Pentium M and P4 question / and a failed test

Josh Scholar joshscholar at
Sat Sep 29 10:29:53 CEST 2007

While I'm busy building GMP on my Pentium M laptop, I thought I'd ask a
question about target processors  (it's been building it for about an hour
so for, though oddly it isn't taking very much processor power to build).

I notice that in the benchmarks, the Pentium M is grouped with the PIII not
the P4, even though the Pentium-M is capable of running the SSE2 and all of
the same instructions as a P4 (as far as I know).

Is there any reason not to compile with a P4 target and optimizations if
you're going to run on a Pentium M?  Is the Pentium M slower running SSE2
than running PIII code?

I'd test it myself, but at this rate it looks like every build takes over an

I have a few other questions too, but I'll RTFM (and read the fine code
too), first.

Josh Scholar

PS. it just finished the build at about 1 1/2 hours.  Now to make the test.
It's a mystery why the processor usage stayed low throughout the build, it
didn't seem that disk bount.  I'll bet it could have built in 1/2 the time.

Ooh, it just failed "make check" (I'm running MSys). From the error message
it looks like it cut off the beginning of a path at a space character.  I
guess I'll have rename the directory structure to get rid of spaces.

libtool: link: cannot find the library `fog/gmp-4.2.2/' or
unhandled argument `fog/gmp-4.2.2/'
make[4]: *** [t-bswap.exe] Error 1
make[4]: Leaving directory `/c/assembler/agner fog/gmp-4.2.2/tests'
make[3]: *** [check-am] Error 2
make[3]: Leaving directory `/c/assembler/agner fog/gmp-4.2.2/tests'
make[2]: *** [check-recursive] Error 1
make[2]: Leaving directory `/c/assembler/agner fog/gmp-4.2.2/tests'
make[1]: *** [check-recursive] Error 1
make[1]: Leaving directory `/c/assembler/agner fog/gmp-4.2.2'
make: *** [check] Error 2

See how the first quoted couple of quoted directories start 'fog...' instead
of '/c/assembler/agner fog'?
-------------- next part --------------
An HTML attachment was scrubbed...

More information about the gmp-discuss mailing list