Skip to content

GlslOptimizerShaderConverter: new plugin #92

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.

Sign up for GitHub

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

Jump to bottom
Draft
mosra wants to merge 1 commit into
base: master
Choose a base branch
from
Draft

GlslOptimizerShaderConverter: new plugin #92

mosra wants to merge 1 commit into from

Conversation

@mosra
Copy link
Owner

@mosra mosra commented Nov 6, 2020

Things to do:

  • not sure about the relevance of this nowadays -- with a quick'n'dumb test on a moderately complex shader containing a bunch of loops, those got all unrolled and the resulting SPIR-V binary was significantly larger than without this, even after passing it through SPIR-V optimizer. Or maybe there's a way to limit the unrolling?
  • needs to expose input version (same as in glslang), otherwise too shitty to use
  • needs to expose preprocessor definitions (~same as in glslang), otherwise too shitty to use
  • even with the https://github.com/jamienicol/glsl-optimizer fork which is based on Mesa 20, it reports that the only supported GLSL versions are <= 1.50 and <= 300 es :(
  • docs, move the cmake subproject setup to a doc snippet
@mosra
[wip] GlslOptimizerShaderConverter: new plugin.
f42ee91 
TODO: not sure about the relevance of this nowadays -- with a
  quick'n'dumb test on a moderately complex shader containing a bunch of
  loops, those got all unrolled and the resulting SPIR-V binary was
  significantly larger than without this, even after passing it through
  SPIR-V optimizer. Or maybe there's a way to limit the unrolling?
TODO: needs to expose input version (same as in glslang), otherwise too
  shitty to use
TODO: needs to expose preprocessor definitions (~same as in glslang),
  otherwise too shitty to use
TODO: even with the https://github.com/jamienicol/glsl-optimizer fork
  which is based on Mesa 20, it reports that the only supported GLSL
  versions are <= 1.50 and <= 300 es :(
TODO: docs, move the cmake subproject setup to a doc snippet
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

1 participant

@mosra