search for: vaddsd

This conflict is with many optimizations incl. copy prop, coalescing, hoisting etc. Each could increase register pressure and with similar impact. Attempts to control the register pressure locally (within an optimization pass) tend to get hard to tune and maintain. Would it be a better way to describe eg in metadata how to undo an optimization? Optimizations that attempt to reduce pressure like

....S -fslp-vectorize-aggressive and loop body looks like: .LBB1_2: # %for.body # =>This Inner Loop Header: Depth=1 cltq vmovsd (%rsi,%rax,8), %xmm0 movq %r9, %r10 sarq $32, %r10 vaddsd (%rdi,%r10,8), %xmm0, %xmm0 vmovsd %xmm0, (%rdi,%r10,8) addq %r8, %r9 addl %ecx, %eax decl %edx jne .LBB1_2 so vector instructions for scalars (vaddsd, vmovsd) were used in the loop and no real gather/scatter emitted. The question is why this...

...> .LBB1_2: # %for.body >> # =>This Inner Loop Header: Depth=1 >> cltq >> vmovsd (%rsi,%rax,8), %xmm0 >> movq %r9, %r10 >> sarq $32, %r10 >> vaddsd (%rdi,%r10,8), %xmm0, %xmm0 >> vmovsd %xmm0, (%rdi,%r10,8) >> addq %r8, %r9 >> addl %ecx, %eax >> decl %edx >> jne .LBB1_2 >> >> so vector instructions for scalars (vaddsd, vmovsd) were used in the loo...

Ah, bad example. This is a general problem for all (maybe most) SSE and AVX SS/SD patterns though, which is why I mentioned Sandybridge. You can swap out VFMADDSD in my example for VADDSD or whatever you like. I have a lion's share of such a change implemented already and performance is greatly affected. If the community is interested in this change, I would be happy to prepare a patch. -Cameron On Thu, Jul 26, 2012 at 2:27 PM, Jan Sjodin <jan_sjodin at yahoo.com> wrote...

...to make that vmovsd has the same stats as well. Michael On Jul 26, 2012, at 11:46 AM, Cameron McInally wrote: > Ah, bad example. This is a general problem for all (maybe most) SSE and AVX SS/SD patterns though, which is why I mentioned Sandybridge. You can swap out VFMADDSD in my example for VADDSD or whatever you like. > > I have a lion's share of such a change implemented already and performance is greatly affected. If the community is interested in this change, I would be happy to prepare a patch. > > -Cameron > > On Thu, Jul 26, 2012 at 2:27 PM, Jan Sjodin <jan...

We're migrating to LLVM 3.1 and trying to use the upstream FMA patterns. Why is VFMADDSD4 defined with vector types? Is this simply because the gcc intrinsic uses vector types? It's quite unnatural if you have a compiler that generates FMAs as opposed to requiring user intrinsics. -Dave

...e stats as well. > > Michael > > On Jul 26, 2012, at 11:46 AM, Cameron McInally wrote: > > Ah, bad example. This is a general problem for all (maybe most) SSE and > AVX SS/SD patterns though, which is why I mentioned Sandybridge. You can > swap out VFMADDSD in my example for VADDSD or whatever you like. > > I have a lion's share of such a change implemented already and performance > is greatly affected. If the community is interested in this change, I would > be happy to prepare a patch. > > -Cameron > > On Thu, Jul 26, 2012 at 2:27 PM, Jan Sjodin...

...same stats as well. > > Michael > > On Jul 26, 2012, at 11:46 AM, Cameron McInally wrote: > >> Ah, bad example. This is a general problem for all (maybe most) SSE and AVX SS/SD patterns though, which is why I mentioned Sandybridge. You can swap out VFMADDSD in my example for VADDSD or whatever you like. >> >> I have a lion's share of such a change implemented already and performance is greatly affected. If the community is interested in this change, I would be happy to prepare a patch. >> >> -Cameron >> >> On Thu, Jul 26, 2012 at 2:2...