[PATCH] T3/T4 sparc shifts, plus more timings
tg at gmplib.org
Mon Apr 1 19:34:11 CEST 2013
David Miller <davem at davemloft.net> writes:
Yes, understood. We have to transpose a few of the shifts with
their neighbouring arithmetic ops in this loop to make it optimal
I found a powered up US2 and run time timing tests. No slowdown there
for the new generic functions.
I suppose mpn/sparc64/ultrasparc1234/[lr]shift.asm are now redundant.
Clearly, the new lshiftc code is not optimal for US1 through US4. It
runs 0.5 c/l slower on them all, compared to what one would hope for
2-way unrolled code.
More information about the gmp-devel