<div dir="ltr"><br><div class="gmail_extra"><br><br><div class="gmail_quote"><br>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div class="gmail_extra"><div class="gmail_quote"><div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote"><div><div class="gmail_extra">Test case name : llvm/projects/test-suite/SingleSource/Benchmarks/Misc/matmul_f64_4x4.c - This is a 4x4 matrix multiplication, we can make small changes to make it a 3x3 matrix multiplication for making things simple to understand .<br>
</div></div></div></div></div></blockquote><div><br></div></div><div>This is one very specific case. How does that behave on all other cases? Normally, every big improvement comes with a cost, and if you only look at the benchmark you're tuning to, you'll never see it. It may be that the cost is small and that we decide to pay the price, but not until we know that the cost is.</div>
<div>
<div><br></div></div></div></div></div></blockquote><div><br><div>I agree that we should approach in whole than in bits and pieces. I was basically comparing performance of clang and gcc
code for benchmarks listed in llvm trunk. I found that wherever there
was floating point ops (specifically floating point multiplication),
performance with clang was bad. On analyzing further those issues, i
came across vmla instruction by gcc. The test cases hit by bad
performance of clang are :<br><br></div><div>Test
Case
No of vmla instructions emitted by gcc (clang does not emit vmla for
cortex-a8)<br>===========
=======================================================<br></div><br>llvm/projects/test-suite/SingleSource/Benchmarks/Misc-C++/Large/sphereflake 55 <br><br>llvm/projects/test-suite/SingleSource/Benchmarks/Misc-C++/Large/ray.cpp 40 <br>
<br>llvm/projects/test-suite/SingleSource/Benchmarks/Misc/ffbench.c 8<br><br>llvm/projects/test-suite/SingleSource/Benchmarks/Misc/matmul_f64_4x4.c 18<br>
<br>llvm/projects/test-suite/SingleSource/Benchmarks/BenchmarkGame/n-body.c 36 <br><br></div><div>With vmul+vadd instruction pair comes extra overhead of load/store ops, as seen in assembly generated. With -mcpu=cortex-a15 option clang performs better, as it emits vmla instructions. <br>
</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div class="gmail_extra"><div class="gmail_quote"><div><div>
</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div class="gmail_extra"><div class="gmail_quote"><div><div class="gmail_extra">
</div><div class="gmail_extra">This was tested on real hardware. Time taken for a 4x4 matrix multiplication:<br></div></div></div></div></div></blockquote><div><br></div></div><div>What hardware? A7? A8? A9? A15?</div></div>
</div></div></blockquote><div><br></div><div>I tested it on A15, i don't have access to A8 rightnow, but i intend to test it for A8 as well. I compiled the code for A8 and as it was working fine on A15 without any crash, i went ahead with cortex-a8 option. I don't think i will get A8 hardware soon, can someone please check it on A8 hardware as well (Sorry for the trouble)? <br>
</div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div class="gmail_extra"><div class="gmail_quote"><div><div>
<br>
</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div dir="ltr"><div class="gmail_extra"><div class="gmail_quote"><div><div class="gmail_extra">
Also, as stated by Renato - "there is a pipeline stall between two sequential VMLAs (possibly due to
the need of re-use of some registers) and this made code much slower
than a sequence of VMLA+VMUL+VADD" , when i use -mcpu=cortex-a15 as option, clang emits vmla instructions back to back(sequential) . Is there something different with cortex-a15 regarding pipeline stalls, that we are ignoring back to back vmla hazards?<br>
</div></div></div></div></div></blockquote><div></div></div></div><br></div><div class="gmail_extra">A8 and A15 are quite different beasts. I haven't read about this hazard in the A15 manual, so I suspect that they have fixed whatever was causing the stall.</div>
</div></blockquote><div><br></div><div>Ok. I couldn't find reference for this. If the pipeline stall issue was fixed in cortex-a15 then LLVM developers will definitely know about this (and hence we are emitting vmla for cortex-a15). I couldn't find any comment related to this in the code. Can someone please point it out? Also, I will be glad to know the code place where we start differentiating between cortex-a8 and cortex-a15 for code generation. <br>
</div></div><br><br clear="all"><br>-- <br>With regards,<br>Suyog Sarda<br>
</div></div>