[linux-dvb] Re: [PATCH] dvb-ttpci+budgetpatch integrated late patch 10

[emard@softhome.net]

> OK, I comitted this part.

Thanx!

> > -	net = alloc_netdev(sizeof(struct dvb_net_priv), "dvb", dvb_net_setup);
> > +	/* "dvb0_0" string must in advance be of the right length */
> > +	net = alloc_netdev(sizeof(struct dvb_net_priv), "dvb0_0", dvb_net_setup);
> 
> I dropped this hunk. struct net_device has name[IFNAMSIZ], and that
> field isn't used before register_netdev() is called. I don't see
> what problem your change could fix.

Maybe I had other disturbances to conclude that. There still might
be rare initialization problem, as in some long runs I caught saa7146 
interrupt status 0x01100000 -> bits 20 and 24 set, meaning RPS1 error 
while RPS1 in UPLOAD (and the PC just halts without any further oops).

My RPS1 doesn't UPLOAD but I obviously shouldn't leave it running after 
rmmod dvb-ttpci,
I guess that problem comes from early initialization workarounds for
bugs in saa7146, where they are overwriting a new RPS1 containing UPLOAD
statements without previously stoping this RPS1. 

This fix stopps my rps1 on rmmod and with this I can leave insmod/rmmod
running.

Emard
diff -pur dvb-kernel.orig/linux/drivers/media/dvb/ttpci/av7110.c dvb-kernel/linux/drivers/media/dvb/ttpci/av7110.c
--- dvb-kernel.orig/linux/drivers/media/dvb/ttpci/av7110.c	2005-01-09 09:41:46.000000000 +0100
+++ dvb-kernel/linux/drivers/media/dvb/ttpci/av7110.c	2005-01-09 09:54:41.000000000 +0100
@@ -2491,6 +2491,10 @@ static int av7110_detach(struct saa7146_
 		return 0;
 
 	if (budgetpatch) {
+		/* Disable RPS1 */
+		saa7146_write(saa, MC1, MASK_29);
+		/* VSYNC LOW (inactive) */
+		saa7146_setgpio(saa, 3, SAA7146_GPIO_OUTLO);
 		saa7146_write(saa, MC1, MASK_20);	/* DMA3 off */
 		SAA7146_IER_DISABLE(saa, MASK_10);
 		SAA7146_ISR_CLEAR(saa, MASK_10);
Only in dvb-kernel/linux/drivers/media/dvb/ttpci: av7110.c~