Displaying 3 results from an estimated 3 matches for "rl5".
Did you mean:
rc5
2007 Jun 18
3
Centos 5 - Setting up yum for ATrpms
I think I found one of my setup problems. I followed instructions at: http://atrpms.net/install.html
And now I see that the information shown there to put into yum.conf (that I put into yum.repo.d/atrpms.repo) is only for FC, not for RL5 (thus Centos 5).
What do I use in my atrmps.repo to get it to access the RL5 directories?
2014 Apr 19
4
[LLVMdev] [NVPTX] Eliminate common sub-expressions in a group of similar GEPs
...address
of
each GEP, wasting tons of registers. e.g., it emits the following PTX for
the
first load and similar PTX for other loads.
mov.u32 %r1, %tid.x;
mov.u32 %r2, %tid.y;
mul.wide.u32 %rl2, %r1, 128;
mov.u64 %rl3, a;
add.s64 %rl4, %rl3, %rl2;
mul.wide.u32 %rl5, %r2, 4;
add.s64 %rl6, %rl4, %rl5;
ld.shared.f32 %f1, [%rl6];
The resultant register pressure causes up to 20% slowdown on some of our
benchmarks.
To reduce register pressure, the optimization implemented in this patch
merges
the common subexpression in a group of GEPs, saving many regi...
2014 Apr 21
2
[LLVMdev] [NVPTX] Eliminate common sub-expressions in a group of similar GEPs
...s the following PTX
> > for the
> > first load and similar PTX for other loads.
> >
> >
> > mov.u32 %r1, %tid.x;
> > mov.u32 %r2, %tid.y;
> > mul.wide.u32 %rl2, %r1, 128;
> > mov.u64 %rl3, a;
> > add.s64 %rl4, %rl3, %rl2;
> > mul.wide.u32 %rl5, %r2, 4;
> > add.s64 %rl6, %rl4, %rl5;
> > ld.shared.f32 %f1, [%rl6];
> >
> >
> > The resultant register pressure causes up to 20% slowdown on some of
> > our
> > benchmarks.
> >
> >
> > To reduce register pressure, the optimization implem...