-
Notifications
You must be signed in to change notification settings - Fork 237
GPU mixed krylov -- WIP #1423
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
base: master
Are you sure you want to change the base?
GPU mixed krylov -- WIP #1423
Changes from 20 commits
31928ae
a8835bb
57d6ebb
542180e
cd69eba
7781cad
5ce7060
91a169f
cb030d2
8bfdbb9
887c4c3
dfea312
18e458a
7c44083
7163b69
1c74ee1
40f8a3a
ac3e7f3
fd1e9e0
366e620
05baa66
489e1fa
3264a49
1fecea4
f052718
34762e3
5bc2ddb
3996555
e20afec
3fd5776
69c34e8
d4fe2a5
a3c320e
49f8e70
905f1b7
feb75f2
File filter
Filter by extension
Conversations
Jump to
Diff view
Diff view
There are no files selected for viewing
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,18 @@ | ||
| hypreCUDAKernel_IJVectorUpdateValues | ||
| hypreGPUKernel_IJMatrixValues_dev1 | ||
| hypreGPUKernel_IJVectorAssemblePar | ||
| hypre_AuxParCSRMatrixStackReallocate | ||
| hypre_IJMatrixAssembleCommunicate | ||
| hypre_IJMatrixAssembleCompressDevice | ||
| hypre_IJMatrixAssembleParCSRDevice | ||
| hypre_IJMatrixAssembleSortAndReduce1 | ||
| hypre_IJMatrixAssembleSortAndReduce2 | ||
| hypre_IJMatrixAssembleSortAndReduce3 | ||
| hypre_IJMatrixGetValuesParCSRDevice | ||
| hypre_IJMatrixSetAddValuesParCSRDevice | ||
| hypre_IJMatrixSetConstantValuesParCSRDevice | ||
| hypre_IJVectorAssembleParDevice | ||
| hypre_IJVectorAssembleSortAndReduce1 | ||
| hypre_IJVectorAssembleSortAndReduce3 | ||
| hypre_IJVectorSetAddValuesParDevice | ||
| hypre_IJVectorUpdateValuesDevice |
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -15,6 +15,8 @@ | |
|
|
||
| scriptdir=`dirname $0` | ||
|
|
||
| BUILD_TYPE=$(echo "$1" | tr '[:lower:]' '[:upper:]') | ||
|
|
||
| # Check if terminal supports colors | ||
| if [ -t 1 ]; then | ||
| # Use colors | ||
|
|
@@ -30,7 +32,16 @@ fi | |
|
|
||
| export LC_COLLATE=C # sort by listing capital letters first | ||
|
|
||
| cat mup.fixed mup.functions mup.methods | sort | uniq > mup_check.old | ||
|
|
||
| if [ "$BUILD_TYPE" = "GPU" ]; then | ||
| cat mup.fixed mup.fixed.gpu \ | ||
|
Contributor
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. You will probably need to modify this if we choose to only include the '.gpu' files in those directories that need them. Maybe a better way to do this is to create a 'FILES' variable that starts out having the standard three, then appends the '.gpu' files if needed, then runs only one 'cat' line at the end. This would also be easily extensible if something else comes up in the future.
Contributor
Author
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. Yes, that's correct. For now, I parsed it to 2>/dev/null to suppress the "file not found messages" so I don't have to check for the files each directory, but your suggestion is probably a better approach. |
||
| mup.functions mup.functions.gpu \ | ||
| mup.methods mup.methods.gpu \ | ||
| | sort | uniq > mup_check.old | ||
| else | ||
| cat mup.fixed mup.functions mup.methods | sort | uniq > mup_check.old | ||
| fi | ||
|
|
||
| $scriptdir/generate_function_list.sh | sort | uniq > mup_check.new | ||
|
|
||
| # Remove functions listed in mup.exclude (if it exists) | ||
|
|
||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Have you looked into the portability of c++flt?
Uh oh!
There was an error while loading. Please reload this page.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, I have thought a bit about this. From what I can tell, it is supported in binutils, macOS, most BSD systems, MinGW/Cygwin and so should be fairly portable. Ideally we would use the demangle option the "-C" for nm, since it is built into the function, but unfortunately work well. There is also --demangle for nm on GNU/linux systems, but c++filt is more portable.