I fixed the bug in tuneup. So here is the list of items for release now: todo: ========== mpir build fails on MinGW due to mpn_addmul_2 symbol defined twice Pavel mpz_powm_ui doc says -ve exponent supported, which is rubbish fix flint bugs in primality testing code interface for mpn_mulmod_2expp1 32 and 64 bit tuning for fft remove old fft tuning params t-next_likely_prime undefined reference to abort and exit fft_combine_bits const warning (see below) ../mpn/generic/gcdext.c:238: warning: integer overflow in expression
const warning: ============== In file included from mul_truncate_sqrt2.c:2: ../fft/mul_truncate_sqrt2.c: In function â__gmpn_mul_truncate_sqrt2â: ../fft/mul_truncate_sqrt2.c:113: warning: passing argument 2 of â__fft_combine_bitsâ from incompatible pointer type ../mpir.h:1860: note: expected âconst mp_limb_t **â but argument is of type âmp_limb_t **â Done: ===== remove old fft tuning code add new fft crossover code to tuneup tuneup fails on mpir-exp branch Wont Fix: ========= Vlad report of MinGW sandybridge combination with FFT assert failure - new fft fixes this Chris report of mpz_get_ux failure on Itanium - this is a compiler bug Next Release: ============= generic build option for sage proper tuning code for mpn_mul_Bexpp1_fft fft/fft_negacyclic.c is misspelled move fft_tuning.c into tuneup Bill. On 8 August 2012 06:10, Bill Hart <goodwillh...@googlemail.com> wrote: > I have now fixed the main big item left for release, namely fixing > tuneup to give crossovers for the new fft. > > However, in doing this I discovered that tuneup causes a double free > when tuning DC_BDIV_QR_THRESHOLD, so this will have to be fixed. > > I have not added tuning code for the mpn_mulmod_Bexpp1 function. That > can get done in the next release. > > I also noted that the name of the file fft/fft_negacyclic is > misspelled. This can't just be changed as everywhere it is referred to > by autotools also needs changing. Someone could volunteer to do the > easy fix, except that they would need access to a very specific > version of autotools. I'll add it to the list of things to be done > (either this release or the next). > > Bill. > > On 8 August 2012 02:15, Bill Hart <goodwillh...@googlemail.com> wrote: >> Well, I was right to be suspicious of the following change: >> >> - sed -n 's/[^G]*GLOBAL_FUNC[:space:]*\(.*\)/\1/p' $tmp_file >> + sed -n 's/^;[ ]*PROLOGUE(\([^,]*\).*)/\1/p' $tmp_file ; >> >> I still don't understand what was intended here. I think the idea was >> to give an alternative sed script for ripping PROLOGUE declarations, >> probably due to some incompatibility on some systems. But instead of >> adding the line it appears to have been written over the top of the >> GLOBAL_FUNC line, which is definitely needed. >> >> Adding it back in causes mpir to build correctly and pass its tests on >> linux. This probably also fixes the issues we were having with MinGW >> on the mpir-exp branch. >> >> Bill. >> >> On 28 July 2012 08:41, Brian Gladman <b...@gladman.plus.com> wrote: >>> -----Original Message----- From: Bill Hart >>> Sent: Saturday, July 28, 2012 1:52 AM >>> >>> To: mpir-devel@googlegroups.com >>> Subject: Re: [mpir-devel] Re: MPIR 2.6 release progress >>> >>> By a process of elimination, this problem occurs on both core2 and >>> k10, so it must be one of the following: >>> >>> U cpuid.c >>> U mpn/x86_64/k8/k10/karasub.asm >>> U mpn/x86_64/core2/karasub.asm >>> U mpn/asm-defs.m4 >>> U config.in >>> U configure.in >>> U gmp-h.in >>> U acinclude.m4 >>> U Makefile.am >>> >>> ================================== >>> I am a bit puzzled about the changes to the karasub subroutines since the >>> merge should have taken these from trunk. >>> >>> I thought that I had watched for that but its possible that I missed some of >>> the assembler versions. >>> >>> It would be worth checking this. >>> >>> >>> Brian >>> >>> -- >>> You received this message because you are subscribed to the Google Groups >>> "mpir-devel" group. >>> To post to this group, send email to mpir-devel@googlegroups.com. >>> To unsubscribe from this group, send email to >>> mpir-devel+unsubscr...@googlegroups.com. >>> For more options, visit this group at >>> http://groups.google.com/group/mpir-devel?hl=en. >>> -- You received this message because you are subscribed to the Google Groups "mpir-devel" group. To post to this group, send email to mpir-devel@googlegroups.com. To unsubscribe from this group, send email to mpir-devel+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/mpir-devel?hl=en.
