Since telegram blockade started a few weeks ago - Telegram Desktop now utilized 100% of CPU for me. Sometimes it is even more than 100% (considering 100% is usually 1 core and I have 4 and they all get utilized at 100%): 400%. This happens right after the start. The only message I see in the terminal is QObject::startTimer: Timers cannot have negative intervals It doesn't matter if I minimize or close the window - the CPU consuming seems quite constant. High CPU usage seems to stop after a few minutes since the application got started.
> High CPU usage seems to stop after a few minutes since the application got started. It seems to get back in a while after.
I can't reproduce this since those IPs are not blocked here. Can you try to grab a stack trace of the busy thread? A possible way to do this: > pgrep telegram <telegram pid> > ps auxH -o lwp | grep <telegram pid> <list of threads> Here, you can find the thread which is consuming 100% cpu. The thread id is listed as the last column. Then, run gdb > attach <telegram pid> > thread <thread id> > where This will give you the backtrace of that thread. This may help in diagnosing the problem. Also, you may report a bug to upstream: https://github.com/telegramdesktop/tdesktop/issues Finally, you may run telegram under ktrace, to find out if telegram is busy calling some system calls, etc.
There already one: https://github.com/telegramdesktop/tdesktop/issues/3640
(In reply to w.schwarzenfeld from comment #3) That ticket describes absolutely different behavior.
You may try the new version of linux-telegram-desktop, version 1.3.0. If that helps, I'll try to speed up the process of updating the native version.