linux/net/core, branch v2.6.13

[NETPOLL]: remove unused variable

2005-08-12T02:28:05Z

Remove unused variable Signed-off-by: Matt Mackall Signed-off-by: David S. Miller

[NETPOLL]: fix initialization/NAPI race

2005-08-12T02:27:43Z

This fixes a race during initialization with the NAPI softirq processing by using an RCU approach. This race was discovered when refill_skbs() was added to the setup code. Signed-off-by: Matt Mackall Signed-off-by: David S. Miller

[NETPOLL]: pre-fill skb pool

2005-08-12T02:26:42Z

we could do one thing (see the patch below): i think it would be useful to fill up the netlogging skb queue straight at initialization time. Especially if netpoll is used for dumping alone, the system might not be in a situation to fill up the queue at the point of crash, so better be a bit more prepared and keep the pipeline filled. [ I've modified this to be called earlier - mpm ] Signed-off-by: Ingo Molnar Signed-off-by: Matt Mackall Signed-off-by: David S. Miller

[NETPOLL]: add retry timeout

2005-08-12T02:25:54Z

Add limited retry logic to netpoll_send_skb Each time we attempt to send, decrement our per-device retry counter. On every successful send, we reset the counter. We delay 50us between attempts with up to 20000 retries for a total of 1 second. After we've exhausted our retries, subsequent failed attempts will try only once until reset by success. Signed-off-by: Matt Mackall Signed-off-by: David S. Miller

[NETPOLL]: netpoll_send_skb simplify

2005-08-12T02:25:11Z

Minor netpoll_send_skb restructuring Restructure to avoid confusing goto and move some bits out of the retry loop. Signed-off-by: Matt Mackall Signed-off-by: David S. Miller

[NETPOLL]: deadlock bugfix

2005-08-12T02:23:50Z

This fixes an obvious deadlock in the netpoll code. netpoll_rx takes the npinfo->rx_lock. netpoll_rx is also the only caller of arp_reply (through __netpoll_rx). As such, it is not necessary to take this lock. Signed-off-by: Jeff Moyer Signed-off-by: Matt Mackall Signed-off-by: David S. Miller

[NETPOLL]: rx_flags bugfix

2005-08-12T02:23:04Z

Initialize npinfo->rx_flags. The way it stands now, this will have random garbage, and so will incur a locking penalty even when an rx_hook isn't registered and we are not active in the netpoll polling code. Signed-off-by: Jeff Moyer Signed-off-by: Matt Mackall Signed-off-by: David S. Miller

[NET] Fix too aggressive backoff in dst garbage collection

2005-07-31T00:47:25Z

The bug is evident when it is seen once. dst gc timer was backed off, when gc queue is not empty. But this means that timer quickly backs off, if at least one destination remains in use. Normally, the bug is invisible, because adding new dst entry to queue cancels the backoff. But it shots deadly with destination cache overflow when new destinations are not released for long time f.e. after an interface goes down. The fix is to cancel backoff when something was released. Signed-off-by: Denis Lunev Signed-off-by: Alexey Kuznetsov Signed-off-by: David S. Miller

[NET]: Fix busy waiting in dev_close().

2005-07-28T19:12:58Z

If the current task has signal_pending(), the loop we have to wait for the __LINK_STATE_RX_SCHED bit to clear becomes a pure busy-loop. Fixed by using msleep() instead of the hand-crafted version. Noticed by Andrew Morton. Signed-off-by: David S. Miller

Merge master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6

2005-07-27T23:37:59Z