SBU will probably not work with low memory (was Re: Collecting SBUs now)

Rui Ferreira ruifmferreira at hotmail.com
Tue May 28 05:36:22 PDT 2002


Hi again

Gerard is now collecting SBU's.
Bash = 70sec (1 SBU)
Binutils = 137sec  = 1.96 (but rounded to 2 SBU's)
Binutils is way larger than 1.96 * Bash.
By your idea one could think something like 8 SBU's
We're yet to find out more about this since both packages fit at a
corner of Gerard's computer memory.
70sec could be 5secCPU + 65secHD
137sec could be 20secCPU + 117secHD
We'll see how good SBU's are when someone reaches the end of a lfs.


----- Original Message -----
From: "Fabio Fracassi" <turiya at linuxfromscratch.org>
To: <lfs-dev at linuxfromscratch.org>
Sent: Tuesday, May 28, 2002 11:04 AM
Subject: SBU will probably not work with low memory (was Re:
Collecting SBUs now)


Hi, just a short note on the SBU.

A machine with low memory (=<64MB) will probably have much greater
variance in
the SBU between small packages and big ones, for example on a
otherwise
similar machine bash may take (purely making up numbers) 1 min and
glibc
takes 30 on the one with much memory, but takes 2 hours on the one
with low
memory.

The reason should be obvious, the bash compile needs less (physical)
memory,
so it will always fit, but glibc has to be swapped out.

Might be that I'm totaly wrong.

This althogh should not be an argument against SBU's, just a point to
keep in
mind when doing the averages.

Fabio

--
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message


-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message



More information about the lfs-dev mailing list