Closed JieyangChen7 closed 2 years ago
@ben-e-whitney Please review my changes to mgard.proto
and other related changes in the CPU version. Thanks!
Looks good to me. Two things:
NO_SHUFFLE
and NOOP
in Encoding::Preprocessor
and Encoding::Compressor
with NOOP_PREPROCESSOR
and NOOP_COMPRESSOR
, respectively. Then I suggest we replace NOOP
in DomainDecomposition::Method
with NOOP_METHOD
for consistency. (It's the numeric values that matter for the buffers, so we aren't locked in to anything here.)Device
message? I want to make sure we don't end up in a situation where, for example, X_HUFFMAN_LZ4
is implemented slightly differently on CUDA
than on HIP
and so we need to refer to Device
to decompress correctly. If we're just storing the backend like you might a timestamp (doesn't change any behavior, but might be good to know), I think that's fine.@ben-e-whitney Thanks for reviewing my changes. I have revised mgard.proto
according to your suggestions. The purpose of the Device
message is just for letting us know which device (backend) was used for compression. It does not change any behavior. This information might be useful for us to quickly locate a backend that potentially contains a bug. Please let me know if you have any more suggestions. I plan to merge this PR early next week. Thanks!
@qliu21 I'm trying to merge this PR. Which merge option do you usually choose Create a merge commit
or Rebase and merge
?
Rebase and merge.
On Tue, Mar 29, 2022 at 9:13 AM Jieyang Chen @.***> wrote:
@qliu21 https://github.com/qliu21 I'm trying to merge this PR. Which merge option do you usually choose Create a merge commit or Rebase and merge?
— Reply to this email directly, view it on GitHub https://github.com/CODARcode/MGARD/pull/188#issuecomment-1081855643, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAYYULGKJJLJONYITDCUDUTVCL6YPANCNFSM5RAYWRIA . You are receiving this because you were mentioned.Message ID: @.***>
-- Dr. Qing Liu Assistant Professor Helen and John C. Hartmann Department of Electrical and Computer Engineering New Jersey Institute of Technology, Newark, NJ Web: https://web.njit.edu/~qliu/ Email: @.*** Phone: 973-596-3526
This should be the last major PR before the March release.