| Commit message (Expand) | Author | Age |
* | Use the predicate methods off of AttributeSet instead of Attribute. | Bill Wendling | 2012-12-30 |
* | Remove the Function::getFnAttributes method in favor of using the AttributeSet | Bill Wendling | 2012-12-30 |
* | Remove intrinsic specific instructions for (V)SQRTPS/PD. Instead lower to tar... | Craig Topper | 2012-12-29 |
* | Merge similar functionality using a nested switch. | Craig Topper | 2012-12-29 |
* | Remove intrinsic specific instructions for SSE/SSE2/AVX floating point max/mi... | Craig Topper | 2012-12-29 |
* | Simplify code, no functionality change. | Jakub Staszak | 2012-12-29 |
* | CostModel: initial checkin for code that estimates the cost of special shuffles. | Nadav Rotem | 2012-12-28 |
* | wrap 80-col lines. | Nadav Rotem | 2012-12-28 |
* | AVX: Move the ZEXT/ANYEXT DAGCo optimizations to the lowering of these optimi... | Nadav Rotem | 2012-12-28 |
* | Reverse the 'if' condition and reduce the indentation. | Nadav Rotem | 2012-12-27 |
* | AVX/AVX2: Move the SEXT lowering code from a target specific DAGco to a lower... | Nadav Rotem | 2012-12-27 |
* | On AVX/AVX2 the type v8i1 is legalized to v8i16, which is an XMM sized | Nadav Rotem | 2012-12-27 |
* | AVX/AVX2: Move the code that lowers vector-trunc from a DAGCo-hook to custom ... | Nadav Rotem | 2012-12-27 |
* | Reformat the docs. | Nadav Rotem | 2012-12-26 |
* | X86: Shave off one shuffle from the pcmpeqq sequence for SSE2 by making use o... | Benjamin Kramer | 2012-12-25 |
* | X86: Custom lower <2 x i64> eq and ne when SSE41 is not available. | Benjamin Kramer | 2012-12-25 |
* | Quiet gcc's -Wparenthesis warning. No functionality change. | Nick Lewycky | 2012-12-24 |
* | whitespace | Nadav Rotem | 2012-12-23 |
* | Loop Vectorizer: Update the cost model of scatter/gather operations and make | Nadav Rotem | 2012-12-23 |
* | X86: Turn mul of <4 x i32> into pmuludq when no SSE4.1 is available. | Benjamin Kramer | 2012-12-22 |
* | X86: Emit vector sext as shuffle + sra if vpmovsx is not available. | Benjamin Kramer | 2012-12-22 |
* | X86: Match pmin/pmax as a target specific dag combine. This occurs during vec... | Benjamin Kramer | 2012-12-21 |
* | X86: Match the SSE/AVX min/max vector ops using a custom node instead of intr... | Benjamin Kramer | 2012-12-21 |
* | Improve the X86 cost model for loads and stores. | Nadav Rotem | 2012-12-21 |
* | Change TargetLowering::getTypeForExtArgOrReturn to take and return | Patrik Hagglund | 2012-12-19 |
* | Change TargetLowering::findRepresentativeClass to take an MVT, instead | Patrik Hagglund | 2012-12-19 |
* | X86ISelLowering.cpp: Fix warnings. [-Wlogical-op-parentheses] | NAKAMURA Takumi | 2012-12-19 |
* | Optimized load + SIGN_EXTEND patterns in the X86 backend. | Elena Demikhovsky | 2012-12-19 |
* | Rename the 'Attributes' class to 'Attribute'. It's going to represent a singl... | Bill Wendling | 2012-12-19 |
* | Reverse order of checking SSE level when calculating compare cost, so we check | Jakub Staszak | 2012-12-18 |
* | Simplify BMI ANDN matching to use patterns instead of a DAG combine. Also add... | Craig Topper | 2012-12-17 |
* | X86: Add a couple of target-specific dag combines that turn VSELECTS into psu... | Benjamin Kramer | 2012-12-15 |
* | TypeLegalizer: Do not generate target specific nodes with illegal types, beca... | Nadav Rotem | 2012-12-14 |
* | Sorry about the churn. One more change to getOptimalMemOpType() hook. Did I | Evan Cheng | 2012-12-12 |
* | - Rename isLegalMemOpType to isSafeMemOpType. "Legal" is a very overloade term. | Evan Cheng | 2012-12-12 |
* | Avoid using lossy load / stores for memcpy / memset expansion. e.g. | Evan Cheng | 2012-12-12 |
* | Revert EVT->MVT changes, r169836-169851, due to buildbot failures. | Patrik Hagglund | 2012-12-11 |
* | Change TargetLowering::getTypeForExtArgOrReturn to take and return | Patrik Hagglund | 2012-12-11 |
* | Change TargetLowering::findRepresentativeClass to take an MVT, instead | Patrik Hagglund | 2012-12-11 |
* | Some enhancements for memcpy / memset inline expansion. | Evan Cheng | 2012-12-10 |
* | - Re-enable population count loop idiom recognization | Shuxin Yang | 2012-12-09 |
* | Revert the patches adding a popcount loop idiom recognition pass. | Chandler Carruth | 2012-12-08 |
* | s/AttrListPtr/AttributeSet/g to better label what this class is going to be i... | Bill Wendling | 2012-12-07 |
* | When we use the BLEND instruction that uses the MSB as a mask, we can remove | Nadav Rotem | 2012-12-07 |
* | X86: Prefer using VPSHUFD over VPERMIL because it has better throughput. | Nadav Rotem | 2012-12-07 |
* | Replace r169459 with something safer. Rather than having computeMaskedBits to | Evan Cheng | 2012-12-06 |
* | Remove unneeded function, since PR8156 was fixed over a year ago. | Jakub Staszak | 2012-12-06 |
* | Simplify code. | Jakub Staszak | 2012-12-06 |
* | Let targets provide hooks that compute known zero and ones for any_extend | Evan Cheng | 2012-12-06 |
* | Simplified BLEND pattern matching for shuffles. | Elena Demikhovsky | 2012-12-05 |