From: Frank Kotler on
santosh wrote:
> volcano wrote:
>
>
>>Hi,
>>
>>http://www.asmcommunity.net/board/index.php?topic=10687.0
>>
>>above is a topic to nasm code about extracting exporting functions
>>from a dll. I have the latest include file AFAIK. The code is working
>>OK for small DLLs, such imagehlp.dll but if bigger, application is
>>freezing.
>>
>>Is nasm so slow or still there are some bugs? I also uploaded my
>>sources for whom wishes.
>>
>>http://www.mediafire.com/?edsd1994stz
>
>
> Your source is compiling fine for me with NASM 0.99.05. Also I don't see
> any noticeable speed difference with NASM 0.98.39, but I didn't time it
> though.

Really? Pretty noticeable on my machine. "time" says about 3.3xx vs
0.3xx. Try it again...

Best,
Frank

From: Herbert Kleebauer on
santosh wrote:

> > http://www.asmcommunity.net/board/index.php?topic=10687.0
> >
> > above is a topic to nasm code about extracting exporting functions
> > from a dll. I have the latest include file AFAIK. The code is working
> > OK for small DLLs, such imagehlp.dll but if bigger, application is
> > freezing.
> >
> > Is nasm so slow or still there are some bugs? I also uploaded my
> > sources for whom wishes.
> >
> > http://www.mediafire.com/?edsd1994stz
>
> Your source is compiling fine for me with NASM 0.99.05. Also I don't see
> any noticeable speed difference with NASM 0.98.39, but I didn't time it
> though.

I think he doesn't speak about the speed of NASM but about the speed
of code generated by NASM. And for sure the code optimizer of NASM
can't compete with the one of a C compiler. But to generate code
which is infinite slow (freeze) shouldn't even happen with an assembler
which uses the crazy Intel syntax.
From: Betov on
Herbert Kleebauer <klee(a)unibwm.de> �crivait
news:4723154A.8404067D(a)unibwm.de:

> I think he doesn't speak about the speed of NASM but about the speed
> of code generated by NASM.

Given the fact that his app hangs pretty well, i fail
to imagine that he would be talking about its speed.


Betov.

< http://rosasm.org >





From: Herbert Kleebauer on
Betov wrote:
> Herbert Kleebauer <klee(a)unibwm.de> �crivait

> > I think he doesn't speak about the speed of NASM but about the speed
> > of code generated by NASM.
>
> Given the fact that his app hangs pretty well, i fail
> to imagine that he would be talking about its speed.

The code is working OK for small DLLs, such imagehlp.dll
but if bigger, application is freezing.

Maybe NASM doesn't optimize his sort algorithm (a well designed
assembler doesn't only look at the instructions the programmer gives
it to assemble but analyses the complete program and replaces a slow
algorithm itself by a faster one) and therefore it gets extremely
slow ("freezing" when compared to the human life time).
From: santosh on
Herbert Kleebauer wrote:

> Betov wrote:
>> Herbert Kleebauer <klee(a)unibwm.de> �crivait
>
>> > I think he doesn't speak about the speed of NASM but about the
>> > speed of code generated by NASM.
>>
>> Given the fact that his app hangs pretty well, i fail
>> to imagine that he would be talking about its speed.
>
> The code is working OK for small DLLs, such imagehlp.dll
> but if bigger, application is freezing.
>
> Maybe NASM doesn't optimize his sort algorithm (a well designed
> assembler doesn't only look at the instructions the programmer gives
> it to assemble but analyses the complete program and replaces a slow
> algorithm itself by a faster one)

OMG! The only reason I use an assembler every now and then is so that I
get the code I specify.

If I want the code generator to do all sorts of transformations behind
my back I'll take an HLL like C, thank you very much.