Hi, 

  for a network monitoring project I'm working on we have to memory map quite 
a lot of data and therefore use vmalloc extensively. A quick google showed me 
that the general assumption is that vmalloc is a lot slower than kmalloc. For 
monitoring high speed network links throughput is of the essence so this made 
me wonder whether I'd taken the correct route.

To test the assumption I created a very simple module that does some reading 
and writing on large buffers. As I'm not too acquainted with the kernel's MM 
this was the best I could do. In any case, 

--> for runtime performance vmalloc'ated memory doesn't appear to perform 
worse than kmalloc'ed. <--

Perhaps the function call itself takes longer (I haven't tested that), but 
that can be handled outside of the datapath of our code.

I thought I'd post the results, including sourcecode, here FYI. Please let me 
know If my test is seriously flawed or you disagree with the outcome for some 
other reason.

cheers,

  Willem

-- 
Willem de Bruijn
wdebruij_at_dds.nl
http://www.wdebruij.dds.nl/

current project : 
Fairly Fast Packet Filter (FFPF)
http://ffpf.sourceforge.net/


-- 
Dit bericht is gescand op virussen en andere gevaarlijke inhoud door ULCN MailScanner en het bericht lijkt schoon te zijn.
This message has been scanned for viruses and dangerous content by ULCN MailScanner, and is believed to be clean.