1 |
Jeremy Huddleston writes: |
2 |
> Ok, you said this was a u2, right? So I assume you have SMP enabled, |
3 |
> and that causes lockups on all 2.6.x kernels. If you don't have SMP |
4 |
> enabled, please let me know as I've been unable to lockup 2.6.7-r17 or |
5 |
> 2.6.8-r11 (not sure if dsd put it in portage yet, but it should be |
6 |
> marked -sparc when he does for testing) on my u10. |
7 |
|
8 |
Yeah, SMP + PREEMPT |
9 |
|
10 |
> So this is more of a qualitative thing, but was it "easier" to lockup |
11 |
> 2.6.7 than it was to lockup 2.6.6? That's what I've been hearing, but |
12 |
> I'd like a side-by-side comparison... |
13 |
|
14 |
when running my stress-test on vanilla 2.6.6, the OOM-killer starts |
15 |
killing things off 'correctly' and it doesn't seem to crash -- at least |
16 |
I haven't run it into the ground yet. I suppose I can try again. |
17 |
|
18 |
gds-2.6.7-r16 dies abruptly without ever calling the OOM-killer. |
19 |
|
20 |
> If that's the case, would you be willing to test out 2.6.7-rc1 and |
21 |
> 2.6.7-rc2 and help narrow down when the lockup was introduced... |
22 |
|
23 |
I assume you're speaking of the vanilla sources? |
24 |
I'll download and compile them.... |