[Git][ghc/ghc][wip/andreask/ci_validate] 2 commits: Rts/elf-linker: Upcast to 64bit to satisfy format string.

Andreas Klebinger gitlab at gitlab.haskell.org
Fri Nov 27 10:40:21 UTC 2020



Andreas Klebinger pushed to branch wip/andreask/ci_validate at Glasgow Haskell Compiler / GHC


Commits:
006ac733 by Andreas Klebinger at 2020-11-27T11:39:52+01:00
Rts/elf-linker: Upcast to 64bit to satisfy format string.

The elf size is 32bit on 32bit builds and 64 otherwise.
We just upcast to 64bits before printing now.

- - - - -
878f4abd by Andreas Klebinger at 2020-11-27T11:39:53+01:00
Use validate flavour for all CI builds.

This also means we compile GHC with -O1 instead of -O2 for some
platforms for CI. As a result a lot of test metrics
got worse which we now have to accept.

-------------------------
Metric Increase:
    ManyAlternatives
    ManyConstructors
    MultiLayerModules
    Naperian
    T10421
    T12150
    T12227
    T12234
    T12425
    T12545
    T12707
    T13035
    T13253
    T13253-spj
    T13379
    T13719
    T14697
    T18282
    T18698a
    T18698b
    T1969
    T3064
    T3294
    T4801
    T5321FD
    T5321Fun
    T5631
    T6048
    T783
    T9020
    T9203
    T9233
    T9872a
    T9872b
    T9872c
    T9872d
    T9961
    haddock.Cabal
    haddock.base
    parsing001
    T5642
    WWRec
-------------------------

- - - - -


2 changed files:

- .gitlab-ci.yml
- rts/linker/Elf.c


Changes:

=====================================
.gitlab-ci.yml
=====================================
@@ -442,7 +442,7 @@ validate-x86_64-darwin:
     LANG: "en_US.UTF-8"
     CONFIGURE_ARGS: "--with-intree-gmp"
     TEST_ENV: "x86_64-darwin"
-    BUILD_FLAVOUR: "perf"
+    BUILD_FLAVOUR: "validate"
   after_script:
     - cp -Rf $HOME/.cabal cabal-cache
     - .gitlab/ci.sh clean
@@ -493,7 +493,7 @@ validate-x86_64-darwin:
   tags:
     - x86_64-linux
   variables:
-    BUILD_FLAVOUR: "perf"
+    BUILD_FLAVOUR: "validate"
   before_script:
       # Build hyperlinked sources for documentation when building releases
     - |


=====================================
rts/linker/Elf.c
=====================================
@@ -904,8 +904,8 @@ ocGetNames_ELF ( ObjectCode* oc )
                    ASSERT(common_used <= common_size);
 
                    IF_DEBUG(linker,
-                            debugBelch("COMMON symbol, size %lu name %s allocated at %p\n",
-                                       symbol->elf_sym->st_size, nm, symbol->addr));
+                            debugBelch("COMMON symbol, size %llu name %s allocated at %p\n",
+                                       (uint64_t) symbol->elf_sym->st_size, nm, symbol->addr));
 
                    /* Pointless to do addProddableBlock() for this area,
                       since the linker should never poke around in it. */



View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e818856f88097a9d58a4997bf6947a16651e2bfb...878f4abdbae9a6e8936aa006940df42def3b148d

-- 
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/e818856f88097a9d58a4997bf6947a16651e2bfb...878f4abdbae9a6e8936aa006940df42def3b148d
You're receiving this email because of your account on gitlab.haskell.org.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/ghc-commits/attachments/20201127/45341266/attachment-0001.html>


More information about the ghc-commits mailing list