Replacing explicit OpenACC with build variable.#9
Open
cponder wants to merge 1 commit intomgduda:atmosphere/gpufrom
Open
Replacing explicit OpenACC with build variable.#9cponder wants to merge 1 commit intomgduda:atmosphere/gpufrom
cponder wants to merge 1 commit intomgduda:atmosphere/gpufrom
Conversation
I think it's a mistake to embed explicit flags like "-ta=tesla:cc60" in the deep Makefile's. It takes some work to figure out what needs to be changed, so it's too easy to lose track of the individual changes, and get the configuration files into an inconsistent state. I recommend adding this $(OPENACC) variable to the Makefile instead. I'll be setting it in my top-level build-script so I can manage it as part of my "thin layer". In the top-level Makefile, you may want to duplicate the "pgi" target as "pgi-openacc" and add this line OPENACC = -acc -Minfo=accel -ta=tesla:cc60,cuda8.0 I recommend cc60 & cuda8.0 for now because we're working mainly with Pascal's and the cuda9.0 is too new to rely on. On the other hand I won't be using the Makefile setting, I'll be controlling the $(OPENACC) variable from my top-level build script.
mgduda
pushed a commit
that referenced
this pull request
Jul 26, 2019
This merging PR #9 to the code
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
(This goes with the analogous change to src/framework/Makefile. I thought the pull-requests were supposed to bundle these changes....)
I think it's a mistake to embed explicit flags like "-ta=tesla:cc60" in the deep Makefile's. It takes some work to figure out what needs to be changed, so it's too easy to lose track of the individual changes, and get the configuration files into an inconsistent state.
I recommend adding this $(OPENACC) variable to the Makefile instead. I'll be setting it in my top-level build-script so I can manage it as part of my "thin layer".
In the top-level Makefile, you may want to duplicate the "pgi" target as "pgi-openacc" and add this line
OPENACC = -acc -Minfo=accel -ta=tesla:cc60,cuda8.0
I recommend cc60 & cuda8.0 for now because we're working mainly with Pascal's and the cuda9.0 is too new to rely on. On the other hand I won't be using the Makefile setting, I'll be controlling the $(OPENACC) variable from my top-level build script.