- ghc-tickets - Haskell.org

Re: [GHC] #7015: Add support for 'static'
by GHC 29 Jan '14

29 Jan '14

#7015: Add support for 'static' -------------------------------------+------------------------------------ Reporter: edsko | Owner: Type: feature request | Status: new Priority: normal | Milestone: 7.8.1 Component: Compiler | Version: 7.4.2 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by mboes): Relevant thread on the GHC Users mailing list: http://www.haskell.org/pipermail/glasgow-haskell- users/2014-January/024571.html. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7015#comment:2> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #703: all binaries built by ghc have executable stacks
by GHC 29 Jan '14

29 Jan '14

#703: all binaries built by ghc have executable stacks ----------------------------+---------------------------------------------- Reporter: duncan | Owner: ezyang Type: merge | Status: merge Priority: normal | Milestone: 6.6.1 Component: | Version: 7.6.3 Compiler (NCG) | Keywords: Resolution: | Architecture: Unknown/Multiple Operating System: Linux | Difficulty: Moderate (less than a day) Type of failure: | Blocked By: None/Unknown | Related Tickets: Test Case: N/A | Blocking: | ----------------------------+---------------------------------------------- Comment (by juhpetersen): Replying to [comment:23 juhpetersen]: > (I noticed that HSbase*.o has executable stack set but I guess that should not affect this, right?) Actually for unpatched ghc-7.6.3 (testing on i686) /usr/lib/ghc-7.6.3/base-4.6.0.1/HSbase-4.6.0.1.o does not seems to have .note.GNU-stack Anyway something must be different in 7.8. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/703#comment:29> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #703: all binaries built by ghc have executable stacks
by GHC 29 Jan '14

29 Jan '14

#703: all binaries built by ghc have executable stacks ----------------------------+---------------------------------------------- Reporter: duncan | Owner: ezyang Type: merge | Status: merge Priority: normal | Milestone: 6.6.1 Component: | Version: 7.6.3 Compiler (NCG) | Keywords: Resolution: | Architecture: Unknown/Multiple Operating System: Linux | Difficulty: Moderate (less than a day) Type of failure: | Blocked By: None/Unknown | Related Tickets: Test Case: N/A | Blocking: | ----------------------------+---------------------------------------------- Changes (by juhpetersen): * cc: simonmar (added) Comment: Replying to [comment:26 ezyang]: > Perhaps some library HSbase is using has an executable stack? Maybe the build should be cleaned and tried again? I am pretty sure I was testing with a fresh build from the buildsystem. I thought HSbase.o is only used by ghci? Is it also used by other parts? (I tried also dynamic linking with unpatched ghc-7.6.3 and it also leads to executable stack.) -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/703#comment:28> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #7629: segmentation fault in compiled program, involves gtk, selinux
by GHC 29 Jan '14

29 Jan '14

#7629: segmentation fault in compiled program, involves gtk, selinux -------------------------+------------------------------------------------- Reporter: | Owner: simonmar wgmitchener | Status: closed Type: bug | Milestone: 7.6.2 Priority: high | Version: 7.4.2 Component: | Keywords: segmentation fault, Runtime System | multithreading, selinux, gtk Resolution: | Architecture: x86 fixed | Difficulty: Unknown Operating System: | Blocked By: Linux | Related Tickets: Type of failure: | Runtime crash | Test Case: | Blocking: | -------------------------+------------------------------------------------- Changes (by juhpetersen): * cc: simonmar (added) Comment: I note for the record that this didn't make ghc-7.6.3 but will be in ghc-7.8. I am finally backporting the patch to Fedora now. wgmitchener: Thank you again for fixing this. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7629#comment:31> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #4114: Add a flag to remove/delete intermediate files generated by GHC
by GHC 29 Jan '14

29 Jan '14

#4114: Add a flag to remove/delete intermediate files generated by GHC -------------------------------------+------------------------------------ Reporter: guest | Owner: iduhetonas Type: feature request | Status: new Priority: low | Milestone: 7.6.2 Component: Compiler | Version: 6.10.4 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: #2258 -------------------------------------+------------------------------------ Changes (by iduhetonas): * owner: => iduhetonas * difficulty: => Unknown -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/4114#comment:12> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #5928: INLINABLE fails to specialize in presence of simple wrapper
by GHC 28 Jan '14

28 Jan '14

#5928: INLINABLE fails to specialize in presence of simple wrapper -------------------------------------+------------------------------------ Reporter: tibbe | Owner: Type: bug | Status: new Priority: normal | Milestone: 7.6.2 Component: Compiler | Version: 7.4.1 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Changes (by crockeea): * cc: ecrockett0@… (added) -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/5928#comment:22> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #1042: Floating point exception
by GHC 28 Jan '14

28 Jan '14

#1042: Floating point exception ------------------------------------------------+-------------------------- Reporter: dons | Owner: igloo Type: merge | Status: closed Priority: normal | Milestone: 6.6.1 Component: Compiler | Version: 6.6 Resolution: fixed | Keywords: Operating System: Unknown/Multiple | Architecture: x86 Type of failure: Incorrect result at runtime | Difficulty: Test Case: | Unknown Blocking: | Blocked By: | Related Tickets: #8695 ------------------------------------------------+-------------------------- Changes (by hvr): * failure: => Incorrect result at runtime * related: => #8695 -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/1042#comment:10> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #7994: Make foldl into a good consumer
by GHC 28 Jan '14

28 Jan '14

#7994: Make foldl into a good consumer -------------------------------------+------------------------------------ Reporter: simonpj | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.6.3 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by Joachim Breitner <mail@…>): In [changeset:"4a8ffcf55c897851b534c700951a0b5bdd43eb97/base"]: {{{ #!CommitTicketReference repository="base" revision="4a8ffcf55c897851b534c700951a0b5bdd43eb97" go-ify foldr2 This helps with the changes in #7994, but might also generally be a good idea (ignore the runtime): -------------------------------------------------------------------------------- Program Size Allocs Runtime Elapsed TotalMem fft2 -0.1% -1.5% 0.07 0.07 +0.0% fibheaps +0.0% -17.2% 0.03 0.03 +0.0% fluid +0.5% -0.7% 0.01 0.01 +0.0% integrate +0.0% -0.9% 0.16 0.16 +0.0% rewrite +0.0% -1.1% 0.02 0.02 +0.0% -------------------------------------------------------------------------------- Min -0.1% -17.2% -1.6% +0.0% +0.0% Max +0.5% +0.0% +107.7% +106.2% +11.3% Geometric Mean +0.0% -0.2% +23.7% +23.9% +0.1% }}} -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7994#comment:14> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #7994: Make foldl into a good consumer
by GHC 28 Jan '14

28 Jan '14

#7994: Make foldl into a good consumer -------------------------------------+------------------------------------ Reporter: simonpj | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.6.3 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by nomeata): JFTR: Simon discussed this with me and encouraged me to pursue the analysis (but don’t merge it before 7.8 has been branched off). But even more so I’d like to record a few things about the `oneShot` approach: * Would also be nice, and can still be revived if we find cases where the analysis fails. * It should always be inlined when passed two arguments. * Could be made to inline very very late, but then it should have a strictness signature. * In that case, one would have to make sure it does not get in the way of CPR or other analysis. -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7994#comment:13> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0

Re: [GHC] #7994: Make foldl into a good consumer
by GHC 28 Jan '14

28 Jan '14

#7994: Make foldl into a good consumer -------------------------------------+------------------------------------ Reporter: simonpj | Owner: Type: bug | Status: new Priority: normal | Milestone: Component: Compiler | Version: 7.6.3 Resolution: | Keywords: Operating System: Unknown/Multiple | Architecture: Unknown/Multiple Type of failure: None/Unknown | Difficulty: Unknown Test Case: | Blocked By: Blocking: | Related Tickets: -------------------------------------+------------------------------------ Comment (by nomeata): I have come up with a stand-alone analysis that tries to find out the “caller arity” of a function. Quite a bit like the usage demand, but different from the demand analyser in a few aspects: It does not bother about the demand put on function arguments by a function (so much less fixed-pointing due to that), and it currently looks only for tail-calls (or almost-tail-calls; some post-processing of return values a after the “tail call” is ok, as long as it does not do any relevant calls). It is an analysis phase that puts the result in the IdInfo, and the simplifier will eta-expand functions where the „caller arity” is higher than the manifest arity. It has zero effect on the current code base. If I implement `foldl` via `foldr` (but ''without'' the use of `oneShot`), I get essentially the same as the result as without the analysis, but with the explicit `oneShot`: {{{ Min -0.2% -74.5% -3.0% -3.0% -50.0% Max +0.9% +1.8% +2.5% +2.5% +14.8% Geometric Mean -0.0% -3.7% -0.3% -0.3% -0.5% }}} It still does not help with `bernoulli`, and probably nothing will: Implementing `foldl` as `foldr` will always yield bad code if list fusion does not work all the way and the `build` (or in this case, `build2`) is not completely gone. But if list fusion works, the results can be very good The `oneShot` change is much simpler, has little risk of introducing bugs and does not increase compilation times. The “caller arity” pass, OTOH, may optimise other code as well (but at least nofib does not seem to contain much such code), and can likely be made more precise. But the conclusion is that `foldl` can be made a good consumer! -- Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/7994#comment:12> GHC <http://www.haskell.org/ghc/> The Glasgow Haskell Compiler

1 0