Commit Graph

9 Commits

Author SHA1 Message Date
Eric Christopher cee313d288 Revert "Temporarily Revert "Add basic loop fusion pass.""
The reversion apparently deleted the test/Transforms directory.

Will be re-reverting again.

llvm-svn: 358552
2019-04-17 04:52:47 +00:00
Eric Christopher a863435128 Temporarily Revert "Add basic loop fusion pass."
As it's causing some bot failures (and per request from kbarton).

This reverts commit r358543/ab70da07286e618016e78247e4a24fcb84077fda.

llvm-svn: 358546
2019-04-17 02:12:23 +00:00
Philip Reames 9ec15faf20 [LICM] Hoist an invariant_start out of loops if there are no stores executed before it
Once the invariant_start is reached, we know that no instruction *after* it can modify the memory. So, if we can prove the location isn't read *between entry into the loop and the execution of the invariant_start*, we can execute the invariant_start before entering the loop.

Differential Revision: https://reviews.llvm.org/D51181

llvm-svn: 340617
2018-08-24 16:24:48 +00:00
Philip Reames a5a8546ac6 [AST] Mark invariant.starts as being readonly
These intrinsics are modelled as writing for control flow purposes, but they don't actually write to any location. Marking these - as we did for guards - allows LICM to hoist loads out of loops containing invariant.starts.

Differential Revision: https://reviews.llvm.org/D50861

llvm-svn: 340245
2018-08-21 00:55:35 +00:00
Anna Thomas 1962621a7e [LICM] Add a diagnostic analysis for identifying alias information
Summary:
Currently, in LICM, we use the alias set tracker to identify if the
instruction (we're interested in hoisting) aliases with instruction that
modifies that memory location.

This patch adds an LICM alias analysis diagnostic tool that checks the
mod ref info of the instruction we are interested in hoisting/sinking,
with every instruction in the loop.  Because of O(N^2) complexity this
is now only a diagnostic tool to show the limitation we have with the
alias set tracker and is OFF by default.

Test cases show the difference with the diagnostic analysis tool, where
we're able to hoist out loads and readonly + argmemonly calls from the
loop, where the alias set tracker analysis is not able to hoist these
instructions out.

Reviewers: reames, mkazantsev, fedor.sergeev, hfinkel

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D50854

llvm-svn: 340026
2018-08-17 13:44:00 +00:00
Philip Reames 684fa57ef7 [MemLoc] Fix a bug causing any use of invariant.end to crash in LICM
The fix is fairly simple, but is says something unpleasant about the usage and testing of invariant.start/end scopes that this went undetected.  To put this in perspective, *any* invariant.end in a loop flowing through LICM crashed.  I haven't bothered to figure out just how far back this goes, but it's not caused by any of the recent changes.  We're probably talking months if not years.  

llvm-svn: 339936
2018-08-16 20:48:55 +00:00
Philip Reames 94b29601ef [LICM] Further strengthen tests for hoisting guards and invariant.starts [NFC]
llvm-svn: 339062
2018-08-06 21:39:43 +00:00
Philip Reames 9d7bb2f700 [LICM] Strengthen invariant.start hoisting tests [NFC]
llvm-svn: 339057
2018-08-06 21:18:34 +00:00
Philip Reames 81c7dc93d2 [LICM] Add tests highlighting missing hoists for intrinsics [NFC]
llvm-svn: 339054
2018-08-06 21:06:15 +00:00