Re: [Xen-devel] [linux-linus test] 25478: regressions - FAIL

[Ian Campbell <Ian.Campbell@xxxxxxxxxx>]

On Mon, 2014-03-17 at 15:36 -0400, Konrad Rzeszutek Wilk wrote:

> > >  will also tell you what type of requests it was.
> > 
> > Do you mean the size? It seems to print that only for certain requests.
> > 
> 
> Right. They aren't that big - and only enough the mptsas is a 4KB one?

s/only/oddly/?

FWIW across serial-gall-mite.log and .0 I see:
      1  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 10072 bytes)
      1  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 12920 bytes)
      1  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 13032 bytes)
      2  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 1448 bytes)
      1  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 14688 bytes)
      1  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 15680 bytes)
      1  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 15928 bytes)
      1  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 21024 bytes)
      1  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 26064 bytes)
      3  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 2760 bytes)
     19  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 2896 bytes)
      1  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 31024 bytes)
      8  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 4344 bytes)
      1  bnx2 0000:02:00.0: swiotlb buffer is full (sz: 8688 bytes)
  25209  mptsas 0000:03:00.0: swiotlb buffer is full
    878  mptsas 0000:03:00.0: swiotlb buffer is full (sz: 4096 bytes)
      1 serial-gall-mite.log:00:03:00.0: swiotlb buffer is full
      1 serial-gall-mite.log:Mar 16 12:53:55.385556 ptsas 0000:03:00.0: swiotlb 
buffer is full
      1 serial-gall-mite.log: mptsas 0000:03:00.0: swiotlb buffer is full
      2 serial-gall-mite.log:mptsas 0000:03:00.0: swiotlb buffer is full
      1 serial-gall-mite.log:sas 0000:03:00.0: swiotlb buffer is full

(where the first column is the number of times it was seen. From:
$ wget 
http://www.chiark.greenend.org.uk/~xensrcts/logs/25558/test-amd64-i386-pair/serial-gall-mite.log.0

        $ grep swiotlb.buffer.is.full serial-gall-mite.log* | cut -f 2 -d \] | 
sort | uniq -c

> 
> > > You might also want to try a larger SWIOTLB buffer, swiotlb=26422 for fun.
> > 
> > Any reason for that particular number?
> 
> It would allocate a larger SWIOTLB space - in case you are using at near its
> capacity.
> > 
> > > I think you are looking at two different issues.
> > 
> > You mean you think the swiotlb issue is unrelated to the slow migration
> > timeouts? I can believe that.
> > 
> > Ian.
> > 



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel