Workaround: Fix test failures on Ubuntu jammy s390x. #63
Conversation
test/zlib/test_zlib.rb
Outdated
| @@ -12,7 +12,18 @@ | |||
| end | |||
|
|
|||
| if defined? Zlib | |||
| class TestZlibDeflate < Test::Unit::TestCase | |||
| class Zlib::TestCase < Test::Unit::TestCase | |||
| def child_env | |||
There was a problem hiding this comment.
The child_env passed to subprocesses is not modified, correct? If so, I recommend this be a frozen constant and not a method. You could define it as Zlib::CHILD_ENV, and skip creating a subclass and modifying the tests to use the subclass.
There was a problem hiding this comment.
Yes, the child_env is not modified. Sure. Let me modify the PR on your way.
There was a problem hiding this comment.
I rebased fixing it on your way. Now what do you think?
junaruga
force-pushed
the
wip/s390x-dfltcc
branch
from
ec99a02
to
532ce51
Compare
|
Note below is the Travis CI Ubuntu jammy s390x result before this PR on my forked repository. It failed as expected. And below is the Travis CI Ubuntu jammy s390x result after this PR. It passed as expected. |
|
@hsbt or @nurse It seems Travis CI for ruby/zlib doesn't start unlike Travis CI for ruby/ruby. Could you enable it? Thanks. |
This commit fixes the test failures on the zlib in Ubuntu jammy s390x. According to the <https://packages.ubuntu.com/jammy-updates/zlib1g> - `zlib_1.2.11.dfsg-2ubuntu9.2.debian.tar.xz`, the zlib deb package is applying the patch 410.patch (madler/zlib#410), and configured by `./configure --dfltcc` on Ubuntu jammy s390x. The `--dfltcc` is to enable the deflate algorithm in hardware. It produces a different (but still valid) compressed byte stream, and causes the test failures in ruby/zlib. As a workaround, set the environment variable `DFLTCC=0` disabling the implementation in zlib on s390x to the failing tests. Note we need to test in a child Ruby process with `assert_separately` to test on the `DFLTCC=0` set by the parent Ruby process.
The only choice to test s390x on pull-request is currently Travis CI as far as I know.
junaruga
force-pushed
the
wip/s390x-dfltcc
branch
from
532ce51
to
cc88b54
Compare
|
I rebased the PR using the |
|
Thanks for your review. Let me merge this PR, though the Travis CI is not enabled for this repository yet. |
|
Nice work figuring out the root cause. |
Thanks. I was definitely not able to fix the issue by the workaround without @iii-i's help and info at #60 (comment). |
|
Note I opened another PR #65 that is rework for this PR. |
This PR fixes #60, the test failures on the zlib in Ubuntu jammy s390x.
This PR has 2 commits.
The 1st commit
According to the https://packages.ubuntu.com/jammy-updates/zlib1g -
zlib_1.2.11.dfsg-2ubuntu9.2.debian.tar.xz, the zlib deb package is applying the patch 410.patch (madler/zlib#410), and configured by./configure --dfltccon Ubuntu jammy s390x. The--dfltccis to enable the deflate algorithm in hardware.It produces a different (but still valid) compressed byte stream, and causes the test failures in ruby/zlib. As a workaround, set the environment variable
DFLTCC=0disabling the implementation in zlib on s390x to the failing tests.Note we need to test in a child Ruby process with
assert_separatelyto test on theDFLTCC=0set by the parent Ruby process.I tested the ruby/zlib unit test on Ruby's s390x Ubuntu jammy server, and confirmed it passed.
The 2nd commit
I want to use Travis CI again for this repository to test the Ubuntu jammy s390x case on CI. Right now the only choice to test s390x on pull-request is Travis CI as far as I know.