LLVM Machine Instruction: Convergent attribute

ref: http://lists.llvm.org/pipermail/llvm-dev/2015-August/089241.html



1. Convergent attribute is useful for SIMT/SPMD programming model.

2. Intended interpretation is that a convergent operation cannot be move either into or out of a conditionally executed region.

3. If you have a convergent instruction A, it islegal to duplicate it to instruction B if (assuming B is after A in program flow) A dominates B and B post-dominates A.

case:

 r1 = texture2D(..., r0, ...)
 if (...) {
   // r0 used as temporary here
   r0 = ...
   r2 = r0 + ...
 } else {
   // only use of r1
   r2 = r1 + ...
}

In this example, various optimizations might try to sink the texture2D operation
into the else block, like so:

if (...) {
  r0 = ...
  r2 = r0 + ...
} else {
  r1 = texture2D(..., r0, ...)
  r2 = r1 + ...
}

In most SPMD/SIMT implementations, the fallout of this races is exposed via
the predicated expression of acyclic control flow:

pred0 <- cmp ...
if (pred0)  r0 = ...                            // thread 1
if (pred0)  r2 = r0 + ...                       // thread 1
if (!pred0) r1 = texture2D(..., r0, ...)        // thread 0 
if (!pred0) r2 = r1 + ...                       // thread 0

If thread 0 takes the else path and perform the texture2D operation, but
its neighbor thread 1 takes the then branch, then the texture2D will fail
because thread 1 has already overwritten its value of r0 before thread 0 has
a chance to read it.