Change Details

Tries to perform (Xshr (add (i2^n X), (i2^n Y)), 2^(n-1)) -> (llvm.uadd.with.overflow (trunc X), (trunc Y)).overflow where - Xshr can be ashr or lshr - X and Y have at least 2^(n-1) leading zeroes. This seems to be a pattern that just comes from OpenCL front-ends, so adding DAG/GISel combines doesn't seem to be worth the complexity. Original patch D107552 by @abinavpp See this issue https://github.com/RadeonOpenCompute/ROCm/issues/488

Tries to perform (Xshr (add (i2^n X), (i2^n Y)), 2^(n-1)) -> (icmp ult (add (trunc X), (trunc Y)), X) where - Only the K leading bits of X and Y can be non-zero. - The add is only used by the shr, or by iK (or narrower) truncates. - The lshr type has more than 2 bits (other types are boolean math). - X > 1 This seems to be a pattern that just comes from OpenCL front-ends, so adding DAG/GISel combines doesn't seem to be worth the complexity. Original patch D107552 by @abinavpp - adapted to use (a + b < a) instead of uaddo following discussion on the review. See this issue https://github.com/RadeonOpenCompute/ROCm/issues/488

Tries to perform (Xshr (add (i2^n X), (i2^n Y)), 2^(n-1)) -> (icmp ult (add (trunc X), (trunc Y)), X) where -> (llvm.uadd.with.overflow (trunc X), (trunc Y)).overflow- Only the K leading bits of X and Y can be non-zero. where- The add is only used by the shr, or by iK (or narrower) truncates. - Xshr can be ashr or lshr- The lshr type has more than 2 bits (other types are boolean math). - X and Y have at least 2^(n-1) leading zeroes.> 1 This seems to be a pattern that just comes from OpenCL front-ends, so adding DAG/GISel combines doesn't seem to be worth the complexity. Original patch D107552 by @abinavpp - adapted to use (a + b < a) instead of uaddo following discussion on the review. See this issue https://github.com/RadeonOpenCompute/ROCm/issues/488