Dan tracking this issue on http://defect.opensolaris.org/bz/show_bug.cgi?id=3835
Not able to reproduce what you saw on a b98 and the build from the gate. Need some more details from you to see if I can repo it. Did turn up a few issues when dtracing it. If you get a chance can you update the bug. Thanks. JR John Rice wrote: > Thanks Dan - if Michal can investigate this today we'll get a patch > submitted for it, otherwise it will need to wait until Tuesday. Sounds > like the updatemanagrnotifier is broken, using gobject.timer_add is the > problem here, it should be canceling the respawned do_next_check and is > not for some reason :( > > JR > > Dan Price wrote: > >> John, Michal, a couple of things. >> >> First, I found that 'make link' doesn't work very well following your >> recent putbacks. There were already some issues but now they are >> worse, and running 'make link' creates turds in the workspace. >> >> I'm leaving for vacation so I am very short on time. But I needed >> this to work for other testing of other stuff. Attached is a patch >> which I think fixes the issue. Do you think you could finish it and get >> it reviewed and fixed? >> >> Please look it over carefully, as I might not have gotten all of >> the directory changes correct. This does work for me, however. >> >> You should also change the permissions on some of the files in the src >> base to add the +x bit. That way when we symlink to them with >> 'make link' they are executable. Like: >> >> /usr/lib/updatemanagernotifer >> >> Second, Could this please be renamed to update-manager-notifier??? >> >> Third, once I started the updatemanagernotifier by hand, it ate a lot of >> CPU. I used -debug and all it printed was: >> >> Called do_next_check >> Called do_next_check >> Called do_next_check >> Called do_next_check >> Called do_next_check >> Called do_next_check >> Called do_next_check >> Called do_next_check >> Called do_next_check >> Called do_next_check >> >> It seems to make from 25,000 syscalls/second to 200,000 syscalls/sec >> and is the top process on my system. This looks like a broken call >> to poll: >> >> # truss -c -p `pgrep update` >> ^C >> syscall seconds calls errors >> ioctl .201 27169 >> pollsys .192 27275 >> -------- ------ ---- >> sys totals: .393 54444 0 >> usr time: .541 >> elapsed: 3.940 >> >> Please fix. >> >> -dp >> >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> pkg-discuss mailing list >> [email protected] >> http://mail.opensolaris.org/mailman/listinfo/pkg-discuss >> > > _______________________________________________ > pkg-discuss mailing list > [email protected] > http://mail.opensolaris.org/mailman/listinfo/pkg-discuss > _______________________________________________ pkg-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
