On Thu, Oct 17, 2024 at 03:42:59PM +0200, Paolo Abeni wrote:
>
>
> On 10/14/24 05:12, Xuan Zhuo wrote:
> > When the frag just got a page, then may lead to regression on VM.
> > Specially if the sysctl net.core.high_order_alloc_disable value is 1,
> > then the frag always get a page when do refill.
> >
> > Which could see reliable crashes or scp failure (scp a file 100M in size
> > to VM):
> >
> > The issue is that the virtnet_rq_dma takes up 16 bytes at the beginning
> > of a new frag. When the frag size is larger than PAGE_SIZE,
> > everything is fine. However, if the frag is only one page and the
> > total size of the buffer and virtnet_rq_dma is larger than one page, an
> > overflow may occur.
> >
> > Here, when the frag size is not enough, we reduce the buffer len to fix
> > this problem.
> >
> > Fixes: f9dac92ba908 ("virtio_ring: enable premapped mode whatever
> > use_dma_api")
> > Reported-by: "Si-Wei Liu" <si-wei....@oracle.com>
> > Signed-off-by: Xuan Zhuo <xuanz...@linux.alibaba.com>
>
> This looks like a fix that should target the net tree, but the following
> patches looks like net-next material. Any special reason to bundle them
> together?
>
> Also, please explicitly include the the target tree in the subj on next
> submissions, thanks!
>
> Paolo
The issue is that net-next is not rebased on net, so if patches 2-5 land
in next and 1 only in net, next will be broken. I think the simplest fix
is just to have 1 on net-next, backport it to net too, and git will
figure it out. Right?
--
MST
