--089e015369f038ea1904ed46d845 Content-Type: text/plain; charset=UTF-8 Great. I will commit the patch once I get a chance then. Unless of course either you or Albert has already done so. On Wed, Dec 11, 2013 at 3:30 AM, Ben Cooksley wrote: > On Wed, Dec 11, 2013 at 2:52 PM, Dawit A wrote: > >> On Tue, Dec 10, 2013 at 1:45 PM, Albert Astals Cid wrote: >> >>> El Dimarts, 10 de desembre de 2013, a les 08:57:47, Dawit A va escriure: >>> > Sorry for the late response. I did not see this email for some reason. >>> None >>> > of the test cases in testRetryDialog hit that assert on my machine. >>> >>> Is your Qt compiled in debug or release? Because as my original e-mail >>> says if >>> it's in release the assert won't trigger. >>> >> >> Ahh... I assumed you were talking about kde-runtime not Qt. >> >>> >>> > If this >>> > unit test fails on your system, can you please enable the debug area >>> for >>> > KPasswdServer, run the unit test only for testRetryDialog only and >>> provide >>> > the output from it? >>> >>> Attached, is that enough? >>> >> >> That is fine. The test is indeed doing the wrong thing. It conflated two >> cases that should be tested separately into one. >> >>> >>> > I am curious to know why it would start failing all of >>> > the sudden since nothing was changed in that code base in a very long >>> time. >>> >>> Because we switched to a debug Qt that now has asserts enabled. >>> >> >> Right. >> >> Can you please check if the attached pass works fine for you? It fixes >> the broken test case for me here. >> > > I have now tested that patch on the CI system. It fixes the issue Albert > mentions. > > Test project /srv/jenkins/workspace/kde-runtime_master/build > Start 1: kpasswdservertest > 1/9 Test #1: kpasswdservertest ................ Passed 0.28 sec > Start 2: testkioarchive > 2/9 Test #2: testkioarchive ................... Passed 0.29 sec > Start 3: testtrash > 3/9 Test #3: testtrash ........................ Passed 1.72 sec > Start 4: testdesktop > 4/9 Test #4: testdesktop ...................... Passed 0.31 sec > Start 5: kurifiltertest > 5/9 Test #5: kurifiltertest ................... Passed 10.07 sec > Start 6: backtraceparsertest > 6/9 Test #6: backtraceparsertest .............. Passed 5.67 sec > Start 7: filetypestest > 7/9 Test #7: filetypestest .................... Passed 1.18 sec > Start 8: sortfiltermodeltest > 8/9 Test #8: sortfiltermodeltest .............. Passed 0.04 sec > Start 9: fullmodelaccesstest > 9/9 Test #9: fullmodelaccesstest .............. Passed 0.01 sec > > 100% tests passed, 0 tests failed out of 9 > > Regards, > Ben > --089e015369f038ea1904ed46d845 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
Great. I will commit the patch once I get a chance then. U= nless of course either you or Albert has already done so.


On Wed, Dec 11, 2013 = at 3:30 AM, Ben Cooksley <bcooksley@kde.org> wrote:
On W= ed, Dec 11, 2013 at 2:52 PM, Dawit A <adawit@kde.org> wrote:
On Tue, Dec 10, 2013 at 1:45 PM, Albert Astals Cid <aaci= d@kde.org> wrote:
El Dimarts, 10 de desembre de 2013, a les 08:57:47, Dawit = A va escriure:
> Sorry for the late response. I did not see this email for some re= ason. None
> of the test cases in testRetryDialog hit that assert on my machine.
Is your Qt compiled in debug or release? Because as my original e-mai= l says if
it's in release the assert won't trigger.

=
Ahh... I assumed you were talking about kde-runtime not Qt= .

> If this
> unit test fails on your system, can you please enable the debug area f= or
> KPasswdServer, run the unit test only for testRetryDialog only and pro= vide
> the output from it?

Attached, is that enough?

T= hat is fine. The test is indeed doing the wrong thing. It conflated two cas= es that should be tested separately into one.=C2=A0

> I am curious to know why it would start failing all of
> the sudden since nothing was changed in that code base in a very long = time.

Because we switched to a debug Qt that now has asserts enabled.

Right.

Can yo= u please check if the attached pass works fine for you? It fixes the broken= test case for me here.

I have now t= ested that patch on the CI system. It fixes the issue Albert mentions.

Test project /srv/jenkins/workspace/kde-runtime_master= /build
=C2=A0 =C2=A0 Start 1: kpasswdservertest
1/9 Test #1: kpassw= dservertest ................ =C2=A0 Passed =C2=A0 =C2=A00.28 sec
= =C2=A0 =C2=A0 Start 2: testkioarchive
2/9 Test #2: testkioarchive= ................... =C2=A0 Passed =C2=A0 =C2=A00.29 sec
=C2=A0 =C2=A0 Start 3: testtrash
3/9 Test #3: testtrash ....= .................... =C2=A0 Passed =C2=A0 =C2=A01.72 sec
=C2=A0 = =C2=A0 Start 4: testdesktop
4/9 Test #4: testdesktop ............= .......... =C2=A0 Passed =C2=A0 =C2=A00.31 sec
=C2=A0 =C2=A0 Start 5: kurifiltertest
5/9 Test #5: kurifilte= rtest ................... =C2=A0 Passed =C2=A0 10.07 sec
=C2=A0 = =C2=A0 Start 6: backtraceparsertest
6/9 Test #6: backtraceparsert= est .............. =C2=A0 Passed =C2=A0 =C2=A05.67 sec
=C2=A0 =C2=A0 Start 7: filetypestest
7/9 Test #7: filetypest= est .................... =C2=A0 Passed =C2=A0 =C2=A01.18 sec
=C2= =A0 =C2=A0 Start 8: sortfiltermodeltest
8/9 Test #8: sortfiltermo= deltest .............. =C2=A0 Passed =C2=A0 =C2=A00.04 sec
=C2=A0 =C2=A0 Start 9: fullmodelaccesstest
9/9 Test #9: full= modelaccesstest .............. =C2=A0 Passed =C2=A0 =C2=A00.01 sec

100% tests passed, 0 tests failed out of 9

<= /div>
Regards,
Ben

--089e015369f038ea1904ed46d845--