screen bug in FC4

[Cimmo]

1) Someone else seen this bug? Seem to appear only with KDE, at random time.
Attachment is very poor because I have to stay under 40 KB, just see all 
the dark part, like window's draw is broken.

2) another thing: I've freshly installed FC4, but the root is /1 instead 
of / that I have specified, any ideas?

my conf:
FC4-x86_64 final
Athlon64 3000+
1 GB of ram
GeForce6600 GT
nv module (no nvidia module installed yet)

thanx
Cimmo

[David Kewley]

On Wednesday 15 June 2005 10:43, Cimmo wrote:
>  1) Someone else seen this bug? Seem to appear only with KDE, at
>  random time. Attachment is very poor because I have to stay under 40
>  KB, just see all the dark part, like window's draw is broken. 
Yes, I see this regularly using KDE under FC4 (fully updated).  I 
assumed hardware since it looks like memory corruption, but an 18-hour 
memtest86 shows no problems.  I figured it might be my video card (even 
though it works for playing Win2k games), but since you see it too, I 
guess not.

Sometimes I'm able to fix it with switching to a text vt & back, but 
sometimes nothing works and I have to hit the reset button.  (I don't 
have another machine at home from which to try to ssh in.)

>  my conf:
>  FC4-x86_64 final
>  Athlon64 3000+
>  1 GB of ram
>  GeForce6600 GT
>  nv module (no nvidia module installed yet) 
mine:
FC4 x86_64 final
  (upgraded, admittedly, from FC4t3, where I also saw this problem)
MSI K8N Neo2 Platinum (Nforce3)
Athlon64 3000+, Venice core
GeForce 6800 (by Gigabyte)
1.5 GB of ram (also happened when I had 1GB)
nv module

I am only just now updating the xorg update packages that were released 
(today?) -- let's see if the problem doesn't reappear.

David

[Kyrre Ness Sjobak]

ons, 15.06.2005 kl. 20.33 skrev David Kewley:
>  On Wednesday 15 June 2005 10:43, Cimmo wrote:
>  > 1) Someone else seen this bug? Seem to appear only with KDE, at
>  > random time. Attachment is very poor because I have to stay under 40
>  > KB, just see all the dark part, like window's draw is broken.
>  
>  Yes, I see this regularly using KDE under FC4 (fully updated).  I 
>  assumed hardware since it looks like memory corruption, but an 18-hour 
>  memtest86 shows no problems.  I figured it might be my video card (even 
>  though it works for playing Win2k games), but since you see it too, I 
>  guess not.
>  
>  Sometimes I'm able to fix it with switching to a text vt & back, but 
>  sometimes nothing works and I have to hit the reset button.  (I don't 
>  have another machine at home from which to try to ssh in.)
>  
>  > my conf:
>  > FC4-x86_64 final
>  > Athlon64 3000+
>  > 1 GB of ram
>  > GeForce6600 GT
>  > nv module (no nvidia module installed yet)
>  
>  mine:
>  FC4 x86_64 final
>    (upgraded, admittedly, from FC4t3, where I also saw this problem)
>  MSI K8N Neo2 Platinum (Nforce3)
>  Athlon64 3000+, Venice core
>  GeForce 6800 (by Gigabyte)
>  1.5 GB of ram (also happened when I had 1GB)
>  nv module
>  
>  I am only just now updating the xorg update packages that were released 
>  (today?) -- let's see if the problem doesn't reappear.
>  
>  David 
I had a similar problem with a vodoo 3 PCI card on fc2 - assumed it was
a defective card, switched it to a gforce 2, and it worked.

It would randomly crash, making the screen flicker (like a wrong
frequency). Most of the time i was able to switch consoles back an forth
to restore, but sometimes i wasn't possible. I was however able to ssh
in from another machine, even if it looked like something was terribly
wrong - the connection ligths on the switch turned on and off at random,
and the floppy/cd drive ligths was blinking like they use to during
kudzu.

On fc3 early days, there was also a bug with the intel gfx drivers that
caused the screen to "hang" - patches of it wasn't redrawn.

Kyrre

[Cimmo]

David Kewley ha scritto:

> I am only just now updating the xorg update packages that were
> released 
> (today?) -- let's see if the problem doesn't reappear.
>   
>  
I have fully updated xorg with 15 of june updated packages and still 
same issue.
Is there some bug filed into bugzilla or should we do? Is really 
annoying bug, KDE is unusable...

bye
Cimmo

[Michael Schwendt]

On Thu, 16 Jun 2005 01:50:09 +0200, Cimmo wrote:

>  David Kewley ha scritto:
>  
>  >I am only just now updating the xorg update packages that were released 
>  >(today?) -- let's see if the problem doesn't reappear.
>  >  
>  >
>  I have fully updated xorg with 15 of june updated packages and still 
>  same issue.
>  Is there some bug filed into bugzilla or should we do? Is really 
>  annoying bug, KDE is unusable... 
Can you reproduce it with GNOME, so you can assume that it is something in
the xorg-x11 drivers?  They have other issues with other graphics chipsets
(maybe due to changes that come with GCC 4.0.0), one more issue would not
be a surprise. ;)

[Cimmo]

Michael Schwendt ha scritto:

> Can you reproduce it with GNOME, so you can assume that it is
> something in
> the xorg-x11 drivers?  They have other issues with other graphics chipsets
> (maybe due to changes that come with GCC 4.0.0), one more issue would not
> be a surprise. ;)
>
>   
>  
Cannot reproduce with Gnome.
I've also reproduced one of the random time that bug appears (it also 
appears in other random ways):

- open KDE
- open control center
- go to system administrator
- go to login manager
- click on asministrator mode

and after inserted root password corruption starts.
Also seen that it appears an image of Windows XP logout with corruption, 
(I've both OS installed), but don't know why!!!

please help!

Cimmo

[Kyrre Ness Sjobak]

fre, 17.06.2005 kl. 16.27 skrev Cimmo:
>  Michael Schwendt ha scritto:
>  
>  >Can you reproduce it with GNOME, so you can assume that it is something in
>  >the xorg-x11 drivers?  They have other issues with other graphics chipsets
>  >(maybe due to changes that come with GCC 4.0.0), one more issue would not
>  >be a surprise. ;)
>  >
>  >  
>  >
>  Cannot reproduce with Gnome.
>  I've also reproduced one of the random time that bug appears (it also 
>  appears in other random ways):
>  
>  - open KDE
>  - open control center
>  - go to system administrator
>  - go to login manager
>  - click on asministrator mode
>  
>  and after inserted root password corruption starts.
>  Also seen that it appears an image of Windows XP logout with corruption, 
>  (I've both OS installed), but don't know why!!! 
Looks like the videocard is
reading from wrong memory? Had a buggy
computer do that once - using the VESA driver

[Cimmo]

Kyrre Ness Sjobak ha scritto:

> Looks like the videocard is reading from wrong memory? Had a buggy
> computer do that once - using the VESA driver
>   
>  
on windows video card works perfectly both in games and normal use.
any other people that can help me?

b.r.
Cimmo

[David Kewley]

On Friday 17 June 2005 09:52, Cimmo wrote:
>  Kyrre Ness Sjobak ha scritto:
>  >Looks like the videocard is reading from wrong memory? Had a buggy
>  >computer do that once - using the VESA driver
>
>  on windows video card works perfectly both in games and normal use.
>  any other people that can help me? 
Hi Cimmo,

Just to let you know where I am: I've not see the screen corruption in 
the past day or so, after updating xorg* and powercycling the computer.  
And today there are KDE updates (not that KDE should conceivably cause 
X screen memory corruption, I think).  So I'm going to wait until I see 
the issue again to give more details and/or open a bug report.

David

[David Kewley]

On Friday 17 June 2005 11:51, David Kewley wrote:
>  On Friday 17 June 2005 09:52, Cimmo wrote:
>  > Kyrre Ness Sjobak ha scritto:
>  > >Looks like the videocard is reading from wrong memory? Had a buggy
>  > >computer do that once - using the VESA driver
>  >
>  > on windows video card works perfectly both in games and normal use.
>  > any other people that can help me?
>
>  Hi Cimmo,
>
>  Just to let you know where I am: I've not see the screen corruption
>  in the past day or so, after updating xorg* and powercycling the
>  computer. And today there are KDE updates (not that KDE should
>  conceivably cause X screen memory corruption, I think).  So I'm going
>  to wait until I see the issue again to give more details and/or open
>  a bug report.
>
>  David 
FYI, to complete this thread (?), I have not seen the screen corruption 
since I updated xorg* and kde (a bunch of rpms), and power-cycled.  I 
don't believe my use patterns now are very different from before, so 
perhaps the source of the problem I saw has been fixed.

Cimmo, are you still seeing problems after the above-mentioned updates & 
power-cycle?

David

[Cimmo]

David Kewley ha scritto:

> FYI, to complete this thread (?), I have not seen the screen
> corruption 
> since I updated xorg* and kde (a bunch of rpms), and power-cycled.  I 
> don't believe my use patterns now are very different from before, so 
> perhaps the source of the problem I saw has been fixed.
>
> Cimmo, are you still seeing problems after the above-mentioned updates & 
> power-cycle?
>
>   
>  
Thanx David,
With new xorg same results, haven't tried yet new kde, because my yum 
and up2date stop working after the first time :(
I'm affected by this bug that it was ignored by red hat because all 
people think it is a nvidia network related bug, but now I think it isn't!

https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=152566

Thanx
Cimmo

[Cimmo]

David Kewley ha scritto:

> FYI, to complete this thread (?), I have not seen the screen corruption 
> since I updated xorg* and kde (a bunch of rpms), and power-cycled.  I 
> don't believe my use patterns now are very different from before, so 
> perhaps the source of the problem I saw has been fixed.
>
> Cimmo, are you still seeing problems after the above-mentioned updates & 
> power-cycle?
>
>   
>  
done a full update and power-cycle.
Same issue :(

Cimmo

[David Kewley]

On Tuesday 21 June 2005 07:54, Cimmo wrote:
>  David Kewley ha scritto:
>  >FYI, to complete this thread (?), I have not seen the screen
>  > corruption since I updated xorg* and kde (a bunch of rpms), and
>  > power-cycled.  I don't believe my use patterns now are very
>  > different from before, so perhaps the source of the problem I saw
>  > has been fixed.
>  >
>  >Cimmo, are you still seeing problems after the above-mentioned
>  > updates & power-cycle?
>
>  done a full update and power-cycle.
>  Same issue :(
>
>  Cimmo 
A quick update: I found out this morning that I had *not* in fact 
updated KDE; I'm doing that now.  I could swear I'd done that, so 
perhaps I did it on my work machine rather than my home machine...  
Sigh.  I found this out because last night I got screen corruption & X 
nonresponse, as before, so I double-checked my installed packages.

This time I let the machine run several hours overnight to see if it 
would ever respond to Ctrl-Alt-F1; it did not, so X at least was locked 
hard.  As before, the mouse pointer moves, but onscreen widgets do not 
update.  Over the past few days, until last night, I had experienced  
no problems, including a few hours of troublefree 3D gaming in Win2k.

Cimmo, how about you and I take this path:

* If you wish to submit a bug report now, go ahead and do so, and reply 
to this thread with a URL to the bug report.  I'll fill in the details 
I have.

* Alternately, wait until I've confirmed whether I still see this bug 
with KE 3.4.1 before submitting a bug report.

Have you searched http://bugzilla.redhat.com/ for similar-sounding 
reports?  I have not yet.

Regarding your network bug report that hadn't seen any action in the 
past three months until you brought it up on this list today, I hear 
your frustration.  But I think that the list members are being very 
helpful to you, both in terms of tracking down the bug & giving you 
guidance about how to get attention to your bugs.  I hope you and the 
bug respondents will together gather all the required facts & find the 
problem.  Remember that it could possibly be hardware, even though you 
also saw it with a PCI network card.  Unless you're a deep, deep 
hardware wizard (I'm good, but not a deep wizard), never 100% rule out 
the possibility of hardware problems. :)

David

[Cimmo]

David Kewley ha scritto:

> Cimmo, how about you and I take this path:
>
> * If you wish to submit a bug report now, go ahead and do so, and reply 
> to this thread with a URL to the bug report.  I'll fill in the details 
> I have.
>   
>  
Just done, have you seen other post in this thread?
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=161239

> * Alternately, wait until I've confirmed whether I still see this
> bug 
> with KE 3.4.1 before submitting a bug report.
>   
>  
kde 3.4.1 results in the same issue

> Have you searched http://bugzilla.redhat.com/ for similar-sounding 
> reports?  I have not yet.
>   
>  
yes, found only a cirrus corrupted screen.

> Regarding your network bug report that hadn't seen any action in
> the 
> past three months until you brought it up on this list today, I hear 
> your frustration.  But I think that the list members are being very 
> helpful to you, both in terms of tracking down the bug & giving you 
> guidance about how to get attention to your bugs.  I hope you and the 
> bug respondents will together gather all the required facts & find the 
> problem.  Remember that it could possibly be hardware, even though you 
> also saw it with a PCI network card.  Unless you're a deep, deep 
> hardware wizard (I'm good, but not a deep wizard), never 100% rule out 
> the possibility of hardware problems. :)
>   
>  
I have just resolved, it is a "NON A BUG" but don't know how it shows

before.

thanx for your reply, please fill your system spec in my bug report if 
you see corrupted screen again.

bye
Cimmo

[Cimmo]

opened a new bug report
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=161239