Re: [Qemu-devel] [PATCHv2] e1000: multi-buffer packet support

["Michael S. Tsirkin" <mst@redhat.com>]

On Tue, Feb 08, 2011 at 03:07:58PM +0100, Kevin Wolf wrote:
> Am 03.02.2011 17:49, schrieb Michael S. Tsirkin:
> > e1000 supports multi-buffer packets larger than rxbuf_size.
> > 
> > This fixes the following (on linux):
> > - in guest: ifconfig eth1 mtu 16110
> > - in host: ifconfig tap0 mtu 16110
> >            ping -s 16082 <guest-ip>
> > 
> > Red Hat bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=602205
> > 
> > Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> > ---
> > 
> > Changes from v1:
> > 	removed dead code
> > 
> >  hw/e1000.c |   29 +++++++++++++++++++----------
> >  1 files changed, 19 insertions(+), 10 deletions(-)
> > 
> > diff --git a/hw/e1000.c b/hw/e1000.c
> > index af101bd..3427ff3 100644
> > --- a/hw/e1000.c
> > +++ b/hw/e1000.c
> > @@ -642,6 +642,8 @@ e1000_receive(VLANClientState *nc, const uint8_t *buf, size_t size)
> >      uint16_t vlan_special = 0;
> >      uint8_t vlan_status = 0, vlan_offset = 0;
> >      uint8_t min_buf[MIN_BUF_SIZE];
> > +    size_t desc_offset;
> > +    size_t desc_size;
> >  
> >      if (!(s->mac_reg[RCTL] & E1000_RCTL_EN))
> >          return -1;
> > @@ -654,12 +656,6 @@ e1000_receive(VLANClientState *nc, const uint8_t *buf, size_t size)
> >          size = sizeof(min_buf);
> >      }
> >  
> > -    if (size > s->rxbuf_size) {
> > -        DBGOUT(RX, "packet too large for buffers (%lu > %d)\n",
> > -               (unsigned long)size, s->rxbuf_size);
> > -        return -1;
> > -    }
> > -
> >      if (!receive_filter(s, buf, size))
> >          return size;
> >  
> > @@ -672,8 +668,15 @@ e1000_receive(VLANClientState *nc, const uint8_t *buf, size_t size)
> >      }
> >  
> >      rdh_start = s->mac_reg[RDH];
> > +    desc_offset = 0;
> >      do {
> > +        desc_size = size - desc_offset;
> > +        if (desc_size > s->rxbuf_size) {
> > +            desc_size = s->rxbuf_size;
> > +        }
> >          if (s->mac_reg[RDH] == s->mac_reg[RDT] && s->check_rxov) {
> > +            /* Discard all data written so far */
> > +            s->mac_reg[RDH] = rdh_start;
> >              set_ics(s, 0, E1000_ICS_RXO);
> >              return -1;
> >          }
> > @@ -684,9 +687,15 @@ e1000_receive(VLANClientState *nc, const uint8_t *buf, size_t size)
> >          desc.status |= (vlan_status | E1000_RXD_STAT_DD);
> >          if (desc.buffer_addr) {
> >              cpu_physical_memory_write(le64_to_cpu(desc.buffer_addr),
> > -                                      (void *)(buf + vlan_offset), size);
> > -            desc.length = cpu_to_le16(size + fcs_len(s));
> > -            desc.status |= E1000_RXD_STAT_EOP|E1000_RXD_STAT_IXSM;
> > +                                      (void *)(buf + desc_offset + vlan_offset),
> > +                                      desc_size);
> > +            desc_offset += desc_size;
> > +            if (desc_offset >= size) {
> > +                desc.length = cpu_to_le16(desc_size + fcs_len(s));
> 
> I think this is not quite right: What happens if desc_size + fcs_len(s)
> > s->rxbuf_size? IIUC, we would seemingly overflow the guests buffer (in
> reality we leave the extra bytes untouched, but we may confuse the guest).
> 
> Kevin

Hmm, good point.