| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | Remove FIXME - if there's a better way to do this it isn't here. | Eric Christopher | 2010-04-02 | 1 | -3/+0 |
| | | | | | llvm-svn: 100176 | ||||
| * | Revert the recent alignment changes. They're broken for -Os because, | Dan Gohman | 2010-04-02 | 1 | -1/+1 |
| | | | | | | | | in particular, they end up aligning strings at 16-byte boundaries, and there's no way for GlobalOpt to check OptForSize. llvm-svn: 100172 | ||||
| * | Teach AnalyzeBranch, RemoveBranch and the branch | Dale Johannesen | 2010-04-02 | 10 | -6/+102 |
| | | | | | | | | folder to be tolerant of debug info following the branch(es) at the end of a block. llvm-svn: 100168 | ||||
| * | Disambiguate conditional expression for newer GCCs. | Chandler Carruth | 2010-04-02 | 1 | -2/+2 |
| | | | | | llvm-svn: 100167 | ||||
| * | Change variables which are exactly 16 bytes to be 16-byte-aligned too. | Dan Gohman | 2010-04-02 | 1 | -1/+1 |
| | | | | | | | This fixes test/Transforms/GlobalOpt/gv-align.ll. llvm-svn: 100161 | ||||
| * | Revert r100143. | Eric Christopher | 2010-04-01 | 1 | -3/+0 |
| | | | | | llvm-svn: 100146 | ||||
| * | In 64-bit mode, use i64 to lower memcpy / memset instead of f64. | Evan Cheng | 2010-04-01 | 1 | -0/+1 |
| | | | | | llvm-svn: 100137 | ||||
| * | Add comments about DstAlign and SrcAlign. | Evan Cheng | 2010-04-01 | 4 | -6/+26 |
| | | | | | llvm-svn: 100132 | ||||
| * | - Avoid using floating point stores to implement memset unless the value is ↵ | Evan Cheng | 2010-04-01 | 4 | -4/+7 |
| | | | | | | | | | zero. - Do not try to infer GV alignment unless its type is sized. It's not possible to infer alignment if it has opaque type. llvm-svn: 100118 | ||||
| * | Fix sdisel memcpy, memset, memmove lowering: | Evan Cheng | 2010-04-01 | 4 | -15/+23 |
| | | | | | | | | | | | | | | 1. Makes it possible to lower with floating point loads and stores. 2. Avoid unaligned loads / stores unless it's fast. 3. Fix some memcpy lowering logic bug related to when to optimize a load from constant string into a constant. 4. Adjust x86 memcpy lowering threshold to make it more sane. 5. Fix x86 target hook so it uses vector and floating point memory ops more effectively. rdar://7774704 llvm-svn: 100090 | ||||
| * | Nehalem unaligned memory access is fast. | Evan Cheng | 2010-04-01 | 3 | -2/+15 |
| | | | | | llvm-svn: 100089 | ||||
| * | Add aeskeygenassist intrinsic and rename all of the aes intrinsics to | Eric Christopher | 2010-04-01 | 1 | -21/+29 |
| | | | | | | | | aes instead of sse4.2. Add a brief todo for a subtarget flag and rework the aeskeygenassist instruction to more closely match the docs. llvm-svn: 100078 | ||||
| * | vml[as] are slow on 1136jf-s also. | Jim Grosbach | 2010-04-01 | 1 | -1/+2 |
| | | | | | llvm-svn: 100066 | ||||
| * | reduce indentation, minor cleanups. | Chris Lattner | 2010-03-31 | 1 | -38/+42 |
| | | | | | llvm-svn: 100042 | ||||
| * | Use spaces, not tabs | Jakob Stoklund Olesen | 2010-03-31 | 1 | -1/+1 |
| | | | | | llvm-svn: 100037 | ||||
| * | Comment the changes for r98218 and friends inside the source code. | Bill Wendling | 2010-03-31 | 1 | -0/+5 |
| | | | | | llvm-svn: 100033 | ||||
| * | Comment the changes for r98218 and friends inside the source code. | Bill Wendling | 2010-03-31 | 2 | -0/+10 |
| | | | | | llvm-svn: 100031 | ||||
| * | Fix PR6750. Don't try to merge a DomainValue with itself. | Jakob Stoklund Olesen | 2010-03-31 | 1 | -0/+2 |
| | | | | | llvm-svn: 100016 | ||||
| * | Replace V_SET0 with variants for each SSE execution domain. | Jakob Stoklund Olesen | 2010-03-31 | 3 | -15/+26 |
| | | | | | llvm-svn: 99975 | ||||
| * | Fix typo. Thank you, valgrind. | Jakob Stoklund Olesen | 2010-03-31 | 1 | -1/+1 |
| | | | | | llvm-svn: 99974 | ||||
| * | Enable -sse-domain-fix by default. Now with tests! | Jakob Stoklund Olesen | 2010-03-30 | 1 | -6/+1 |
| | | | | | llvm-svn: 99954 | ||||
| * | V_SETALLONES is an integer instruction. | Jakob Stoklund Olesen | 2010-03-30 | 1 | -1/+1 |
| | | | | | | | | Since it is just a pxor in disguise, we should probably expand it to a full polymorphic triple. llvm-svn: 99953 | ||||
| * | Renumber SSE execution domains for better code size. | Jakob Stoklund Olesen | 2010-03-30 | 3 | -24/+21 |
| | | | | | | | | | SSEDomainFix will collapse to the domain with the lower number when it has a choice. The SSEPackedSingle domain often has smaller instructions, so prefer that. llvm-svn: 99952 | ||||
| * | Revert Mon Ping's change 99928, since it broke all the llvm-gcc buildbots. | Bob Wilson | 2010-03-30 | 6 | -19/+15 |
| | | | | | llvm-svn: 99948 | ||||
| * | Revert "Enable -sse-domain-fix by default. What could possibly go wrong?" | Jakob Stoklund Olesen | 2010-03-30 | 1 | -1/+6 |
| | | | | | | | Not running 'make check-all' before committing is a bad idea. llvm-svn: 99933 | ||||
| * | Enable -sse-domain-fix by default. What could possibly go wrong? | Jakob Stoklund Olesen | 2010-03-30 | 1 | -6/+1 |
| | | | | | llvm-svn: 99931 | ||||
| * | Added support for address spaces and added a isVolatile field to memcpy, ↵ | Mon P Wang | 2010-03-30 | 6 | -15/+19 |
| | | | | | | | | | | memmove, and memset, e.g., llvm.memcpy.i32(i8*, i8*, i32, i32) -> llvm.memcpy.p0i8.p0i8.i32(i8*, i8*, i32, i32, i1) A update of langref will occur in a subsequent checkin. llvm-svn: 99928 | ||||
| * | Add cross-block inference to SSEDomainFix. | Jakob Stoklund Olesen | 2010-03-30 | 1 | -29/+85 |
| | | | | | llvm-svn: 99916 | ||||
| * | Remove the pmulld intrinsic and autoupdate it as a vector multiply. | Eric Christopher | 2010-03-30 | 2 | -3/+22 |
| | | | | | | | | Rewrite the pmulld patterns, and make sure that they fold in loads of arguments into the instruction. llvm-svn: 99910 | ||||
| * | Rip out the 'is temporary' nonsense from the MCContext interface to | Chris Lattner | 2010-03-30 | 7 | -21/+18 |
| | | | | | | | | | | create symbols. It is extremely error prone and a source of a lot of the remaining integrated assembler bugs on x86-64. This fixes rdar://7807601. llvm-svn: 99902 | ||||
| * | XFAIL some PIC16 tests when running under valgrind-leaks. I don't expect these | Benjamin Kramer | 2010-03-30 | 1 | -0/+2 |
| | | | | | | | to be fixed any time soon. llvm-svn: 99888 | ||||
| * | PIC16: Plug a leak in PIC16Section by allocating name & address strings in the | Benjamin Kramer | 2010-03-30 | 2 | -13/+20 |
| | | | | | | | MCContext. There is still one leak left in PIC16Section (the Items vector). llvm-svn: 99887 | ||||
| * | Add FIXME for operand promotion. | Eric Christopher | 2010-03-30 | 1 | -0/+4 |
| | | | | | llvm-svn: 99859 | ||||
| * | Be gentle to MSVC. C++ is hard, after all. | Jakob Stoklund Olesen | 2010-03-30 | 1 | -2/+3 |
| | | | | | llvm-svn: 99855 | ||||
| * | Basic implementation of SSEDomainFix pass. | Jakob Stoklund Olesen | 2010-03-29 | 3 | -57/+399 |
| | | | | | | | Cross-block inference is primitive and wrong, but the pass is working otherwise. llvm-svn: 99848 | ||||
| * | Make isInt?? and isUint?? template specializations of the generic versions. This | Benjamin Kramer | 2010-03-29 | 9 | -40/+42 |
| | | | | | | | | makes calls a little bit more consistent and allows easy removal of the specializations in the future. Convert all callers to the templated functions. llvm-svn: 99838 | ||||
| * | We'll never match these as instructions, just as intrinsics so remove | Eric Christopher | 2010-03-29 | 1 | -16/+10 |
| | | | | | | | the SDNodes. llvm-svn: 99835 | ||||
| * | Remove a bunch of integer width predicate functions in favor of MathExtras. | Benjamin Kramer | 2010-03-29 | 4 | -139/+11 |
| | | | | | | | | Most of these were unused, some of them were wrong and unused (isS16Constant<short>, isS10Constant<short>). llvm-svn: 99827 | ||||
| * | From Kalle Raiskila: | Chris Lattner | 2010-03-29 | 4 | -14/+158 |
| | | | | | | | | | "the bigstack patch for SPU, with testcase. It is essentially the patch committed as 97091, and reverted as 97099, but with the following additions: -in vararg handling, registers are marked to be live, to not confuse the register scavenger -function prologue and epilogue are not emitted, if the stack size is 16. 16 means it is empty - there is only the register scavenger emergency spill slot, which is not used as there is no stack." llvm-svn: 99819 | ||||
| * | add a note. | Chris Lattner | 2010-03-29 | 1 | -0/+3 |
| | | | | | llvm-svn: 99815 | ||||
| * | Add NVTBLFrm to represent A8.6.406 VTBL, VTBX Vector Table Lookup Instructions. | Johnny Chen | 2010-03-29 | 2 | -8/+9 |
| | | | | | | | | | | | | | | These instructions use byte index in a control vector (M:Vm) to lookup byte values in a table and generate a new vector (D:Vd). The table is specified via a list of vectors, which can be: {Dn} {Dn D<n+1>} {Dn D<n+1> D<n+2>} {Dn D<n+1> D<n+2> D<n+3>} llvm-svn: 99789 | ||||
| * | zap an extra line that Eli noticed! | Chris Lattner | 2010-03-28 | 1 | -1/+0 |
| | | | | | llvm-svn: 99770 | ||||
| * | fix a type contradition: XCoreISD::RETSP has one argument, not zero. | Chris Lattner | 2010-03-28 | 1 | -1/+1 |
| | | | | | llvm-svn: 99760 | ||||
| * | remove a pattern with no testcase that doesn't appear to be | Chris Lattner | 2010-03-28 | 1 | -2/+0 |
| | | | | | | | matchable: it seems like it would always constant fold. llvm-svn: 99758 | ||||
| * | fix integer negates to use the proper type for the zero vectors, | Chris Lattner | 2010-03-28 | 1 | -11/+14 |
| | | | | | | | | this also depends on the new "bitconvert dropping" behavior just added to tblgen. llvm-svn: 99757 | ||||
| * | fix a typo, bitconvert from node to itself isn't valid. | Chris Lattner | 2010-03-28 | 1 | -1/+1 |
| | | | | | llvm-svn: 99755 | ||||
| * | fix vnot matching to explicitly specify the type of the | Chris Lattner | 2010-03-28 | 1 | -10/+16 |
| | | | | | | | | | | input to be v8i8 or v16i8, which buildvectors get canonicalized to. This allows the patterns that were previously using a bare 'vnot' to match, before they couldn't. llvm-svn: 99754 | ||||
| * | fix up vnot matching, eliminating a dead pattern, correcting a couple of | Chris Lattner | 2010-03-28 | 1 | -6/+11 |
| | | | | | | | | patterns that would never match because of bitcast, and eliminating use of vnot_conv. llvm-svn: 99753 | ||||
| * | stop using vnot_conv | Chris Lattner | 2010-03-28 | 1 | -3/+8 |
| | | | | | llvm-svn: 99750 | ||||
| * | revert r99743, this is saying that the repmovs instructinos have an | Chris Lattner | 2010-03-28 | 1 | -2/+4 |
| | | | | | | | *input* of other type, which is the VT. llvm-svn: 99749 | ||||

