1 |
Further adventures: |
2 |
|
3 |
After wiping and reinstalling from scratch, the freeze still occurs. |
4 |
I seem to have cut down on the frequency, though; I was able to get a good |
5 |
several hours of running time before the lockup. I also tried immediately to |
6 |
restart X once the computer came back up, and the next freeze was very |
7 |
quick. |
8 |
This confuses me - that pattern makes me think this is heat, but |
9 |
under Windows I can play any game I want for a good long time and not have |
10 |
the system overheat. I've been careful to include the ACPI drivers in my |
11 |
kernel, and if I understand correctly, that should make sure overheating |
12 |
doesn't occur, more or less. |
13 |
It's also not related to the computer's workload. I ran glxgears and |
14 |
several applications without any hitches - when the crash finally occurred, |
15 |
it was at relatively low system load. |
16 |
|
17 |
As far as the scope of the crash, it's total. I tried pinging the |
18 |
machine after it froze, and got nothing back, which also scratches the |
19 |
remote access idea. |
20 |
|
21 |
I also threw in a Fedora Core 2 install, just to see what happened, |
22 |
and I didn't get a crash. I can't put too much stock in that, given the long |
23 |
running time I got here, but I'm also wondering what might happen if I put a |
24 |
recent version of XFree86 on this thing instead of X.org's server. |
25 |
|
26 |
The threading idea is also something I've poked at: the laptop has a |
27 |
Pentium 4, and I've seen the HyperThreading or whatever it's called cause |
28 |
some hell, at least under Windows. |
29 |
I'm running the development-sources, and I've tried both with and |
30 |
without SMP in the kernel - and I was also sure to disable the APIC option |
31 |
mentioned earlier in this thread. The crash still occurred. |
32 |
|
33 |
Right now what I'm going to try and do is use X.org 6.8.0 and the |
34 |
kernel radeon driver with DRM. |
35 |
|
36 |
Another idea - might this be some odd option in the kernel that's |
37 |
causing this? I've seen lots of xorg.conf files posted in the forums for |
38 |
Inspiron XPS laptops, but I don't have a kernel .config to test against. |
39 |
|
40 |
And, thank you again for all the suggestions. |
41 |
|
42 |
Gabe. |
43 |
|
44 |
|
45 |
|
46 |
> -----Original Message----- |
47 |
> From: Stuart Stegall [mailto:stuart@×××××××××××××.com] |
48 |
> Sent: Monday, October 25, 2004 12:43 PM |
49 |
> To: 'Paul de Vrieze'; gentoo-desktop@l.g.o |
50 |
> Subject: RE: [gentoo-desktop] X freezing? |
51 |
> |
52 |
> To add further fun to this equation, Dell laptops have notoriously |
53 |
> broken bios' and is probably part of the problem. I use an Inspiron |
54 |
> 1100 and depending on what version of X Windows I wish to |
55 |
> use, I have to |
56 |
> flash a particular bios (and 6.8.1 is currently not working with any |
57 |
> version). |
58 |
> |
59 |
> Thanks, |
60 |
> Stuart |
61 |
> |
62 |
> > -----Original Message----- |
63 |
> > From: Paul de Vrieze [mailto:pauldv@g.o] |
64 |
> > Sent: Monday, October 25, 2004 4:39 AM |
65 |
> > To: gentoo-desktop@l.g.o |
66 |
> > Subject: Re: [gentoo-desktop] X freezing? |
67 |
> > |
68 |
> > |
69 |
> > On Monday 25 October 2004 00:56, A. Gabriel W. Daleson wrote: |
70 |
> > > Hm. I've emerged openoffice already, and also given it |
71 |
> > long lists of |
72 |
> > > other packages to merge, both while X doesn't run. It |
73 |
> > doesn't seem to |
74 |
> > > be related to the machine load in the that sense. |
75 |
> > > |
76 |
> > > Also on this note, I'm dual booting the system - Windows seems to |
77 |
> > > have no trouble with it at all, and every hardware check I've run |
78 |
> > > comes out fine. If it's a hardware problem, it's a real |
79 |
> subtle one. |
80 |
> > |
81 |
> > Windows behaviour is normally not very good in detecting |
82 |
> > broken hardware. |
83 |
> > Windows is rather conservative in it's use of instructions |
84 |
> > (it must run |
85 |
> > on the whole range, while a gentoo system doesn't need to). |
86 |
> > Further there |
87 |
> > are normally enough other sources of crashes to make it hard |
88 |
> > to see that |
89 |
> > a crash is hardware related. |
90 |
> > |
91 |
> > Paul |
92 |
> > |
93 |
> > ps. Do note that the X drivers are not entirely bugfree |
94 |
> > either, and can |
95 |
> > and will lock up your screen, keyboard and mouse. Remote access is |
96 |
> > normally still possible. |
97 |
> > |
98 |
> > pps. There is another good source of seemingly random |
99 |
> > crashes, and that |
100 |
> > has to do with race conditions and other thread related issues. |
101 |
> > Programming safely for multiple threads is difficult. |
102 |
> > Especially if those |
103 |
> > threads are executed on more than one processor. |
104 |
> > |
105 |
> > The causes of such behaviour could be anywhere. Given that |
106 |
> > kde is normally |
107 |
> > stable on multiprocessor systems I would certainly not exclude the |
108 |
> > mainboard or graphics card bioses. (Yes, graphic cards have a |
109 |
> > bios too, |
110 |
> > and it can normally be updated) |
111 |
> > |
112 |
> > -- |
113 |
> > Paul de Vrieze |
114 |
> > Gentoo Developer |
115 |
> > Mail: pauldv@g.o |
116 |
> > Homepage: http://www.devrieze.net |
117 |
> > |
118 |
> |
119 |
> |
120 |
> |
121 |
> -- |
122 |
> gentoo-desktop@g.o mailing list |
123 |
> |
124 |
|
125 |
|
126 |
-- |
127 |
gentoo-desktop@g.o mailing list |