New thresholds in table

Niels Möller nisse at
Mon Nov 14 10:28:10 CET 2011

Torbjorn Granlund <tg at> writes:

> Have you made a sanity check plot of the various higher level functions
> after having compiled-in new thresholds?  Any bumps might indicate a
> spot for improvement.

I tried that. Some small bumps in the timing for gcdext, and some narrow
spikes (which for now I attribute to measurement error). So it looks
reasonably sane.

I have tweaked the tuneup parameters a bit. I removed min_size for
HGCD_APPR_THRESHOLD (was 30, now uses the default of 10). This threshold
is definitely lower than I had expected. On, it's 10 on one machine
(itanium2), and 11 or 12 on several 64-bit machines (x86_64 and power).

I also increased max_size and step_factor for HGCD_REDUCE_THRESHOLD.
That might have broken tuneup a bit, I see results are missing for a
large number of machines for last night. Typical values seem to be a few
thousand limbs.


Niels Möller. PGP-encrypted email is preferred. Keyid C0B98E26.
Internet email is subject to wholesale government surveillance.

More information about the gmp-devel mailing list