linux-kernel - Atom N270 more efficent with int 0x80 than with syscalls

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <1277544011.24284.14.camel@lithium.local.net>
Date:	Sat, 26 Jun 2010 10:20:11 +0100
From:	Alex Buell <alex.buell@...ted.org.uk>
To:	Mailing Lists - Kernel Developers <linux-kernel@...r.kernel.org>
Subject: Atom N270 more efficent with int 0x80 than with syscalls

Recently I rewrote a small assembler program to test timings and print
out timing informations for various assembly routines. Then I thought to
add tests for timings on int 0x80 and calling via syscalls. 

Now I know int 0x80 calls are expensive on processors such as the
Pentium IV but someone has tested my assembly program with their Atom
N270 1.6GHz system, and int 0x80s calls are actually much quicker than
syscalls on these processors. 

Here's the test program that exposed the timing differences with these
int 0x80 calls and syscalls:
http://www.munted.org.uk/programming/newtimer.zip

Perhaps the kernel could be greatly speeded up on certain processors if
it dynamically adapts to use the quickest type of system calls? 
-- 
http://www.munted.org.uk

One very high maintenance cat living here.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/