tty/n_gsm: avoid fifo overflow in gsm_dlci_data_output

n_gsm use a simple approach: every writing to fifo correspond exactly one
reading from fifo. There are no problem in this approach until we read
less bytes then we write. As result fifo may owerflow. This leads to packet
loss and very slow responce.

For example, this happens with ping packets (about 96 byte each) and default
gsm->mtu = 64. As result we get 50 sec ping timeout and 20% packet loss.

Fix the problem by reading and sending all data from the fifo

Signed-off-by: Mikhail Kshevetskiy <mikhail.kshevetskiy@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This commit is contained in:
Mikhail Kshevetskiy 2011-09-23 19:22:56 +04:00 committed by Greg Kroah-Hartman
parent f37ac5a144
commit 268e526b93

View File

@ -809,12 +809,14 @@ static int gsm_dlci_data_output(struct gsm_mux *gsm, struct gsm_dlci *dlci)
{
struct gsm_msg *msg;
u8 *dp;
int len, size;
int len, total_size, size;
int h = dlci->adaption - 1;
total_size = 0;
while(1) {
len = kfifo_len(dlci->fifo);
if (len == 0)
return 0;
return total_size;
/* MTU/MRU count only the data bits */
if (len > gsm->mtu)
@ -838,8 +840,10 @@ static int gsm_dlci_data_output(struct gsm_mux *gsm, struct gsm_dlci *dlci)
}
WARN_ON(kfifo_out_locked(dlci->fifo, dp , len, &dlci->lock) != len);
__gsm_data_queue(dlci, msg);
total_size += size;
}
/* Bytes of data we used up */
return size;
return total_size;
}
/**