[Gmp-commit] /var/hg/www: One more Sparc item done.
mercurial at gmplib.org
mercurial at gmplib.org
Tue Apr 30 00:18:51 CEST 2013
details: /var/hg/www/rev/569753dddda9
changeset: 60:569753dddda9
user: Torbjorn Granlund <tege at gmplib.org>
date: Tue Apr 30 00:18:47 2013 +0200
description:
One more Sparc item done.
diffstat:
devel/sparc.html | 8 ++++----
1 files changed, 4 insertions(+), 4 deletions(-)
diffs (25 lines):
diff -r c004de8197ac -r 569753dddda9 devel/sparc.html
--- a/devel/sparc.html Tue Apr 30 00:16:59 2013 +0200
+++ b/devel/sparc.html Tue Apr 30 00:18:47 2013 +0200
@@ -89,10 +89,6 @@
the loop. This should be quite easy; just bubble mulx/umulxhi dependees
downwards a few lines.
</li>
- <li> We have two new mpn addmul_1 variants, one 2-way unrolled and one 4-way
- unrolled. Commit the one which gives best performance for the critical
- operand sizes.
- </li>
<li> Explore using the mpmul instruction for mpn_mul_basecase. Since mpmul
handles just same-size operands, a GMP {up,un} × {vp,vn} multiply (where
un ≥ vn) will require an initial {up,vn} × {vp,vn} multiply, then
@@ -119,6 +115,10 @@
DONE:
<p>
<ul>
+ <li> We have two new mpn addmul_1 variants, one 2-way unrolled and one 4-way
+ unrolled. Commit the one which gives best performance for the critical
+ operand sizes.
+ </li>
<li> Rewrite mpn_submul_1 for speedup 5.8 c/l → 4.5 c/l.
</li>
<li> Finish the cnd_aors_n.asm code.
More information about the gmp-commit
mailing list