Re: Torture test for Nautilus

[Alexander Larsson <alexl redhat com>]

On Mon, 2006-10-09 at 17:17 -0500, Federico Mena Quintero wrote:

> - Nautilus doesn't survive very long if made to display the ~/torture
> directory :)  There are some assertion failures and some "real" crashes.

How long did you have to run it to get a failure? I ran if for a while
and didn't see anything (HEAD). I'll start a longer run now to test.

> - gnome-vfs-monitor performs very badly.  For a torture directory with
> only 400 files in it, gnome-vfs-monitor already had a list of callbacks
> with over 12,000 elements.  It spent all its time in g_list_append() and
> in get_min_delay(), apparently freezing the GUI.

I made some performance fixes to gnome-vfs-monitor.c to make this less
total ass performance-wise, but the fact is that you're running an
application that creates file notification events as quickly as the cpu
can generate them. We can't really do anything but spend 100% of the
time handling incoming events. 

The "ideal" thing to do in a case like this is to detect it and switch
over to a mode where we reload the directory once every few seconds. I'm
not sure this is an important case to spend a lot of time on though.

> The torture program supports a "--seed" option that you can use to seed
> the random number generator.  This is so that the sequence of "random"
> torture steps will be reproducible.

However, it doesn't print the seed used on startup, so you don't know
what seed to use the second time.

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 Alexander Larsson                                            Red Hat, Inc 
                   alexl redhat com    alla lysator liu se 
He's an oversexed neurotic cat burglar searching for his wife's true killer. 
She's an artistic wisecracking femme fatale with the power to bend men's 
minds. They fight crime!