Linux Operating System Source Code: IBM Patches

One of the largest places for BKL contention in 2.4 is ext2_get_block(). On my 2-way, dbench causes around 4% of combined CPU time just spinning on the lock_kernel() in ext2_get_block(). On bigger machines it just gets worse.

Al Viro produced patch for 2.5 to address this problem

I've done a fairly straightforward backport to the 2.4 tree. But, I'm not sure how it interacts with the VFS patches which came earlier in 2.5. It is quite stable on my 2-way; definitely no deadlocks. I'll be testing it on larger machines soon. I don't expect it to go into the mainline kernel right away, but I would like the secondary tree maintainers to pick it up and let people pound on it.

You can see the lockmeter data here: http://www.sr71.net/ibm/bkl-ext2-dbench/ You may notice that the data from my patch show a slight drop in throughput (~2%). However, my testing conditions were less than ideal and the drop can probably be attributed to statistical error.