Mail Archives: djgpp/1997/11/27/22:46:29
Eli Zaretskii wrote:
>
> On 22 Nov 1997, Ed Leiby wrote:
>
> > The problem is that it seems to run at two different speeds: As fast as
> > Borland, and 2-3 times slower than Borland. I have tried using
> > cwsdpmi,cwsdpr0 (rev 0.90+) and win95 DPMI server.
>
> Does this happen on all platforms (i.e., plain DOS+CWSDPMI and Windows
> 95)?
When running plain DOS ( version 4.00.950 ) I have never seen the faster
performance. This is probably an important clue.
>
> > Any ideas ????
>
> Compile and link with -pg and compare the profiles of fast and slow
> runs. Maybe you will see what function takes longer, and that would
> give a clue.
I tried the -pg/gprof angle and was able to identify where a general
purpose floating pt. complex correlation routine was showing a total
ms/call of 0.04 ( slow case ) compared to 5.11 total us/call. Assuming
ms=milliseconds and us =microseconds this is a ~8:1 delta. This routine
is simple and seems to give the same results whether its running slow or
fast.
I was wondering whether or not alignment might not be a factor in this.
Is there anything that win95 could do at run-time to change the
alignment of float data? If so, what is the difference in machine cycles
between a float access for a "good" alignment and a "bad" alignment?
Would this affect both reads/fetches and writes? If a misaligned access
is cached, is there a subsequent cycle hit on all further fetches?
Sorry for all the questions, hope they're close enough to topic. I also
hope somebody will answer them.
Thanks,
Ed Leiby
Ed Leiby
- Raw text -