Lunderberg commented on code in PR #15809:
URL: https://github.com/apache/tvm/pull/15809#discussion_r1339126163
##########
python/tvm/relax/vm_build.py:
##########
@@ -310,6 +310,7 @@ def foo(x: Tensor((3, 4), "float32"), y: Tensor((3, 4),
"float32")):
passes.append(relax.transform.RemovePurityChecking())
passes.append(relax.transform.CallTIRRewrite())
passes.append(relax.transform.StaticPlanBlockMemory())
+ passes.append(relax.transform.LowerAllocTensor())
if
tvm.transform.PassContext.current().config.get("relax.backend.use_cuda_graph",
False):
passes.append(relax.transform.RewriteCUDAGraph())
Review Comment:
This bug ended up being trickier to track down than I had expected. It will
be much simpler to solve after https://github.com/apache/tvm/pull/15810 lands,
since the `.kill_*` methods won't be inserted yet. For now, I've re-ordered
the passes so the `LowerAllocTensor` occurs after `RewriteCUDAGraph`.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
