Skip to content
This repository has been archived by the owner on Jul 4, 2023. It is now read-only.

ghc 7.8.2 with --devel #29084

Closed
wants to merge 1 commit into from
Closed

ghc 7.8.2 with --devel #29084

wants to merge 1 commit into from

Conversation

glehmann
Copy link
Contributor

@glehmann glehmann commented May 9, 2014

ghc 7.8 uses clang internally. clang is used to build the ghc 7.8 on mavericks
64 bit, and gcc is used on older OS and 32 bit builds. Also advise to use
cabal-install in the caveat.

ghc 7.8 uses clang internally.
clang is used to build the ghc 7.8 on mavericks 64 bit, and gcc is used on
older OS and 32 bit builds.
Also advise to use cabal-install in the caveat.
@glehmann glehmann mentioned this pull request May 9, 2014
@@ -144,11 +137,17 @@ def install
system "make"
# -j1 fixes an intermittent race condition
system "make", "-j1", "install"
if build.devel?
# use clang, even when gcc was used to build ghc
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hmm, seems potentially weird. Can clang not be used to build too?

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

no, because a binary ghc 7.8 is required to booststrap the build, and it is available to download only for mavericks 64 bit

@MikeMcQuaid
Copy link
Member

Cheers for doing this!

@Josh-Tilles
Copy link
Contributor

By the way, when I built GHC 7.8.2 on my machine (Mavericks) with the --tests flag, I got the following outcome:

Unexpected results from:
TEST="T8958 T8256 T3064 T4801 T6048"

OVERALL SUMMARY for test run started at Wed May  7 17:26:50 2014 EDT
 0:08:11 spent to go through
    3639 total tests, which gave rise to
   14083 test cases, of which
   10738 were skipped

      43 had missing libraries
    3261 expected passes
      36 expected failures

       0 caused framework failures
       0 unexpected passes
       5 unexpected failures

Unexpected failures:
   codeGen/should_run    T8256 [exit code non-0] (normal)
   perf/compiler         T3064 [stat not good enough] (normal)
   perf/compiler         T4801 [stat not good enough] (normal)
   perf/compiler         T6048 [stat not good enough] (optasm)
   roles/should_compile  T8958 [stderr mismatch] (normal)

(I don't know whether that's alarming or not.)

In addition, I was able to build pandoc and shellcheck from source with GHC 7.8.2. Both formulae also passed brew test.

Caveat: I used my version of the formula because glehmann's didn't yet exist.

@Josh-Tilles
Copy link
Contributor

Heads up: GHC 7.8.2 will fail to build cabal-install from source. For details, check out:

The workaround could be as simple as adding --no-doc to the cabal-install.rb line system "sh", "bootstrap.sh". Or perhaps when GHC 7.8 gets moved out of --devel, we can simultaneously change the cabal-install formula to use a wrapper script that injects --ghc-options=-optP-P.

@adamv
Copy link
Contributor

adamv commented May 11, 2014

Given issues with 7.8.x at large, is it OK to merge this as a devel version now, or is there more work to do first?

@MikeMcQuaid
Copy link
Member

@adamv 👍

@adamv adamv closed this in 9651028 May 11, 2014
@Homebrew Homebrew locked and limited conversation to collaborators Feb 17, 2016
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants