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

cxgbe/iw_cxgbe: Do not start the EP timer if soaccept fails.

Browse Source 
This fixes a panic that would occur when the timer tried to close a
stale socket.

Submitted by:	Krishnamraju Eraparaju @ Chelsio
MFC after:	1 week
Sponsored by:	Chelsio Communications
This commit is contained in:
np 2020-04-15 03:40:33 +00:00
parent fb9f2a1abb
commit b90518c84c
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
sys/dev/cxgbe/iw_cxgbe/cm.c
View File

@ -1013,7 +1013,6 @@ process_newconn(struct c4iw_listen_ep *master_lep, struct socket *new_so)
c4iw_get_ep(&real_lep->com);
init_timer(&new_ep->timer);
new_ep->com.state = MPA_REQ_WAIT;
START_EP_TIMER(new_ep);
setiwsockopt(new_so);
ret = soaccept(new_so, (struct sockaddr **)&remote);
@ -1023,7 +1022,6 @@ process_newconn(struct c4iw_listen_ep *master_lep, struct socket *new_so)
__func__, master_lep->com.so, new_so, ret);
if (remote != NULL)
free(remote, M_SONAME);
uninit_iwarp_socket(new_so);
soclose(new_so);
c4iw_put_ep(&new_ep->com);
c4iw_put_ep(&real_lep->com);
@ -1031,6 +1029,8 @@ process_newconn(struct c4iw_listen_ep *master_lep, struct socket *new_so)
}
free(remote, M_SONAME);
START_EP_TIMER(new_ep);
/* MPA request might have been queued up on the socket already, so we
* initialize the socket/upcall_handler under lock to prevent processing
* MPA request on another thread(via process_req()) simultaniously.