On Wed, Nov 17 2004, Alan Cox wrote:

On Mer, 2004-11-17 at 15:37, Jens Axboe wrote:

quoted

quoted

-		HWIF(drive)->OUTB(WIN_PACKETCMD, IDE_COMMAND_REG);
+		spin_lock_irqsave(&ide_lock, flags);
+		HWIF(drive)->OUTBSYNC(WIN_PACKETCMD, IDE_COMMAND_REG);
+		ndelay(400);
+		spin_unlock_irqsave(&ide_lock, flags);
 		return (*handler) (drive);
 	}
 }

What good does the lock do?

The same as in ide_execute_command - make sure we don't take an IDE
interrupt that tries to read the state during the delay. That is the old
2.4 "may drives shared IRQ random fails" fix and why the lock is taken
in ide_execute_command.

Ah I see, makes sense. I didn't think about the shared irq case.

-- 
Jens Axboe