[all-commits] [llvm/llvm-project] 32bf46: [Polly] Fix -polly-opt-isl -analyze

Michael Kruse via All-commits all-commits at lists.llvm.org
Fri Jul 10 14:42:55 PDT 2020


  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 32bf46842025b740de870dbcd446e4b0f203c9aa
      https://github.com/llvm/llvm-project/commit/32bf46842025b740de870dbcd446e4b0f203c9aa
  Author: Michael Kruse <llvm-project at meinersbur.de>
  Date:   2020-07-10 (Fri, 10 Jul 2020)

  Changed paths:
    M polly/lib/Transform/ScheduleOptimizer.cpp
    M polly/test/ScheduleOptimizer/mat_mul_pattern_data_layout_2.ll
    M polly/test/ScheduleOptimizer/pattern-matching-based-opts_3.ll
    M polly/test/ScheduleOptimizer/tile_after_fusion.ll

  Log Message:
  -----------
  [Polly] Fix -polly-opt-isl -analyze

The member LastSchedule was never set, such that printScop would always
print "n/a" instead of the last schedule.

To ensure that the isl_ctx lives as least as long as the stored
schedule, also store a shared_ptr.

Also set the schedule tree output style to ISL_YAML_STYLE_BLOCK to avoid
printing everything on a single line.

`opt -polly-opt-isl -analyze` will be used in the next commit.


  Commit: c0bc995429c417c1e206841d6b9727218fab3f73
      https://github.com/llvm/llvm-project/commit/c0bc995429c417c1e206841d6b9727218fab3f73
  Author: Michael Kruse <llvm-project at meinersbur.de>
  Date:   2020-07-10 (Fri, 10 Jul 2020)

  Changed paths:
    M polly/lib/Transform/ScheduleOptimizer.cpp
    A polly/test/ScheduleOptimizer/focaltech_test_detail_threshold-7bc17e.ll

  Log Message:
  -----------
  [Polly] Fix prevectorization of fused loops.

The schedule of a fused loop has one isl_space per statement, such that
a conversion to a isl_map fails. However, the prevectorization is
interested in the schedule space only: Converting to the non-union
representation only after extracting the schedule range fixes the problem.

This fixes llvm.org/PR46578


Compare: https://github.com/llvm/llvm-project/compare/1d68a780b34e...c0bc995429c4


More information about the All-commits mailing list