osa1 (Ömer Sinan Ağacan)
User

Projects

User does not belong to any projects.

User Details

User Since
Jun 6 2014, 2:22 AM (224 w, 3 d)

Recent Activity

Fri, Sep 21

osa1 committed rGHC29f1c556dd56: Remove redundant slop zeroing (authored by osa1).
Remove redundant slop zeroing
Fri, Sep 21, 2:32 AM
osa1 committed rGHC66c17293648f: Fix slop zeroing for AP_STACK eager blackholes in debug build (authored by osa1).
Fix slop zeroing for AP_STACK eager blackholes in debug build
Fri, Sep 21, 2:04 AM
osa1 closed D5165: Fix slop zeroing for AP_STACK eager blackholes in debug build.
Fri, Sep 21, 2:04 AM
osa1 updated the diff for D5165: Fix slop zeroing for AP_STACK eager blackholes in debug build.
  • Fix closure size
Fri, Sep 21, 1:17 AM

Thu, Sep 20

osa1 planned changes to D5165: Fix slop zeroing for AP_STACK eager blackholes in debug build.

I did some refactoring but haven't slow validated yet. Marking as "changes planned" to avoid merging in the meantime.

Thu, Sep 20, 9:53 AM
osa1 updated the diff for D5165: Fix slop zeroing for AP_STACK eager blackholes in debug build.
  • Comments and fixes
Thu, Sep 20, 9:53 AM
osa1 updated the diff for D5165: Fix slop zeroing for AP_STACK eager blackholes in debug build.
  • Comments
  • Revert AP_STACK_NOUPD change
  • Implement a variant of OVERWRITING_CLOSURE that takes a size arg
Thu, Sep 20, 8:53 AM
osa1 updated the diff for D5165: Fix slop zeroing for AP_STACK eager blackholes in debug build.
  • Revert AP_STACK_NOUPD change
Thu, Sep 20, 8:37 AM
osa1 committed rGHC077b92fa3983: Remove -Waggregate-return when building RTS (authored by osa1).
Remove -Waggregate-return when building RTS
Thu, Sep 20, 8:06 AM
osa1 closed D5164: Remove -Waggregate-return when building RTS.
Thu, Sep 20, 8:06 AM
osa1 updated the summary of D5164: Remove -Waggregate-return when building RTS.
Thu, Sep 20, 8:04 AM
osa1 retitled D5164: Remove -Waggregate-return when building RTS from Fix slow validate build to Remove -Waggregate-return when building RTS.
Thu, Sep 20, 8:02 AM
osa1 updated the diff for D5164: Remove -Waggregate-return when building RTS.

Update

Thu, Sep 20, 8:02 AM
osa1 added a comment to D5164: Remove -Waggregate-return when building RTS.

Sure!

Thu, Sep 20, 7:59 AM
osa1 updated the diff for D5165: Fix slop zeroing for AP_STACK eager blackholes in debug build.
  • Comments
Thu, Sep 20, 5:51 AM
osa1 updated the test plan for D5165: Fix slop zeroing for AP_STACK eager blackholes in debug build.
Thu, Sep 20, 5:40 AM
osa1 created D5165: Fix slop zeroing for AP_STACK eager blackholes in debug build.
Thu, Sep 20, 5:38 AM
osa1 created D5164: Remove -Waggregate-return when building RTS.
Thu, Sep 20, 5:34 AM
osa1 accepted D5162: GHCi does not honour implicit `module Main (main) where` for re-exported `main`s.

Thanks! LGTM but added an inline comment.

Thu, Sep 20, 1:29 AM

Wed, Sep 19

osa1 added inline comments to D5153: Add RubbishLit for absent bindings of UnliftedRep.
Wed, Sep 19, 6:14 AM

Tue, Sep 18

osa1 added a comment to D5145: Don't shortcut SRTs for static functions (#15544).

Sorry for the noise -- I tried again and this seems to fix the segfaults. I guess I used an older binary of the app before.

Tue, Sep 18, 10:25 AM
osa1 added a comment to D5145: Don't shortcut SRTs for static functions (#15544).

Just a note: the reproducer still segfaults with this patch.

Tue, Sep 18, 1:46 AM

Mon, Sep 17

osa1 committed rGHC4edc6d64d1bc: Users guide: EmptyDataDecls on by default (authored by chris-martin).
Users guide: EmptyDataDecls on by default
Mon, Sep 17, 11:52 PM

Sun, Sep 16

osa1 committed rGHCc6bff5261236: Fix for #13862: Optional "-v" not allowed with :load in GHCi (authored by RolandSenn).
Fix for #13862: Optional "-v" not allowed with :load in GHCi
Sun, Sep 16, 11:27 AM
osa1 closed D5122: Fix for #13862: Optional "-v" not allowed with :load in GHCi.
Sun, Sep 16, 11:27 AM
osa1 accepted D5122: Fix for #13862: Optional "-v" not allowed with :load in GHCi.

Thanks! Merging...

Sun, Sep 16, 11:24 AM
osa1 requested changes to D5122: Fix for #13862: Optional "-v" not allowed with :load in GHCi.

Indeed this introduces 14 failures. Marking as "requested changes" to avoid merging.

Sun, Sep 16, 3:18 AM
osa1 accepted D5122: Fix for #13862: Optional "-v" not allowed with :load in GHCi.

Thanks. This looks good to me. There are some test suite failures so we should check those are existing errors or introduced with this patch.

Sun, Sep 16, 2:04 AM

Fri, Sep 14

osa1 added inline comments to D5145: Don't shortcut SRTs for static functions (#15544).
Fri, Sep 14, 3:24 AM
osa1 added inline comments to D5145: Don't shortcut SRTs for static functions (#15544).
Fri, Sep 14, 2:29 AM

Wed, Sep 12

osa1 added a comment to D5145: Don't shortcut SRTs for static functions (#15544).

I updated the ticket -- there are still a lot of segfaults and assertion errors.

Wed, Sep 12, 4:01 AM
osa1 added inline comments to D5145: Don't shortcut SRTs for static functions (#15544).
Wed, Sep 12, 2:52 AM
osa1 added a comment to D5145: Don't shortcut SRTs for static functions (#15544).

This seems to fix the segfault when I apply this with the STM patch, but the reproducer now fails with another error (see the ticket).

Wed, Sep 12, 2:28 AM
osa1 committed rGHCce23451c2c77: Refactor info table entry error messages (authored by osa1).
Refactor info table entry error messages
Wed, Sep 12, 1:13 AM
osa1 closed D5143: Refactor info table entry error messages.
Wed, Sep 12, 1:12 AM

Tue, Sep 11

osa1 created D5144: Revert incorrect STM wakeup optimisation.
Tue, Sep 11, 7:08 AM
osa1 updated the test plan for D5143: Refactor info table entry error messages.
Tue, Sep 11, 6:51 AM
osa1 added a comment to D5143: Refactor info table entry error messages.

@simonmar This seems to work fine for stg_GCD_CAF but I don't understand Cmm calling conventions too well so I don't know if this change makes sense. In particular I assumed that the object that we entered is always in R1. Could you take a look?

Tue, Sep 11, 6:49 AM
osa1 created D5143: Refactor info table entry error messages.
Tue, Sep 11, 6:47 AM

Mon, Sep 10

osa1 committed rGHCd36b1ffac996: Build debugged prof runtimes (authored by osa1).
Build debugged prof runtimes
Mon, Sep 10, 5:54 AM
osa1 closed D5140: Build debugged prof runtimes.
Mon, Sep 10, 5:54 AM
osa1 committed rGHC3cc3edf30f37: Update UnsafeReenter test (authored by osa1).
Update UnsafeReenter test
Mon, Sep 10, 3:19 AM
osa1 closed D5136: Update UnsafeReenter test.
Mon, Sep 10, 3:19 AM
osa1 created D5140: Build debugged prof runtimes.
Mon, Sep 10, 1:43 AM

Sun, Sep 9

osa1 added a comment to D5136: Update UnsafeReenter test.

@simonmar anything else?

Sun, Sep 9, 3:59 AM
osa1 updated the diff for D5136: Update UnsafeReenter test.
  • Update
Sun, Sep 9, 3:59 AM
osa1 updated the diff for D5136: Update UnsafeReenter test.
  • Update
Sun, Sep 9, 3:57 AM

Sat, Sep 8

osa1 added a comment to D5122: Fix for #13862: Optional "-v" not allowed with :load in GHCi.

Introducing the new DynFlag isInteractive is a first step in this direction

Sat, Sep 8, 9:57 AM
osa1 committed rGHC510c5f4f22ac: Avoid creating unevaluated Int thunks when iterating in GHC.Foreign (authored by ndmitchell).
Avoid creating unevaluated Int thunks when iterating in GHC.Foreign
Sat, Sep 8, 4:17 AM
osa1 committed rGHC62cd44013eaa: Refactor Foreign.Marshal modules for more modern style (authored by osa1).
Refactor Foreign.Marshal modules for more modern style
Sat, Sep 8, 1:57 AM
osa1 closed D5124: Refactor Foreign.Marshal modules for more modern style.
Sat, Sep 8, 1:57 AM
osa1 added a comment to D5124: Refactor Foreign.Marshal modules for more modern style.

So unless I'm wildly misunderstanding here

Sat, Sep 8, 1:55 AM

Fri, Sep 7

osa1 added a comment to D5122: Fix for #13862: Optional "-v" not allowed with :load in GHCi.

Why not simply change the error message to

Fri, Sep 7, 6:28 AM
osa1 added a comment to D5136: Update UnsafeReenter test.

@simonmar I deleted the ghci test as ghci is also using threaded runtime which causes this test to hang.

Fri, Sep 7, 6:17 AM
osa1 created D5136: Update UnsafeReenter test.
Fri, Sep 7, 6:09 AM
osa1 committed rGHCd9a26c7e8d9b: Various RTS bug fixes: (authored by osa1).
Various RTS bug fixes:
Fri, Sep 7, 1:59 AM
osa1 closed D5134: Various RTS bug fixes:.
Fri, Sep 7, 1:59 AM
osa1 added a comment to D5124: Refactor Foreign.Marshal modules for more modern style.

One of the reasons why I wanted reviews for this patch is because it changes visible arities of some definitions (e.g. pooledReallocArray previously had arity 0, now it has arity 3) and that may effect inlining decisions. I don't think it'll cause any problems downstream but I'm not sure. Any opinions @dfeuer @RyanGlScott @bgamari ?

Fri, Sep 7, 1:34 AM
osa1 added inline comments to D5126: Implement feature #3427.
Fri, Sep 7, 1:27 AM

Thu, Sep 6

osa1 committed rGHCc6fbac6a6a69: Fix a race between GC threads in concurrent scavenging (authored by osa1).
Fix a race between GC threads in concurrent scavenging
Thu, Sep 6, 7:53 AM
osa1 closed D5115: Fix a race between GC threads in concurrent scavenging.
Thu, Sep 6, 7:53 AM
osa1 updated the summary of D5115: Fix a race between GC threads in concurrent scavenging.
Thu, Sep 6, 7:52 AM
osa1 updated the diff for D5115: Fix a race between GC threads in concurrent scavenging.
  • Remove comment
Thu, Sep 6, 7:47 AM
osa1 created D5134: Various RTS bug fixes:.
Thu, Sep 6, 7:41 AM
osa1 committed rGHC16bc7ae8b191: Remove an incorrect assertion in threadPaused: (authored by osa1).
Remove an incorrect assertion in threadPaused:
Thu, Sep 6, 3:40 AM
osa1 closed D5133: Remove an incorrect assertion in threadPaused:.
Thu, Sep 6, 3:40 AM
osa1 updated the test plan for D5133: Remove an incorrect assertion in threadPaused:.
Thu, Sep 6, 3:39 AM
osa1 updated the summary of D5133: Remove an incorrect assertion in threadPaused:.
Thu, Sep 6, 3:39 AM
osa1 added inline comments to D5126: Implement feature #3427.
Thu, Sep 6, 2:51 AM
osa1 abandoned D5130: Fix -po RTS flag.

Ah! I see now that -po does not imply -p. That's not too clear in the man page but it's clear in the user manual. Sorry for the noise.

Thu, Sep 6, 2:46 AM
osa1 created D5133: Remove an incorrect assertion in threadPaused:.
Thu, Sep 6, 2:37 AM
osa1 requested changes to D5126: Implement feature #3427.

Thanks for doing this! I haven't reviewed this in detail but you can't make constructor as it's currently a valid identifier and making it a keyword breaks programs that use it as an identifier. In fact, I can't even boot GHC with this patch as some of the ghc libraries use constructor as an identifier.

Thu, Sep 6, 2:07 AM

Wed, Sep 5

osa1 accepted D5010: testsuite: Use bools for booleans, not ints.

Thanks!

Wed, Sep 5, 6:37 AM
osa1 added inline comments to D5010: testsuite: Use bools for booleans, not ints.
Wed, Sep 5, 6:24 AM
osa1 accepted D4958: testsuite: Add test for #15368.

Thanks!

Wed, Sep 5, 6:24 AM
osa1 created D5130: Fix -po RTS flag.
Wed, Sep 5, 6:08 AM

Tue, Sep 4

osa1 accepted D5093: Compiler panic on invalid syntax (unterminated pragma).

Thanks!

Tue, Sep 4, 6:52 AM
osa1 committed rGHCfa3143c76ac7: Fix typos in -Wsimplifiable-class-constraints flag docs (authored by sergv).
Fix typos in -Wsimplifiable-class-constraints flag docs
Tue, Sep 4, 3:13 AM

Mon, Sep 3

osa1 created D5124: Refactor Foreign.Marshal modules for more modern style.
Mon, Sep 3, 7:08 AM
osa1 committed rGHCd1514e8f0e14: Remove duplicate "since" field in glasgow_exts.rst (authored by JoshPrice247).
Remove duplicate "since" field in glasgow_exts.rst
Mon, Sep 3, 6:07 AM

Fri, Aug 31

osa1 updated the Trac tickets for D5119: Skip eventlog tests in GHCi way.
Fri, Aug 31, 4:50 AM
osa1 created D5119: Skip eventlog tests in GHCi way.
Fri, Aug 31, 4:48 AM
osa1 updated the test plan for D5115: Fix a race between GC threads in concurrent scavenging.
Fri, Aug 31, 4:12 AM
osa1 updated the diff for D5115: Fix a race between GC threads in concurrent scavenging.
  • Add prototype
Fri, Aug 31, 2:59 AM

Thu, Aug 30

osa1 created D5115: Fix a race between GC threads in concurrent scavenging.
Thu, Aug 30, 6:04 AM
osa1 updated the Trac tickets for D5059: testsuite: Save performance metrics in git notes..
Thu, Aug 30, 6:00 AM
osa1 added a comment to D5093: Compiler panic on invalid syntax (unterminated pragma).

Ah so you wanted to update languagePragParseError too? Why? To be clear, your wording is fine (but note that I'm also not a native speaker ;-), my concern is about consistency and not the choice of words of tense.

Thu, Aug 30, 2:14 AM
osa1 added a comment to D5093: Compiler panic on invalid syntax (unterminated pragma).

Thanks @RolandSenn! Added a minor comment about the error message.

Thu, Aug 30, 1:39 AM

Wed, Aug 29

osa1 committed rGHC36c1431d9d2d: Fixed typo in exponent example (authored by chris-bacon <chrisbacon2009@hotmail.com>).
Fixed typo in exponent example
Wed, Aug 29, 4:34 AM

Aug 22 2018

osa1 committed rGHCab55b4ddb717: Revert "Properly tag fun field of PAPs generated by ap_0_fast" (authored by osa1).
Revert "Properly tag fun field of PAPs generated by ap_0_fast"
Aug 22 2018, 1:05 AM
osa1 added a reverting change for rGHC2693eb11f55f: Properly tag fun field of PAPs generated by ap_0_fast: rGHCab55b4ddb717: Revert "Properly tag fun field of PAPs generated by ap_0_fast".
Aug 22 2018, 1:05 AM
osa1 added a comment to D5051: Properly tag fun field of PAPs generated by ap_0_fast.

I reverted this commit with ab55b4ddb7.

Aug 22 2018, 1:05 AM
osa1 added a comment to D5051: Properly tag fun field of PAPs generated by ap_0_fast.

@bgamari this should be reverted as it breaks another program (not caught by the test suite -- see my comments above). Sorry I should have marked this as "changes planned".

Aug 22 2018, 12:51 AM

Aug 20 2018

osa1 added a comment to D5078: Expose the StableName constructor.

Found it, for reference: https://github.com/ghc-proposals/ghc-proposals/blob/master/proposals/0021-unlifted-array.rst

Aug 20 2018, 2:23 AM
osa1 accepted D5078: Expose the StableName constructor.

Once we have UnliftedArray#

Aug 20 2018, 2:09 AM

Aug 17 2018

osa1 committed rGHC23f6f31dd66d: Document default value of +RTS -N in user's guide (authored by osa1).
Document default value of +RTS -N in user's guide
Aug 17 2018, 2:57 AM
osa1 closed D5065: Document default value of +RTS -N in user's guide.
Aug 17 2018, 2:57 AM

Aug 14 2018

osa1 updated the diff for D5065: Document default value of +RTS -N in user's guide.
  • Reword
Aug 14 2018, 1:22 AM

Aug 13 2018

osa1 retitled D5065: Document default value of +RTS -N in user's guide from Document default value of +RTS -N to Document default value of +RTS -N in user's guide.
Aug 13 2018, 10:56 AM
osa1 created D5065: Document default value of +RTS -N in user's guide.
Aug 13 2018, 10:55 AM