FrozenGene commented on a change in pull request #6211:
URL: https://github.com/apache/incubator-tvm/pull/6211#discussion_r465442663
##########
File path: python/tvm/target/target.py
##########
@@ -188,7 +188,8 @@ def arm_cpu(model='unknown', options=None):
"p20": ["-model=kirin970", "-mtriple=arm64-linux-android",
"-mattr=+neon"],
"p20pro": ["-model=kirin970", "-mtriple=arm64-linux-android",
"-mattr=+neon"],
"rasp3b": ["-model=bcm2837", "-mtriple=armv7l-linux-gnueabihf",
"-mattr=+neon"],
- "rasp4b": ["-model=bcm2711", "-mtriple=arm-linux-gnueabihf",
"-mattr=+neon"],
+ "rasp4b": ["-model=bcm2711", "-mtriple=armv8l-linux-gnueabihf",
"-mattr=+neon"],
+ "rasp4b64": ["-model=bcm2711", "-mtriple=aarch64-linux-gnu",
"-mattr=+neon"],
Review comment:
I think we should add `-mcpu=cortex-a72` for rasp4. We should get better
performance after it than generic cpu instruction.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
