aacid added a comment.
In https://phabricator.kde.org/D9328#198101, @rkflx wrote: > In https://phabricator.kde.org/D9328#197788, @aacid wrote: > > > In https://phabricator.kde.org/D9328#196555, @rkflx wrote: > > > > > >> 5. `ASSERT: "page()" in file okular/core/generator_p.cpp, line 129` > > > > > > Ok, i found how to reproduce too and this should be fixed now. > > > Can confirm, and I just noticed that simply running `make test` would've found this problem too (I kinda assumed you ran this, so did not check again). > > However, the autotests still have lots of issues (at least for me). As an example, `./autotests/parttest testReload` results in > > Thread 1 "parttest" received signal SIGSEGV, Segmentation fault. > 0x00007ffff74c1930 in Okular::TextPageGenerationThread::abortExtraction (this=0x0) > at okular/core/generator_p.cpp:113 > > > Could you recheck? Wops, the code assumed that there would be always a text generation thread but that's not necessarily true on the tests. fixed. Also fixed a different crash that happened on the tests (and at least this one seems like a real bug). So good thing we have tests.... i should run them more often :D > > > --- > >>>>> 6. Text page generation not cancellable >> >> […] > > Ok, thanks for the explanation, makes sense. It's not ideal, but good enough for now in my book. Maybe add a note to the commit message? Ok, added a small blurb. REPOSITORY R223 Okular REVISION DETAIL https://phabricator.kde.org/D9328 To: aacid, ervin, rkflx Cc: #okular, michaelweghorn, ngraham, aacid
