| Commit message (Expand) | Author | Age | Files | Lines |
|---|
| * | Add support for gcc __builtin_ia32_ptest{z,c,nzc} intrinsics. Lower | Eric Christopher | 2009-07-29 | 1 | -2/+11 |
| * | Update insertps handling based on feedback. Move to a v4f32 style | Eric Christopher | 2009-07-24 | 1 | -14/+10 |
| * | Support insertps via the intrinsic and add a couple of simple | Eric Christopher | 2009-07-23 | 1 | -2/+14 |
| * | Fix for PR2484: add an SSE1 pattern for a shuffle we normally prefer to | Eli Friedman | 2009-06-19 | 1 | -0/+6 |
| * | Fix an obvious typo. | Eli Friedman | 2009-06-06 | 1 | -1/+1 |
| * | The MONITOR and MWAIT instructions have insufficient information for | Bill Wendling | 2009-05-28 | 1 | -2/+2 |
| * | Fix MOVMSKPDrr encoding. | Evan Cheng | 2009-05-28 | 1 | -1/+1 |
| * | Fix PSIGND encoding bug. Patch by Sean Callanan. | Evan Cheng | 2009-05-28 | 1 | -1/+1 |
| * | "The instructions MMX_PSADBWrm and MMX_PSADBWrr have opcode 0b11100000 (e0), but | Bill Wendling | 2009-05-28 | 1 | -1/+1 |
| * | Fix sfence jit encoding. Patch by Sean Callanan. | Evan Cheng | 2009-05-27 | 1 | -1/+1 |
| * | 80 col violations. | Evan Cheng | 2009-05-12 | 1 | -2/+4 |
| * | Fix infinite recursion in the C++ code which handles movddup by making it unn... | Nate Begeman | 2009-04-29 | 1 | -3/+13 |
| * | 2nd attempt, fixing SSE4.1 issues and implementing feedback from duncan. | Nate Begeman | 2009-04-27 | 1 | -319/+225 |
| * | Revert 69952. Causes testsuite failures on linux x86-64. | Rafael Espindola | 2009-04-24 | 1 | -225/+319 |
| * | PR2957 | Nate Begeman | 2009-04-24 | 1 | -319/+225 |
| * | Re-apply 68552. | Rafael Espindola | 2009-04-08 | 1 | -4/+4 |
| * | Temporarily revert r68552. This was causing a failure in the self-hosting LLVM | Bill Wendling | 2009-04-07 | 1 | -4/+4 |
| * | Reduce code duplication on the TLS implementation. | Rafael Espindola | 2009-04-07 | 1 | -4/+4 |
| * | ADDS{D|S}rr_Int and MULS{D|S}rr_Int are not commutable. The users of these in... | Evan Cheng | 2009-02-26 | 1 | -8/+4 |
| * | Generate better code for v8i16 shuffles on SSE2 | Nate Begeman | 2009-02-23 | 1 | -0/+8 |
| * | Handle llvm.x86.sse2.maskmov.dqu in 64-bit. | Evan Cheng | 2009-02-10 | 1 | -0/+5 |
| * | A few more isAsCheapAsAMove. | Evan Cheng | 2009-02-05 | 1 | -0/+4 |
| * | The memory alignment requirement on some of the mov{h|l}p{d|s} patterns are 1... | Evan Cheng | 2009-01-28 | 1 | -18/+16 |
| * | Whitespace and other minor adjustments to make SSE instructions have | Dan Gohman | 2009-01-09 | 1 | -26/+32 |
| * | Fixed x86 code generation of multiple for v2i64. It was incorrect for SSE4.1. | Mon P Wang | 2008-12-18 | 1 | -4/+2 |
| * | Rename isSimpleLoad to canFoldAsLoad, to better reflect its meaning. | Dan Gohman | 2008-12-03 | 1 | -16/+16 |
| * | Mark x86's V_SET0 and V_SETALLONES with isSimpleLoad, and teach X86's | Dan Gohman | 2008-12-03 | 1 | -2/+6 |
| * | Fix lfence and mfence encoding. These look like MRM5r and MRM6r instructions ... | Evan Cheng | 2008-10-17 | 1 | -2/+2 |
| * | Fix the predicate for memop64 to be a regular load, not just | Dan Gohman | 2008-10-16 | 1 | -1/+1 |
| * | Now that predicates can be composed, simplify several of | Dan Gohman | 2008-10-15 | 1 | -20/+8 |
| * | Fix SSE4.1 roundss, roundsd. While the instructions have | Dale Johannesen | 2008-10-10 | 1 | -38/+52 |
| * | Certain patterns involving the "movss" instruction were marked as requiring S... | Anders Carlsson | 2008-10-07 | 1 | -3/+3 |
| * | "The original bug was a complaint that _mm_srli_si128 mis-compiled when passed | Bill Wendling | 2008-10-02 | 1 | -0/+4 |
| * | Implement "punpckldq %xmm0, $xmm0" as "pshufd $0x50, %xmm0, %xmm" unless op... | Evan Cheng | 2008-09-26 | 1 | -4/+25 |
| * | unpckhps requires sse1, punpckhdq requires sse2. | Evan Cheng | 2008-09-26 | 1 | -4/+4 |
| * | With sse3 and when the source is a load or has multiple uses, favors movddup ... | Evan Cheng | 2008-09-25 | 1 | -8/+25 |
| * | pmovsxbq etc. requires sse4.1. | Evan Cheng | 2008-09-25 | 1 | -6/+6 |
| * | Fix patterns for SSE4.1 move and sign extend instructions. Also add instructi... | Evan Cheng | 2008-09-24 | 1 | -6/+76 |
| * | Rename ConstantSDNode::getValue to getZExtValue, for consistency | Dan Gohman | 2008-09-12 | 1 | -1/+1 |
| * | Fix for PR2687: Add patterns to match sint_to_fp and fp_to_sint for <2 x | Eli Friedman | 2008-09-05 | 1 | -0/+4 |
| * | FsFLD0S{S|D} and V_SETALLONES are as cheap as moves. | Evan Cheng | 2008-08-28 | 1 | -3/+3 |
| * | Tablegen generated code already tests the opcode value, so it's not | Dan Gohman | 2008-08-20 | 1 | -20/+16 |
| * | Add an EXTRACTPSmr pattern to match the pattern that | Dan Gohman | 2008-08-08 | 1 | -0/+7 |
| * | Fix PR2620: Fix X86cmppd selection code so it expects operands to be v2f64. | Evan Cheng | 2008-08-05 | 1 | -2/+2 |
| * | Fix a typo in last commit | Nate Begeman | 2008-07-17 | 1 | -1/+1 |
| * | SSE codegen for vsetcc nodes | Nate Begeman | 2008-07-17 | 1 | -36/+64 |
| * | Fix for PR2472. Use movss to set lower 32-bits of a zero XMM vector. | Evan Cheng | 2008-07-10 | 1 | -0/+2 |
| * | Horizontal-add instructions are not commutative. | Evan Cheng | 2008-06-16 | 1 | -4/+4 |
| * | mpsadbw is commutable. | Evan Cheng | 2008-06-16 | 1 | -1/+1 |
| * | Disable some DAG combiner optimizations that may be | Duncan Sands | 2008-06-13 | 1 | -0/+2 |
