lkml.org 
[lkml]   [1998]   [Nov]   [17]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: Linux-asm (was A patch for linux 2.1.127)
From
Date

Sorry, I can't resist...

> I won't consume any more bandwidth on this list until some
> interested programmers have reviewed my paper and checked out
> the code.
>
> Typical output from the test program, testing a procedure that
> is essential in kernel programming (unnamed here) is:
>
>
> Verifying ASM code
> Counting C loops for 2 seconds
> Counting ASM loops for 2 seconds
> C routine : 44263
> AS routine : 539274
> Change : 12.18 times faster
> AS clocks : 573 clocks/byte : 0.28
> C clocks : 16017 clocks/byte : 7.82

What are you trying to prove ? That assembly code can be faster than C code ?
True, but it is often faster to write fast code in C than in assembler, because
it is easier :

Verifying ASM code
Counting C loops for 2 seconds
Counting ASM loops for 2 seconds
C routine : 446593
AS routine : 366775
Change : 0.82 times faster <------- ;-)
AS clocks : 8422 clocks/byte : 4.11
C clocks : 7111 clocks/byte : 3.47

Here is a patch to your chksum.c, to improve your checksum.S :

begin 644 chksum.c.diff.gz
M'XL("-8P4#8"`T$`C9/+;H,P$$77X2MF"7&<8,`D@!+Q`7TLNJQ8-.``*@7$
MHU%5]=]K0T4@@LJSL.29X[EW)`_&&,+DO6X_MN&VSHKKZJ7-X:GX!$)!-US+
M<JD-Q'$."D)H0%>/Q1_%BQS9NY3VE.\#-NG&!L1/TP#?5X!'QG(XG8Y`O.Y:
MPA'4-J_3.&<1U$E1-;#6X-Q7N02OZP^>@OCMFJ094[L&7$Z#[RXKHF)Q6C>L
M@J%35N2Q>&YLQ&EZ`RI:HK&F(#4H7XW@!@D-S#W:DW<&S+TCP00R9R$]D'!@
MCJ"I[BQNWPO/4E1&V0H\"<']@K\%92(UM",E?9":E9`%APNXW-C$E!,W9,8E
M=,YB*>"[WR92O:G1#_[IUF)8!(PU!?Z77)<(]<N47M1+]L93D]BMX3F*X/S5
2,!P6;<[7;Z?\`G41%%P-!```
`
end


-- Benoit

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:45    [from the cache]
©2003-2011 Jasper Spaans