RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-21 03:53:04 +02:00
Code Issues Projects Releases Wiki Activity
38,651 Commits 4 Branches 5 Tags 704 MiB
646f3458c4
Commit Graph

38651 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Bill Wendling
646f3458c4 Constify the machine instruction passed into the 
"is{Trivially,Really}ReMaterializable" methods.

llvm-svn: 51001
 2008-05-12 20:54:26 +00:00
Nate Begeman
2ae55cecc6 Initial X86 codegen support for VSETCC. 
llvm-svn: 51000
 2008-05-12 20:34:32 +00:00
Dale Johannesen
25896f8e93 Further rework of tail merge algorithm. Not quite 
semantically identical, but little difference in
either results or execution speed; but it's much
easier to read, at least IMO.

llvm-svn: 50999
 2008-05-12 20:33:57 +00:00
Nate Begeman
a3a9e89b61 Simplify some checks 
llvm-svn: 50998
 2008-05-12 20:33:52 +00:00
Dan Gohman
efa0925915 Fix a copy+paste bug; pseudo-instructions shouldn't have 
encoding information.

llvm-svn: 50997
 2008-05-12 20:22:45 +00:00
Nate Begeman
e9476924d9 Pointer comparisons should use icmp, not vicmp 
llvm-svn: 50996
 2008-05-12 20:16:55 +00:00
Owen Anderson
0256b368f8 Go back to passing the analyses around as parameters. 
llvm-svn: 50995
 2008-05-12 20:15:55 +00:00
Nate Begeman
b72a398339 Pointer comparisons should be handled by icmp, not vicmp :) 
llvm-svn: 50994
 2008-05-12 20:11:05 +00:00
Evan Cheng
228cce5b2d Forgot this. 
llvm-svn: 50993
 2008-05-12 20:08:05 +00:00
Nate Begeman
6258a36fe3 Hard code CmpInst back to i1 for now while I go track down what in the bitcode reader/writer is assuming i1 
This was breaking a bunch of tests

llvm-svn: 50992
 2008-05-12 20:01:56 +00:00
Evan Cheng
c7e9acfed7 Refactor isConsecutiveLoad from X86 to TargetLowering so DAG combiner can make use of it. 
llvm-svn: 50991
 2008-05-12 19:56:52 +00:00
Bill Wendling
f3c4b7fbb5 Revert the previous commit. Go ahead and hoist rematerializable instructions. 
llvm-svn: 50990
 2008-05-12 19:47:18 +00:00
Nate Begeman
11c0772a30 Add support for vicmp/vfcmp codegen, more legalize support coming. 
This is necessary to unbreak the build.

llvm-svn: 50988
 2008-05-12 19:40:03 +00:00
Bill Wendling
74cabebb44 One real change - don't hoist something that's trivially rematerializable. It's 
possible for it to produce worse code than before.

The rest of this patch is code cleanup.

llvm-svn: 50987
 2008-05-12 19:38:32 +00:00
Nate Begeman
e2b25610bb Fix build breakage 
llvm-svn: 50986
 2008-05-12 19:23:22 +00:00
Nate Begeman
b5b1e1353b Add two new instructions to the llvm IR, vicmp and vfcmp. see updated LangRef 
for details.  CodeGen support coming in a follow up patch

llvm-svn: 50985
 2008-05-12 19:01:56 +00:00
Dan Gohman
4224373cc8 Make firstEightPowers const. 
llvm-svn: 50975
 2008-05-12 16:38:14 +00:00
Dan Gohman
4f0d9d18c3 Update comments. 
llvm-svn: 50974
 2008-05-12 16:34:30 +00:00
Mikhail Glushenkov
18508c5df4 Filter option names to escape symbols not allowed as C++ identifiers. 
Makes it possible to use options with names like "Wa,".
Also fixes the -Wall option handling as a side-effect.

llvm-svn: 50973
 2008-05-12 16:33:06 +00:00
Mikhail Glushenkov
0cc6bb7a70 Make it possible to choose between different compilation graph definitions at compile-time. 
llvm-svn: 50972
 2008-05-12 16:32:24 +00:00
Mikhail Glushenkov
5005376b0c Fix some error messages; Make LLVMC pass through the exit code of a failed tool. 
llvm-svn: 50971
 2008-05-12 16:31:42 +00:00
Dan Gohman
8212eaa43a Fix a compile error on compilers that still want a return value 
in a non-void function that calls abort.

llvm-svn: 50969
 2008-05-12 16:17:19 +00:00
Dan Gohman
6fcb44e191 Fix a missing break in the ISD::FLT_ROUNDS_ handling. Patch by giuma! 
llvm-svn: 50967
 2008-05-12 16:07:15 +00:00
Duncan Sands
45b1810980 Testcase for PR2264. 
llvm-svn: 50965
 2008-05-12 13:01:19 +00:00
Owen Anderson
3f2d12126f Move the various analyses used by GVN into static variables so we don't have to keep passing them around or refetching them. 
llvm-svn: 50963
 2008-05-12 08:15:27 +00:00
Chris Lattner
0eb3dfb3b0 prune #includes. 
llvm-svn: 50962
 2008-05-12 01:13:53 +00:00
Chris Lattner
f459fcde61 Add a new SparsePropagation analysis utility, which allows you to do 
SCCP like sparse lattice analysis with relative ease.  Just pick your
lattice function and implement the transfer function and you're good.
Just make sure you don't break monotonicity ;-)

llvm-svn: 50961
 2008-05-12 01:12:24 +00:00
Nick Lewycky
6bc321fbdf Make constructors target-specific. This fixes problems where the path would 
include backslashes on Windows. This should fix llvm-ld problems on win32.

llvm-svn: 50960
 2008-05-11 17:37:40 +00:00
Anton Korobeynikov
ad83aeb489 Add note 
llvm-svn: 50959
 2008-05-11 14:33:15 +00:00
Chris Lattner
b700ffd977 Fix various DOUTs to not call the extremely expensive Value::getName() 
method.  DOUT statements are disabled when assertions are off, but the
side effects of getName() are still evaluated.  Just call getNameSTart,
which is close enough and doesn't cause heap traffic.

llvm-svn: 50958
 2008-05-11 01:55:59 +00:00
Chris Lattner
858c3299c1 Simplify code by using SwitchInst::findCaseValue instead of reimplementing it. 
llvm-svn: 50957
 2008-05-10 23:56:54 +00:00
Bill Wendling
cc084a6ba1 Remove warnings when using -Wshorten-64-to-32. 
llvm-svn: 50952
 2008-05-10 19:59:59 +00:00
Duncan Sands
15622620d3 Testcase for PR2303. 
llvm-svn: 50951
 2008-05-10 16:43:10 +00:00
Anton Korobeynikov
0f39a96feb Fix linking of internal aliases 
llvm-svn: 50950
 2008-05-10 14:41:43 +00:00
Bill Wendling
cc8838799d Prevent -W64-to-32-shortened warnings. 
llvm-svn: 50947
 2008-05-10 11:26:52 +00:00
Bill Wendling
adfb940f3a Prevent warnings from the -Wshorten-64-to-32 flag. 
llvm-svn: 50946
 2008-05-10 10:58:07 +00:00
Gabor Greif
724556e521 remove commented-out code, it is subsumed by DECLARE_TRANSPARENT_OPERAND_ACCESSORS 
llvm-svn: 50944
 2008-05-10 09:36:58 +00:00
Gabor Greif
49bf1a4cf6 merge of use-diet branch to trunk 
llvm-svn: 50943
 2008-05-10 08:32:32 +00:00
Owen Anderson
e34787f5c0 Re-enable loop deletion by default. 
llvm-svn: 50941
 2008-05-10 07:10:24 +00:00
Evan Cheng
c19c639ad7 When transforming a vector_shuffle to a load, the base address must not be an undef. 
llvm-svn: 50940
 2008-05-10 06:46:49 +00:00
Tanya Lattner
a449142a0a Set to 2.4 and regenerate configure. 
llvm-svn: 50935
 2008-05-10 04:20:38 +00:00
Evan Cheng
fc4b8e1d96 Add nounwind. 
llvm-svn: 50931
 2008-05-10 02:22:25 +00:00
Dan Gohman
4b23d9e60a For now, abort when an ISD::VAARG is encountered on x86-64, rather 
than silently generate invalid code.

llvm-gcc does not currently use VAArgInst; it lowers va_arg in the
front-end.

llvm-svn: 50930
 2008-05-10 01:26:14 +00:00
Evan Cheng
6a3fa28b38 Some clean up. 
llvm-svn: 50929
 2008-05-10 00:59:18 +00:00
Evan Cheng
79230955a8 If movl top bits are undef, let it be selected to movlps, etc. 
llvm-svn: 50928
 2008-05-10 00:58:41 +00:00
Evan Cheng
cf4d5567d5 If all sources of a PHI node are defined by an implicit_def, just emit an implicit_def instead of a copy. 
llvm-svn: 50927
 2008-05-10 00:17:50 +00:00
Bill Wendling
89c91f81ae Cosmetic changes: 
- Comment fixes.
 - Moar whitespace.
 - Made ivars "private" by default.
No functionality change.

llvm-svn: 50926
 2008-05-10 00:12:52 +00:00
Evan Cheng
2adea48f7e Add a pattern to do move the low element of a v4f32 and zero extend the rest. 
llvm-svn: 50922
 2008-05-09 23:37:55 +00:00
Dale Johannesen
a422bcbec2 Remove an evil vector bool. Cosmetic refactoring, 
no functional change.

llvm-svn: 50921
 2008-05-09 23:28:24 +00:00
Evan Cheng
3493e43afd Handle a few more cases of folding load i64 into xmm and zero top bits. 
Note, some of the code will be moved into target independent part of DAG combiner in a subsequent patch.

llvm-svn: 50918
 2008-05-09 21:53:03 +00:00