Commits · f14d18c7a9544ea5fbea736210d05e938a986033 · Roger Ferrer / llvm-epi · GitLab

Jun 03, 2022
- [SLP]Improve shuffles cost estimation where possible. · cac60940
  Alexey Bataev authored Dec 09, 2021
  cac60940
- [coro async] Add code to support dynamic aligment of over-aligned types in async frames · 5c902af5
  Arnold Schwaighofer authored May 27, 2022
  5c902af5
- [VPlan] Silence another unused variable warning in release builds · a8d2a381
  Benjamin Kramer authored Jun 03, 2022
  
  a8d2a381
- [VPlan] Inline variable into assertion. NFC. · 6b7c1863
  Benjamin Kramer authored Jun 03, 2022
  6b7c1863
- [VPlan] Replace CondBit with BranchOnCond VPInstruction. · a5bb4a3b
  Florian Hahn authored Jun 03, 2022
  a5bb4a3b
- Revert "[SLP]Improve shuffles cost estimation where possible." · df0f30dc
  Fangrui Song authored Jun 03, 2022
  df0f30dc
- Revert "[LoopInterchange] New cost model for loop interchange" · f1940a58
  Daniil Suchkov authored Jun 03, 2022
  f1940a58
- [LoopInterchange] New cost model for loop interchange · 00633447
  Congzhe Cao authored Jun 02, 2022
  00633447
Jun 02, 2022
- [InstCombine] make pattern matching more consistent; NFC · 8689463b
  Sanjay Patel authored Jun 02, 2022
  8689463b
- [SLP]Improve shuffles cost estimation where possible. · 9980c997
  Alexey Bataev authored Dec 09, 2021
  9980c997
- [llvm][ModuleInliner] Refactor InlineSizePriority and PriorityInlineOrder · 14e8add9
  Liqiang Tao authored Jun 02, 2022
  14e8add9
- Revert "[llvm][ModuleInliner] Refactor InlineSizePriority and PriorityInlineOrder" · 5c6ed60c
  Liqiang Tao authored Jun 02, 2022
  5c6ed60c
- [llvm][ModuleInliner] Refactor InlineSizePriority and PriorityInlineOrder · 50de7f1e
  Liqiang Tao authored Jun 02, 2022
  50de7f1e
- [VPlan] Remove dead VPlan-native special case from BranchOnCount (NFC). · 4f1c86e3
  Florian Hahn authored Jun 02, 2022
  4f1c86e3
- [LSR] Early exit for RateFormula when it is already losing. NFC · 6eab5cad
  eopXD authored Jun 01, 2022
  6eab5cad
- Revert "[SLP]Improve shuffles cost estimation where possible." · 73020b45
  Alexey Bataev authored Jun 01, 2022
  73020b45
Jun 01, 2022
- [LV] Update var name to Exiting, in line with terminology (NFC) · 08482830
  Florian Hahn authored Jun 01, 2022
  08482830
- [SLP]Improve shuffles cost estimation where possible. · fd5a6ce9
  Alexey Bataev authored Dec 09, 2021
  fd5a6ce9
- [SLP]Fix PR55796: insert point for extractelements from different basic blocks. · fe494994
  Alexey Bataev authored May 31, 2022
  fe494994
- Revert "[InstCombine] Combine instructions of type or/and where AND masks can be combined." · aa98e7e1
  Alexander Kornienko authored Jun 01, 2022
  aa98e7e1
- [VPlan] Use region for each loop in native path. · 05776122
  Florian Hahn authored Jun 01, 2022
  05776122
- [VPlan] Remove unused native utilities incompatible with nested regions. · d1570194
  Florian Hahn authored Jun 01, 2022
  d1570194
- [LoopIdiom] Fix bailout for aliasing in memcpy transform. · abdf0da8
  Eli Friedman authored May 31, 2022
  abdf0da8
May 31, 2022
- [InstCombine] fold icmp of sext bool based on limited range · 2bf6123f
  Sanjay Patel authored May 31, 2022
  2bf6123f
- BuildLibCalls: infer allockind attributes on relevant functions · 73f66460
  Augie Fackler authored Mar 29, 2022
  73f66460
- attributes: introduce allockind attr for describing allocator fn behavior · 42861faa
  Augie Fackler authored Mar 29, 2022
  42861faa
- [InstCombine] Fix inbounds preservation when swapping GEPs (PR44206) · 36cbdaa1
  Nikita Popov authored May 30, 2022
  36cbdaa1
- [NFC] Change LoopVectorizationCostModel::useOrderedReductions() to be a const function. · b0fc7653
  Mel Chen authored May 31, 2022
  b0fc7653
- [InstCombine] Fix const folding of switches with default case · 4fb3fd7d
  Danila Malyutin authored May 30, 2022
  4fb3fd7d
- [InstCombine] Fix inbounds preservation when merging GEPs (PR55722) · 872d69e5
  Nikita Popov authored May 31, 2022
  872d69e5
May 30, 2022
- [InstCombine] fold shift-right-by-constant with shift-right-of-constant operand · a0c3c607
  Sanjay Patel authored May 30, 2022
  a0c3c607
- [InstCombine] remove unnecessary one-use check from (C2 << X) << C1 fold · c5d942a4
  Sanjay Patel authored May 30, 2022
  c5d942a4
- [Support] Make report_fatal_error respect its GenCrashDiag argument so it... · 80b3dcc0
  Nuno Lopes authored May 30, 2022
  80b3dcc0
- [InstCombine] When swapping GEPs, only keep inbounds if both are · a770f534
  Nikita Popov authored May 30, 2022
  a770f534
- [InstCombine] Always create new GEPs when swapping GEPs · 2d7bab66
  Nikita Popov authored May 30, 2022
  2d7bab66
- [Local] Don't remove invoke of non-willreturn function · 2e101cca
  Nikita Popov authored May 30, 2022
  2e101cca
- [InstCombine] Fold a mul with bool value into and · 3e6ba890
  zhongyunde authored May 30, 2022
  3e6ba890
- [GVN] Enable enable-split-backedge-in-load-pre option by default · 1721ff1d
  Nikita Popov authored May 25, 2022
  1721ff1d
- [JumpThreading][NFCI] Reuse existing DT instead of recomputation · 503d5771
  Max Kazantsev authored May 30, 2022
  503d5771
- [InstCombine] bitcast (extractelement <1 x elt>, dest) -> bitcast(<1 x elt>, dest) · ef256ed5
  Chenbing Zheng authored May 30, 2022
  ef256ed5