[PATCH] Add NetMos 9835 to parport_serial

[PATCH] Add NetMos 9835 to parport_serial

[Zwane Mwaikambo]

Tested simultaneous serial i/o and parallel port. The base_baud of 115200 
i just chose as a safe default, i didn't try pushing it any further. 
During testing of parport interrupt sharing, i noticed an oddity whereupon 
parport was allowed to register the same irq even though it didn't specify 
SA_SHIRQ and serial did a request_irq before parport. But then again it 
was late saturday and the booze was plenty...

Regards,
	Zwane Mwaikambo

--- linux-2.4.19-pre7-ac3/drivers/parport/parport_serial.c.orig	Sun May  5 14:24:36 2002
+++ linux-2.4.19-pre7-ac3/drivers/parport/parport_serial.c	Sun May  5 14:38:57 2002
@@ -41,6 +41,7 @@
 	avlab_2s1p,
 	avlab_2s1p_650,
 	avlab_2s1p_850,
+	netmos_9835
 };
 
 
@@ -74,6 +75,7 @@
 	/* avlab_2s1p     */		{ 1, { { 2, 3}, } },
 	/* avlab_2s1p_650 */		{ 1, { { 2, 3}, } },
 	/* avlab_2s1p_850 */		{ 1, { { 2, 3}, } },
+	/* netmos_9835 */		{ 1, { { 2, 3}, } },
 };
 
 static struct pci_device_id parport_serial_pci_tbl[] __devinitdata = {
@@ -92,6 +94,8 @@
 	{ 0x14db, 0x2160, PCI_ANY_ID, PCI_ANY_ID, 0, 0, avlab_2s1p},
 	{ 0x14db, 0x2161, PCI_ANY_ID, PCI_ANY_ID, 0, 0, avlab_2s1p_650},
 	{ 0x14db, 0x2162, PCI_ANY_ID, PCI_ANY_ID, 0, 0, avlab_2s1p_850},
+	{ PCI_VENDOR_ID_NETMOS, PCI_DEVICE_ID_NETMOS_9835,
+	  PCI_ANY_ID, PCI_ANY_ID, 0, 0, netmos_9835 },
 	{ 0, } /* terminate list */
 };
 MODULE_DEVICE_TABLE(pci,parport_serial_pci_tbl);
@@ -129,6 +133,7 @@
 /* avlab_2s1p (n/t) */	{ SPCI_FL_BASE0 | SPCI_FL_BASE_TABLE, 2, 115200 },
 /* avlab_2s1p_650 (nt)*/{ SPCI_FL_BASE0 | SPCI_FL_BASE_TABLE, 2, 115200 },
 /* avlab_2s1p_850 (nt)*/{ SPCI_FL_BASE0 | SPCI_FL_BASE_TABLE, 2, 115200 },
+/* netmos_9835 */	{ SPCI_FL_BASE0 | SPCI_FL_BASE_TABLE, 2, 115200 },
 };
 
 struct parport_serial_private {
@@ -266,6 +271,10 @@
                                         "hi" as an offset (see SYBA
                                         def.) */
 		/* TODO: test if sharing interrupts works */
+
+		/* not with the netmos card i tested with, due to the
+		 * parport ISR methinks -Zwane
+		*/
 		printk (KERN_DEBUG "PCI parallel port detected: %04x:%04x, "
 			"I/O at %#lx(%#lx)\n",
 			parport_serial_pci_tbl[i].vendor,

-- 
http://function.linuxpower.ca
		

Re: [PATCH] Add NetMos 9835 to parport_serial

[Tim Waugh]

[-- Attachment #1: Type: text/plain, Size: 592 bytes --]

On Mon, May 06, 2002 at 08:17:52AM +0200, Zwane Mwaikambo wrote:

> +	/* netmos_9835 */		{ 1, { { 2, 3}, } },

Are you sure these values are correct?  They are different to the ones
in ftp://people.redhat.com/twaugh/patches/linux25/linux-netmos.patch.

That patch seems to work for some people but not for others, and I
have no idea why; until that's sorted out I'm quite reluctant to
submit any NetMos support to the mainstream kernel.  The failure mode
is a complete lock-up. :-(

Perhaps you could chase the oddity you found and see if you can figure
out what's going on?

Thanks,
Tim.
*/

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]

Re: [PATCH] Add NetMos 9835 to parport_serial

[Zwane Mwaikambo]

On Mon, 6 May 2002, Tim Waugh wrote:

> On Mon, May 06, 2002 at 08:17:52AM +0200, Zwane Mwaikambo wrote:
> 
> > +	/* netmos_9835 */		{ 1, { { 2, 3}, } },
> 
> Are you sure these values are correct?  They are different to the ones
> in ftp://people.redhat.com/twaugh/patches/linux25/linux-netmos.patch.
> 
> That patch seems to work for some people but not for others, and I
> have no idea why; until that's sorted out I'm quite reluctant to
> submit any NetMos support to the mainstream kernel.  The failure mode
> is a complete lock-up. :-(

All the patches i've seen thus far were for some other chip (forgot the 
ID), but for that 9835 i needed it desperately so i tested it quite a lot. 

+	/* netmos_9835 (not tested) */	{ 1, { { 2, -1 }, } },

I'm not sure about the others, but i doubt that one would work. Where 
there conflicting success/failure reports for the same devices?

> Perhaps you could chase the oddity you found and see if you can figure
> out what's going on?

I'll definately do that this evening.

Regards,
	Zwane

-- 
http://function.linuxpower.ca
		

Re: [PATCH] Add NetMos 9835 to parport_serial

[Zwane Mwaikambo]

On Mon, 6 May 2002, Tim Waugh wrote:

> Well, if { 2, 3 } works then { 2, -1 } will surely work, although
> without ECP support.  I didn't realise that NetMos cards had ECP
> support at the time I wrote the above code.

My ignorance rears itself ;) I got the numbers from messing with 
pci_resource_start, and got ECP etc detected, although with 
PARPORT_IRQ_NONE, we might as well do -1 anyway since parport_pc didn't 
detect ECP with no irq.

	Zwane

-- 
http://function.linuxpower.ca
		

Re: [PATCH] Add NetMos 9835 to parport_serial

[Tim Waugh]

[-- Attachment #1: Type: text/plain, Size: 657 bytes --]

On Mon, May 06, 2002 at 02:14:25PM +0200, Zwane Mwaikambo wrote:

> All the patches i've seen thus far were for some other chip (forgot the 
> ID), but for that 9835 i needed it desperately so i tested it quite a lot. 
> 
> +	/* netmos_9835 (not tested) */	{ 1, { { 2, -1 }, } },
> 
> I'm not sure about the others, but i doubt that one would work.

Well, if { 2, 3 } works then { 2, -1 } will surely work, although
without ECP support.  I didn't realise that NetMos cards had ECP
support at the time I wrote the above code.

> Where there conflicting success/failure reports for the same
> devices?

I'm sorry, I don't recall.

Tim.
*/

[-- Attachment #2: Type: application/pgp-signature, Size: 232 bytes --]