Re: 2.6.0 performance problems

[Andrew Morton <akpm@osdl.org>]

Thomas Molina <tmolina@cablespeed.com> wrote:
>
>  On Tue, 30 Dec 2003, Linus Torvalds wrote:
>  > Ok. This looks much closer to the 2.4.x numbers you reported:
>  > 
>  > 	real    13m50.198s
>  > 	user    0m33.780s
>  > 	sys     0m15.390s
>  > 
>  > so I assume that we can consider this problem largely solved? There's 
>  > still some difference, that could be due to just VM tuning.. 
>  > 
>  > I suspect that what happened is:
>  >  - slab debugging adds a heavy CPU _and_ it also makes all the slab caches 
>  >    much less dense.
>  >  - as a result, you see much higher system times, and you also end up 
>  >    needing much more memory for things like the dentry cache, so your 
>  >    memory-starved machine ended up swapping a lot more too.
> 
>  So you are telling me that I am paying the price for running development 
>  kernels and enabling all the debugging.

CONFIG_DEBUG_PAGEALLOC really does hurt on small machines.  Mainly because
it rounds the size of all slab object which are >= 128 bytes up to a full
4k.  So things like inodes and dentries take vastly more memory.

The other debug options are less costly.