[Git][ghc/ghc][wip/T22121] ci: Attempt using normal submodule cloning strategy

[Matthew Pickering (@mpickering)]

Matthew Pickering pushed to branch wip/T22121 at Glasgow Haskell Compiler / GHC


Commits:
e3678d47 by Matthew Pickering at 2022-08-29T14:07:23+01:00
ci: Attempt using normal submodule cloning strategy

We do not use any recursively cloned submodules, and this protects us
from flaky upstream remotes.

Fixes #22121

- - - - -


2 changed files:

- .gitlab-ci.yml
- .gitlab/ci.sh


Changes:

=====================================
.gitlab-ci.yml
=====================================
@@ -17,7 +17,7 @@ variables:
   # Overridden by individual jobs
   CONFIGURE_ARGS: ""
 
-  GIT_SUBMODULE_STRATEGY: "recursive"
+  GIT_SUBMODULE_STRATEGY: "normal"
 
   # Makes ci.sh isolate CABAL_DIR
   HERMETIC: "YES"


=====================================
.gitlab/ci.sh
=====================================
@@ -377,8 +377,8 @@ function cleanup_submodules() {
     # On Windows submodules can inexplicably get into funky states where git
     # believes that the submodule is initialized yet its associated repository
     # is not valid. Avoid failing in this case with the following insanity.
-    git submodule sync --recursive || git submodule deinit --force --all
-    git submodule update --init --recursive
+    git submodule sync || git submodule deinit --force --all
+    git submodule update --init
     git submodule foreach git clean -xdf
   else
     info "Not cleaning submodules, not in a git repo"



View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e3678d47079559e17f082b618a6ea12f521b64fc

-- 
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/e3678d47079559e17f082b618a6ea12f521b64fc
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/20220829/318cacf7/attachment-0001.html>