Closed GoogleCodeExporter closed 9 years ago
[deleted comment]
First thing that comes to mind when reading this:
Are you monitoring your video card's temperature? Does this happen with any
other
game after long periods of constant playing?
You may want to try watching your video card's temperature, or at least
checking it
after such an occurrence to see if it somehow got up to ~170*F or so. At this
temperature, the newer lead-free solder compounds tend to get a little less...
solid.
I suggest the tool "CPUCool" - they have a trial version available for free.
You also could, on general principle, remove your video card and blow out the
dust
with compressed air. Good to do on a regular basis anyway, might as well blow
out
your computer as well (pay particular attention to the processor fan / heatsink
and
the video card fan / heatsink.)
Original comment by chris.paiano@gmail.com
on 15 Jul 2009 at 9:44
(oops, you're running Linux. I'm sure they have a temperature monitoring tool
available somewhere too...)
Original comment by chris.paiano@gmail.com
on 15 Jul 2009 at 9:45
Yes, in the nvidia-settings utility that comes with the driver. You can find a
GPU
Temp monitor there.
Original comment by lysdestic@gmail.com
on 15 Jul 2009 at 11:00
Correct, as lysdestic stated, there is a GPU temp monitor in my
Nvidia-settings. I
haven't ever watched it while playing at the same time since I play fullscreen.
I
will have to run FoFix in a window and have the nvidia thermal monitor open at
the
same time to monitor how high it gets.
I do exit the game and open up the nvidia tool right after to see what it is.
The
highest it has ever been is 51*c. The temp threshold is 122*c. And it is
currently
at 45*c (which is where it is normally).
I'm not too sure this is temp related though. It happens with multiple events
and
messages on the screen at once. I will try fofix in a window and monitor temp
at
same time to see though and report back.
Original comment by mtanne...@yahoo.com
on 16 Jul 2009 at 1:40
Also, I'm sure you aren't but just to check: Is compositing disabled?
I would love to attempt to recreate this for you this weekend, as I am running
Kubuntu with KDE 4.3 RC1, an NVIDIA 8400GS, python2.6 /and/ the 185.18.14
driver.
Talk about a near-perfect machine to replicate this on, unfortunately, since I
updated [some-package] I've had a pygame segfault with shaders enabled.
Check your compositing, tweak some settings in nvidia-settings, roll back video
drivers, monitor the temp, and if at all possible, spend 40+ minutes in another
game
that uses some form of shaders. Hopefully this can be tracked down, whether it
be
from FoFiX or your machine.
Original comment by lysdestic@gmail.com
on 16 Jul 2009 at 7:31
Yes, I am running compositing...just the basic KDE 4.3 kwin compositing. No
compiz
stuff...
That is funny you have almost the same setup. What CPU? I have P4 2.5 Ghz,
1gig
ram, and the 8400GS has 512mb. It is a PCI card.
I have noticed when the popup messages appear...they lag. They are in slow
motion to
fly up/jump up and down a few times, and then slowly fly away. And when these
messages appear, the notes and lyrics also slow down. It is not as bad as it
was in
previous revisions. But it is not as fast as what I see in some of the videos
other
users post. I just figured it was my setup hindering the performance.
As far as other games to try with shaders...This is the only game that I play.
So I
don't really know of any other linux games. Have any recommendations???
Original comment by mtanne...@yahoo.com
on 16 Jul 2009 at 12:55
I just got done playing FoFix for the past 40-60 minutes with about 10 songs
played.
I played it in a window and I did have the nvidia thermal monitor open to monitor
the temp. I left compositing ON. This time, the shader blanked to white while
loading a song...after the 3 second countdown.
My observation, I don't see this being a temp issue. Before playing the game,
it was
46. And while playing it fluctuated between 49-50*c with 51*c being the
highest.
When the shader blanked, it was at 50*c.
There was a few times that it went up to 55*c...but this was after playing a
song and
it was loading the screens after (good job message, scores...).
I was also monitoring the CPU/Ram/Swap usage. CPU seemed to be pegged at 100%.
The
ram started off at 55% used. But then the more and more I played, the ram usage
finally went up to 100%. And then swap started being used. I didn't have the
system
monitor up while the shader turned white to see what it was. But I believe it
was
around 100% CPU/ 100% Ram / and 20% Swap. My swap is 1.1 gig.
I don't know if there is a correlation between ram at 100% and swap going up to
20%
to make the shader turn white. But it seems odd that as there isn't any more
ram, it
has to use the hard drive making the game slower eventually shutting off the
stage
shader. Again, not sure if this is what the cause is. Just plain observation.
I did take an actual screenshot this time. It wasn't the exact moment it
happened.
But it shows the temp (shows 51*c...it was at 50*c when it happened). It shows
the
shader blanked out to white. And it also shows the system monitor overlay at
the
bottom (not real time when it happened). This was just to show what the
cpu/ram/and
swap was at around the time it happened.
Hopefully you are able to reproduce this lysdestic. I am still looking for
another
linux game that uses shaders.
Original comment by mtanne...@yahoo.com
on 16 Jul 2009 at 2:36
Attachments:
That sounds like the old memory leak problem that was meant to be fixed in
Python 2.6.
Original comment by evilbobt...@gmail.com
on 16 Jul 2009 at 4:19
re: "memory leak..."
Are you sure its not a memory leak related to the stage shader itself? It
seems odd
that the stage shader is the only one that blanks out like this. And you are
still
able to play the game. I will have to turn off the stage shader and play for 40
minutes and see if my ram is used up leading to the swap usage. If it is a leak
caused by the shader, then with it off would not use up all the ram and start
using
swap.
Original comment by mtanne...@yahoo.com
on 20 Jul 2009 at 3:51
Any news on this?
Any luck?
Original comment by evily...@gmail.com
on 25 Sep 2009 at 3:55
Issue set to invalid due to one or more of the following reasons:
1) Age. This issue is considered abandoned.
2) Unrelated. The issue isn't related to FoFiX.
3) Fluke. The issue cannot be reproduced or was fixed long ago. (AKA: It's just
you.)
Feel free to reply if you are certain the issue still exists and is relevant to
FoFiX.
Original comment by fuzio...@gmail.com
on 2 Sep 2010 at 1:39
Original issue reported on code.google.com by
mtanne...@yahoo.com
on 15 Jul 2009 at 9:16Attachments: