Ben Vanik
Ben Vanik
That's believable - the LLVMGPU lowerings have always been sketchy. Today I strongly suspect any difference between the dispatch and the dispatchee will not work correctly. It should, but that...
that's odd as SPIR-V should definitely allow that since we're using Vulkan/SPIR-V's binding model - if there's one codegen target I'd expect to work it'd be that one :P
the cost should only be in execution time as the same vmfbs are used - I thought we weren't bounded by execution time? having them nightly/on releases would still be...
and yeah, we should probably always have debug info on the bots - it has zero runtime overhead but does introduce a text printing step to compilation - probably worth...
The other thing we can do is make it opt-in for particular benchmarks. The major benefit to me is looking for improvement/regressions over time that show up in end-to-end performance,...
yeah I don't think we care about the pixel phones anymore - I am mostly concerned with losing the ability to look at what fusion decisions/whole architecture decisions we are...
opt-in on PR via a label and then continuously run on main SGTM
I do occasionally, but am fine with telling people that if they want me to investigate a performance issue they need to get me a trace so 🤷
(all for having easier tracy bumps!)
try --trace_execution?