fbeeb130d8
Revert r111375, "move gep decomposition out of ValueTracking into BasicAA. The
...
form of", it doesn't pass tests.
llvm-svn: 111385
2010-08-18 18:43:08 +00:00
0d323aef46
Improve whitespace.
...
llvm-svn: 111384
2010-08-18 18:41:13 +00:00
817e857b13
Marked with ATTRIBUTE_USED so that clang doesn't complain.
...
llvm-svn: 111383
2010-08-18 18:40:57 +00:00
208636fa33
Inform LazyValueInfo whenever a block is deleted, to avoid dangling pointer issues.
...
llvm-svn: 111382
2010-08-18 18:39:01 +00:00
8e92d9b68d
MC/ELF: Allow null values in virtual sections, ELF doesn't use special
...
directives for putting contents in .bss, for example.
llvm-svn: 111376
2010-08-18 18:22:37 +00:00
54fe883203
move gep decomposition out of ValueTracking into BasicAA. The form of
...
decomposition that it is doing is very basicaa specific and is only used
by basicaa.
llvm-svn: 111375
2010-08-18 18:22:17 +00:00
e0e9b3013f
Add hook for re-using virtual base registers for local stack slot access.
...
Nothing fancy, just ask the target if any currently available base reg
is in range for the instruction under consideration and use the first one
that is. Placeholder ARM implementation simply returns false for now.
ongoing saga of rdar://8277890
llvm-svn: 111374
2010-08-18 17:57:37 +00:00
7ba6f22b47
Eliminate some redundancy by relying on raw_fd_ostream to handle "-"
...
properly.
llvm-svn: 111373
2010-08-18 17:55:15 +00:00
083330a8f2
Allow the -analyze option to follow the -o option, which defaults to
...
standard output, instead of just hardcoding outs().
llvm-svn: 111372
2010-08-18 17:42:59 +00:00
b01aed1cb2
Don't translate "-" to outs() manually; raw_ostream does that automatically.
...
llvm-svn: 111371
2010-08-18 17:40:10 +00:00
80f9de4bb5
Don't translate "-" to outs() manually; raw_ostream does that
...
automatically.
llvm-svn: 111370
2010-08-18 17:26:50 +00:00
6111d7c05b
Don't register stdout to be deleted on a signal.
...
llvm-svn: 111368
2010-08-18 17:21:42 +00:00
3d051ff2bc
Remove an unused command-line option.
...
llvm-svn: 111367
2010-08-18 17:20:51 +00:00
952a621d93
Preserve subregs on PHI source operands. Patch by Krister Wombell!
...
llvm-svn: 111366
2010-08-18 16:09:47 +00:00
e60b5161d1
Fix a bug with insertelement on SPU.
...
The previous algorithm in LowerVECTOR_SHUFFLE
didn't check all requirements for "monotonic" shuffles.
llvm-svn: 111361
2010-08-18 10:20:29 +00:00
ab49360f59
Remove all traces of v2[i,f]32 on SPU.
...
The "half vectors" are now widened to full size by the legalizer.
The only exception is in parameter passing, where half vectors are
expanded. This causes changes to some dejagnu tests.
llvm-svn: 111360
2010-08-18 10:04:39 +00:00
f3984d1ef6
Change SPU C calling convention to match that described in
...
"SPU Application Binary Interface Specification, v1.9" by
IBM.
Specifically: use r3-r74 to pass parameters and the return value.
llvm-svn: 111358
2010-08-18 09:50:30 +00:00
a25c05ed15
fix a buggy test
...
llvm-svn: 111354
2010-08-18 04:55:12 +00:00
a33edcb56c
fix PR7589: In brief:
...
gep P, (zext x) != gep P, (sext x)
DecomposeGEPExpression was getting this wrong, confusing
basicaa.
llvm-svn: 111352
2010-08-18 04:28:19 +00:00
c8e38eb60b
filecheckize and detrivialize.
...
llvm-svn: 111350
2010-08-18 04:25:43 +00:00
3c603024bb
Fix PR7755: knowing something about an inval for a pred
...
from the LHS should disable reconsidering that pred on the
RHS. However, knowing something about the pred on the RHS
shouldn't disable subsequent additions on the RHS from
happening.
llvm-svn: 111349
2010-08-18 03:14:36 +00:00
f0b5b67ba5
fit in 80 cols
...
llvm-svn: 111348
2010-08-18 03:13:35 +00:00
2edfdd289f
remove some dead code.
...
llvm-svn: 111345
2010-08-18 02:42:11 +00:00
b45de95345
remove some dead code.
...
llvm-svn: 111344
2010-08-18 02:41:56 +00:00
3e3e63efe1
remove some code that is dead now that lea's are modeled with segment registers.
...
llvm-svn: 111343
2010-08-18 02:40:44 +00:00
6aabb66139
remove dead prototype.
...
llvm-svn: 111342
2010-08-18 02:37:06 +00:00
fb7eaff759
Expand ZERO_EXTEND operations for NEON vector types.
...
Testcase from Nick Lewycky.
llvm-svn: 111341
2010-08-18 01:45:52 +00:00
a610304ed9
Fix the "Finished Creating" messages for aliases to print the
...
right name.
llvm-svn: 111340
2010-08-18 01:35:53 +00:00
44790e7aa7
Revert r111321. This doesn't fix a problem.
...
llvm-svn: 111339
2010-08-18 01:34:52 +00:00
b91c903f46
stomp some more undefined behavior, PR7775.
...
llvm-svn: 111337
2010-08-18 00:33:47 +00:00
816191bcad
don't emit zero bit fields with Emit, fixing undefined behavior,
...
PR7778
llvm-svn: 111336
2010-08-18 00:29:18 +00:00
03e17f8387
Don't pass in a null pointer to std::string's ctor, an empty string
...
ref should produce an empty std::string. This fixes PR7879.
llvm-svn: 111332
2010-08-18 00:11:25 +00:00
249b976ed2
fix a broken link
...
llvm-svn: 111326
2010-08-17 23:26:04 +00:00
7aa9eb16fd
include config.h to get config params, hopefully unbreaking mingw builder.
...
llvm-svn: 111325
2010-08-17 23:22:10 +00:00
ce3b2c3f77
Fix the rest of rdar://8318441 which happens when a raw_fd_ostream
...
(e.g. errs()) fails in close() due to (e.g.) a broken pipe. As
previously written, the had_error() flag would get set and then
the raw_ostream dtor would report a fatal error. There is nothing
the client can do about this and we have no way to report the error,
so just eat it.
llvm-svn: 111321
2010-08-17 23:11:56 +00:00
6217082dc3
report_fatal_error can't use errs(), because errs() can call
...
into report_fatal_error. Just blast the string to stderr with write(2)
and hope for the best! Part of rdar://8318441
llvm-svn: 111320
2010-08-17 23:03:53 +00:00
51edc7b7e1
Temporarily revert r110987 as it's causing some miscompares in
...
vector heavy code. I'll re-enable when we've tracked down the problem.
llvm-svn: 111318
2010-08-17 22:55:27 +00:00
ed2b005842
Tweak IVUsers' concept of "interesting" to exclude add recurrences
...
where the step value is an induction variable from an outer loop, to
avoid trouble trying to re-expand such expressions. This effectively
hides such expressions from indvars and lsr, which prevents them
from getting into trouble.
llvm-svn: 111317
2010-08-17 22:50:37 +00:00
3cf08661f4
Add materialization of virtual base registers for frame indices allocated into
...
the local block. Resolve references to those indices to a new base register.
For simplification and testing purposes, a new virtual base register is
allocated for each frame index being resolved. The result is truly horrible,
but correct, code that's good for exercising the new code paths.
Next up is adding thumb1 support, which should be very simple. Following that
will be adding base register re-use and implementing a reasonable ARM
heuristic for when a virtual base register should be generated at all.
llvm-svn: 111315
2010-08-17 22:41:55 +00:00
efdc74ea59
Add nounwind.
...
llvm-svn: 111312
2010-08-17 22:35:20 +00:00
09b0c78918
CrashRecovery: Clear the current context on the first crash, to avoid re-entering it if the cleanup code crashes.
...
llvm-svn: 111309
2010-08-17 22:32:39 +00:00
b30266ed20
CrashRecovery: Add CrashRecoveryContext::GetCurrent(), so clients can find the active context from anywhere.
...
llvm-svn: 111308
2010-08-17 22:32:37 +00:00
ff329942cd
CrashRecovery: Make CrashRecoveryContext static methods thread safe.
...
llvm-svn: 111307
2010-08-17 22:32:34 +00:00
16f96445c3
Make fast scheduler handle asm clobbers correctly.
...
PR 7882. Follows suggestion by Amaury Pouly, thanks.
llvm-svn: 111306
2010-08-17 22:17:24 +00:00
88c09879c7
Revert part of one of the prev. patches - tailjmp will follow later.
...
llvm-svn: 111291
2010-08-17 21:08:28 +00:00
231ab847ca
More fixes for win64:
...
- Do not clobber al during variadic calls, this is AMD64 ABI-only feature
- Emit wincall64, where necessary
Patch by Cameron Esfahani!
llvm-svn: 111289
2010-08-17 21:06:07 +00:00
cd78af6e3c
Enable more win64 calls folding opportunities.
...
Patch by Cameron Esfahani!
llvm-svn: 111288
2010-08-17 21:06:01 +00:00
14be4dff8e
Add some win64 coff goodness.
...
Patch by Cameron Esfahani!
llvm-svn: 111287
2010-08-17 21:05:54 +00:00
16bfe5b0f5
PHI elimination shouldn't require machineloopinfo since it's used at -O0. Move the requirement to LiveIntervalAnalysis instead. Note this does not change the number of times machineloopinfo is computed.
...
llvm-svn: 111285
2010-08-17 21:00:37 +00:00
e0db9d01d9
Machine CSE preserves CFG. Pass manager was freeing machineloopinfo after machine cse before.
...
llvm-svn: 111281
2010-08-17 20:57:42 +00:00
Daniel Dunbar