[GHC] #13817: Simplifier and SpecConstr performance regression with 8.2.1
#13817: Simplifier and SpecConstr performance regression with 8.2.1 -------------------------------------+------------------------------------- Reporter: albertov | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1-rc1 Keywords: | Operating System: Linux Architecture: | Type of failure: Compile-time Unknown/Multiple | performance bug Test Case: | Blocked By: Blocking: | Related Tickets: Differential Rev(s): | Wiki Page: -------------------------------------+------------------------------------- A propietary application I've upgraded to compile with 8.2.1-rc1 makes GHC manifest a performance regression when producing optimized ({{{-O2}}}) code. After a crude analysis of a compile log produced with {{{-show-passes}}} it appears that the biggest slowdown is in the Simplifier and SpecConstr phases when compiling the "main" modules. Theres a conversation at GHC Users with more details: https://mail.haskell.org/pipermail/glasgow-haskell- users/2017-June/026543.html. The data, the results and the script to to analyse it live at: https://gist.github.com/albertov/145ac5c01bfbadc5c9ff55e9c5c2e50e I'll try to update this ticket with a way to to reproduce it as soon as I can (if I can) -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/13817 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler
#13817: Simplifier and SpecConstr performance regression with 8.2.1 -------------------------------------+------------------------------------- Reporter: albertov | Owner: (none) Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 8.2.1-rc1 Resolution: | Keywords: Operating System: Linux | Architecture: Type of failure: Compile-time | Unknown/Multiple performance bug | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Comment (by simonpj): How big is the slow-down? On the thread you also mentioned that execution time has improved significantly. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/13817#comment:1 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler
#13817: Simplifier and SpecConstr performance regression with 8.2.1 -------------------------------------+------------------------------------- Reporter: albertov | Owner: (none) Type: bug | Status: infoneeded Priority: normal | Milestone: Component: Compiler | Version: 8.2.1-rc1 Resolution: | Keywords: Operating System: Linux | Architecture: Type of failure: Compile-time | Unknown/Multiple performance bug | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Changes (by dfeuer): * status: new => infoneeded Old description:
A propietary application I've upgraded to compile with 8.2.1-rc1 makes GHC manifest a performance regression when producing optimized ({{{-O2}}}) code.
After a crude analysis of a compile log produced with {{{-show-passes}}} it appears that the biggest slowdown is in the Simplifier and SpecConstr phases when compiling the "main" modules.
Theres a conversation at GHC Users with more details: https://mail.haskell.org/pipermail/glasgow-haskell- users/2017-June/026543.html. The data, the results and the script to to analyse it live at: https://gist.github.com/albertov/145ac5c01bfbadc5c9ff55e9c5c2e50e
I'll try to update this ticket with a way to to reproduce it as soon as I can (if I can)
New description: A proprietary application I've upgraded to compile with 8.2.1-rc1 makes GHC manifest a performance regression when producing optimized ({{{-O2}}}) code. After a crude analysis of a compile log produced with {{{-show-passes}}} it appears that the biggest slowdown is in the Simplifier and SpecConstr phases when compiling the "main" modules. Theres a conversation at GHC Users with more details: https://mail.haskell.org/pipermail/glasgow-haskell- users/2017-June/026543.html. The data, the results and the script to to analyse it live at: https://gist.github.com/albertov/145ac5c01bfbadc5c9ff55e9c5c2e50e I'll try to update this ticket with a way to to reproduce it as soon as I can (if I can). -- Comment: I don't think we can do anything about this without a proper reproduction. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/13817#comment:2 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler
participants (1)
-
GHC