Re: pata_serverworks oopses in latest -git

[Jeff Garzik <jeff@garzik.org>]

[-- Attachment #1: Type: text/plain, Size: 675 bytes --]

Alan Cox wrote:
> Ar Maw, 2006-09-26 am 14:00 +0200, ysgrifennodd Diego Calleja:
>> When trying to test the new libata PATA drivers in the latest -git tree, I 
>> got this when udev tried to load the module:
> 
> Yes something seems to be very ill in the -git tree but I'm not sure
> what has changed in the libata core to trigger all this at the moment.

One thing I notice in pata_serverworks.c is the lack of an ->irq_clear() 
hook, which is called unconditionally in ata_device_add().  It's a bit 
strange that it wasn't caught before, since ->irq_clear() has been 
called unconditionally for a great many kernel releases.

Diego, does the attached patch help?

	Jeff




[-- Attachment #2: patch --]
[-- Type: text/plain, Size: 697 bytes --]

diff --git a/drivers/ata/pata_serverworks.c b/drivers/ata/pata_serverworks.c
index af45611..9c1719c 100644
--- a/drivers/ata/pata_serverworks.c
+++ b/drivers/ata/pata_serverworks.c
@@ -356,6 +356,8 @@ static struct ata_port_operations server
 	.data_xfer	= ata_pio_data_xfer,
 
 	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
 	.port_start	= ata_port_start,
 	.port_stop	= ata_port_stop,
 	.host_stop	= ata_host_stop
@@ -389,6 +391,8 @@ static struct ata_port_operations server
 	.data_xfer	= ata_pio_data_xfer,
 
 	.irq_handler	= ata_interrupt,
+	.irq_clear	= ata_bmdma_irq_clear,
+
 	.port_start	= ata_port_start,
 	.port_stop	= ata_port_stop,
 	.host_stop	= ata_host_stop