On Tue, Apr 07, 2020 at 11:54:01AM +0200, Claudio Jeker wrote:
> Unsure about this one here. I would prefer if the panic remained for now
> (mainly because of the XXXSMP NET_UNLOCK() dance just above). I wonder if the
> order of this could not be modified so that the NET_LOCK is released after
> the RBT_REMOVE.
Well, let's clenup these dances before. Diff below just breaks giant
NET_LOCK() in pppxioctl() to many small.
Index: net/if_pppx.c
===================================================================
RCS file: /cvs/src/sys/net/if_pppx.c,v
retrieving revision 1.81
diff -u -p -r1.81 if_pppx.c
--- net/if_pppx.c 7 Apr 2020 07:11:22 -0000 1.81
+++ net/if_pppx.c 7 Apr 2020 10:46:34 -0000
@@ -433,7 +433,6 @@ pppxioctl(dev_t dev, u_long cmd, caddr_t
struct pppx_dev *pxd = pppx_dev2pxd(dev);
int error = 0;
- NET_LOCK();
switch (cmd) {
case PIPEXSMODE:
/*
@@ -447,46 +446,59 @@ pppxioctl(dev_t dev, u_long cmd, caddr_t
break;
case PIPEXASESSION:
+ NET_LOCK();
error = pppx_add_session(pxd,
(struct pipex_session_req *)addr);
+ NET_UNLOCK();
break;
case PIPEXDSESSION:
+ NET_LOCK();
error = pppx_del_session(pxd,
(struct pipex_session_close_req *)addr);
+ NET_UNLOCK();
break;
case PIPEXCSESSION:
+ NET_LOCK();
error = pppx_config_session(pxd,
(struct pipex_session_config_req *)addr);
+ NET_UNLOCK();
break;
case PIPEXGSTAT:
+ NET_LOCK();
error = pppx_get_stat(pxd,
(struct pipex_session_stat_req *)addr);
+ NET_UNLOCK();
break;
case PIPEXGCLOSED:
+ NET_LOCK();
error = pppx_get_closed(pxd,
(struct pipex_session_list_req *)addr);
+ NET_UNLOCK();
break;
case PIPEXSIFDESCR:
+ NET_LOCK();
error = pppx_set_session_descr(pxd,
(struct pipex_session_descr_req *)addr);
+ NET_UNLOCK();
break;
case FIONBIO:
break;
case FIONREAD:
+ NET_LOCK();
*(int *)addr = mq_hdatalen(&pxd->pxd_svcq);
+ NET_UNLOCK();
break;
default:
error = ENOTTY;
break;
}
- NET_UNLOCK();
return (error);
}
