RE: compile problems

[prev] [thread] [next] [Date index for 2004/12/29]

From: Barksdale, Ray
Subject: RE: compile problems
Date: 16:59 on 29 Dec 2004
 

> -----Original Message-----
> From: Joe Schaefer [mailto:joe+gmane@xxxxxxxxxx.xxx] 
> Sent: Tuesday, December 28, 2004 7:30 PM
> To: mod_perl user list
> Subject: Re: compile problems
> 
> "Barksdale, Ray" <barksdale@xxxx.xxxxx.xx.xx> writes:
> 
> [...]
> 
> >> >>From this:
> >> > Total:     608169K ( 580M) size, 145465K ( 139M) approx 
> >> real size (-shared)
> >> > 
> >> > down to this:
> >> > Total:     600129K ( 572M) size, 131588K ( 125M) approx 
> >> real size (-shared)
> 
> 
> Are you sure this isn't just some wackiness in the virtual 
> memory reporting on amd64?   I've been using debian amd64 
> for a while, and the VSZ numbers have been consistently absurd 
> for all my processes (eg they often total to considerably more 
> than my available RAM + swap).  For instance here's my emacs 
> on debian woody (32bit):
> 
>        % emacs -nw &
>        [1] 13029
>        % ps u
>        USER       PID %CPU %MEM   VSZ  RSS ...
>        joe      13029  0.0  0.3  8020 3208 ...
> 
> and on amd64:
> 
>        % ps u
>        USER       PID %CPU %MEM    VSZ   RSS ...
>        joe      21248  0.0  0.2  29144  4336 ...
> 
> 
> The libraries emacs links to are basically the same:
> 
>  % ldd `which emacs` | perl -alpe '$a += `du -bL $F[2]`} $_=$a;{'
> 
>  =>
> 
>    5060944 bytes for 64bit, 
>    4063232 bytes for 32bit.
> 
> So the libraries are ~25% larger, the emacs binary
> is ~60% larger (3985408 vs 6633208), but ps reports
> almost a four-fold increase in VSZ.  Go figure.
> 
> Here's a possibly relevant link I found while googling
> for an answer:
> 
>  
> https://www.redhat.com/archives/amd64-list/2004-September/msg0
> 0003.html

I'm glad you found this. I was going to build a new gcc and friends
this afternoon. 

Looks like that's it. I'm seeing a 5-fold difference in reported usage
between the 32-bit and 64-bit boxes for perl and CPAN shell. 
Haven't seen any httpd's killed off but the box has 4 Gigs and a light load.

The 25% figure is what I had expected based on 
http://gcc.fyxm.net/summit/2003/Porting%20gcc%20to%20the%20amd64.pdf



> 
> -- 
> Joe Schaefer
> 
> -- 
> Report problems: http://perl.apache.org/bugs/
> Mail list info: http://perl.apache.org/maillist/modperl.html
> List etiquette: http://perl.apache.org/maillist/email-etiquette.html
> 


*****CONFIDENTIALITY NOTICE*****
This e-mail and any files or attachments may contain confidential and
privileged information.  If you have received this message in error, please
notify the sender at the above e-mail address and delete it and all copies
from your system.



        -- 
        Report problems: http://perl.apache.org/bugs/
Mail list info: http://perl.apache.org/maillist/modperl.html
List etiquette: http://perl.apache.org/maillist/email-etiquette.html

(message missing)

RE: compile problems
Barksdale, Ray 16:57 on 17 Dec 2004

Re: compile problems
Joe Schaefer 18:27 on 17 Dec 2004

Re: compile problems
Stas Bekman 18:38 on 17 Dec 2004

RE: compile problems
Barksdale, Ray 19:05 on 17 Dec 2004

Re: compile problems
Stas Bekman 19:26 on 17 Dec 2004

RE: compile problems
Barksdale, Ray 21:05 on 17 Dec 2004

Re: compile problems
Stas Bekman 21:19 on 17 Dec 2004

RE: compile problems
Barksdale, Ray 23:38 on 17 Dec 2004

Re: compile problems
Stas Bekman 23:44 on 17 Dec 2004

RE: compile problems
Barksdale, Ray 15:19 on 22 Dec 2004

Re: compile problems
Stas Bekman 20:27 on 22 Dec 2004

RE: compile problems
Barksdale, Ray 20:37 on 22 Dec 2004

Re: compile problems
Stas Bekman 21:16 on 22 Dec 2004

RE: compile problems
Barksdale, Ray 22:01 on 22 Dec 2004

Re: compile problems
Stas Bekman 22:22 on 22 Dec 2004

RE: compile problems
Barksdale, Ray 23:07 on 22 Dec 2004

Re: compile problems
Stas Bekman 00:01 on 23 Dec 2004

RE: compile problems
Barksdale, Ray 22:50 on 28 Dec 2004

Re: compile problems
Stas Bekman 23:01 on 28 Dec 2004

RE: compile problems
Barksdale, Ray 23:15 on 28 Dec 2004

Re: compile problems
Stas Bekman 23:17 on 28 Dec 2004

RE: compile problems
Barksdale, Ray 23:50 on 28 Dec 2004

Re: compile problems
Joe Schaefer 01:29 on 29 Dec 2004

RE: compile problems
Barksdale, Ray 16:59 on 29 Dec 2004

Re: compile problems
Stas Bekman 17:21 on 29 Dec 2004

Generated at 12:16 on 16 Jan 2005 by mariachi v0.52