httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Greg Stein <gst...@lyra.org>
Subject Re: filtered I/O memory from stack or pools.
Date Thu, 01 Jun 2000 00:44:20 GMT
On Wed, 31 May 2000 rbb@covalent.net wrote:
>...
> Okay, that's one area where I misunderstood the design then.  They way the
> code looks to me, anytime we modify the data, we will need to pass it down
> the stack.  Especially since it is possible for the data as passed to
> ap_rwrite to be const'ed data.  If we write everything in one big buffer
> (again the way the psued-code looks to me), this does imply big stack
> usage.  I'll need to review the psuedo code again tomorrow.

The buffer can come from a pool; it doesn't have to be on the stack.

Note that if "readonly" flags are introduced into the hook/iovec approach,
then the two designs have similar memory usage patterns.

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/


Mime
View raw message