You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@tvm.apache.org by GitBox <gi...@apache.org> on 2021/09/15 17:26:33 UTC

[GitHub] [tvm] manupa-arm edited a comment on issue #8978: [Bug][VTA][OpenCL] If allowed to allocate in stack, VTA multiple target test will fail

manupa-arm edited a comment on issue #8978:
URL: https://github.com/apache/tvm/issues/8978#issuecomment-920219633


   Thanks for flagging this!
   
   IIUC,
   
   This is problematic for micro because now the constant sized allocates are forced to be placed on stack (bypassing TVMPlatformAllocate abstraction) because that is how the codegen_c lowers the allocate.
   
   https://github.com/apache/tvm/blob/1b99adcd8f4b7bd4ad2e333da82b2204803438ee/src/target/source/codegen_c.cc#L860-L877
   
   I dont think is desired.
   
   Can someone explain why the guarded TVMBackendAllocWorkspace calls to allocate memory is problematic ? -- which are the ones I see generated in TIR prior to the last rever : #8274 ?


-- 
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: commits-unsubscribe@tvm.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org