ide_free_irq()

Simple patch. The current drivers/ide/ide.c calls free_irq() instead of ide_free_irq(). We depend on these alternate routines to deal with our semi-broken hardware, but since ide-probe.c calls ide_reqest_irq(), it seems logical that this should also be ide_free_irq().

Thanks,
Will Jhun

*** drivers/ide/ide.c.orig Mon Feb 25 11:37:57 2002
--- drivers/ide/ide.c Mon Mar 4 14:05:41 2002
***************
*** 2115,2121 ****
g = g->next;
} while (g != hwgroup->hwif);
if (irq_count == 1)
! free_irq(hwif->irq, hwgroup);

/*
* Note that we only release the standard ports,
--- 2115,2121 ----
g = g->next;
} while (g != hwgroup->hwif);
if (irq_count == 1)
! ide_free_irq(hwif->irq, hwgroup);

/*
* Note that we only release the standard ports,

Sure but only one arch I see that is effected and they do not use this storage class ...

asm-s390/ide.h:#define ide_free_irq(irq,dev_id) do {} while (0)
asm-s390x/ide.h:#define ide_free_irq(irq,dev_id) do {} while (0)

Is there another ?

Andre Hedrick
Linux Disk Certification Project Linux ATA Development

Well, our MIPS-based platform (whose changes haven't been submitted yet). We have some broken PCMCIA hardware that requires us to actually simulate interrupts with a hi-res timer (by calling ide_intr() directly). Yes, it's ugly, but it's our only solution. This change would enable us to put our device-specific interrupt "simulation" code into the ide_* routines only.

It's not a must right now, but it looked like something that was left out, since the rest of the ide code uses the ide_* generic #defines...

Thanks,
Will Jhun