While using IPoIB over EHCA (rc6 bits), unregister_netdev hangs with
the message: "waiting for ib2 to become free. Usage count = -515276",
etc.
The problem is that the poll handler does netif_rx_complete (which
does a dev_put) followed by netif_rx_reschedule() to schedule for
more receives (which again does a dev_put). This reduces refcount to
< 0 (depending on how many times netif_rx_complete followed by
netif_rx_reschedule was called).
The following patch fixes the bug, but I don't know if there is some
specific IB issue that prevents this approach.
Signed-off-by: Krishna Kumar <[EMAIL PROTECTED]>
---
ipoib_ib.c | 11 ++++-------
1 files changed, 4 insertions(+), 7 deletions(-)
diff -ruNp org/drivers/infiniband/ulp/ipoib/ipoib_ib.c
new1/drivers/infiniband/ulp/ipoib/ipoib_ib.c
--- org/drivers/infiniband/ulp/ipoib/ipoib_ib.c 2007-09-18 15:50:09.000000000
+0530
+++ new1/drivers/infiniband/ulp/ipoib/ipoib_ib.c 2007-09-18
16:14:20.000000000 +0530
@@ -291,7 +291,6 @@ int ipoib_poll(struct napi_struct *napi,
done = 0;
-poll_more:
while (done < budget) {
int max = (budget - done);
@@ -316,12 +315,10 @@ poll_more:
}
if (done < budget) {
- netif_rx_complete(dev, napi);
- if (unlikely(ib_req_notify_cq(priv->cq,
- IB_CQ_NEXT_COMP |
- IB_CQ_REPORT_MISSED_EVENTS)) &&
- netif_rx_reschedule(napi))
- goto poll_more;
+ if (likely(!ib_req_notify_cq(priv->cq,
+ IB_CQ_NEXT_COMP |
+ IB_CQ_REPORT_MISSED_EVENTS)))
+ netif_rx_complete(dev, napi);
}
return done;
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
