Re: Flaw in ide_unregister()

["Richard Purdie" <rpurdie@rpsys.net>]

Alan Cox:
[Re: ide_unregister and problems with ide-cs.c]
> In 2.6.9-ac and 2.6.10-ac the ide_unregister_hwif calls return an error
> if the drive is in use. At this point the ide-cs code still throws it
> away. The -ac code IDE also adds "removed_hwif_iops" so the bits are
> there for the correct result which is something like
>
> if(ide_unregister_hwif(hwif) < 0 {
> printk("Whine whine...");
> removed_hwif_ops(hwif);
> while(ide_unregister_hwif(hwif) < 0)
> msleep(1000);
> }

I've just tried 2.6.10-ac8 on my target (an arm PDA) and it doesn't help 
ide-cs.c. When I pull out a mounted CF out the socket, the kernel prints 
"Unregister 0 fail 1 0" repeatedly on the console showing the usage count 
permanently stays at 1. This is the same problem as under 2.6.10.

I haven't investigated it yet but I suspect the usage count is held by 
ide-disk as the CF card has a mounted filesystem. As previously mentioned 
and for reference, this patch has the changes I had to make to get standard 
2.6.10 to work:

http://www.rpsys.net/openzaurus/2.6.10/ide_fixes-r1.patch

Richard 



-- 
No virus found in this outgoing message.
Checked by AVG Anti-Virus.
Version: 7.0.300 / Virus Database: 265.6.9 - Release Date: 06/01/2005