div_qr_1 interface

[Niels Möller]

Torbjorn Granlund <tg at gmplib.org> writes:

> But I lack unit testing code for the function, making hacking quite
> cumbersome.  I don't feel safe hacking *any* GMP assembly code without
> tests/devel/try.c's function and access checks.

tests/mpn/t-div.c includes tests for mpn_div_qr_1, including single
guard limbs before and after the q area. That provides for some
reasonable testing, and I was running while GMP_CHECK_RANDOMIZE=0
./t-div ; : ; done over night in the weekend.

But sure, support also in try.c would be good.

> (1) Shorten a dep chain, and avoid keeping CF live over an inc
>     instruction.  The cmov doesn't really depend on sbb, since the
>     latter insn never really changes carry.  (This might btw be useful
>     to teach loppmixer!)

Right, using cmov (instead of masking) allows some more reordering.

> (2) Reallocate Q2 to an "old" register (not r8-r15) and then use the
>     32-bit form of "adc $0,reg".  That form is shorter.

Clever!

> (3) Offet UP to avoid the offset in the loop.

Should be straight forward. Or one could offset UN, but I think it's
nice to be able to use dec / jnz for the loop. Or is the obscure "loop"
decrement and branch instruction useful?

Regards,
/Niels

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