aboutsummaryrefslogtreecommitdiff
path: root/llvm/lib/Target/NVPTX
AgeCommit message (Expand)AuthorFilesLines
2024-03-29Reland "[NVPTX] Use .common linkage for common globals" (#86824)Alex MacLean1-7/+9
2024-03-26[NFC][NVPTX] remove truncating c-style cast (#85889)Alex MacLean1-1/+1
2024-03-23[NewPM][NVPTX] Add NVPTXPassRegistry.def NFCI (#86246)paperchalice2-39/+42
2024-03-19[NVPTX][DebugInfo] avoid emitting extra .loc directives (#84584)Alex MacLean1-2/+5
2024-03-19[NVPTX] Use PTX prmt for llvm.bswap (#85545)Alex MacLean2-3/+19
2024-03-19[NFC][RemoveDIs] Use iterators for insertion at various call-sites (#84736)Jeremy Morse3-9/+9
2024-03-15Revert "[NVPTX] Use .common linkage for common globals (#84416)"Sterling Augustine1-9/+7
2024-03-15[NVPTX] support dynamic allocas with PTX alloca instruction (#84585)Alex MacLean4-14/+63
2024-03-15Reland "[NVPTX] Add support for atomic add for f16 type" (#85197)Adrian Kuegel2-1/+19
2024-03-14[NVPTX] Use .common linkage for common globals (#84416)Alex MacLean1-7/+9
2024-03-12Revert "[NVPTX] Add support for atomic add for f16 type" (#84918)Danial Klimkin2-18/+0
2024-03-12[NVPTX] Add support for atomic add for f16 type (#84295)Adrian Kuegel2-0/+18
2024-03-05[NVPTX] Remove sub.s16x2 instructionBenjamin Kramer2-2/+0
2024-02-25[CodeGen] Port AtomicExpand to new Pass Manager (#71220)Rishabh Bali1-1/+1
2024-02-22[NVPTX] fixup support for unaligned parameters and returns (#82562)Alex MacLean3-20/+271
2024-02-21[NVPTX] Correctly guard int -> bf16 on PTX version and SM versionDavid Majnemer2-10/+12
2024-02-21[NVPTX] Simplify handling of ISD::BF16_TO_FPDavid Majnemer1-4/+1
2024-02-21[NVPTX] Set ISD::FP_{EXTEND,ROUND} to Custom for more typesDavid Majnemer1-2/+4
2024-02-21Correctly round FP -> BF16 when SDAG expands such nodes (#82399)David Majnemer3-5/+87
2024-02-13[LLVM] Add `__builtin_readsteadycounter` intrinsic and builtin for realtime c...Joseph Huber3-1/+3
2024-02-12[NVPTX] Fix the error in a pattern match in v4i8 comparisons. (#81308)Artem Belevich1-23/+69
2024-02-12[NVPTX] Implement `__builtin_readcyclecounter` on NVPTX (#81344)Joseph Huber2-0/+6
2024-02-12Fix use after free error in NVVMReflect (#81471)Petr1-5/+13
2024-02-12Do not use PerformEXTRACTCombine for v8i8 types (#81242)Adrian Kuegel1-3/+4
2024-02-09[NVVMReflect] Improve folding inside of the NVVMReflect pass (#81253)Joseph Huber1-53/+17
2024-02-08[NVVMReflect][Reland] Force dead branch elimination in NVVMReflect (#81189)Joseph Huber1-0/+65
2024-02-08Revert "[NVVMReflect] Force dead branch elimination in NVVMReflect (#81189)"Joseph Huber1-62/+0
2024-02-08[NVVMReflect] Force dead branch elimination in NVVMReflect (#81189)Joseph Huber1-0/+62
2024-02-08[NVPTX] Add support for calling aliases (#81170)Alex MacLean2-30/+29
2024-02-08[NVPTX][NFC] cleanup dead vars, use MAKE_CASE (#81161)Alex MacLean2-621/+428
2024-02-01[TTI] Use Register in isLoadFromStackSlot and isStoreToStackSlot [nfc] (#80339)Philip Reames1-2/+2
2024-01-31[NVPTX] improve Boolean ISel (#80166)Alex MacLean1-0/+10
2024-01-29Revert "Disable incorrect peephole optimizations" (#79916)Justin Fargnoli1-0/+13
2024-01-29Disable incorrect peephole optimizationsJustin Fargnoli1-13/+0
2024-01-29[NVPTX] Add builtin support for 'globaltimer' (#79765)Joseph Huber1-0/+2
2024-01-29[NVPTX] Add builtin for 'exit' handling (#79777)Joseph Huber1-0/+3
2024-01-29[NVPTX] Add builtin support for 'nanosleep' PTX instrunction (#79888)Joseph Huber1-0/+6
2024-01-29[NVPTX] Add 'activemask' builtin and intrinsic support (#79768)Joseph Huber2-1/+7
2024-01-26[NVPTX] improve identifier renaming for PTX (#79459)Alex MacLean1-3/+6
2024-01-25[llvm] Move CodeGenTypes library to its own directory (#79444)Nico Weber2-2/+2
2024-01-24[NVPTX] use incomplete aggregate initializers (#79062)Alex MacLean1-2/+13
2024-01-19[LLVM][NVPTX] Add cp.async.bulk.commit/wait intrinsics (#78698)Durgadoss R1-0/+16
2024-01-18[NVPTX][NFC] Remove unused parameter of getArgumentAlignment (#78604)Alex MacLean2-7/+6
2024-01-17[NVPTX] extend type support for nvvm.{min,max,mulhi,sad} (#78385)Alex MacLean1-1/+12
2024-01-17[NVPTX] Add tex.grad.cube{array} intrinsics (#77693)Alex MacLean5-0/+129
2024-01-16[NVPTX] Fix generating permute bytes from register pair when the initial valu...mmoadeli1-2/+4
2024-01-14[Target] Use getConstantOperandVal (NFC)Kazu Hirata1-3/+1
2024-01-13[LLVM][NVPTX]: Add aligned versions of cluster barriers (#77940)Durgadoss R1-0/+10
2024-01-10[Target] Use getConstantOperandAPInt (NFC)Kazu Hirata1-3/+1
2024-01-09[LLVM][NVPTX]: Add intrinsic for setmaxnreg (#77289)Durgadoss R2-0/+16
generated by cgit v1.1 at 2024-07-31 23:53:09 +0000