On Sun, Jul 06, 2014 at 06:48:49PM +0200, Guillaume Morin wrote: > On 02 Jul 22:54, Florian Westphal wrote: > > > > Short version: Since 3.4.92, when forwarding, > > GRO aggregated packets are no longer forwarded because > > kernel believes segments exceed path mtu. > > > > This is because kernels pre 3.11 lack > > > > commit 21d1196a35f5686c4323e42a62fdb4b23b0ab4a3 > > Author: Eric Dumazet <edumazet@xxxxxxxxxx> > > ipv4: set transport header earlier > > > > [ this commit was already picked up by 3.10.y tree ] > > > > The backport of 895162b1101b3ea5db08ca6822ae9672717efec0 > > exposed this problem. > > > > Original report quoted below. > > > > ----- Forwarded message from Florian Westphal <fw@xxxxxxxxx> ----- > > > > Date: Fri, 27 Jun 2014 11:05:25 +0200 > > From: Florian Westphal <fw@xxxxxxxxx> > > To: jungwon park <jwpark2@xxxxxxxxxxxxxxx> > > Cc: netdev@xxxxxxxxxxxxxxx > > Subject: Re: GRO issue with kernel 3.4.94 (icmp fragmentation needed) > > > > jungwon park <jwpark2@xxxxxxxxxxxxxxx> wrote: > > > When using the linux router is turned on GRO, router send the 'fragmentation > > > needed' packets to the sender. > > > > Indeed 8-( > > > > > When I turned off GRO, the router operate normally, and there is no problem. > > > and with 3.4.91 kernel, the router has no problem. > > > > > > I doubt 'ipv4: ip_forward: fix inverted local_df test' patch. > > > (http://patchwork.ozlabs.org/patch/345509/) > > > When I revert this patch, the router has no problem. > > > > Can you please cherry-pick following patch on top of vanilla 3.4.92? > > https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=21d1196a35f5686c4323e42a62fdb4b23b0ab4a3 > > > > commit 21d1196a35f5686c4323e42a62fdb4b23b0ab4a3 > > Author: Eric Dumazet <edumazet@xxxxxxxxxx> > > ipv4: set transport header earlier > > > > I think that should fix this bug, it should apply cleanly on top of > > 3.4.y tree. > > > > [ patch is in 3.11, also backported to 3.10.y tree ] > > > > The problem is that, when dealing with GRO packets, we try to determine > > the size of the individual packets. To do this, we rely on the > > transport header. > > > > Unfortunately the transport header is not set for the forward path in 3.4, > > so we look at the network header instead. > > > > ----- End forwarded message ----- > > Greg, > > Any reason this is not included in 3.4.97 or you're planning to queue it > for .98? Actually, looking at this, it's a network patch, which needs to go through the networking maintainer to get merged into a stable kernel tree. Please go ask about this on the netdev mailing list and it should come in to me that way. thanks, greg k-h -- To unsubscribe from this list: send the line "unsubscribe stable" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html