- User Since
- Jul 28 2014, 8:50 AM (269 w, 3 h)
Feb 19 2018
Feb 16 2018
Nov 9 2017
To be conservative it would really need to be the same binary.
I didn't need this anymore after having my use case, inline-java, use GHC plugins.
Didn't need this in the end since the use case I had, inline-java, moved to using ghc plugins.
Sep 13 2017
Sep 11 2017
Where was the previous check, the one that TH bypassed?
Sep 8 2017
Aug 21 2017
Dropping the patch as recommended in https://ghc.haskell.org/trac/ghc/ticket/14090#comment:15
Aug 18 2017
- Mention addCorePlugin in the user guide.
Aug 14 2017
There seems to be more than one place where the bindings are dropped. simpleOptPgm is one such place, but later on the simplifier also does drop the bindings prematurely.
- Be explicit on the non-traversed CoreExpr constructors.
Aug 12 2017
- Improve comments.
- Improve comments.
Aug 4 2017
Jun 5 2017
Besides the problem with parsing of patterns, this patch wouldn't work for splices within brackets. We'd need a better design.
May 24 2017
Mar 7 2017
Mar 6 2017
https://ghc.haskell.org/trac/ghc/ticket/13366 talks about controlling compilation flags, but it doesn't look like this patch is addressing that. Perhaps update the ticket to explain the plan?
Feb 10 2017
@nomeata please, could you remove the branches fd/fix-travis and fd/fix-travis2? Some server hook is preventing me from doing so.
Relaxed even more the test so it works with clang.
Travis uses an older gcc version which doesn't print the offending line in the error message. https://phabricator.haskell.org/rGHCb9bebd8cedcc
=====> TH_addCStub1(normal) 4122 of 5714 [0, 0, 0]--- /tmp/ghctest-u7xmp5/test spaces/./th/TH_addCStub2.run/TH_addCStub2.stderr.normalised 2017-02-09 17:41:03.087318367 +0000 +++ /tmp/ghctest-u7xmp5/test spaces/./th/TH_addCStub2.run/TH_addCStub2.comp.stderr.normalised 2017-02-09 17:41:03.087318367 +0000 @@ -1,6 +1,4 @@
Checking the stderr of the compiler and returning the right exit code is non-trivial in the current test framework. Couldn't tell either if it will work in all platforms. Travis build passes with a former version of this fix and I hope it will do it again with this change. The ongoing build is here https://travis-ci.org/ghc/ghc/builds/200324037
Feb 9 2017
Feb 8 2017
Addressed Ben's comments.
Did minor touch ups to tests.
Reverted unintended change in GHCi.Message.
Jan 26 2017
Thanks for taking a look. We can manage without this. Let's wait until we can discuss it then.
Jan 25 2017
I was aiming towards merging this before the freeze deadline for ghc-8.2.1. We might still be able to address the open issues and new comments on time, provided we get some feedback and we find no major blockers. That said, I understand everybody might be just too busy for this plan to work.
Jan 24 2017
Remove uses of unsafeCoerce by changing the ModFinalizers type.
Jan 23 2017
This patch addresses the intended issue but has the problems discussed in Trac #12778.
Jan 13 2017
Jan 11 2017
Moved StaticPtrInfo construction to StaticPtrTable.hs.
bgamari, I agree with you. The point would be that the two open discussions do not look like major concerns, and if they were the patch surely wouldn't be accepted.
Jan 10 2017
I'm thinking of landing this last thing on Friday, given that it is accepted. Can address requests after that, but let me know if I should address any more concerns in advance.
Jan 9 2017
Address more comments.
Addressed comments by RyanGlScott and some of simonpj's.
Jan 6 2017
hm, some misuse of arc land, I suppose.
I hope the trac ticket provides the necessary context to motivate the patch. But let me know if the commit message needs to be more descriptive.
This approach is favored https://phabricator.haskell.org/D2930 over the current patch.
Dec 20 2016
Address review comments.
Fixed up test after rebasing master.
Nov 19 2016
Nov 16 2016
Address review comments.
Ben, please merge at your convenience. If I don't address the comments on time, I will push them to master.
Nov 15 2016
I hate withdrawing an accepted PR, but I earned more insight today that could make disabling unpacking of StaticPtrs unnecessary. See the discussion on the related ticket for details.
Note taken. I'll address comments after @simonpj feedback.
Further improve comments.
Nov 14 2016
Nov 1 2016
This was landed as commit 0b70ec0c3b72a7f87776743e64b47b65ef0ca4a5 to master.