CppUnit SdMiscTest::testTdf119956 Failing on Windows with 150% scale factor

classic Classic list List threaded Threaded
3 messages Options
slacka slacka
Reply | Threaded
Open this post in threaded view
|

CppUnit SdMiscTest::testTdf119956 Failing on Windows with 150% scale factor

While testing various scaling factors for the starmath issue,
http://document-foundation-mail-archive.969070.n3.nabble.com/CppunitTest-starmath-qa-cppunit-Failing-on-Windows-with-HiDPI-screens-tp4261772.html

I discovered that  SdMiscTest::testTdf119956 is also failing.

Any ideas why this test would also be dpi dependent?

The build log is below. Would you like any additional debugging info?


-Luke

[build CUT] sd_misc_tests
SdMiscTest::testTdf67248 finished in: 1086ms
D:/core/sd/qa/unit/misc-tests.cxx:777:SdMiscTest::testTdf119956
equality assertion failed
- Expected: 3
- Actual  : 0

SdMiscTest::testTdf119956 finished in: 1851ms
SdMiscTest::testTdf120527 finished in: 500ms
D:/core/sd/qa/unit/misc-tests.cxx(777) : error : Assertion
Test name: SdMiscTest::testTdf119956
equality assertion failed
- Expected: 3
- Actual  : 0

Failures !!!
Run: 15   Failure total: 1   Failures: 1   Errors: 0

Error: a unit test failed, please do one of:
make CppunitTest_sd_misc_tests CPPUNITTRACE=TRUE # which is a shortcut for the following line

_______________________________________________
LibreOffice mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/libreoffice
Regina Henschel Regina Henschel
Reply | Threaded
Open this post in threaded view
|

Re: CppUnit SdMiscTest::testTdf119956 Failing on Windows with 150% scale factor

Hi Luke,

I get curious results.

If I run the single test using
/opt/lo/bin/make CppunitTest_sd_misc_tests
CPPUNIT_TEST_NAME="testTdf119956" CPPUNITTRACE=TRUE
or
/opt/lo/bin/make CppunitTest_sd_misc_tests CPPUNIT_TEST_NAME="testTdf119956"

it does _not_ fail.

If I run the test among other tests using
$ /opt/lo/bin/make sd.check
it fails.

I have tested it with Version: 6.3.0.0.alpha1+ (x64)
Build ID: 115ab48f86d4e3c6eede49767df1ee5a82b4ab22
CPU threads: 8; OS: Windows 10.0; UI render: default; VCL: win;
Locale: en-US (en_US); UI-Language: en-US
Calc: threaded

I have set Windows 10 to 150% scale factor. Larger factors make apps
unusable on my monitor.

I do not remember, whether I had tested it with 150% at the time I wrote
the test. In the meantime the tab bar has been changed. The tabs are
much wider now than before. So it might be, that the tab center, that
should get a click, is no longer visible.

If I change the window width from 1024 to 1600 in line 145 in
SdMiscTest::Load(), then the test does not fail in 150% screen setting.

But that does not explain, why the test works here as single test and
not in the series of tests.

Kind regards
Regina



Luke Benes schrieb am 20-May-19 um 22:30:

> While testing various scaling factors for the starmath issue,
> http://document-foundation-mail-archive.969070.n3.nabble.com/CppunitTest-starmath-qa-cppunit-Failing-on-Windows-with-HiDPI-screens-tp4261772.html
>
> I discovered that  SdMiscTest::testTdf119956 is also failing.
>
> Any ideas why this test would also be dpi dependent?
>
> The build log is below. Would you like any additional debugging info?
>
>
> -Luke
>
> [build CUT] sd_misc_tests
> SdMiscTest::testTdf67248 finished in: 1086ms
> D:/core/sd/qa/unit/misc-tests.cxx:777:SdMiscTest::testTdf119956
> equality assertion failed
> - Expected: 3
> - Actual  : 0
>
> SdMiscTest::testTdf119956 finished in: 1851ms
> SdMiscTest::testTdf120527 finished in: 500ms
> D:/core/sd/qa/unit/misc-tests.cxx(777) : error : Assertion
> Test name: SdMiscTest::testTdf119956
> equality assertion failed
> - Expected: 3
> - Actual  : 0
>
> Failures !!!
> Run: 15   Failure total: 1   Failures: 1   Errors: 0
>
> Error: a unit test failed, please do one of:
> make CppunitTest_sd_misc_tests CPPUNITTRACE=TRUE # which is a shortcut for the following line
>
>

_______________________________________________
LibreOffice mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/libreoffice
Jan-Marek Glogowski Jan-Marek Glogowski
Reply | Threaded
Open this post in threaded view
|

Re: CppUnit SdMiscTest::testTdf119956 Failing on Windows with 150% scale factor



Am 21. Mai 2019 00:48:40 MESZ schrieb Regina Henschel <[hidden email]>:

>I get curious results.
>
>If I run the single test using
>/opt/lo/bin/make CppunitTest_sd_misc_tests
>CPPUNIT_TEST_NAME="testTdf119956" CPPUNITTRACE=TRUE
>or
>/opt/lo/bin/make CppunitTest_sd_misc_tests
>CPPUNIT_TEST_NAME="testTdf119956"
>
>it does _not_ fail.
>
>If I run the test among other tests using
>$ /opt/lo/bin/make sd.check
>it fails.
>
...
>
>But that does not explain, why the test works here as single test and
>not in the series of tests.

There is a general problem with LO unit tests: a test may change the state of the application, which eventually influences the next test. Some stuff is cached.
In theory you would have to restart LO for each test to reset any internal state, but that would be very slow.

The other, more probable possibility, I have also faced a few times, is memory corruption. This means your failing test is actually valid, but a previous test has passed but corrupted some memory. You should run the whole test in some memory checker.

Jan-Marek
_______________________________________________
LibreOffice mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/libreoffice