[PATCH] Adds Cortex-A53 and Cortex-A57 subtargets.

Dave Estes cestes at codeaurora.org
Tue Feb 25 07:39:47 PST 2014


>    I think we'd better disable it first and enable it later on all at once after you complete all, and otherwise I expect some test cases added to verify the scheduling effect.
>
>    The document you mentioned is not publicly available to everybody in community, so we should avoid mentioning it in the code.

Thanks, Jiangning!

I'll make sure the reference is removed in the next patchset. I'll just 
put your email address there in it's place. *joking*

This model will only be used when the -mcpu=cortex-a53 flag is set. At 
this point, that flag does little to nothing else, and so I imagined 
that people would not even be using it. Do you think that this is this 
enough of an "off by default" behavior. The alternative would be to turn 
off the use of the machine model (schedmodel) completely for the AArch64 
backend? While it has a larger effect, it's still not a big deal since 
none of the AArch64 subtargets have machine models yet.

Thanks...
-Dave

-- 
Employee of Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, 
hosted by The Linux Foundation



More information about the llvm-commits mailing list