Description of problem: pidgin with purple facebook gets stuck consuming 100% of CPU. This only happens in my local hosted Xorg. If I use my new intel NUC running fedora 31 x64 and run pidgin via ssh -X, everything is fine. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3.
So the looping occurs if everything is run locally under Mageia, but not if pidgin (+ purple-facebook) is run via remote login to the same system? Can you say what process is hogging the processor? Since you cite x11-server, perhaps you know it is that.
Severity: critical => majorCC: (none) => lewyssmith
(In reply to Lewis Smith from comment #1) > So the looping occurs if everything is run locally under Mageia, but not if > pidgin (+ purple-facebook) is run via remote login to the same system? yes, pidgin also behaves when run inside Xephyr (= nested x) that is running jwm. > Can you say what process is hogging the processor? Since you cite > x11-server, perhaps you know it is that. the process that hogs the processor is "pidgin" and it stops accepting keyboard input.
Thanks for that info. I should have asked this before also: does the error only happen with the addition of 'purple-facebook' to 'pidgin'? Trying to pin down the culprit. Not the same maintainer for the two elements.
Created attachment 11428 [details] Attempted fix for pango
(In reply to Lewis Smith from comment #3) > Thanks for that info. > I should have asked this before also: does the error only happen with the > addition of 'purple-facebook' to 'pidgin'? Trying to pin down the culprit. > Not the same maintainer for the two elements. I no longer seem to have any non-facebook contacts there
Also see https://gitlab.gnome.org/GNOME/pango/merge_requests/167 .
Thank you for coming back on this. And all your work - a fix! Assign to you or the pango maintainer? [ovitters] Should the bug title be changed - to what? On thin ice here: Do we mark this 'upstream' with the URL as in the previous comment? My understanding being that your fix proposed chez pango will eventually filter back down. OR Should this bug be re-classified as 'pango' SRPM and your patch applied directly by the relevant maintainer? > I no longer seem to have any non-facebook contacts there Oh dear. I asked because I recall that when QA has to test pidgin, there is always a problem finding someone with a suitable account to do so.
(In reply to Lewis Smith from comment #7) > Thank you for coming back on this. And all your work - a fix! You are welcome. > Assign to you or the pango maintainer? [ovitters] perhaps CC him. > Should the bug title be changed - to what? I think it should not. > > On thin ice here: Do we mark this 'upstream' with the URL as in the previous > comment? My understanding being that your fix proposed chez pango will Dont know. > eventually filter back down. > OR > Should this bug be re-classified as 'pango' SRPM and your patch applied > directly by the relevant maintainer? it may be a problem in an upper layer of that function call. > > > I no longer seem to have any non-facebook contacts there > Oh dear. I asked because I recall that when QA has to test pidgin, there is > always a problem finding someone with a suitable account to do so. ah.
This is a pig. I have installed Pidgin and got it to work on IRC #mageia channels. So this bug is not generic Pidgin. * Unsure that the bug occurs *without* 'purple facebook' [see above]; * Why should the X server be relevant? It seems to be; * A fix is proposed for 'pango'! So is it X, pidgin, purple-facebook, pango? No choice but to assign this globally; CC'ing Olav for the pango patch. purple-facebook has mixed committers. wally has done all recent Pidgin commits, but it is not really his baby.
CC: (none) => mageia, olavAssignee: bugsquad => pkg-bugsURL: (none) => https://gitlab.gnome.org/GNOME/pango/merge_requests/167
Fixed upstream, included in pango-1.45.2-2.mga8.
Status: NEW => RESOLVEDResolution: (none) => FIXED