[arch-general] Why is clock_gettime called when the mouse moves?

[Ralph Corderoy]

Hi Piscium,

Second attempt, the first went to the old list address.

> What I posted before was just a short snippet. In
> the link below there is a trace for leafpad, available for 7 days.
> https://we.tl/t-BAbO60lnGN
>
> > I suggest using the simplest X client you can which shows the issue,
> > e.g. xev(1).
...
> Good suggestion. I tried xev and I cannot reproduce the issue. The
> simplest app that I tried that shows the issue is leafpad.
>
> > The second, if you read the two other emails in the thread, fixes the
> > problem by changing the kernel's clock source which may differ between
> > your two test machines?
>
> I am not sure what you are referring to, maybe a link for the second
> page is missing?

Sorry.
https://www.mail-archive.com/debian-bugs-dist@lists.debian.org/msg1823248.html

> There are tons of GUI apps. I have tested just a handful of them.
> Have the clock_gettime issue: leafpad, mousepad, emacs, audacious, konsole.
> Don't have the issue: xev, xterm.
>
> I tried to figure out if there is a common library used by all apps
> that have the problem but so far have not found it. A way to figure
> this out might be using gdb and putting a breakpoint on clock_gettime,
> but i am reluctant to do this as it would take a lot of time

It may be perf_events would help you analyse the frequency of difference
clock_gettime() callers, say by maintaining frequency counts of the
previous three functions in the call stack, but I don't know enough to
suggest how.  https://www.brendangregg.com/perf.html

-- 
Cheers, Ralph.