Re: [PATCH] Scalable page cache - take two

• Next message: Rik van Riel: "Re: [PATCH: NEW ARCHITECTURE FOR 2.5] support for NCR voyager 343x/345x/4100/51xx architecture"
• Previous message: David S. Miller: "Re: [PATCH] - 2.4.17 - if_arp.h - Add the Prism2 ARP type"
• In reply to: Momchil Velikov: "[PATCH] Scalable page cache - take two"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

 
Hi,

> This is the second mutation of the scalable page cache patch. It:

I ran the dbench test over it just to annoy Andrew:

http://samba.org/~anton/linux/pagecache_locking/2/

This is the still 12 way ppc64, the results are similar to the last
test. It confirms that the current pagecache locking is a bottleneck
for larger SMP machines.

We quickly hit the following spinlocks with the pagecache lock out of
the way:

ext2_get_block: BKL
refile_buffer: lru_list_lock
remove_from_queues: lru_list_lock
try_to_free_buffers: lru_list_lock
d_lookup: dcache_lock
ext2_discard_prealloc: BKL

Is someone working on removing the BKL from ext2 in 2.5?

Anton
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

• Next message: Rik van Riel: "Re: [PATCH: NEW ARCHITECTURE FOR 2.5] support for NCR voyager 343x/345x/4100/51xx architecture"
• Previous message: David S. Miller: "Re: [PATCH] - 2.4.17 - if_arp.h - Add the Prism2 ARP type"
• In reply to: Momchil Velikov: "[PATCH] Scalable page cache - take two"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

This archive was generated by hypermail 2b29 : Sun Dec 23 2001 - 21:00:31 EST