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
MSL: Use of MSLVertexAttr #1050
Comments
Yup, looking at MSLVertexAttr, it does not seem to be relevant anymore. I know the format is used to do fixups, but is that all? I'm fine changing the API in the C++ backend, but I cannot break any API in the C wrapper. We can deprecate and add a better interface though, as long as we can express the old APIs with the new. As I understand it, most of the fields in MSLVertexAttr are completely ignored anyways, so we don't have to care about those when moving to a new API. |
Understood about the C ABI...although I wonder if anyone is using the C equivalents to this...since the C++ struct was driven by MoltenVK originally. But we can just stub the references out if we change this struct I guess. @cdavis5e Since the remaining use of |
@billhollings @cdavis5e Any further comments? |
As I mention...I'm fine to trim it (preferred) or remove it from the C++ API and replace with something simpler. The C ABI could stay the same since it uses a parallel As far as I can tell...SPIRV-Cross only uses the @cdavis5e Most of those three are used by code introduced with tess...so you'd be better able to remark on whether we can elide any of those elements too. |
The use of
MSLVertexAttr
has evolved over time. Currently...almost none of the information it contains is used at all within SPIRV-Cross...I assume because we rely completely on thestage_in
construct (originally we supported bothstage_in
and indexed vertex access.Like a good government worker...MoltenVK is currently diligently filling out the contents of this struct...even though most of it serves no purpose.
I'm wondering if we should trim this struct...either remove it entirely in favour of other API options...or at least trim (and maybe rename) it for it's current use as a re-mapper of tessellation formats and builtins.
@HansKristian-Work @cdavis5e ?
The text was updated successfully, but these errors were encountered: