panic: Duplicate free of item 0xc433e420 from zone 0xc1045c00(g_bio)

[Søren Schmidt]

Garance A Drosihn wrote:
> I got another panic.  This time I got a clean core-dump, and I also
> have a kernel with all the debugging symbols if that would help.
> Here is some of the info.  The system had been up about 18 hours
> before the panic.  The system was basically idle at the time of the
> panic.  This is a snapshot of current which I started to build at
> about 2pm EST on January 20th.

The duplicate free on failed ATA retry is known. I have a fix here 
locally but it doesnt apply to whats in -current..

However, -current has severe problems that provokes those timeouts for 
no good reason and in all my cases locks up hard. I just vasted last 
nights sleep hunting for what I thought was a bug in my WIP. I had 
PREEMPTION turned of etc, but it sitll failed in odd ways.
Going back to Jan 10 00:00 UTC sources make things "just work" with 
PREEMPTION and all. My testbox has now moved around some 800 GB's of 
data with no problems with above backstep, on -current it fails within 
the first few Mbytes..

> Note that I have no idea when those WRITE_DMA messages came out, so
> I do not know if they are related to the panic, or if those messages
> were written hours earlier.
> 
> Let me know if anyone wants to pursue this further.
> 
> ad4: TIMEOUT - WRITE_DMA retrying (2 retries left) LBA=42098457
> ad4: TIMEOUT - WRITE_DMA retrying (2 retries left) LBA=42098457
> ad4: FAILURE - WRITE_DMA timed out

-- 

-Søren