avr-gcc-list
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [avr-gcc-list] asm/C


From: David Brown
Subject: Re: [avr-gcc-list] asm/C
Date: Thu, 13 May 2004 13:28:11 +0200

Hi,

Sometimes there are good reasons for using assembly, but this does not
strike me as an assembly-friendly problem.  You will probably get smaller
and faster 32-bit arithmetic code using C and letting the compiler do the
optomisation, rather than using assembly and perhaps using sub-optimal code
in the interests of writing understandable and maintainable code.  I
certainly can't see how memory size (code or data) could be an issue with
this sort of problem on a Mega128.  I'm sure it would be possible to write
faster assembly than that generated by the compiler, but you really have to
think whether it is worth the effort, and whether you can be sure that it is
equally reliable.  Even then, I'd do the prototype in C and look at
hand-optomising the inner loops later.

mvh.,

David.

----- Original Message -----
From: "Muhammed" <address@hidden>
To: <address@hidden>
Sent: Tuesday, May 11, 2004 12:59 PM
Subject: [avr-gcc-list] asm/C


want to ask some more questions with your permission. Here they are :)
   Is there a way to get these numbers into registers as an argumentative
way? I mean by using keyboard like console applications in C or should it be
given in the asm start code.
   Is it possible to open an AVR asm.obj file directly in gcc? If not, is
there a tool to convert AVR asm code to inline assembly code in a C file?
  Actually, what I need is having the capability of using 32 bit math for
calculating a hash value
for a security application by using AtMega128 8-bit microcontroller. Due to
memory, power constraints I need to do these things by using inline
assembly. Thanks in advance,

_______________________________________________
avr-gcc-list mailing list
address@hidden
http://www.avr1.org/mailman/listinfo/avr-gcc-list





reply via email to

[Prev in Thread] Current Thread [Next in Thread]