Skip to content
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

Fix layout fragment type mismatch #80

Merged
merged 1 commit into from
Nov 9, 2021
Merged

Conversation

smnbl
Copy link
Contributor

@smnbl smnbl commented Nov 5, 2021

This fixes an issue causing a type mismatch in the matmul_pipelined kernel between
the return type of Layout.load (inferred by the compiler) and the element type of the {a,b}_fragment arrays (hinted by the fragtype function) that came up when working with non-Float16 input matrices.

place of mismatch:

@inbounds a_fragment[i, j] = Layout.load(conf.global_a_layout, a, translate_base(thread_tile, (M = block_i, K = 0)))

this issue caused a type mismatch in the matmul_pipelined kernel between
the return type of Layout.load (inferred by the compiler) and the element type
(hinted by the fragtype function)
@maleadt
Copy link
Member

maleadt commented Nov 9, 2021

Rebased, so CI should work now.

What about a test for this case?
EDIT: can't be tested yet, will be covered in #81.

@codecov
Copy link

codecov bot commented Nov 9, 2021

Codecov Report

Merging #80 (d9c5783) into master (7595233) will not change coverage.
The diff coverage is 0.00%.

Impacted file tree graph

@@           Coverage Diff           @@
##           master      #80   +/-   ##
=======================================
  Coverage   42.31%   42.31%           
=======================================
  Files           9        9           
  Lines         423      423           
=======================================
  Hits          179      179           
  Misses        244      244           
Impacted Files Coverage Δ
src/layout.jl 16.21% <0.00%> (ø)

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 7595233...d9c5783. Read the comment docs.

@maleadt maleadt merged commit 2f0cc6d into JuliaGPU:master Nov 9, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants