Dehao Chen via llvm-dev
2017-May-18 22:30 UTC
[llvm-dev] Enable vectorizer-maximize-bandwidth by default?
Hi,
I'm proposing to make vectorizer-maximize-bandwidth on by default for loop
vectorizer because it should generally help performance.
I've tested the performance impact on Intel sandybridge machine with
speccpu benchmarks:
Benchmark Base:Reference (1)
-------------------------------------------------------
spec/2006/fp/C++/444.namd 26.84 -0.31%
spec/2006/fp/C++/447.dealII 46.19 +0.89%
spec/2006/fp/C++/450.soplex 42.92 -0.44%
spec/2006/fp/C++/453.povray 38.57 -2.25%
spec/2006/fp/C/433.milc 24.54 -0.76%
spec/2006/fp/C/470.lbm 41.08 +0.26%
spec/2006/fp/C/482.sphinx3 47.58 -0.99%
spec/2006/int/C++/471.omnetpp 22.06 +1.87%
spec/2006/int/C++/473.astar 22.65 -0.12%
spec/2006/int/C++/483.xalancbmk 33.69 +4.97%
spec/2006/int/C/400.perlbench 33.43 +1.70%
spec/2006/int/C/401.bzip2 23.02 -0.19%
spec/2006/int/C/403.gcc 32.57 -0.43%
spec/2006/int/C/429.mcf 40.35 +0.27%
spec/2006/int/C/445.gobmk 26.96 +0.06%
spec/2006/int/C/456.hmmer 24.4 +0.19%
spec/2006/int/C/458.sjeng 27.91 -0.08%
spec/2006/int/C/462.libquantum 57.47 -0.20%
spec/2006/int/C/464.h264ref 46.52 +1.35%
geometric mean +0.29%
Scores are benchmark specific.
We do have regression on 453.povray, but it's due to secondary effects as
all hot functions are the same. I've also tested the code size impact, it
does not change for tested speccpu benchmarks.
I've prepared https://reviews.llvm.org/D33341 to do this.
I really appreciate if the community can help test the performance impact
of this change on other architectures so that we can decide if this should
go target-dependent.
Any comments/concerns?
Thanks,
Dehao
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20170518/0a56b7de/attachment.html>
陳韋任 via llvm-dev
2017-May-19 13:18 UTC
[llvm-dev] Enable vectorizer-maximize-bandwidth by default?
Besides speccpu, any other real-world applications benefit from this option? Regards, chenwj 2017-05-19 6:30 GMT+08:00 Dehao Chen via llvm-dev <llvm-dev at lists.llvm.org>:> Hi, > > I'm proposing to make vectorizer-maximize-bandwidth on by default for > loop vectorizer because it should generally help performance. > > I've tested the performance impact on Intel sandybridge machine with > speccpu benchmarks: > > Benchmark Base:Reference (1) > ------------------------------------------------------- > spec/2006/fp/C++/444.namd 26.84 -0.31% > spec/2006/fp/C++/447.dealII 46.19 +0.89% > spec/2006/fp/C++/450.soplex 42.92 -0.44% > spec/2006/fp/C++/453.povray 38.57 -2.25% > spec/2006/fp/C/433.milc 24.54 -0.76% > spec/2006/fp/C/470.lbm 41.08 +0.26% > spec/2006/fp/C/482.sphinx3 47.58 -0.99% > spec/2006/int/C++/471.omnetpp 22.06 +1.87% > spec/2006/int/C++/473.astar 22.65 -0.12% > spec/2006/int/C++/483.xalancbmk 33.69 +4.97% > spec/2006/int/C/400.perlbench 33.43 +1.70% > spec/2006/int/C/401.bzip2 23.02 -0.19% > spec/2006/int/C/403.gcc 32.57 -0.43% > spec/2006/int/C/429.mcf 40.35 +0.27% > spec/2006/int/C/445.gobmk 26.96 +0.06% > spec/2006/int/C/456.hmmer 24.4 +0.19% > spec/2006/int/C/458.sjeng 27.91 -0.08% > spec/2006/int/C/462.libquantum 57.47 -0.20% > spec/2006/int/C/464.h264ref 46.52 +1.35% > > geometric mean +0.29% > > Scores are benchmark specific. > > We do have regression on 453.povray, but it's due to secondary effects as > all hot functions are the same. I've also tested the code size impact, it > does not change for tested speccpu benchmarks. > > I've prepared https://reviews.llvm.org/D33341 to do this. > > I really appreciate if the community can help test the performance impact > of this change on other architectures so that we can decide if this should > go target-dependent. > > Any comments/concerns? > > Thanks, > Dehao > > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev > >-- Wei-Ren Chen (陳韋任) Homepage: https://people.cs.nctu.edu.tw/~chenwj -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170519/d2d66943/attachment.html>
Matthew Simpson via llvm-dev
2017-May-19 15:56 UTC
[llvm-dev] Enable vectorizer-maximize-bandwidth by default?
This sounds good to me. Enabling this by default has been mentioned a few times already. I've tested this feature in the past on AArch64 (Kryo and Falkor) and found it to be beneficial for mixed-type loops. Thanks! On Thu, May 18, 2017 at 6:30 PM, Dehao Chen via llvm-dev < llvm-dev at lists.llvm.org> wrote:> Hi, > > I'm proposing to make vectorizer-maximize-bandwidth on by default for > loop vectorizer because it should generally help performance. > > I've tested the performance impact on Intel sandybridge machine with > speccpu benchmarks: > > Benchmark Base:Reference (1) > ------------------------------------------------------- > spec/2006/fp/C++/444.namd 26.84 -0.31% > spec/2006/fp/C++/447.dealII 46.19 +0.89% > spec/2006/fp/C++/450.soplex 42.92 -0.44% > spec/2006/fp/C++/453.povray 38.57 -2.25% > spec/2006/fp/C/433.milc 24.54 -0.76% > spec/2006/fp/C/470.lbm 41.08 +0.26% > spec/2006/fp/C/482.sphinx3 47.58 -0.99% > spec/2006/int/C++/471.omnetpp 22.06 +1.87% > spec/2006/int/C++/473.astar 22.65 -0.12% > spec/2006/int/C++/483.xalancbmk 33.69 +4.97% > spec/2006/int/C/400.perlbench 33.43 +1.70% > spec/2006/int/C/401.bzip2 23.02 -0.19% > spec/2006/int/C/403.gcc 32.57 -0.43% > spec/2006/int/C/429.mcf 40.35 +0.27% > spec/2006/int/C/445.gobmk 26.96 +0.06% > spec/2006/int/C/456.hmmer 24.4 +0.19% > spec/2006/int/C/458.sjeng 27.91 -0.08% > spec/2006/int/C/462.libquantum 57.47 -0.20% > spec/2006/int/C/464.h264ref 46.52 +1.35% > > geometric mean +0.29% > > Scores are benchmark specific. > > We do have regression on 453.povray, but it's due to secondary effects as > all hot functions are the same. I've also tested the code size impact, it > does not change for tested speccpu benchmarks. > > I've prepared https://reviews.llvm.org/D33341 to do this. > > I really appreciate if the community can help test the performance impact > of this change on other architectures so that we can decide if this should > go target-dependent. > > Any comments/concerns? > > Thanks, > Dehao > > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev > >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170519/942fa225/attachment.html>
Dehao Chen via llvm-dev
2017-May-19 17:35 UTC
[llvm-dev] Enable vectorizer-maximize-bandwidth by default?
Yes, we do see performance benefits for this change on some google internal benchmarks. Dehao On Fri, May 19, 2017 at 6:18 AM, 陳韋任 <chenwj.cs97g at g2.nctu.edu.tw> wrote:> Besides speccpu, any other real-world applications benefit from this > option? > > Regards, > chenwj > > 2017-05-19 6:30 GMT+08:00 Dehao Chen via llvm-dev <llvm-dev at lists.llvm.org > >: > >> Hi, >> >> I'm proposing to make vectorizer-maximize-bandwidth on by default for >> loop vectorizer because it should generally help performance. >> >> I've tested the performance impact on Intel sandybridge machine with >> speccpu benchmarks: >> >> Benchmark Base:Reference (1) >> ------------------------------------------------------- >> spec/2006/fp/C++/444.namd 26.84 -0.31% >> spec/2006/fp/C++/447.dealII 46.19 +0.89% >> spec/2006/fp/C++/450.soplex 42.92 -0.44% >> spec/2006/fp/C++/453.povray 38.57 -2.25% >> spec/2006/fp/C/433.milc 24.54 -0.76% >> spec/2006/fp/C/470.lbm 41.08 +0.26% >> spec/2006/fp/C/482.sphinx3 47.58 -0.99% >> spec/2006/int/C++/471.omnetpp 22.06 +1.87% >> spec/2006/int/C++/473.astar 22.65 -0.12% >> spec/2006/int/C++/483.xalancbmk 33.69 +4.97% >> spec/2006/int/C/400.perlbench 33.43 +1.70% >> spec/2006/int/C/401.bzip2 23.02 -0.19% >> spec/2006/int/C/403.gcc 32.57 -0.43% >> spec/2006/int/C/429.mcf 40.35 +0.27% >> spec/2006/int/C/445.gobmk 26.96 +0.06% >> spec/2006/int/C/456.hmmer 24.4 +0.19% >> spec/2006/int/C/458.sjeng 27.91 -0.08% >> spec/2006/int/C/462.libquantum 57.47 -0.20% >> spec/2006/int/C/464.h264ref 46.52 +1.35% >> >> geometric mean +0.29% >> >> Scores are benchmark specific. >> >> We do have regression on 453.povray, but it's due to secondary effects as >> all hot functions are the same. I've also tested the code size impact, it >> does not change for tested speccpu benchmarks. >> >> I've prepared https://reviews.llvm.org/D33341 to do this. >> >> I really appreciate if the community can help test the performance impact >> of this change on other architectures so that we can decide if this should >> go target-dependent. >> >> Any comments/concerns? >> >> Thanks, >> Dehao >> >> _______________________________________________ >> LLVM Developers mailing list >> llvm-dev at lists.llvm.org >> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev >> >> > > > -- > Wei-Ren Chen (陳韋任) > Homepage: https://people.cs.nctu.edu.tw/~chenwj >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170519/38f2bda1/attachment.html>
Adam Nemet via llvm-dev
2017-May-19 23:01 UTC
[llvm-dev] Enable vectorizer-maximize-bandwidth by default?
> On May 18, 2017, at 3:30 PM, Dehao Chen via llvm-dev <llvm-dev at lists.llvm.org> wrote: > > Hi, > > I'm proposing to make vectorizer-maximize-bandwidth on by default for loop vectorizer because it should generally help performance. > > I've tested the performance impact on Intel sandybridge machine with speccpu benchmarks: > > Benchmark Base:Reference (1) > ------------------------------------------------------- > spec/2006/fp/C++/444.namd 26.84 -0.31% > spec/2006/fp/C++/447.dealII 46.19 +0.89% > spec/2006/fp/C++/450.soplex 42.92 -0.44% > spec/2006/fp/C++/453.povray 38.57 -2.25% > spec/2006/fp/C/433.milc 24.54 -0.76% > spec/2006/fp/C/470.lbm 41.08 +0.26% > spec/2006/fp/C/482.sphinx3 47.58 -0.99% > spec/2006/int/C++/471.omnetpp 22.06 +1.87% > spec/2006/int/C++/473.astar 22.65 -0.12% > spec/2006/int/C++/483.xalancbmk 33.69 +4.97% > spec/2006/int/C/400.perlbench 33.43 +1.70% > spec/2006/int/C/401.bzip2 23.02 -0.19% > spec/2006/int/C/403.gcc 32.57 -0.43% > spec/2006/int/C/429.mcf 40.35 +0.27% > spec/2006/int/C/445.gobmk 26.96 +0.06% > spec/2006/int/C/456.hmmer 24.4 +0.19% > spec/2006/int/C/458.sjeng 27.91 -0.08% > spec/2006/int/C/462.libquantum 57.47 -0.20% > spec/2006/int/C/464.h264ref 46.52 +1.35% > > geometric mean +0.29% > > Scores are benchmark specific. > > We do have regression on 453.povray, but it's due to secondary effects as all hot functions are the same. I've also tested the code size impact, it does not change for tested speccpu benchmarks.Can you please describe the config for the runs (optimization level, PGO/no-PGO, etc). It would be good to provide analysis for the changes >1%. I.e. we need to make sure that the improvements are not noise either ;).> > I've prepared https://reviews.llvm.org/D33341 <https://reviews.llvm.org/D33341> to do this. > > I really appreciate if the community can help test the performance impact of this change on other architectures so that we can decide if this should go target-dependent.I will run it on Cyclone/AArch64 next week. Adam> > Any comments/concerns? > > Thanks, > Dehao > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170519/74e7eb11/attachment-0001.html>
Dehao Chen via llvm-dev
2017-May-22 16:57 UTC
[llvm-dev] Enable vectorizer-maximize-bandwidth by default?
On Fri, May 19, 2017 at 4:01 PM, Adam Nemet <anemet at apple.com> wrote:> > On May 18, 2017, at 3:30 PM, Dehao Chen via llvm-dev < > llvm-dev at lists.llvm.org> wrote: > > Hi, > > I'm proposing to make vectorizer-maximize-bandwidth on by default for > loop vectorizer because it should generally help performance. > > I've tested the performance impact on Intel sandybridge machine with > speccpu benchmarks: > > Benchmark Base:Reference (1) > ------------------------------------------------------- > spec/2006/fp/C++/444.namd 26.84 -0.31% > spec/2006/fp/C++/447.dealII 46.19 +0.89% > spec/2006/fp/C++/450.soplex 42.92 -0.44% > spec/2006/fp/C++/453.povray 38.57 -2.25% > spec/2006/fp/C/433.milc 24.54 -0.76% > spec/2006/fp/C/470.lbm 41.08 +0.26% > spec/2006/fp/C/482.sphinx3 47.58 -0.99% > spec/2006/int/C++/471.omnetpp 22.06 +1.87% > spec/2006/int/C++/473.astar 22.65 -0.12% > spec/2006/int/C++/483.xalancbmk 33.69 +4.97% > spec/2006/int/C/400.perlbench 33.43 +1.70% > spec/2006/int/C/401.bzip2 23.02 -0.19% > spec/2006/int/C/403.gcc 32.57 -0.43% > spec/2006/int/C/429.mcf 40.35 +0.27% > spec/2006/int/C/445.gobmk 26.96 +0.06% > spec/2006/int/C/456.hmmer 24.4 +0.19% > spec/2006/int/C/458.sjeng 27.91 -0.08% > spec/2006/int/C/462.libquantum 57.47 -0.20% > spec/2006/int/C/464.h264ref 46.52 +1.35% > > geometric mean +0.29% > > Scores are benchmark specific. > > We do have regression on 453.povray, but it's due to secondary effects as > all hot functions are the same. I've also tested the code size impact, it > does not change for tested speccpu benchmarks. > > > Can you please describe the config for the runs (optimization level, > PGO/no-PGO, etc). >This is O2 build without PGO.> > It would be good to provide analysis for the changes >1%. I.e. we need to > make sure that the improvements are not noise either ;). >Good point. I just examined all benchmarks with >1% "improvement". Turns out they are all noises: the hot functions (with >1% total cycles) are all identical. So the conclusion is: this change does not affect speccpu2006 performance. Thanks, Dehao> > > I've prepared https://reviews.llvm.org/D33341 to do this. > > I really appreciate if the community can help test the performance impact > of this change on other architectures so that we can decide if this should > go target-dependent. > > > I will run it on Cyclone/AArch64 next week. > > Adam > > > Any comments/concerns? > > Thanks, > Dehao > _______________________________________________ > LLVM Developers mailing list > llvm-dev at lists.llvm.org > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev > > >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170522/8f4148fa/attachment.html>
Chandler Carruth via llvm-dev
2017-May-30 07:58 UTC
[llvm-dev] Enable vectorizer-maximize-bandwidth by default?
On Fri, May 19, 2017 at 4:01 PM Adam Nemet via llvm-dev < llvm-dev at lists.llvm.org> wrote:> I will run it on Cyclone/AArch64 next week. >FYI, we're still waiting on these Adam... -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20170530/7cb390ca/attachment.html>
Apparently Analagous Threads
- Debug info interacting with optimization and code generation
- (RFC) Adjusting default loop fully unroll threshold
- (RFC) Encoding code duplication factor in discriminator
- (RFC) Adjusting default loop fully unroll threshold
- (RFC) Encoding code duplication factor in discriminator