[hashcash] Re: benchmark
- From: Atom 'Smasher' <atom@xxxxxxxxxxxxxx>
- To: hashcash@xxxxxxxxxxxxx
- Date: Mon, 13 Sep 2004 04:35:59 -0400 (EDT)
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
On Mon, 13 Sep 2004, Adam Back wrote:
Does compiling with make x86 work on BSD?
=========
that doesn't get far, at all... pretty much the same error with gmake...
# make x86
make "CFLAGS=-DREGEXP_POSIX -O3 -funroll-loops -march=i386 -mcpu=pentium -mmmx
-D_REENTRANT -D_THREAD_SAFE -fPIC " build
cc -DREGEXP_POSIX -O3 -funroll-loops -march=i386 -mcpu=pentium -mmmx
-D_REENTRANT -D_THREAD_SAFE -fPIC -c hashcash.c
cc1: Invalid option `mmx'
*** Error code 1
Stop in /usr/home/alpha/dmz/hashcash-1.08.
*** Error code 1
Stop in /usr/home/alpha/dmz/hashcash-1.08.
~~~~~~~~~~~~~~~~~~~~~~~
i can't help notice that it specifies "-mcpu=pentium". does it matter that
i'm using an amd chip?
might it make a difference if i enable this in my kernel?
# CPU_ENABLE_SSE enables SSE/MMX2 instructions support.
i should probably have that enabled anyway...
looking for "mmx" in dmesg gives me this (does this help?):
$ dmesg | egrep -iB2 mmx
CPU: AMD Duron(tm) Processor (1194.92-MHz 686-class CPU)
Origin = "AuthenticAMD" Id = 0x671 Stepping = 1
Features=0x383f9ff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,MMX,FXSR,SSE>
That will enable the MMX assembler code and be a fair bit faster.
=============
nonetheless, i'm impressed with the improved speed since 1.02.
...atom
_________________________________________
PGP key - http://atom.smasher.org/pgp.txt
762A 3B98 A3C3 96C9 C6B7 582A B88D 52E4 D9F5 7808
-------------------------------------------------
"All animals are equal, but some more than others."
-- George Orwell
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.3.6 (FreeBSD)
Comment: What is this gibberish?
Comment: http://atom.smasher.org/links/#digital_signatures
iQEcBAEBCAAGBQJBRVv1AAoJEAx/d+cTpVciK5MH/383jAKsfwUuHZfdw6sDEoAX
0uHYLQHzXt1Tjbx+iBVYUoAdfEiwcfcSL6Atcb6kzycgaULqNOomIIMiqpzcaGvK
BBx8qLXGgz1ZUKv0g58gAqJCdebWasT0jIk4jw9/ppactMgN6EAERbpwdxzrItsH
D/qhx+2EH81eP90C58zhvMQD47GOXs/e3X+0t4IFB74DCLBsno0LU2umwEnctyf8
qoX95m2Efh50SnjAw6Nbk+MHQ3rC+jCjY2ql2Zc5StzIUZMu8pfZZzsJXF8GLwH9
OSsVEhdt9NsSg9hEzkrlvToQlMgzVsVr253KEhLPMJNEDPFLtsUypdFF0UqAWb4=
=N/8K
-----END PGP SIGNATURE-----
- Follow-Ups:
- [hashcash] Re: benchmark
- From: Adam Back
- References:
- [hashcash] benchmark
- From: Atom 'Smasher'
- [hashcash] Re: benchmark
- From: Adam Back
Other related posts:
- » [hashcash] benchmark
- » [hashcash] Re: benchmark
- » [hashcash] Re: benchmark
- » [hashcash] Re: benchmark
- » [hashcash] Re: benchmark
- » [hashcash] Re: benchmark
- » [hashcash] Re: benchmark
- » [hashcash] Re: benchmark
- » [hashcash] Re: benchmark
- » [hashcash] Re: benchmark
# make x86 make "CFLAGS=-DREGEXP_POSIX -O3 -funroll-loops -march=i386 -mcpu=pentium -mmmx -D_REENTRANT -D_THREAD_SAFE -fPIC " build cc -DREGEXP_POSIX -O3 -funroll-loops -march=i386 -mcpu=pentium -mmmx -D_REENTRANT -D_THREAD_SAFE -fPIC -c hashcash.c cc1: Invalid option `mmx' *** Error code 1
...atom
- [hashcash] Re: benchmark
- From: Adam Back
- [hashcash] benchmark
- From: Atom 'Smasher'
- [hashcash] Re: benchmark
- From: Adam Back