GH-50575: [CI][Dev] Fix shellcheck errors in the ci/scripts/python_wheel_xlinux_build.sh#50577
Conversation
…hon_wheel_xlinux_build.sh
|
|
| -Dxsimd_SOURCE=BUNDLED \ | ||
| ${ARROW_EXTRA_CMAKE_FLAGS} \ | ||
| -G ${CMAKE_GENERATOR} \ | ||
| "${ARROW_EXTRA_CMAKE_FLAGS}" \ |
There was a problem hiding this comment.
ARROW_EXTRA_CMAKE_FLAGS defined here.
The name ARROW_EXTRA_CMAKE_FLAGS looks plural, but it seems to hold just a single flag
if [[ "$(uname -m)" == arm* ]] || [[ "$(uname -m)" == aarch* ]]; then
# Build jemalloc --with-lg-page=16 in order to make the wheel work on both
# 4k and 64k page arm64 systems. For more context see
# https://github.com/apache/arrow/issues/10929
export ARROW_EXTRA_CMAKE_FLAGS="-DARROW_JEMALLOC_LG_PAGE=16"
: ${ARROW_JEMALLOC:=OFF}There was a problem hiding this comment.
Could you keep accepting multiple flags?
We may use multiple flags later.
There was a problem hiding this comment.
@kou Thank you for your comment. What do you think this change?
+ARROW_EXTRA_CMAKE_FLAGS_ARRAY=()
if [[ "$(uname -m)" == arm* ]] || [[ "$(uname -m)" == aarch* ]]; then
# Build jemalloc --with-lg-page=16 in order to make the wheel work on both
# 4k and 64k page arm64 systems. For more context see
# https://github.com/apache/arrow/issues/10929
- export ARROW_EXTRA_CMAKE_FLAGS="-DARROW_JEMALLOC_LG_PAGE=16"
+ ARROW_EXTRA_CMAKE_FLAGS_ARRAY+=("-DARROW_JEMALLOC_LG_PAGE=16")
: "${ARROW_JEMALLOC:=OFF}"
else
: "${ARROW_JEMALLOC:=ON}"
@@ -146,7 +148,7 @@ cmake \
-DVCPKG_MANIFEST_MODE=OFF \
-DVCPKG_TARGET_TRIPLET="${VCPKG_TARGET_TRIPLET}" \
-Dxsimd_SOURCE=BUNDLED \
- "${ARROW_EXTRA_CMAKE_FLAGS}" \
+ "${ARROW_EXTRA_CMAKE_FLAGS_ARRAY[@]}" \There was a problem hiding this comment.
We don't need _ARRAY suffix. It's redundant.
There was a problem hiding this comment.
Thanks. removed _ARRAY suffix
…hon_wheel_xlinux_build.sh
|
@github-actions crossbow submit python-wheel-manylinux-2-28 python-wheel-musllinux-1-2 |
|
|
@github-actions crossbow submit python-wheel-manylinux-2-28* python-wheel-musllinux-1-2* |
|
|
@github-actions crossbow submit wheel-manylinux-2-28* wheel-musllinux-1-2* |
|
Revision: 1a78357 Submitted crossbow builds: ursacomputing/crossbow @ actions-9720cb735d |
…hon_wheel_xlinux_build.sh
|
@raulcd Thank you for your help! |
|
@github-actions crossbow submit wheel-manylinux-2-28* wheel-musllinux-1-2* |
|
Revision: 8336ea1 Submitted crossbow builds: ursacomputing/crossbow @ actions-efd09d92b1 |
…hon_wheel_xlinux_build.sh
|
@github-actions crossbow submit wheel-manylinux-2-28* wheel-musllinux-1-2* |
|
Revision: be94644 Submitted crossbow builds: ursacomputing/crossbow @ actions-f639e5d2da |
|
I’m not sure why two tests are still failing.
|
|
The failures are unrelated, they are also failing on main for the nightly builds, it is tracked here: |
|
|
||
| if [[ -f visible_symbols.log && `cat visible_symbols.log | wc -l` -eq 0 ]]; then | ||
| # Return early if the log file exists but is empty. | ||
| if [[ -f visible_symbols.log && ! -s visible_symbols.log ]]; then |
There was a problem hiding this comment.
I've just learnt about the -f && ! -s combo, nice!
|
After merging your PR, Conbench analyzed the 4 benchmarking runs that have been run so far on merge-commit 7667f41. There was 1 benchmark result indicating a performance regression:
The full Conbench report has more details. It also includes information about 208 possible false positives for unstable benchmarks that are known to sometimes produce them. |
Rationale for this change
This is the sub issue #44748.
ls.What changes are included in this PR?
lsiteration with globsAre these changes tested?
Yes.
Are there any user-facing changes?
No.