Uday Bondhugula
Uday Bondhugula
Same issue with heat-1dp, heat-3dp
This is the transformation found for 'tiled'. ../../polycc ssymm.c --noparallel --codegen-context=100 -o ssymm.tiled.c [pluto] compute_deps (isl) [pluto] Number of statements: 3 [pluto] Total number of loops: 8 [pluto] Number of...
The output of 'par' and 'tiled' match, but they both differ from that of 'orig'.
This is a precision related issue likely associated with value unsafe optimizations. The error goes away with -O1. And even at -O3, the difference is for a small number of...
Sorry about the extreme delay in responding to this. The 'relocation truncated to fit' errors are due to excessive global variable storage sizes! You'll have to reduce the problem size...
Thanks for reporting this. Sorry about the slow response - I'll be able to look at this by this weekend. Pluto has no support for detecting private / local vars....
Can I know which version of Pluto you are using? $ ./polycc -v
Sure. I'm open to this, but I myself have no experience with maintaining a cmake build system nor the time to get started on this until Mar 2019. Note that...
Your loops only include the ones traversing the data space. You don't plan to model the time loop as well? In the code pasted above, you have a reduction here...
Sorry about the delay - had missed this. Would reducing the number of make jobs help? You can lose the "-j4" in the line below. Also, at which particular step...