llvm/lib/CodeGen/SelectionDAG, branch release_24http://llvm.orghttps://git.amat.us/llvm/atom/lib/CodeGen/SelectionDAG?h=release_242008-10-23T05:03:22ZMerge from mainline.2008-10-23T05:03:22ZTanya Lattnertonic@nondot.org2008-10-23T05:03:22Zurn:sha1:e11adfd26fd843f9b037cb7063f8d9c4fe28b6d4
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_24@58032 91177308-0d34-0410-b5e6-96231b3b80d8
Merge from mainline.2008-10-17T17:53:06ZTanya Lattnertonic@nondot.org2008-10-17T17:53:06Zurn:sha1:6025e4b7e6b207cffe2d0766fb159ba581798239
Change CALLSEQ_BEGIN and CALLSEQ_END to take TargetConstant's as
parameters instead of raw Constants. This prevents the constants from
being selected by the isel pass, fixing PR2735.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/branches/release_24@57700 91177308-0d34-0410-b5e6-96231b3b80d8
Correctly handle calls with no return values. This fixes2008-10-07T00:12:37ZDan Gohmangohman@apple.com2008-10-07T00:12:37Zurn:sha1:b5cc34d098149a9606dacb6d421471d787253907
2006-01-23-UnionInit on x86-64 when inlining is not enabled.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57223 91177308-0d34-0410-b5e6-96231b3b80d8
wrap some long lines and expand i32 mul's to libcalls, inspired by a2008-10-04T21:27:46ZChris Lattnersabre@nondot.org2008-10-04T21:27:46Zurn:sha1:31d7161206d7502e97a01a51a76445a8d9ac6ea5
patch by Mikael Lepisto!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57077 91177308-0d34-0410-b5e6-96231b3b80d8
Fix fast-isel's handling of atomic instructions. They may2008-10-04T00:56:36ZDan Gohmangohman@apple.com2008-10-04T00:56:36Zurn:sha1:241f464d24a6c22721607841069bbeb17b3f71e6
expand to multiple basic blocks, in which case fast-isel
needs to informed of which block to use as it resumes
inserting instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57040 91177308-0d34-0410-b5e6-96231b3b80d8
Pass MemOperand through for 64-bit atomics on 32-bit,2008-10-03T19:41:08ZDale Johannesendalej@apple.com2008-10-03T19:41:08Zurn:sha1:1b54c7fe9b034a6450af8596231ad5368b617725
incidentally making the case where the memop is a
pointer deref work. Fix cmp-and-swap regression.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57027 91177308-0d34-0410-b5e6-96231b3b80d8
Use -1ULL instead of uint64_t(-1), at Anton's suggestion.2008-10-03T17:56:45ZDan Gohmangohman@apple.com2008-10-03T17:56:45Zurn:sha1:ca0a5d93569d728659352ca5fbfe17daeed5f6ac
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@57021 91177308-0d34-0410-b5e6-96231b3b80d8
The result of getSetCCResultType (eg: i32) may be larger2008-10-03T07:41:46ZDuncan Sandsbaldrick@free.fr2008-10-03T07:41:46Zurn:sha1:e47561ccff9c7a76f9392b26ae6202e30f1247ed
than the type an i1 is promoted to (eg: i8). Account
for this. Noticed by Tilmann Scheller on CellSPU; he
will hopefully take care of fixing this in LegalizeDAG
and adding a testcase!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56997 91177308-0d34-0410-b5e6-96231b3b80d8
Implement fast-isel support for zero-extending from i1.2008-10-03T01:28:47ZDan Gohmangohman@apple.com2008-10-03T01:28:47Zurn:sha1:91b6f97ce4273fee5516692e3f27cd76d67986fc
It turns out that this is a fairly common operation,
and it's easy enough to handle.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56990 91177308-0d34-0410-b5e6-96231b3b80d8
Optimize conditional branches in X86FastISel. This replaces2008-10-02T22:15:21ZDan Gohmangohman@apple.com2008-10-02T22:15:21Zurn:sha1:d98d6203e429b2d7208b6687931e9079e85e95ec
sequences like this:
sete %al
testb %al, %al
jne LBB11_1
with this:
je LBB11_1
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@56969 91177308-0d34-0410-b5e6-96231b3b80d8