Aren
/
NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Send data for as long as there is new data available. Otherwise 

there was a danger of smb_iod_recvall() blocking, hence releasing
the kernel lock, new data creeping into the queue, and a wakeup
being missed (well, there's still a race, but since it's theoretical
enough for me to never have encountered it, I'll rather solve it
by periodic wakeups).
This commit is contained in:
pooka 2009-09-04 16:12:45 +00:00
parent a3879acf53
commit 56df608091
1 changed files with 19 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
sys/netsmb/smb_iod.c
View File

@ -1,4 +1,4 @@
/* $NetBSD: smb_iod.c,v 1.32 2009/07/06 11:46:49 njoly Exp $ */
/* $NetBSD: smb_iod.c,v 1.33 2009/09/04 16:12:45 pooka Exp $ */
/*
* Copyright (c) 2000-2001 Boris Popov
@ -35,7 +35,7 @@
*/
#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: smb_iod.c,v 1.32 2009/07/06 11:46:49 njoly Exp $");
__KERNEL_RCSID(0, "$NetBSD: smb_iod.c,v 1.33 2009/09/04 16:12:45 pooka Exp $");
#include <sys/param.h>
#include <sys/systm.h>
@ -66,7 +66,7 @@ static MALLOC_DEFINE(M_SMBIOD, "SMBIOD", "SMB network io daemon");
static int smb_iod_next;
static void smb_iod_sendall(struct smbiod *iod);
static bool smb_iod_sendall(struct smbiod *iod);
static int smb_iod_disconnect(struct smbiod *iod);
static void smb_iod_thread(void *);
@ -555,11 +555,12 @@ smb_iod_waitrq(struct smb_rq *rqp)
}
static void
static bool
smb_iod_sendall(struct smbiod *iod)
{
struct smb_rq *rqp;
int herror;
bool sentany = false;
herror = 0;
/*
@ -580,11 +581,14 @@ smb_iod_sendall(struct smbiod *iod)
if (__predict_false(herror != 0))
break;
sentany = true;
}
}
SMB_IOD_RQUNLOCK(iod);
if (herror == ENOTCONN)
smb_iod_dead(iod);
return sentany;
}
/*
@ -647,8 +651,19 @@ smb_iod_main(struct smbiod *iod)
}
}
#endif
/*
* Do a send/receive cycle once and then as many times
* afterwards as we can send out new data. This is to make
* sure we got all data sent which might have ended up in the
* queue during the receive phase (which might block releasing
* the kernel lock).
*/
smb_iod_sendall(iod);
smb_iod_recvall(iod);
while (smb_iod_sendall(iod)) {
smb_iod_recvall(iod);
}
}
void