Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Conversation

kainino0x
Copy link
Contributor

@kainino0x kainino0x commented Dec 13, 2019

It's awkward to have to specify an imageHeight when you're not doing an array/3d texture copy (i.e. almost always).
Make it default to 0 - and I'd have this be valid unless copySize.depth > 1.

... also, I still don't really like the name "imageHeight"...


Preview | Diff

It's [awkward](https://github.com/gpuweb/cts/pull/95/files#diff-d79e5d51a5b1c3e1d586625d3f7aff62R77) to have to specify an `imageHeight` when you're not doing an array/3d texture copy (i.e. almost always).
Make it default to 0 - and I would have this be valid unless `copySize.depth > 1` 

... also, I still don't really like this name...
Copy link
Contributor

@Kangz Kangz left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you add a sentence or two of spec language to explain what the default means and when it is valid?

@kainino0x
Copy link
Contributor Author

Done (sketchily)

@kdashg
Copy link
Contributor

kdashg commented Dec 16, 2019

I think I like having 0 indicate tightly-packed src, a la GL. No strong feelings, but I think that's easier to use.

@kainino0x
Copy link
Contributor Author

I'm pretty neutral, leaning very slightly toward "0 is only valid if unused".

@Kangz
Copy link
Contributor

Kangz commented Jan 6, 2020

This was discussed in the 2020-01-06 meeting

@kainino0x
Copy link
Contributor Author

kainino0x commented Jan 6, 2020

Resolution: update to make 0 mean tightly-packed

@kainino0x
Copy link
Contributor Author

Discussed today. Given #537, I would advocate to land this PR as-is, instead of trying to encode that limitation into the definition of "tightly packed". @jdashg @kvark any objections?

@grorg
Copy link
Contributor

grorg commented Feb 3, 2020

Discussed again at the 2020-02-03 teleconference

@kainino0x
Copy link
Contributor Author

Let's talk about this at the F2F:
Should we make imagePitch=0 (and even rowPitch=0) really, semantically mean 0? I.e. the same data would be copied into every (layer/slice, row, respectively).

action item @kainino0x: look up whether any native APIs allow this.

@kainino0x
Copy link
Contributor Author

Another alternative is to make 0 just mean "naively tightly packed", and error if that violates other restrictions. I think(?) that's what Vulkan does.

@kainino0x
Copy link
Contributor Author

Done: #519 (comment)

@kainino0x
Copy link
Contributor Author

kainino0x commented Feb 12, 2020

Discussed at F2F. To be resolved by spec editors based on the input there.

@kainino0x
Copy link
Contributor Author

kainino0x commented Mar 9, 2020

Actually, this particular item was resolved at the F2F. Updated and landing, then I'll do the rest of it as a new PR.

@kainino0x kainino0x merged commit 6867df9 into master Mar 9, 2020
@kainino0x kainino0x deleted the kainino0x-patch-1 branch March 9, 2020 22:25
litherum pushed a commit to litherum/gpuweb that referenced this pull request Mar 10, 2020
It's awkward to have to specify an `imageHeight` when you're not doing an array/3d texture copy (i.e. almost always).

Make it default to zero. Zero is valid iff `copySize.depth > 1`
JusSn pushed a commit to JusSn/gpuweb that referenced this pull request Jun 8, 2020
It's awkward to have to specify an `imageHeight` when you're not doing an array/3d texture copy (i.e. almost always).

Make it default to zero. Zero is valid iff `copySize.depth > 1`
JusSn pushed a commit to JusSn/gpuweb that referenced this pull request Jun 8, 2020
It's awkward to have to specify an `imageHeight` when you're not doing an array/3d texture copy (i.e. almost always).

Make it default to zero. Zero is valid iff `copySize.depth > 1`
ben-clayton pushed a commit to ben-clayton/gpuweb that referenced this pull request Sep 6, 2022
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.

5 participants