diff options
| author |  Mark Heffernan <meheff@google.com> | 2018-06-25 23:05:07 -0700 |
|---|---|---|
| committer |  TensorFlower Gardener <gardener@tensorflow.org> | 2018-06-25 23:08:42 -0700 |
| commit | 22c06a62b7dcc75f6f717cea63a947b21805afda (patch) | |
| tree | 020fd0b07dad4994217170ba13624f15a3709feb /tensorflow/compiler/xla/service/hlo_evaluator.h | |
| parent | 5968eb87766d5536feefaaf8a9ee2f30aeadb354 (diff) | |
Rename HLO opcode kGenerateToken to kAfterAll.
Long term I think we want to require kAfterAll to take at least one token as operand so it cannot generate a token out of thin air, so kGenerateToken is no longer an appropriate name. Instead, a primordial token would be supplied some how in the entry computation, perhaps as a parameter, and then threaded to any side-effecting ops.
NFC.
PiperOrigin-RevId: 202079040
Diffstat (limited to 'tensorflow/compiler/xla/service/hlo_evaluator.h')
| -rw-r--r-- | tensorflow/compiler/xla/service/hlo_evaluator.h | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/tensorflow/compiler/xla/service/hlo_evaluator.h b/tensorflow/compiler/xla/service/hlo_evaluator.h index fc2fc9437b..2ad56080d8 100644 --- a/tensorflow/compiler/xla/service/hlo_evaluator.h +++ b/tensorflow/compiler/xla/service/hlo_evaluator.h @@ -174,7 +174,7 @@ class HloEvaluator : public DfsHloVisitorWithDefault { Status HandleBroadcast(HloInstruction* broadcast) override; - Status HandleGenerateToken(HloInstruction* token) override; + Status HandleAfterAll(HloInstruction* token) override; // Returns the already-evaluated literal result for the instruction. // A Constant instruction is considered evaluated and its literal will be |