Elias wrote:
On Thursday, November 15, 2012 5:34:33 PM UTC+1, leif wrote:
So what does
$ gcc-4.6 -march=native -v -c -x c /dev/null -o /dev/null 2>&1 |
grep -w cc1
/usr/lib/gcc/x86_64-linux-gnu/4.6/cc1 -quiet -v -imultilib .
-imultiarch x86_64-linux-gnu /dev/null -march=corei7-avx -mno-cx16
-msahf -mno-movbe -mno-aes -mno-pclmul -mno-popcnt -mno-abm -mno-lwp
-mno-fma -mno-fma4 -mno-xop -mno-bmi -mno-tbm -mno-avx -mno-sse4.2
-mno-sse4.1 -mno-rdrnd -mno-f16c -mno-fsgsbase --param l1-cache-size=32
--param l1-cache-line-size=64 --param l2-cache-size=6144
-mtune=corei7-avx -quiet -dumpbase null -auxbase-strip /dev/null
-version -fstack-protector -o /tmp/ccdLHpEj.s
Thanks. So GCC is pretty smart detecting the features actually
supported... (Note the many '-mno-...' parameters.)
I guess MPIR just passed '-march=corei7-avx' (Sandy Bridge), which led
GCC to emitting instructions supported by the CPU, but not VirtualBox
(or whatever virtual machine you're using).
FWIW, /proc/cpuinfo seems to correctly reflect what the VM is capable of:
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush mmx fxsr sse sse2 syscall nx rdtscp lm constant_tsc up
rep_good nopl pni monitor ssse3 lahf_lm
-leif
give (especially the '-m...' parameters passed to 'cc1'), in
contrast to
what MPIR used (shown by 'configure', the lines after 'using
ABI="...") ?
By the way, Ubuntu's "GCC 4.6" is presumably Ubuntu's GCC 4.6.3, which
is quite different to FSF's GCC 4.6.3, i.e., besides some other changes
contains parts from the 4.7 branch.
This is a very good point. I didn't notice.
-- Elias
--
() The ASCII Ribbon Campaign
/\ Help Cure HTML E-Mail
--
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.
