Re: dentry bloat.

[Dipankar Sarma <dipankar@in.ibm.com>]

On Fri, May 14, 2004 at 04:03:23PM +0530, Raghavan wrote:
> 
> Environment - 2-way P4 Xeon 2.4MHz SMP box with 4.5GB RAM.
> Tests were run for 10 iterations to calculate the milliseconds/iteration
> and then mean and deviation were calculated.

I think it is microseconds/iteration. Lesser the better.

> Kernel version              Mean              Standard Deviation
> ---------------             ----              ------------------
> 
> 2.6.6-rc3(baseline)         10578             221
> 
> 2.6.6                       10280             110

So alignment changes helped.

> 
> 2.6.6-bk                    10862             30

Hash function changes regressed.


> 2.6.6-mm1                   10626             36

dentry size change patchset helps.

> To find out if the huge performance dip between the 2.6.6
> and 2.6.6-bk is because of the hash changes, I removed the hash patch
> from 2.6.6-bk and applied it to 2.6.6.
> 
> 2.6.6-bk with old hash      10685             34
> 
> 2.6.6 with new hash         10496             125
> 
> Looks like the new hashing function has brought down the performance.
> Also some code outside dcache.c and inode.c seems to have pushed down
> the performance in 2.6.6-bk.

OK, I am confused. These numbers show that the new hash function
is better. It contradicts your conclusion. And why are you 
comparing 2.6.6-bk+old has with 2.6.6+new hash ?  Why not 
2.6.6-bk vs. 2.6.6-bk-with-old-hash ? 

Thanks
Dipankar