On Monday, August 25, 2003, at 06:04  PM, Andrew Brown wrote:

>>> I wonder why the /dev/zero case is slower. I guess there's not much
>>> algorithmic improvement possible there; it's probably the worst case
>>> for the rbtree algorithm. But, it's also an uncommon case, so the
>>> regular file map speedup seems to clearly outweigh that one.
>>
>> I guess the question is: does it hurt MAP_ANON performance the same 
>> way
>> it hurts mmap of /dev/zero performance?
>
> isn't that a rhetorical question, since we have:

Yes, I realize what we have, which is why I asked it.  Frank's comment 
that mmaping of /dev/zero wasn't very common is true, but MAP_ANON 
mappings are fairly common, and so it is important not to degrade their 
performance.

         -- Jason R. Thorpe <thorpej@wasabisystems.com>