[PATCH 28/30] db: enable metadump on CRC filesystems

Christoph Hellwig hch at infradead.org
Mon Nov 4 03:18:28 CST 2013


On Wed, Oct 30, 2013 at 03:31:19PM +1100, Dave Chinner wrote:
> While there, add IO stack unwinding to xfs_db to prevent metadump
> and other db programs from exiting with unreleased buffers in the IO
> stack and hence emitting warnings like:
> 
> cache_purge: shake on cache 0x69e4f0 left 7 nodes!?

Seems like that should be a separate patch.


> @@ -202,7 +204,8 @@ close_devices:
>  	 * make sure that we pop the last buffer context we held so that the
>  	 * buffer is released before purge the caches during unmount.
>  	 */
> -	pop_cur();
> +	while (iocur_sp > start_iocur_sp)
> +		pop_cur();

The comment above this needs an update.


Otherwise looks fine.



More information about the xfs mailing list