[GHC] #9246: GHC generates poor code for repeated uses of min/max
GHC
ghc-devs at haskell.org
Sun Jun 29 19:37:31 UTC 2014
#9246: GHC generates poor code for repeated uses of min/max
--------------------------------------------+------------------------------
Reporter: arotenberg | Owner:
Type: feature request | Status: new
Priority: normal | Milestone: 7.10.1
Component: Compiler | Version: 7.8.2
Resolution: | Keywords:
Operating System: Windows | Architecture: x86_64
Type of failure: Runtime performance bug | (amd64)
Test Case: | Difficulty: Unknown
Blocking: | Blocked By:
| Related Tickets: #6135
--------------------------------------------+------------------------------
Comment (by arotenberg):
On the contrary, replacing the NOINLINEs with INLINEs results in
essentially the original Core, local lambdas and all. I explicitly put the
NOINLINEs in because that was the only way I could find to force GHC to
not expand the branches and add the unnecessary lambdas and lazy lets.
Perhaps this issue should be split into two tickets - something like "Add
branchless min/max primops" and "Min/max-like functions cause unnecessary
closure creation"?
--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9246#comment:4>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler
More information about the ghc-tickets
mailing list