This is an automated email from the ASF dual-hosted git repository.
sanirudh pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tvm.git
The following commit(s) were added to refs/heads/main by this push:
new 8543cec133 [Hexagon] Remove "c" as aot_host_target
tvm/contrib/hexagon/pytest_pl… (#14997)
8543cec133 is described below
commit 8543cec1334945f2d807d405e03ecf1eff2667d6
Author: Krzysztof Parzyszek <[email protected]>
AuthorDate: Thu Jun 1 00:48:27 2023 -0500
[Hexagon] Remove "c" as aot_host_target tvm/contrib/hexagon/pytest_pl…
(#14997)
[Hexagon] Remove "c" as aot_host_target tvm/contrib/hexagon/pytest_plugin.py
After PR#14918 (changes to SplitHostDevice), LowerIntrin code generates
fma, which then becomes the LLVM intrinsic fmuladd. Problem is, the code
then goes to the C codegen, which knows nothing about LLVM intrinsics.
The result is an abort at compile-time:
```
E File ".../src/target/source/codegen_c.cc", line 611
E TVMError: Unresolved call Op(tir.call_llvm_pure_intrin)
```
Why this didn't happen before is unclear at the moment, but using C
codegen with Hexagon code doesn't sound like a winning combination.
Hexagon-related code assumes LLVM codegen, and the above crash is the
result of that.
---
python/tvm/contrib/hexagon/pytest_plugin.py | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/python/tvm/contrib/hexagon/pytest_plugin.py
b/python/tvm/contrib/hexagon/pytest_plugin.py
index 585a6cc3c5..91a01ac56d 100644
--- a/python/tvm/contrib/hexagon/pytest_plugin.py
+++ b/python/tvm/contrib/hexagon/pytest_plugin.py
@@ -309,7 +309,7 @@ def terminate_rpc_servers():
os.system("ps ax | grep tvm_rpc_x86 | awk '{print $1}' | xargs kill")
-aot_host_target = tvm.testing.parameter("c", HEXAGON_AOT_LLVM_TARGET)
+aot_host_target = tvm.testing.parameter(HEXAGON_AOT_LLVM_TARGET)
@tvm.testing.fixture
