I'm working with the new EH in the libata from the 2.6.18-rc4, and I'm
curious what the proper way is to signal the EH code that a hotplugin
event has occurred, or that a hotunplug event has occurred.
(this is for the sata_promise driver which does its own interrupt
handling, with special register flags for each drive's hotplugin and
hotunplug events). Here is what I have so far (seems to work):
if( plugdata & (1 << i) ) {
printk( KERN_DEBUG "ata%d: Hot-unplug interrupt\n", ap->id );
/*
* Set up drive as already missing so it is not
* re-probed multiple times.
*/
struct ata_device *dev = &ap->device[0];
dev->flags |= ATA_DFLAG_DETACH;
if( !(ap->pflags & ATA_PFLAG_FROZEN) )
ata_port_freeze(ap);
handled++;
} else if( plugdata & (1 << (i+4)) ) {
printk( KERN_DEBUG "ata%d: Hot-plugin interrupt\n", ap->id );
/*
* Just 'freezing' the drive works... it is
* reset and re-probed automatically!
*/
if( !(ap->pflags & ATA_PFLAG_FROZEN) )
ata_port_freeze(ap);
handled++;
}
The only problem that I've noticed so far is that in the plugin case,
the bus scanning never seems to run the ata_wait_spinup routine. Is
there something more I should do in this case to let everything know
that we've got a new device and should wait for it to spinup?
Also, I added the 'dev->flags |= ATA_DFLAG_DETACH' on a hunch...
before this on an unplug event it would try to scan the device 3 times
before actually removing it, and this isn't strictly necessary. Is
this wrong, or is there a better way to tell everything that the drive
is gone and shouldn't be rescanned?
--
Jim Ramsay
"Me fail English? That's unpossible!"
-
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
