Re: Saturated operations
Michael Hope wrote on 28.02.2012 00:04:17: > Hi Ulrich. The saturated work you're doing at the moment - is it the > saturated add/subtract QADD/QSUB/QDADD/QDSUB or the ARMv6 > saturate-word-with-shift SSAT/USAT? Hi Michael, it's the latter (SSAT/USAT). > I've updated the old blueprint and added a new one for SSAT/USAT: > https://blueprints.launchpad.net/gcc-linaro/+spec/armv6-saturation- > instructions > https://blueprints.launchpad.net/gcc-linaro/+spec/armv5-saturated-ops > > and was wondering which to put against your name. I've added myself to the armv6-saturation-instructions blueprint and updated its status. B.t.w. note that there are also USAT16/SSAT16 (which operate on vectors of two 16-bit integers). My current patch does *not* address those; I'd suggest this to be done in a follow-on blueprint (if at all), possibly in conjunction with other core-register vector operations. Mit freundlichen Gruessen / Best Regards Ulrich Weigand -- Dr. Ulrich Weigand | Phone: +49-7031/16-3727 STSM, GNU compiler and toolchain for Linux on System z and Cell/B.E. IBM Deutschland Research & Development GmbH Vorsitzende des Aufsichtsrats: Martina Koederitz | Geschäftsführung: Dirk Wittkopp Sitz der Gesellschaft: Böblingen | Registergericht: Amtsgericht Stuttgart, HRB 243294 ___ linaro-toolchain mailing list linaro-toolchain@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-toolchain
Spec 2000 broken?
Hi Michael, I ran the benchmark tests on GCC 4.7, but the spec2000 run seems to have failed. This is both on the baseline, and on my patch. See here: http://ex.seabright.co.nz/benchmarks/gcc-linaro-4.7%2bbzr114968~ams-codesourcery~arm-64-bit-shifts-4.7/logs/armv7l-natty-cbuild254-tcpanda06-cortexa9r1/ Andrew ___ linaro-toolchain mailing list linaro-toolchain@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-toolchain
Re: Spec 2000 broken?
On Wed, Feb 29, 2012 at 3:04 AM, Andrew Stubbs wrote: > Hi Michael, > > I ran the benchmark tests on GCC 4.7, but the spec2000 run seems to have > failed. This is both on the baseline, and on my patch. > > See here: > http://ex.seabright.co.nz/benchmarks/gcc-linaro-4.7%2bbzr114968~ams-codesourcery~arm-64-bit-shifts-4.7/logs/armv7l-natty-cbuild254-tcpanda06-cortexa9r1/ I noticed similar yesterday. SPEC 'relocate' uses the build directory in a regex which fails when it hits the '+' and other random characters we use in the path name. Once past that, it hits the 80 character interpreter path limit and dies. I've fixed both. Please resubmit any SPEC jobs. -- Michael ___ linaro-toolchain mailing list linaro-toolchain@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-toolchain
Re: Spec 2000 broken?
On Wed, Feb 29, 2012 at 8:21 AM, Michael Hope wrote: > On Wed, Feb 29, 2012 at 3:04 AM, Andrew Stubbs > wrote: >> Hi Michael, >> >> I ran the benchmark tests on GCC 4.7, but the spec2000 run seems to have >> failed. This is both on the baseline, and on my patch. >> >> See here: >> http://ex.seabright.co.nz/benchmarks/gcc-linaro-4.7%2bbzr114968~ams-codesourcery~arm-64-bit-shifts-4.7/logs/armv7l-natty-cbuild254-tcpanda06-cortexa9r1/ > > I noticed similar yesterday. SPEC 'relocate' uses the build directory > in a regex which fails when it hits the '+' and other random > characters we use in the path name. Once past that, it hits the 80 > character interpreter path limit and dies. > > I've fixed both. Please resubmit any SPEC jobs. I've spawned spec2000 runs of gcc-linaro-4.7+bzr114968~ams-codesourcery~arm-64-bit-shifts-4.7 and the ancestor gcc-linaro-4.7+bzr114965. -- Michael ___ linaro-toolchain mailing list linaro-toolchain@lists.linaro.org http://lists.linaro.org/mailman/listinfo/linaro-toolchain
