On Fri, 3 Oct 2014, Geert Uytterhoeven wrote:
On Thu, Oct 2, 2014 at 8:56 AM, Finn Thain <fthain@xxxxxxxxxxxxxxxxxxx>
wrote:
Oak scsi doesn't use any IRQ, but it sets irq = IRQ_NONE rather than
SCSI_IRQ_NONE. Problem is, the core NCR5380 driver expects SCSI_IRQ_NONE
if it is to issue IDENTIFY commands that prevent target disconnection.
Other drivers, when they can't get an IRQ or can't use one, will set
host->irq = SCSI_IRQ_NONE (that is, 255). But when they exit they will
attempt to free IRQ 255 which was never requested.
Fix these bugs by using IRQ_NONE in place of SCSI_IRQ_NONE. This means
IRQ 0 is no longer probed by ISA drivers but I don't think this matters.
IRQ_NONE is part of enum irqreturn. I guess you meant NO_IRQ?
But NO_IRQ is deprecated, and not available on all architectures.
The recommended way is to just use 0, like in "if (instance->irq) ...".
Note that some drivers do
#ifndef NO_IRQ
#define NO_IRQ (-1)
#endif
and others do
#ifndef NO_IRQ
#define NO_IRQ 0
#endif
Well, the question becomes, is it better to replace SCSI_IRQ_NONE with 0
or with NO_IRQ?
I guess drivers use #ifndef in case the architecture brings its own
definition of NO_IRQ (presumably because it can't use 0).
Since NCR5380 drivers cover a variety of architectures (ARM, m68k, ISA,
PCI...) it seems that the more prudent option is,
#ifndef NO_IRQ
#define NO_IRQ 0
#endif
--
--
To unsubscribe from this list: send the line "unsubscribe linux-m68k" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
