[Git][ghc/ghc][wip/fix-ci-clean] 4 commits: ci: Try using multi repl in ghc-in-ghci test

8 Dec 2025


      Cheng Shao pushed to branch wip/fix-ci-clean at Glasgow Haskell Compiler / GHC


Commits:
3bfe7aa2 by Matthew Pickering at 2025-12-07T12:18:57-05:00
ci: Try using multi repl in ghc-in-ghci test

This should be quite a bit faster than the ./hadrian/ghci command as it
doesn't properly build all the dependencies.

- - - - -
2ef1601a by Rodrigo Mesquita at 2025-12-07T12:19:38-05:00
Stack.Decode: Don't error on bitmap size 0

A RET_BCO may have a bitmap with no payload.
In that case, the bitmap = 0.

One can observe this by using -ddump-bcos and interpreting

```
main = pure ()
```

Observe, for instance, that the BCO for this main function has size 0:

```
ProtoBCO Main.main#0:
   \u []
       breakmain:Main,0() GHC.Internal.Base.pure
                                GHC.Internal.Base.$fApplicativeIO GHC.Internal.Tuple.()
   bitmap:  0 []
   BRK_FUN <breakarray> main:Main 0 <cc>
   PACK     () 0
   PUSH_G   GHC.Internal.Base.$fApplicativeIO
   PUSH_APPLY_PP
   PUSH_G   GHC.Internal.Base.pure
   ENTER
```

Perhaps we never tried to decode a stack in which a BCO like this was
present. However, for the debugger, we want to decode stacks of threads
stopped at breakpoints, and these kind of BCOs do get on a stack under
e.g. `stg_apply_interp_info` frames.

See the accompanying test in the next commit for an example to trigger
the bug this commit fixes.

Fixes #26640

- - - - -
747153d2 by Rodrigo Mesquita at 2025-12-07T12:19:38-05:00
Add test for #26640

- - - - -
70f76c29 by Cheng Shao at 2025-12-08T10:37:39+01:00
ci: fix "ci.sh clean" to address frequent out of space error on windows runners

This patch fixes the `ci.sh clean` logic to address frequent out of
space error on windows runners; previously it didn't clean up the
inplace mingw blobs, which is the largest source of space leak on
windows runners. See added comment for detailed explanation.

- - - - -


8 changed files:

- .gitlab-ci.yml
- .gitlab/ci.sh
- .gitlab/generate-ci/gen_ci.hs
- libraries/ghc-internal/src/GHC/Internal/Stack/Decode.hs
- + testsuite/tests/bytecode/T26640.hs
- + testsuite/tests/bytecode/T26640.script
- + testsuite/tests/bytecode/T26640.stdout
- testsuite/tests/bytecode/all.T


Changes:

=====================================
.gitlab-ci.yml
=====================================
@@ -416,7 +416,7 @@ hadrian-ghc-in-ghci:
     # workaround for docker permissions
     - sudo chown ghc:ghc -R .
   variables:
-    GHC_FLAGS: -Werror
+    GHC_FLAGS: -Werror -Wwarn=unused-imports
   tags:
     - x86_64-linux
   script:
@@ -428,7 +428,7 @@ hadrian-ghc-in-ghci:
     - "echo '  ghc-options: -Werror'  >> hadrian/cabal.project.local"
     # Load ghc-in-ghci then immediately exit and check the modules loaded
     - export CORES="$(mk/detect-cpu-count.sh)"
-    - echo ":q" | HADRIAN_ARGS=-j$CORES hadrian/ghci -j$CORES | tail -n2 | grep "Ok,"
+    - echo ":q" | HADRIAN_ARGS=-j$CORES hadrian/ghci-multi -j$CORES | tail -n2 | grep "Ok,"
   after_script:
     - .gitlab/ci.sh save_cache
     - cat ci_timings.txt


=====================================
.gitlab/ci.sh
=====================================
@@ -275,7 +275,7 @@ function setup() {
 
 function fetch_ghc() {
   local should_fetch=false
-  
+
   if [ ! -e "$GHC" ]; then
     if [ -z "${FETCH_GHC_VERSION:-}" ]; then
       fail "GHC not found at '$GHC' and FETCH_GHC_VERSION is not set"
@@ -292,7 +292,7 @@ function fetch_ghc() {
       fi
     fi
   fi
-  
+
   if [ "$should_fetch" = true ]; then
       local v="$FETCH_GHC_VERSION"
 
@@ -887,8 +887,28 @@ function save_cache () {
 }
 
 function clean() {
-  rm -R tmp
-  run rm -Rf _build
+  # When CI_DISPOSABLE_ENVIRONMENT is not true (e.g. using shell
+  # executor on windows/macos), the project directory is not removed
+  # by gitlab runner automatically after each job. To mitigate the
+  # space leak, other than periodic cleaning on the runner host, we
+  # also must aggressively cleanup build products, otherwise we run
+  # into out of space errors too frequently.
+  #
+  # When CI_DISPOSABLE_ENVIRONMENT is true (using docker executor on
+  # linux), the runner will do proper cleanup, so no need to do
+  # anything here.
+  #
+  # The exclude list are the artifacts that we do expect to be
+  # uploaded. Keep in sync with `jobArtifacts` in
+  # `.gitlab/generate-ci/gen_ci.hs`!
+  if [[ "${CI_DISPOSABLE_ENVIRONMENT:-}" != true ]]; then
+    git submodule foreach --recursive git clean -xdf
+    git clean -xdf \
+      --exclude=ci_timings.txt \
+      --exclude=ghc-*.tar.xz \
+      --exclude=junit.xml \
+      --exclude=unexpected-test-output.tar.gz
+  fi
 }
 
 function run_hadrian() {


=====================================
.gitlab/generate-ci/gen_ci.hs
=====================================
@@ -889,6 +889,8 @@ job arch opsys buildConfig = NamedJob { name = jobName, jobInfo = Job {..} }
       , if testsuiteUsePerf buildConfig then "RUNTEST_ARGS" =: "--config perf_path=perf" else mempty
       ]
 
+    -- Keep in sync with the exclude list in `function clean()` in
+    -- `.gitlab/ci.sh`!
     jobArtifacts = Artifacts
       { junitReport = "junit.xml"
       , expireIn = "2 weeks"


=====================================
libraries/ghc-internal/src/GHC/Internal/Stack/Decode.hs
=====================================
@@ -269,7 +269,7 @@ decodeLargeBitmap getterFun# stackSnapshot# index relativePayloadOffset = do
     cWordArrayToList ptr size = mapM (peekElemOff ptr) [0 .. (size - 1)]
 
     usedBitmapWords :: Int -> Int
-    usedBitmapWords 0 = error "Invalid large bitmap size 0."
+    usedBitmapWords 0 = 0
     usedBitmapWords size = (size `div` fromIntegral wORD_SIZE_IN_BITS) + 1
 
     bitmapWordsPointerness :: Word -> [Word] -> [Pointerness]


=====================================
testsuite/tests/bytecode/T26640.hs
=====================================
@@ -0,0 +1,4 @@
+-- Main.hs
+module Main where
+main = pure ()
+


=====================================
testsuite/tests/bytecode/T26640.script
=====================================
@@ -0,0 +1,6 @@
+:l T26640
+:break 3
+main
+import GHC.Conc
+import GHC.Stack.CloneStack
+() <- mapM_ (\ix -> do !_ <- decode =<< cloneThreadStack ix; return ()) =<< listThreads


=====================================
testsuite/tests/bytecode/T26640.stdout
=====================================
@@ -0,0 +1,3 @@
+Breakpoint 0 activated at T26640.hs:3:8-14
+Stopped in Main.main, T26640.hs:3:8-14
+_result :: IO () = _


=====================================
testsuite/tests/bytecode/all.T
=====================================
@@ -8,6 +8,7 @@ test('T25975', extra_ways(ghci_ways), compile_and_run,
 
 test('T26565', extra_files(["T26565.hs"]), ghci_script, ['T26565.script'])
 test('T23973', extra_files(["T23973.hs"]), ghci_script, ['T23973.script'])
+test('T26640', extra_files(["T26640.hs"]), ghci_script, ['T26640.script'])
 
 # Nullary data constructors
 test('T26216', extra_files(["T26216_aux.hs"]), ghci_script, ['T26216.script'])



View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/72eab1d557a0c17d931156a244188fc...

-- 
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/72eab1d557a0c17d931156a244188fc...
You're receiving this email because of your account on gitlab.haskell.org.

    

Cheng Shao (＠TerrorJack)

tags

participants (1)