d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix some stability problems:

Browse Source 
- Normally, the driver allocates an mbuf cluster for each receive
  descriptor. This is because we have to be prepared to accomodate up to
  1500 bytes (a cluster buffer can hold up to 2K). However, using up a
  whole cluster buffer for a tiny packet is a bit of a waste. Also,
  it seems to me that sometimes mbufs will linger in the kernel for
  a while after being passed out of the driver, which means we might
  drain the mbuf cluster pool. The cluster pool is smaller than the
  mbuf pool in general, so we do the following: if the packet is less
  that MINCLSIZE bytes, then we copy it into a small mbuf chain and
  leave the mbuf cluster in place for another go-round. This saves
  mbuf clusters in some cases while still allowing them to be used
  for heavy traffic exchanges with lots of full-sized frames.

- The transmit descriptor has a bit in the control word which allows
  the driver to request that a 'TX OK' interrupt be generated when
  a frame has been completed. Sometimes, a frame can be fragmented
  across several descriptors. The manual for the real DEC 21140A says
  that if this happens, the 'TX interrupt request' bit is only valid
  in the descriptor of the last fragment. With the ASIX chip, it seems
  the 'TX interrupt request' bit is only valid in the descriptor of
  the _first_ fragment. Actually, the manual contains conflicting
  information, but I think it's supposed to be the first fragment.
  To play it safe, set the bit in both the first and last fragment to
  be sure that we get a TX OK interrupt. Without this fix, the driver
  can sometimes be late in releasing mbufs from the transmit queue
  after transmission.
This commit is contained in:
wpaul 1999-01-16 06:19:38 +00:00
parent 56728d4e23
commit 9617128953
1 changed files with 22 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
sys/pci/if_ax.c
View File

@ -29,7 +29,7 @@
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
* THE POSSIBILITY OF SUCH DAMAGE.
*
* $Id: if_ax.c,v 1.6 1999/01/08 19:40:59 wpaul Exp $
* $Id: if_ax.c,v 1.7 1999/01/16 01:23:56 wpaul Exp $
*/
/*
@ -87,7 +87,7 @@
#ifndef lint
static const char rcsid[] =
"$Id: if_ax.c,v 1.6 1999/01/08 19:40:59 wpaul Exp $";
"$Id: if_ax.c,v 1.7 1999/01/16 01:23:56 wpaul Exp $";
#endif
/*
@ -1446,6 +1446,17 @@ static void ax_rxeof(sc)
total_len -= ETHER_CRC_LEN;
if (total_len < MINCLSIZE) {
m = m_devget(mtod(cur_rx->ax_mbuf, char *),
total_len, 0, ifp, NULL);
cur_rx->ax_ptr->ax_status = AX_RXSTAT;
cur_rx->ax_ptr->ax_ctl = (MCLBYTES - 1);
if (m == NULL) {
ifp->if_ierrors++;
continue;
}
} else {
m = cur_rx->ax_mbuf;
/*
* Try to conjure up a new mbuf cluster. If that
* fails, it means we have an out of memory condition and
@ -1453,17 +1464,18 @@ static void ax_rxeof(sc)
* result in a lost packet, but there's little else we
* can do in this situation.
*/
if (ax_newbuf(sc, cur_rx) == ENOBUFS) {
ifp->if_ierrors++;
cur_rx->ax_ptr->ax_status = AX_RXSTAT;
cur_rx->ax_ptr->ax_ctl = (MCLBYTES - 1);
continue;
if (ax_newbuf(sc, cur_rx) == ENOBUFS) {
ifp->if_ierrors++;
cur_rx->ax_ptr->ax_status = AX_RXSTAT;
cur_rx->ax_ptr->ax_ctl = (MCLBYTES - 1);
continue;
}
m->m_pkthdr.rcvif = ifp;
m->m_pkthdr.len = m->m_len = total_len;
}
ifp->if_ipackets++;
eh = mtod(m, struct ether_header *);
m->m_pkthdr.rcvif = ifp;
m->m_pkthdr.len = m->m_len = total_len;
#if NBPFILTER > 0
/*
* Handle BPF listeners. Let the BPF user see the packet, but
@ -1746,6 +1758,7 @@ static int ax_encap(sc, c, m_head)
c->ax_mbuf = m_head;
c->ax_lastdesc = frag - 1;
AX_TXCTL(c) |= AX_TXCTL_LASTFRAG|AX_TXCTL_FINT;
c->ax_ptr->ax_frag[0].ax_ctl |= AX_TXCTL_FINT;
AX_TXNEXT(c) = vtophys(&c->ax_nextdesc->ax_ptr->ax_frag[0]);
return(0);
}