This splits the full MR test run, with short running tests on a side and now executed for both Win64 and WOW64, roughly 5min each, and the long running tests executed only for Win64, taking roughly 20min. All are executed unconditionally, but we could maybe add some early exit if that feels useful.
From: Rémi Bernon rbernon@codeweavers.com
--- tools/gitlab/test.yml | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tools/gitlab/test.yml b/tools/gitlab/test.yml index d2b1966ad6e..b08e7966bf8 100644 --- a/tools/gitlab/test.yml +++ b/tools/gitlab/test.yml @@ -42,7 +42,7 @@ - wine wineboot.exe -u - wineserver -w
-test-linux-32: +test-linux-64: extends: .wine-test variables: EXCLUDE_TESTS: "d3d10core:d3d10core d3d11:d3d11 d3d8:device d3d8:visual d3d9:d3d9ex d3d9:device d3d9:visual" @@ -52,7 +52,7 @@ test-linux-32: - job: build-linux script: - export WINETEST_COLOR=1 - - wine usr/local/lib/wine/i386-windows/winetest.exe -q -q -o - -t gitlab -u $CI_JOB_URL -n $EXCLUDE_TESTS + - wine usr/local/lib/wine/x86_64-windows/winetest.exe -q -q -o - -t gitlab -u $CI_JOB_URL -n $EXCLUDE_TESTS
debian-32: extends: .wine-test
From: Rémi Bernon rbernon@codeweavers.com
--- tools/gitlab/test.yml | 24 +++++++++++++++++++++--- 1 file changed, 21 insertions(+), 3 deletions(-)
diff --git a/tools/gitlab/test.yml b/tools/gitlab/test.yml index b08e7966bf8..48cc722f526 100644 --- a/tools/gitlab/test.yml +++ b/tools/gitlab/test.yml @@ -8,6 +8,16 @@ GIT_STRATEGY: none GECKO_VER: 2.47.3 MONO_VER: 7.4.0 + EXTRA_TESTS: | + dinput:hid dwrite:layout kernel32:debugger riched20:richole twain_32:dsm wbemdisp:wbemdisp mmdevapi:render + msi:install msi:msi quartz:vmr7 user32:winstation comdlg32:filedlg kernel32:sync mshtml:htmldoc qmgr:job + dsound:dsound kernel32:pipe ntdll:threadpool quartz:videorenderer reg.exe:add kernel32:loader mscoree:mscoree + mshtml:script winmm:timer regedit.exe:regedit user32:input adsldp:ldap quartz:vmr9 reg.exe:import + uxtheme:system dwrite:font user32:clipboard kernel32:process msi:action winmm:wave quartz:dsoundrender + dwrite:analyzer quartz:filtergraph wsdapi:discovery riched20:editor xaudio2_7:xaudio2 kernel32:console + amstream:amstream explorerframe:nstc d2d1:d2d1 winmm:mci ddraw:ddraw1 dxgi:dxgi shell32:shlexec ddraw:ddraw2 + ddraw:ddraw4 user32:win ddraw:ddraw7 winhttp:winhttp cmd.exe:batch user32:msg ws2_32:sock + EXCLUDE_TESTS: d3d10core:d3d10core d3d11:d3d11 d3d8:device d3d8:visual d3d9:d3d9ex d3d9:device d3d9:visual cache: - key: wine-gecko-$GECKO_VER paths: @@ -44,15 +54,23 @@
test-linux-64: extends: .wine-test - variables: - EXCLUDE_TESTS: "d3d10core:d3d10core d3d11:d3d11 d3d8:device d3d8:visual d3d9:d3d9ex d3d9:device d3d9:visual" rules: - if: $CI_PIPELINE_SOURCE == 'merge_request_event' needs: - job: build-linux script: - export WINETEST_COLOR=1 - - wine usr/local/lib/wine/x86_64-windows/winetest.exe -q -q -o - -t gitlab -u $CI_JOB_URL -n $EXCLUDE_TESTS + - wine usr/local/lib/wine/x86_64-windows/winetest.exe -q -q -o - -t gitlab -u $CI_JOB_URL -n $EXCLUDE_TESTS $EXTRA_TESTS + +test-linux-64-extra: + extends: .wine-test + rules: + - if: $CI_PIPELINE_SOURCE == 'merge_request_event' + needs: + - job: build-linux + script: + - export WINETEST_COLOR=1 + - wine usr/local/lib/wine/x86_64-windows/winetest.exe -q -q -o - -t gitlab -u $CI_JOB_URL $EXTRA_TESTS
debian-32: extends: .wine-test
From: Rémi Bernon rbernon@codeweavers.com
--- tools/gitlab/test.yml | 10 ++++++++++ 1 file changed, 10 insertions(+)
diff --git a/tools/gitlab/test.yml b/tools/gitlab/test.yml index 48cc722f526..0dfc37fe99b 100644 --- a/tools/gitlab/test.yml +++ b/tools/gitlab/test.yml @@ -52,6 +52,16 @@ - wine wineboot.exe -u - wineserver -w
+test-linux-32: + extends: .wine-test + rules: + - if: $CI_PIPELINE_SOURCE == 'merge_request_event' + needs: + - job: build-linux + script: + - export WINETEST_COLOR=1 + - wine usr/local/lib/wine/i386-windows/winetest.exe -q -q -o - -t gitlab -u $CI_JOB_URL -n $EXCLUDE_TESTS $EXTRA_TESTS + test-linux-64: extends: .wine-test rules:
Hi,
It looks like your patch introduced the new failures shown below. Please investigate and fix them before resubmitting your patch. If they are not new, fixing them anyway would help a lot. Otherwise please ask for the known failures list to be updated.
The tests also ran into some preexisting test failures. If you know how to fix them that would be helpful. See the TestBot job for the details:
The full results can be found at: https://testbot.winehq.org/JobDetails.pl?Key=125700
Your paranoid android.
=== debian11 (32 bit report) ===
d3d8: device.c:3365: Test failed: Expected message 0x1c for window 0x1, but didn't receive it
dinput: force_feedback.c:5796: Test failed: got status 0 force_feedback.c:5796: Test failed: GetResults returned 0x8000000e force_feedback.c:5796: Test failed: got result 0
wmvcore: wmvcore.c:2997: Test failed: Wait timed out.
FWIW I took all the tests that took more than 5s, on some random Gitlab run, for the "extra" list.
Well, that took a bit longer than what I saw during my testing. Maybe it's a bit too soon but I think that splitting the tests like this can help parallelizing, especially when we have more runners.
Then I think it'd be better anyway to run the 64-bit tests instead of wow64, because it otherwise doesn't run most dinput tests for instance. I'll open a separate MR.
This merge request was closed by Rémi Bernon.