OSDN Git Service
Chris Lattner [Tue, 19 Feb 2008 17:33:52 +0000 (17:33 +0000)]
rename tests to avoid a test- prefix when they aren't related to the test instruction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47329
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Tue, 19 Feb 2008 09:28:48 +0000 (09:28 +0000)]
Fix a compiler warning.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47328
91177308-0d34-0410-b5e6-
96231b3b80d8
Tanya Lattner [Tue, 19 Feb 2008 08:07:33 +0000 (08:07 +0000)]
Remove llvm-upgrade and update tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47325
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 19 Feb 2008 07:58:11 +0000 (07:58 +0000)]
dead pass
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47324
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 19 Feb 2008 07:49:17 +0000 (07:49 +0000)]
remove the LowerSelect pass. The last client was the old Sparc backend, which is long dead by now.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47323
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 19 Feb 2008 07:39:17 +0000 (07:39 +0000)]
remove the lower packed pass. It can never work and even the parts that
could work don't work fully. This fixes PR1705. Oh yeah, we don't have
packed types anymore either ;-)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47322
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Tue, 19 Feb 2008 07:07:51 +0000 (07:07 +0000)]
Refactor this method a bit, and correct a test that was completely wrong but happened to work out anyways. :-)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47321
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 19 Feb 2008 06:53:20 +0000 (06:53 +0000)]
isa+cast -> dyncast.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47320
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 19 Feb 2008 06:52:38 +0000 (06:52 +0000)]
simplify this code again, try 2 :)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47319
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Tue, 19 Feb 2008 06:51:23 +0000 (06:51 +0000)]
Fix a comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47318
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Tue, 19 Feb 2008 06:47:18 +0000 (06:47 +0000)]
In addition to arguments passed to it, memcpy (and all other calls) can ModRef pointers that alias their arguments as well. This fixes PR2057.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47317
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Tue, 19 Feb 2008 06:35:43 +0000 (06:35 +0000)]
Major improvements to yesterday's return slot optimization. Remove some unneccessary constraints,
and add some others that should have been in from the first place. Document the whole thing better.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47315
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 19 Feb 2008 06:22:12 +0000 (06:22 +0000)]
Fix some minor issues folding undef, PR2052
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47314
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 19 Feb 2008 06:12:18 +0000 (06:12 +0000)]
fdiv/frem of undef can produce undef, because the undef operand
can be a SNaN. We could be more aggressive and turn this into
unreachable, but that is less nice, and not really worth it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47313
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 19 Feb 2008 04:36:25 +0000 (04:36 +0000)]
regenerate
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47312
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Tue, 19 Feb 2008 04:36:07 +0000 (04:36 +0000)]
Fix PR2060 by rejecting invalid types for integer constants.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47311
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Tue, 19 Feb 2008 03:27:34 +0000 (03:27 +0000)]
Factor the profitability check for return slot optimization out into a static function.
At some point in the future, this check will become smarter.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47310
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Tue, 19 Feb 2008 03:15:29 +0000 (03:15 +0000)]
An sret parameter is required to be the first parameter, so there's no need to loop
over all the parameters of the callee looking for it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47309
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Tue, 19 Feb 2008 03:11:47 +0000 (03:11 +0000)]
Don't spew stats to stderr.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47308
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Tue, 19 Feb 2008 03:09:45 +0000 (03:09 +0000)]
Cleanup some of my patches from yesterday. Refactor the check for which xform
to apply to a memcpy into processInstruction. Also, fix a bug in the check due to
missing braces.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47307
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Tue, 19 Feb 2008 02:58:36 +0000 (02:58 +0000)]
Fix up the run line for this new test.
llc: for the -info-output-file option: requires a value!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47306
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Tue, 19 Feb 2008 02:53:23 +0000 (02:53 +0000)]
Fix Transforms/GVN/memcpy.ll, which Chris broke in r47275 by reordering the branches. memcpy's are a kind of CallInst.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47305
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Tue, 19 Feb 2008 02:09:58 +0000 (02:09 +0000)]
New test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47302
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Tue, 19 Feb 2008 02:09:37 +0000 (02:09 +0000)]
PR1909: Tail merging pass ran wild. It makes no sense to merge blocks in order to save a single instruction since a branch will be inserted for each BB.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47301
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Tue, 19 Feb 2008 02:05:16 +0000 (02:05 +0000)]
Me not like duplicated comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47300
91177308-0d34-0410-b5e6-
96231b3b80d8
Tanya Lattner [Tue, 19 Feb 2008 01:44:26 +0000 (01:44 +0000)]
Remove llvm-upgrade and update tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47297
91177308-0d34-0410-b5e6-
96231b3b80d8
Tanya Lattner [Tue, 19 Feb 2008 01:41:04 +0000 (01:41 +0000)]
Remove llvm-upgrade and update tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47296
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Mon, 18 Feb 2008 23:04:32 +0000 (23:04 +0000)]
- When DAG combiner is folding a bit convert into a BUILD_VECTOR, it should check if it's essentially a SCALAR_TO_VECTOR. Avoid turning (v8i16) <10, u, u, u> to <10, 0, u, u, u, u, u, u>. Instead, simply convert it to a SCALAR_TO_VECTOR of the proper type.
- X86 now normalize SCALAR_TO_VECTOR to (BIT_CONVERT (v4i32 SCALAR_TO_VECTOR)). Get rid of X86ISD::S2VEC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47290
91177308-0d34-0410-b5e6-
96231b3b80d8
Nick Lewycky [Mon, 18 Feb 2008 22:48:05 +0000 (22:48 +0000)]
Correctly fold divide-by-constant, even when faced with overflow.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47287
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Mon, 18 Feb 2008 19:34:53 +0000 (19:34 +0000)]
Chris pointed out that it's not necessary to set i64 MUL to Expand
on x86-32 since i64 itself is not a Legal type. And, update some
comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47282
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Mon, 18 Feb 2008 18:56:31 +0000 (18:56 +0000)]
- Remove the previous check which broke coalescer-commute3.ll
- For now, conservatively ignore copy MI whose source is a physical register. Commuting its def MI can cause a physical register live interval to be live through a loop (since we know it's live coming into the def MI).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47281
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 18 Feb 2008 18:46:39 +0000 (18:46 +0000)]
upgrade some tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47280
91177308-0d34-0410-b5e6-
96231b3b80d8
Nate Begeman [Mon, 18 Feb 2008 18:39:23 +0000 (18:39 +0000)]
Add a note
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47279
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 18 Feb 2008 18:30:13 +0000 (18:30 +0000)]
Add a note about sext from i1 plus flags use.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47278
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Mon, 18 Feb 2008 17:55:26 +0000 (17:55 +0000)]
Don't mark scalar integer multiplication as Expand on x86, since x86
has plain one-result scalar integer multiplication instructions.
This avoids expanding such instructions into MUL_LOHI sequences that
must be special-cased at isel time, and avoids the problem with that
code that provented memory operands from being folded.
This fixes PR1874, addressesing the most common case. The uncommon
cases of optimizing multiply-high operations will require work
in DAGCombiner.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47277
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 18 Feb 2008 17:50:16 +0000 (17:50 +0000)]
Transforming -A + -B --> -(A + B) isn't safe for FP, thanks
to Dale for noticing this!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47276
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 18 Feb 2008 17:47:29 +0000 (17:47 +0000)]
minor code simplification, no functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47275
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 18 Feb 2008 17:33:10 +0000 (17:33 +0000)]
make this just a bit more strict.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47274
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Mon, 18 Feb 2008 17:32:13 +0000 (17:32 +0000)]
Simplify caller updating using a CallSite, as
requested by Chris. While there, do the same
for an existing function committed by someone
called "lattner" :)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47273
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 18 Feb 2008 17:28:21 +0000 (17:28 +0000)]
don't bother calling getUnderlyingObject for non-pointers.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47272
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Mon, 18 Feb 2008 17:15:45 +0000 (17:15 +0000)]
Fix a missing space in the description of the find-bugs option.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47271
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Mon, 18 Feb 2008 10:11:00 +0000 (10:11 +0000)]
Since we're not checking for the more general AllocationInst first, we need to explicitly check
that Object is an Argument before casting it to one.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47268
91177308-0d34-0410-b5e6-
96231b3b80d8
Roman Levenstein [Mon, 18 Feb 2008 09:35:30 +0000 (09:35 +0000)]
New helper function getMBBFromIndex() that given an index in any instruction of an MBB returns a pointer the MBB. Reviewed by Evan.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47267
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Mon, 18 Feb 2008 09:24:53 +0000 (09:24 +0000)]
Add support to GVN for performing sret return slot optimization. This means that, if an sret function tail calls
another sret function, it should pass its own sret parameter to the tail callee, allowing it to fill in the correct
return value. llvm-gcc does not emit this by default. Instead, it allocates space in the caller for the sret of
the tail call and then uses memcpy to copy the result into the caller's sret parameter. This optimization detects
and optimizes that case.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47265
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Mon, 18 Feb 2008 09:22:21 +0000 (09:22 +0000)]
I got the predicate backwards in my last patch. The comment is correct, the code was not.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47264
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Mon, 18 Feb 2008 09:11:02 +0000 (09:11 +0000)]
This check is not correct for mallocs, so exclude them earlier.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47263
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Mon, 18 Feb 2008 08:40:53 +0000 (08:40 +0000)]
For now, avoid commuting def MI for copy MI's whose source is not killed. That simply trade a live interval for another and because only the non-two-address operands can be folded into loads, may end up pessimising code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47262
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 18 Feb 2008 07:42:56 +0000 (07:42 +0000)]
switch simplifycfg from using vectors for most things to smallvectors,
this speeds it up 2.3% on eon.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47261
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 18 Feb 2008 06:12:38 +0000 (06:12 +0000)]
optimize away stackrestore calls that have no intervening alloca or call.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47258
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 18 Feb 2008 06:11:00 +0000 (06:11 +0000)]
upgrade this test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47257
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Mon, 18 Feb 2008 04:19:38 +0000 (04:19 +0000)]
Clarify that 'sret' only applies to pointers, and
only applies to the first parameter.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47256
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Mon, 18 Feb 2008 04:09:01 +0000 (04:09 +0000)]
Make the definition of the noalias attribute clearer.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47255
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Mon, 18 Feb 2008 04:06:26 +0000 (04:06 +0000)]
Duncan pointed out that we can fast fail here, because the sret parameter of
a function must be the first parameter.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47254
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Mon, 18 Feb 2008 03:52:21 +0000 (03:52 +0000)]
Fix a comment, and a bug where we weren't applying the tail call logic in cases that failed the first test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47253
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Mon, 18 Feb 2008 02:31:23 +0000 (02:31 +0000)]
Fix bugs that Chris noticed in my last patch.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47252
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 18 Feb 2008 02:18:25 +0000 (02:18 +0000)]
simplify some code, BreakUpSubtract always returns nonnull now.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47251
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Mon, 18 Feb 2008 02:11:28 +0000 (02:11 +0000)]
bitcasts of pointers are always pointers.
If we see a memcpy of a pointer, make sure to check later
uses of the pointer as well.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47250
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Mon, 18 Feb 2008 00:10:55 +0000 (00:10 +0000)]
Add support for setting parameters to CallSite.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47249
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Sun, 17 Feb 2008 23:22:28 +0000 (23:22 +0000)]
Add a predicate to Argument to check for the StructRet attribute.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47248
91177308-0d34-0410-b5e6-
96231b3b80d8
Owen Anderson [Sun, 17 Feb 2008 21:29:08 +0000 (21:29 +0000)]
Teach getModRefInfo that memcpy, memmove, and memset don't "capture" memory addresses.
Also, noalias arguments are be considered "like" stack allocated ones for this purpose, because
the only way they can be modref'ed is if they escape somewhere in the current function.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47247
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 17 Feb 2008 21:03:36 +0000 (21:03 +0000)]
Fold (-x + -y) -> -(x+y) which promotes better association, fixing
the second half of PR2047
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47244
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 17 Feb 2008 20:54:40 +0000 (20:54 +0000)]
fix pasto
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47242
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 17 Feb 2008 20:51:26 +0000 (20:51 +0000)]
Split up subtracts into add+negate if they have a reassociable use or operand
that is also a subtract. This implements PR2047 and Transforms/Reassociate/subtest2.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47241
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 17 Feb 2008 20:48:43 +0000 (20:48 +0000)]
upgrade and simplify this test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47240
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 17 Feb 2008 20:44:51 +0000 (20:44 +0000)]
make the logic for breaking up subtracts more explicit, no
functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47239
91177308-0d34-0410-b5e6-
96231b3b80d8
Tanya Lattner [Sun, 17 Feb 2008 20:02:20 +0000 (20:02 +0000)]
Remove llvm-upgrade.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47238
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 17 Feb 2008 19:43:57 +0000 (19:43 +0000)]
move PR2053 to here.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47237
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 17 Feb 2008 00:15:25 +0000 (00:15 +0000)]
fix this test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47232
91177308-0d34-0410-b5e6-
96231b3b80d8
Tanya Lattner [Sun, 17 Feb 2008 00:15:09 +0000 (00:15 +0000)]
Remove llvm-upgrade
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47231
91177308-0d34-0410-b5e6-
96231b3b80d8
Tanya Lattner [Sun, 17 Feb 2008 00:13:09 +0000 (00:13 +0000)]
Remove llvm-upgrade.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47230
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 17 Feb 2008 00:12:03 +0000 (00:12 +0000)]
this test isn't useful since we added @ notation for globals.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47229
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Sun, 17 Feb 2008 00:09:08 +0000 (00:09 +0000)]
this line was commented out.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47228
91177308-0d34-0410-b5e6-
96231b3b80d8
Tanya Lattner [Sat, 16 Feb 2008 23:55:46 +0000 (23:55 +0000)]
Remove llvm-upgrade and update tests.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47227
91177308-0d34-0410-b5e6-
96231b3b80d8
Tanya Lattner [Sat, 16 Feb 2008 23:27:24 +0000 (23:27 +0000)]
Removing llvm upgrade, so remove tests specific to llvm-upgrade and update the tests that used it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47225
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Sat, 16 Feb 2008 20:56:04 +0000 (20:56 +0000)]
Remove any 'nest' parameter attributes if the function
is not passed as an argument to a trampoline intrinsic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47220
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Sat, 16 Feb 2008 20:53:06 +0000 (20:53 +0000)]
Some micro-optimizations.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47219
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrew Lenharth [Sat, 16 Feb 2008 14:46:26 +0000 (14:46 +0000)]
I cannot find a libgcc function for this builtin. Therefor expanding it to a noop (which is how it use to be treated). If someone who knows the x86 backend better than me could tell me how to get a lock prefix on an instruction, that would be nice to complete x86 support.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47213
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Sat, 16 Feb 2008 10:29:26 +0000 (10:29 +0000)]
Teach LegalizeTypes how to expand the operands of
br_cc. This fixes 5 "make check" failures.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47212
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Sat, 16 Feb 2008 09:47:41 +0000 (09:47 +0000)]
Try to clarify which compilers can be used for the
build.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47211
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Sat, 16 Feb 2008 02:32:17 +0000 (02:32 +0000)]
Refactor some code; check if commuteInstruction is able to commute the instruction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47208
91177308-0d34-0410-b5e6-
96231b3b80d8
Andrew Lenharth [Sat, 16 Feb 2008 01:24:58 +0000 (01:24 +0000)]
llvm.memory.barrier, and impl for x86 and alpha
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47204
91177308-0d34-0410-b5e6-
96231b3b80d8
Bill Wendling [Sat, 16 Feb 2008 01:09:25 +0000 (01:09 +0000)]
Fix typos.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47200
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Sat, 16 Feb 2008 00:36:48 +0000 (00:36 +0000)]
Rename CountMemOperands to ComputeMemOperandsEnd to reflect what
it actually does. Simplify CountOperands a little by reusing
ComputeMemOperandsEnd. And reword some comments for both.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47198
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Sat, 16 Feb 2008 00:25:40 +0000 (00:25 +0000)]
Revert 47177, which was incorrect.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47196
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Fri, 15 Feb 2008 23:58:25 +0000 (23:58 +0000)]
upgrade this test, which wasn't testing the right thing since llvm-upgrade came around.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47194
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Fri, 15 Feb 2008 23:51:48 +0000 (23:51 +0000)]
rename llx -> ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47192
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Fri, 15 Feb 2008 23:06:21 +0000 (23:06 +0000)]
This test is not interesting.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47189
91177308-0d34-0410-b5e6-
96231b3b80d8
Scott Michel [Fri, 15 Feb 2008 23:05:48 +0000 (23:05 +0000)]
Make tblgen a little smarter about constants smaller than i32. Currently,
tblgen will complain if a sign-extended constant does not fit into a
data type smaller than i32, e.g., i16. This causes a problem when certain
hex constants are used, such as 0xff for byte masks or immediate xor
values.
tblgen will try the sign-extended value first and, if the sign extended
value would overflow, it tries to see if the unsigned value will fit.
Consequently, a software developer can now safely incant:
(XORHIr16 R16C:$rA, 0xffff)
which is somewhat clearer and more informative than incanting:
(XORHIr16 R16C:$rA, (i16 -1))
even if the two are bitwise equivalent.
Tblgen also outputs the 64-bit unsigned constant in the generated ISel code
when getTargetConstant() is invoked.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47188
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Fri, 15 Feb 2008 22:57:17 +0000 (22:57 +0000)]
simplify this.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47187
91177308-0d34-0410-b5e6-
96231b3b80d8
Dale Johannesen [Fri, 15 Feb 2008 22:05:15 +0000 (22:05 +0000)]
Modify test to expect improved code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47182
91177308-0d34-0410-b5e6-
96231b3b80d8
Evan Cheng [Fri, 15 Feb 2008 21:36:51 +0000 (21:36 +0000)]
The copy instruction being coalesced will be removed, it is not a kill.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47179
91177308-0d34-0410-b5e6-
96231b3b80d8
Ted Kremenek [Fri, 15 Feb 2008 21:12:46 +0000 (21:12 +0000)]
Fixed bug in FoldingSetIteratorImpl where we did not correctly check if
we had reached the "fake bucket" after the last bucket, allowing the iterator
in some cases to run off the end of the hashtable.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47178
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Fri, 15 Feb 2008 20:59:17 +0000 (20:59 +0000)]
Skip over the defs and start at the uses when looking for operands
with the TIED_TO attribute.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47177
91177308-0d34-0410-b5e6-
96231b3b80d8
Dan Gohman [Fri, 15 Feb 2008 20:50:13 +0000 (20:50 +0000)]
Use the TargetInstrDescr to determine the number of operands
that should be checked for the TIED_TO attribute instead of
using CountOperands.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47176
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Fri, 15 Feb 2008 19:42:13 +0000 (19:42 +0000)]
Fix this test on linux, which returns S242
using sret.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47173
91177308-0d34-0410-b5e6-
96231b3b80d8
Duncan Sands [Fri, 15 Feb 2008 19:34:17 +0000 (19:34 +0000)]
Teach LegalizeTypes how to promote the flags
in a ret node. These are created as i32 constants
but on some platforms i32 is not legal. This
fixes 26 "make check" failures, for example
Alpha/2005-07-12-TwoMallocCalls.ll.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47172
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Fri, 15 Feb 2008 19:04:54 +0000 (19:04 +0000)]
Handle \n's in value names for more targets. The asm printers
really really really need refactoring :(
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47171
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Fri, 15 Feb 2008 18:56:05 +0000 (18:56 +0000)]
If the llvm name contains an unprintable character, don't print it in
the global comment. This prevents printing things like:
... # foo
bar
when the name is "foo\nbar".
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47170
91177308-0d34-0410-b5e6-
96231b3b80d8
Chris Lattner [Fri, 15 Feb 2008 18:54:56 +0000 (18:54 +0000)]
targets that support quotes for mangled names still need to escape newlines
when they occur in the name, just like " is escaped.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47169
91177308-0d34-0410-b5e6-
96231b3b80d8
Dale Johannesen [Fri, 15 Feb 2008 18:40:53 +0000 (18:40 +0000)]
Cosmetics.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@47168
91177308-0d34-0410-b5e6-
96231b3b80d8