Hi,
Further to my previous email, it looks like the interrupt is getting through
but is ignored because the card is busy. I've added the code below to ide.c
(function drive_is_ready()). So the IDE code is fine and it doesn't get the
interrupt 'before it is ready' of course. Stupid of me to imagine that.
But there are two problems:
1. This code should only be used when the unexpected interrupt problem
appears.
2. This code shouldn't be needed at all.
I did find a similar problem reported with a PC laptop and a CF card (late
2000), so it has happened before and not just on ARM.
Any ideas?
Regards,
if (IDE_CONTROL_REG)
stat = GET_ALTSTAT();
else
#endif /* CONFIG_IDEPCI_SHARE_IRQ */
#ifdef CONFIG_SA1100_ELF
{
int i;
/* this technique was borrowed from ide_wait_stat() as the SanDisk
SDCFB-16 seems to issue an interrupt before it is ready. Not
completely sure. The result of this hack is a few unexpected
interrupts when transferring large files (2-3MB), but nothing
worse. What is going on here? */
for (i = 0; i < 10000; i++) {
stat = GET_STAT();/* Note: this may clear a pending IRQ!! */
if (!(stat & BUSY_STAT))
break;;
udelay(1);
}
}
#else
stat = GET_STAT(); /* Note: this may clear a pending IRQ!! */
#endif
if (stat & BUSY_STAT)
return 0; /* drive busy: definitely not interrupting */
_______________________________________________
http://lists.arm.linux.org.uk/mailman/listinfo/linux-arm
Please visit the above address for information on this list.
