similar to: [PATCH v2] virtio_net: fix PAGE_SIZE > 64k

On Tue, Jan 24, 2017 at 08:07:40PM -0800, Alexei Starovoitov wrote: > On Tue, Jan 24, 2017 at 7:48 PM, John Fastabend > <john.fastabend at gmail.com> wrote: > > > > It is a concern on my side. I want XDP and Linux stack to work > > reasonably well together. > > btw the micro benchmarks showed that page per packet approach > that xdp took in mlx4 should be

On Tue, Jan 24, 2017 at 03:53:31PM -0500, David Miller wrote: > From: "Michael S. Tsirkin" <mst at redhat.com> > Date: Tue, 24 Jan 2017 22:45:37 +0200 > > > On Tue, Jan 24, 2017 at 03:09:59PM -0500, David Miller wrote: > >> From: "Michael S. Tsirkin" <mst at redhat.com> > >> Date: Tue, 24 Jan 2017 21:53:13 +0200 > >> >

On Tue, Jan 24, 2017 at 03:53:31PM -0500, David Miller wrote: > From: "Michael S. Tsirkin" <mst at redhat.com> > Date: Tue, 24 Jan 2017 22:45:37 +0200 > > > On Tue, Jan 24, 2017 at 03:09:59PM -0500, David Miller wrote: > >> From: "Michael S. Tsirkin" <mst at redhat.com> > >> Date: Tue, 24 Jan 2017 21:53:13 +0200 > >> >

On Tue, Jan 24, 2017 at 04:10:46PM -0500, David Miller wrote: > This works in the regimen that XDP packets always live in exactly one > page. That will be needed to mmap the RX ring into userspace, and it > helps make adjust_header trivial as well. I think the point was to avoid resets across xdp attach/detach. If we are doing resets now, we could do whatever buffering we want. We

On Tue, Jan 24, 2017 at 04:10:46PM -0500, David Miller wrote: > This works in the regimen that XDP packets always live in exactly one > page. That will be needed to mmap the RX ring into userspace, and it > helps make adjust_header trivial as well. I think the point was to avoid resets across xdp attach/detach. If we are doing resets now, we could do whatever buffering we want. We

On Tue, Jan 24, 2017 at 03:09:59PM -0500, David Miller wrote: > From: "Michael S. Tsirkin" <mst at redhat.com> > Date: Tue, 24 Jan 2017 21:53:13 +0200 > > > I didn't realise. Why can't we? I thought that adjust_header is an > > optional feature that userspace can test for, so no rush. > > No, we want the base set of XDP features to be present in

On Tue, Jan 24, 2017 at 03:09:59PM -0500, David Miller wrote: > From: "Michael S. Tsirkin" <mst at redhat.com> > Date: Tue, 24 Jan 2017 21:53:13 +0200 > > > I didn't realise. Why can't we? I thought that adjust_header is an > > optional feature that userspace can test for, so no rush. > > No, we want the base set of XDP features to be present in

On Tue, Jan 24, 2017 at 02:42:27PM -0500, David Miller wrote: > From: "Michael S. Tsirkin" <mst at redhat.com> > Date: Mon, 23 Jan 2017 21:37:52 +0200 > > > I don't have any guests with PAGE_SIZE > 64k but the > > code seems to be clearly broken in that case > > as PAGE_SIZE / MERGEABLE_BUFFER_ALIGN will need > > more than 8 bit and so the

On Tue, Jan 24, 2017 at 02:42:27PM -0500, David Miller wrote: > From: "Michael S. Tsirkin" <mst at redhat.com> > Date: Mon, 23 Jan 2017 21:37:52 +0200 > > > I don't have any guests with PAGE_SIZE > 64k but the > > code seems to be clearly broken in that case > > as PAGE_SIZE / MERGEABLE_BUFFER_ALIGN will need > > more than 8 bit and so the

Merry Xmas and a Happy New year to all: This series tries to fixes several issues for virtio-net XDP which could be categorized into several parts: - fix several issues during XDP linearizing - allow csumed packet to work for XDP_PASS - make EWMA rxbuf size estimation works for XDP - forbid XDP when GUEST_UFO is support - remove big packet XDP support - add XDP support or small buffer Please

Merry Xmas and a Happy New year to all: This series tries to fixes several issues for virtio-net XDP which could be categorized into several parts: - fix several issues during XDP linearizing - allow csumed packet to work for XDP_PASS - make EWMA rxbuf size estimation works for XDP - forbid XDP when GUEST_UFO is support - remove big packet XDP support - add XDP support or small buffer Please

CC: John Fastabend <john.fastabend at gmail.com> Signed-off-by: Jason Wang <jasowang at redhat.com> --- drivers/net/virtio_net.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 511f833..a0ef4b0 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -373,7 +373,6 @@ static struct sk_buff

CC: John Fastabend <john.fastabend at gmail.com> Signed-off-by: Jason Wang <jasowang at redhat.com> --- drivers/net/virtio_net.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 511f833..a0ef4b0 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -373,7 +373,6 @@ static struct sk_buff

Add support for XDP adjust head by allocating a 256B header region that XDP programs can grow into. This is only enabled when a XDP program is loaded. In order to ensure that we do not have to unwind queue headroom push queue setup below bpf_prog_add. It reads better to do a prog ref unwind vs another queue setup call. TBD merge with Jason Wang's fixes, do a bit more testing, note

Add support for XDP adjust head by allocating a 256B header region that XDP programs can grow into. This is only enabled when a XDP program is loaded. In order to ensure that we do not have to unwind queue headroom push queue setup below bpf_prog_add. It reads better to do a prog ref unwind vs another queue setup call. TBD merge with Jason Wang's fixes, do a bit more testing, note

I don't have any guests with PAGE_SIZE > 64k but the code seems to be clearly broken in that case as PAGE_SIZE / MERGEABLE_BUFFER_ALIGN will need more than 8 bit and so the code in mergeable_ctx_to_buf_address does not give us the actual true size. Cc: John Fastabend <john.fastabend at gmail.com> Signed-off-by: Michael S. Tsirkin <mst at redhat.com> --- changes from v1: fix

I don't have any guests with PAGE_SIZE > 64k but the code seems to be clearly broken in that case as PAGE_SIZE / MERGEABLE_BUFFER_ALIGN will need more than 8 bit and so the code in mergeable_ctx_to_buf_address does not give us the actual true size. Cc: John Fastabend <john.fastabend at gmail.com> Signed-off-by: Michael S. Tsirkin <mst at redhat.com> --- changes from v1: fix

From: "Michael S. Tsirkin" <mst at redhat.com> Date: Tue, 24 Jan 2017 23:07:51 +0200 > On Tue, Jan 24, 2017 at 03:53:31PM -0500, David Miller wrote: >> From: "Michael S. Tsirkin" <mst at redhat.com> >> Date: Tue, 24 Jan 2017 22:45:37 +0200 >> >> > On Tue, Jan 24, 2017 at 03:09:59PM -0500, David Miller wrote: >> >> From:

On 16-12-23 06:37 AM, Jason Wang wrote: > When XDP_PASS were determined for linearized packets, we try to get > new buffers in the virtqueue and build skbs from them. This is wrong, > we should create skbs based on existed buffers instead. Fixing them by > creating skb based on xdp_page. > > With this patch "ping 192.168.100.4 -s 3900 -M do" works for XDP_PASS. >

On 16-12-23 06:37 AM, Jason Wang wrote: > When XDP_PASS were determined for linearized packets, we try to get > new buffers in the virtqueue and build skbs from them. This is wrong, > we should create skbs based on existed buffers instead. Fixing them by > creating skb based on xdp_page. > > With this patch "ping 192.168.100.4 -s 3900 -M do" works for XDP_PASS. >