summaryrefslogtreecommitdiffstats
path: root/llvm/lib/Bitcode
diff options
context:
space:
mode:
authorJustin Lebar <jlebar@google.com>2016-03-14 23:15:34 +0000
committerJustin Lebar <jlebar@google.com>2016-03-14 23:15:34 +0000
commit6827de19b2aedfa8679aa85e74de0989d4559d7c (patch)
tree322c59afe8f60e30acc6336b5947302a2778f90c /llvm/lib/Bitcode
parentfa99425667e4e12f453ed791f1638154a9195609 (diff)
downloadbcm5719-llvm-6827de19b2aedfa8679aa85e74de0989d4559d7c.tar.gz
bcm5719-llvm-6827de19b2aedfa8679aa85e74de0989d4559d7c.zip
[LoopUnroll] Respect the convergent attribute.
Summary: Specifically, when we perform runtime loop unrolling of a loop that contains a convergent op, we can only unroll k times, where k divides the loop trip multiple. Without this change, we'll happily unroll e.g. the following loop for (int i = 0; i < N; ++i) { if (i == 0) convergent_op(); foo(); } into int i = 0; if (N % 2 == 1) { convergent_op(); foo(); ++i; } for (; i < N - 1; i += 2) { if (i == 0) convergent_op(); foo(); foo(); }. This is unsafe, because we've just added a control-flow dependency to the convergent op in the prelude. In general, runtime unrolling loops that contain convergent ops is safe only if we don't have emit a prelude, which occurs when the unroll count divides the trip multiple. Reviewers: resistor Subscribers: llvm-commits, mzolotukhin Differential Revision: http://reviews.llvm.org/D17526 llvm-svn: 263509
Diffstat (limited to 'llvm/lib/Bitcode')
0 files changed, 0 insertions, 0 deletions
OpenPOWER on IntegriCloud