From 9ffdbbda305450364c27ab0600be145a748543f4 Mon Sep 17 00:00:00 2001 From: Kevin Ryde Date: Wed, 7 Nov 2001 23:46:54 +0100 Subject: Add tal-notreent.c keep block allocated. Add cray inline shifts. --- doc/tasks.html | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) (limited to 'doc') diff --git a/doc/tasks.html b/doc/tasks.html index ee0f9c098..ff49e4fc7 100644 --- a/doc/tasks.html +++ b/doc/tasks.html @@ -34,7 +34,7 @@ Copyright 2000, 2001 Free Software Foundation, Inc.
- This file current as of 5 Nov 2001. An up-to-date version is available at + This file current as of 8 Nov 2001. An up-to-date version is available at http://www.swox.com/gmp/tasks.html. Please send comments about this page to bug-gmp@gnu.org. @@ -304,6 +304,12 @@ Copyright 2000, 2001 Free Software Foundation, Inc. improvement for the generic C (which is used on UltraSPARC/64 for instance). GCC recognises the generic C ul*vh and vl*uh are identical, but does two separate additions to the rest of the result. +
  • tal-notreent.c could keep a block of memory permanently allocated. + Currently the last nested TMP_FREE releases all memory, so + there's an allocate and free every time a top-level function using + TMP is called. Would need + mp_set_memory_functions to tell tal-notreent.c to release + any cached memory when changing allocation functions though. @@ -451,6 +457,8 @@ Copyright 2000, 2001 Free Software Foundation, Inc. -hpipeline3 seems promising. We should at least up -O to -O2 or -O3.
  • Cray: mpn_com_n and mpn_and_n etc very probably wants a pragma like MPN_COPY_INCR. +
  • Cray vector systems: mpn_lshift and mpn_rshift + are nice and small and could be inlined to avoid function calls.
  • Cray: Variable length arrays seem to be faster than the tal-notreent.c scheme. Not sure why, maybe they merely give the compiler more information about aliasing (or the lack thereof). Would like to modify -- cgit v1.2.1