Re: Is: SKB_MAX_LEN bites again. Was: Re: bug disabling guest interface

[Steven Haigh <netwiz@crc.id.au>]

[-- Attachment #1.1: Type: text/plain, Size: 2416 bytes --]

On 9/03/2013 1:57 PM, Ian Campbell wrote:
>>> - change MAX_SKB_FRAGS to 19 to accommodate all guests
>
> Changing MAX_SKB_FRAGS is *not* an option upstream. This might be a
> useful local hack but we need to drop the idea as a long term fix.

I agree. Its a hack that seems to work until we have something else to 
offer though. I'll be honest, the internals of the kernel interactions 
like this is a bit beyond my knowledge - but I get quite a few emails of 
stuff going strange. This one has increased dramatically in the past week.

>> Ugh. The negotiations between host and guest is probably the best
>> choice. The issues you are going to hit are that you might need
>> to redo the skbs to match what the frontend's max is.
>
> IMHO the right fix is for netback to coalesce as it copies from the
> frontend if it needs to do so, it is copying anyway so it should be
> cheap enough. I thought we had discussed this and someone was working on
> implementing it. If not Annie then perhaps it was Matt or Siva (both now
> CC'd)

I did see some talk about it on the xen-devel lists quite some time ago 
- however it seemed to die with no outcome that I could find. Maybe its 
a good time for a nudge on the matter :)

> If necessary netback could even allocate a larger order head in order to
> accommodate very large packets, but I don't expect that to be required
> to fix the immediate issue we are seeing (but gives flexibility)
>
> This should get us past the immediate issue of the upstream change from
> 18->16  frags thing. Longer term the negotiation will allow us to avoid
> future incompatible changes in guest and host network stacks, as well as
> allowing frontends on other OSes (in particular Windows) to havea better
> chance of to DTRT.

Whatever the fix, it has to be on the Dom0 kernel - as stated by others 
in the past, it isn't a feasible fix to include changes on the client. 
There would be too many different guest OSes that would make 
implementation a nightmare - or in fact impossible.

>> Annie, Wei, Ian - were there some RFC patches floating around
>> for this?

I didn't stumble across any in the list archives that I hunted for. Not 
to say they don't exist, but I didn't find them if they do exist.

-- 
Steven Haigh

Email: netwiz@crc.id.au
Web: https://www.crc.id.au
Phone: (03) 9001 6090 - 0412 935 897
Fax: (03) 8338 0299


[-- Attachment #1.2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4240 bytes --]

[-- Attachment #2: Type: text/plain, Size: 126 bytes --]

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