl 20(%esp), %eax
        movl 16(%esp), %ecx
        cmpl $4, %eax
        setg %al
        testl %ecx, %ecx
        setne %cl
        testb %cl, %al
        jne LBB1_2      #UnifiedReturnBlock
LBB1_1: #cond_true
        call L_bar$stub
        addl $12, %esp
        ret
LBB1_2: #UnifiedReturnBlock
        addl $12, %esp
        ret

And on ppc to:

        cmpwi cr0, r29, 5
        blt cr0, LBB1_1 ;cond_true
LBB1_3: ;entry
        cmplwi cr0, r30, 0
        bne cr0, LBB1_2 ;UnifiedReturnBlock

instead of:

        cmpwi cr7, r4, 4
        mfcr r2
        addic r4, r3, -1
        subfe r30, r4, r3
        rlwinm r29, r2, 30, 31, 31
        and r2, r29, r30
        cmplwi cr0, r2, 0
        bne cr0, LBB1_2 ;UnifiedReturnBlock

llvm-svn: 31230
 2006-10-27 21:36:01 +00:00
Chris Lattner
38e3f37f1c new testcase 
llvm-svn: 31229
 2006-10-27 21:32:10 +00:00
Evan Cheng
090e9abaee Fixed a significant bug where unpcklpd is incorrectly used to extract element 1 from a v2f64 value. 
llvm-svn: 31228
 2006-10-27 21:08:32 +00:00
Evan Cheng
8861459dfc Add a new vextract test case. 
llvm-svn: 31227
 2006-10-27 21:05:18 +00:00
Bill Wendling
dde9793f0c Forgot a } 
llvm-svn: 31226
 2006-10-27 20:22:04 +00:00
Bill Wendling
b8441e6cce Re-added the part where it tries to remove all global variables first. 
llvm-svn: 31225
 2006-10-27 20:18:06 +00:00
Jim Laskey
d3d001e012 SmallVector append not insert. 
llvm-svn: 31224
 2006-10-27 19:38:32 +00:00
Jim Laskey
904dfc49e6 Grrr. 
llvm-svn: 31223
 2006-10-27 19:20:12 +00:00
Jim Laskey
398ce23c4d Temp patch for missing functionality. 
llvm-svn: 31222
 2006-10-27 19:14:16 +00:00
Reid Spencer
dd294b507f Make the Value and Type methods print a newline so it prints nicely in gdb 
llvm-svn: 31221
 2006-10-27 18:58:54 +00:00
Evan Cheng
a1ce4523e5 Fix for PR968: expand vector sdiv, udiv, srem, urem. 
llvm-svn: 31220
 2006-10-27 18:49:08 +00:00
Bill Wendling
7b6f145c8e MathExtras isn't in the llvm/ADT directory but in the llvm/Support directory. 
llvm-svn: 31219
 2006-10-27 18:47:29 +00:00
Jim Laskey
3d1116e32c Apply editorials. 
llvm-svn: 31218
 2006-10-27 18:05:12 +00:00
Reid Spencer
f8a163421d Improve cvs ignoring of test results. 
llvm-svn: 31217
 2006-10-27 16:43:34 +00:00
Jim Laskey
69b57830c2 Breakout folding hash set from SelectionDAGCSEMap. 
llvm-svn: 31215
 2006-10-27 16:16:16 +00:00
Reid Spencer
79e316703d Initialize CStringSection member var. 
llvm-svn: 31214
 2006-10-27 16:14:06 +00:00
Reid Spencer
18fcfbb563 Add a run line to this test. 
llvm-svn: 31211
 2006-10-27 05:30:23 +00:00
Evan Cheng
9048e1010b Change load PatFrag to ignore indexed load. 
llvm-svn: 31210
 2006-10-26 21:55:50 +00:00
Evan Cheng
88c050f772 getPreIndexedLoad -> getIndexedLoad. 
llvm-svn: 31209
 2006-10-26 21:53:40 +00:00
Evan Cheng
d2d6b83b83 Indexed load / store changes. 
llvm-svn: 31208
 2006-10-26 21:52:24 +00:00
Evan Cheng
65e78f3d08 Place cstrings in .cstring section. 
llvm-svn: 31207
 2006-10-26 21:48:57 +00:00
Evan Cheng
a42e1492fd Speed up isCString() 
llvm-svn: 31206
 2006-10-26 21:48:03 +00:00
Devang Patel
ac5109eec8 Save temp. bc files when saveTemps flag is true. Use final output file 
name supplied by linker to construct temp bc file names.

Remove tabs.

llvm-svn: 31205
 2006-10-26 20:46:22 +00:00
Reid Spencer
4665cb220d Simplify code a bit by changing instances of: 
InsertNewInstBefore(new CastInst(Val, ValTy, Val->GetName()), I)
into:
   InsertCastBefore(Val, ValTy, I)

llvm-svn: 31204
 2006-10-26 19:19:06 +00:00
Evan Cheng
3144b839f0 Put cstrings in .cstring section when compiling for Mac OS X. 
llvm-svn: 31203
 2006-10-26 19:18:18 +00:00
Evan Cheng
5963dcb6e8 Added CStringSection. 
llvm-svn: 31202
 2006-10-26 19:16:20 +00:00
Evan Cheng
9d0bfecda6 Add isCString() - returns true if a ConstantArray is a CString. 
llvm-svn: 31201
 2006-10-26 19:15:05 +00:00
Chris Lattner
53cc483549 Fix Transforms/InstCombine/2006-10-26-VectorReassoc.ll 
llvm-svn: 31200
 2006-10-26 18:27:26 +00:00
Chris Lattner
8377ebcec3 new testcase, should not reassoc fp vector 
llvm-svn: 31199
 2006-10-26 18:26:53 +00:00
Chris Lattner
257998ec14 Add isFPOrFPVector() method, which indicates if a type is either FP or a 
vector of FP types.

llvm-svn: 31198
 2006-10-26 18:22:45 +00:00
Rafael Espindola
99322ef58c initial support for frame pointers 
llvm-svn: 31197
 2006-10-26 13:31:26 +00:00
Reid Spencer
b3409f52ef Enclose a case in { and } so that the pickier compilers don't complain. 
llvm-svn: 31196
 2006-10-26 06:17:40 +00:00
Reid Spencer
6833ffe8b8 For PR950: 
Make necessary changes to support DIV -> [SUF]Div. This changes llvm to
have three division instructions: signed, unsigned, floating point. The
bytecode and assembler are bacwards compatible, however.

llvm-svn: 31195
 2006-10-26 06:15:43 +00:00
Nick Lewycky
e7580b4a17 Fix 2006-10-25-AddSetCC. A relational operator (like setlt) can never 
produce an EQ property.

llvm-svn: 31193
 2006-10-26 02:35:18 +00:00
Nick Lewycky
312b457350 Handling of setlt/le/gt/ge is broken at the moment; predsimplify removes 
the call the pass2.

llvm-svn: 31192
 2006-10-26 02:33:51 +00:00
Reid Spencer
917bba3135 Make the makefile tell us when Intrinsics.gen is being updated. 
llvm-svn: 31191
 2006-10-26 01:42:23 +00:00
Nick Lewycky
5076440f74 This testcase was logically wrong! Fixing, and including the test for the 
mistake that I made in the file too.

llvm-svn: 31190
 2006-10-26 00:51:58 +00:00
Nick Lewycky
be9829c45f Resurrect r1.25. 
Fix and comment the "or", "and" and "xor" transformations.

llvm-svn: 31189
 2006-10-25 23:48:24 +00:00
Chris Lattner
2258020df2 simplify code 
llvm-svn: 31188
 2006-10-25 22:21:37 +00:00
Evan Cheng
ac5b12519b New shuffle test case. 
llvm-svn: 31187
 2006-10-25 21:59:00 +00:00
Evan Cheng
1abe8bd233 During vector shuffle lowering, we sometimes commute a vector shuffle to try 
to match MOVL (movss, movsd, etc.). Don't forget to commute it back and try
unpck* and shufp* if that doesn't pan out.

llvm-svn: 31186
 2006-10-25 21:49:50 +00:00
Evan Cheng
034305a2e8 X86ISD::PEXTRW 3rd operand type is always target pointer type. 
llvm-svn: 31185
 2006-10-25 21:35:05 +00:00
Chris Lattner
f4a5fcbb3a hide symbols properly 
llvm-svn: 31184
 2006-10-25 21:14:31 +00:00