Re: [PATCH] gianfar: Fix TX ring processing on SMP machines

On Mar 4, 2010, at 2:41 AM, David Miller wrote:

From: Anton Vorontsov <redacted>
Date: Wed, 3 Mar 2010 21:18:58 +0300
=20

quoted

Starting with commit a3bc1f11e9b867a4f49505 ("gianfar: Revive SKB
recycling") gianfar driver sooner or later stops transmitting any
packets on SMP machines.
=20
start_xmit() prepares new skb for transmitting, generally it does
three things:
=20
1. sets up all BDs (marks them ready to send), except the first one.
2. stores skb into tx_queue->tx_skbuff so that clean_tx_ring()
  would cleanup it later.
3. sets up the first BD, i.e. marks it ready.
=20
Here is what clean_tx_ring() does:
=20
1. reads skbs from tx_queue->tx_skbuff
2. checks if the *last* BD is ready. If it's still ready [to send]
  then it it isn't transmitted, so clean_tx_ring() returns.
  Otherwise it actually cleanups BDs. All is OK.
=20
Now, if there is just one BD, code flow:
=20
- start_xmit(): stores skb into tx_skbuff. Note that the first BD
 (which is also the last one) isn't marked as ready, yet.
- clean_tx_ring(): sees that skb is not null, *and* its lstatus
 says that it is NOT ready (like if BD was sent), so it cleans
 it up (bad!)
- start_xmit(): marks BD as ready [to send], but it's too late.
=20
We can fix this simply by reordering lstatus/tx_skbuff writes.
=20
Reported-by: Martyn Welch <redacted>
Bisected-by: Paul Gortmaker [off-list ref]
Signed-off-by: Anton Vorontsov <redacted>
Tested-by: Paul Gortmaker <redacted>
Tested-by: Martyn Welch <redacted>

=20
Applied.

Anton,

Once this makes it into Linus's tree can you make sure we get it added =
to -stable.

- k=