| Summary: | CPU lock with nouveau_fan_update | ||||||
|---|---|---|---|---|---|---|---|
| Product: | xorg | Reporter: | ddamienn | ||||
| Component: | Driver/nouveau | Assignee: | Nouveau Project <nouveau> | ||||
| Status: | RESOLVED INVALID | QA Contact: | Xorg Project Team <xorg-team> | ||||
| Severity: | major | ||||||
| Priority: | medium | CC: | ddamienn | ||||
| Version: | unspecified | ||||||
| Hardware: | Other | ||||||
| OS: | Linux (All) | ||||||
| Whiteboard: | |||||||
| i915 platform: | i915 features: | ||||||
| Attachments: |
|
||||||
nouveau_fan_update:
spin_lock_irqsave(&fan->lock, flags);
/* schedule next fan update, if not at target speed already */
if (list_empty(&fan->alarm.head) && target != duty) {
u16 bump_period = fan->bios.bump_period;
u16 slow_down_period = fan->bios.slow_down_period;
...
ptimer->alarm(ptimer, delay * 1000 * 1000, &fan->alarm);
If delay is somehow 0, the ->alarm will cause nouveau_fan_update to get called immediately. Can you add a printk to that function that shows the values? (This may end up totally flooding your dmesg too... but I think the values may be the same across prints.)
e.g.
diff --git a/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c b/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
index c728380..9453afd 100644
--- a/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
+++ b/drivers/gpu/drm/nouveau/core/subdev/therm/fan.c
@@ -88,7 +88,7 @@ nouveau_fan_update(struct nouveau_fan *fan, bool immediate, int target)
delay = min(bump_period, slow_down_period) ;
else
delay = bump_period;
-
+ nv_info(therm, "Scheduling fan update in %d (slow: %d, bump: %d)\n", delay, slow_down_period, bump_period);
ptimer->alarm(ptimer, delay * 1000 * 1000, &fan->alarm);
}
(In reply to comment #1) > If delay is somehow 0, the ->alarm will cause nouveau_fan_update to get > called immediately. Can you add a printk to that function that shows the > values? (This may end up totally flooding your dmesg too... but I think the > values may be the same across prints.) Thanks for your help. I will try to add the printk. The lock has made the machine inaccessible remotely, so it'll be a couple of days until I can post the results. I've had the debug statement active for the last few days, but the error hasn't re-occurred. I discovered that my video card's fan was only working intermittently, which a clean seems to have fixed. Since the inactive fan likely caused the error I was seeing, I am closing this bug report as invalid. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.
Created attachment 80540 [details] kernel log Kernel logs report "Watchdog detected hard LOCKUP on cpu 1". It has occurred once per day for the last three days. Each time, it happens seemingly randomly - monitor is turned off, I'm connected remotely via SSH but system is idle. This is on ArchLinux with xf86-video-nouveau 1.0.7. The machine is ~2.5 years old and has been very solid until now. My video card is: 01:00.0 VGA compatible controller: NVIDIA Corporation GT216 [GeForce GT 220] (rev a2). I have attached the kernel log. I also note that I don't know why the 'PTHERM' events would be occurring - this was while the connected monitor was off and I was only connected remotely via SSH.