Messages in this thread |  | | | Date | Thu, 26 Mar 2009 11:40:59 +0800 | | From | Herbert Xu <> | | Subject | Re: Revert "gro: Fix legacy path napi_complete crash", |
| |
On Wed, Mar 25, 2009 at 08:20:50PM -0700, David Miller wrote: > > There is still a difference compared to your fix Herbert. Jarek's > patch flushes GRO first before the unlink. > > I still believe that's critical, although like you I can't pinpoint > why. > > I know that GRO ought to be disabled here, but what if for some reason > it isn't? :-)
Sure, I can accept that somehow someone has enabled GRO :) But I'd still like to know why flushing the GRO afterwards would lead to a hang.
> Adam Richter has successfully tested Jarek's variant, and if Ingo's > tests show that it makes his problem go away too then I'm definitely > going with Jarek's patch.
I don't have a problem with that.
I've asked Adam to test my patch as well. So far the only failure case against it has been on Ingo's machine, where the process_backlog path isn't involved. Yes it's used for loopback, but he's seeing the hang on eth0, which with that config uses netif_receive_skb so it should continue to work even if process_backlog completely seizes up.
Cheers, -- Visit Openswan at http://www.openswan.org/ Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
|  |