Join GitHub today
GitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.
Sign upSend an IpcSharedMemory in tex_image_2d and tex_sub_image_2d #22225
Conversation
highfive
commented
Nov 19, 2018
|
Heads up! This PR modifies the following files:
|
highfive
commented
Nov 19, 2018
|
@bors-servo try |
Send an IpcSharedMemory in tex_image_2d and tex_sub_image_2d
|
|
|
This should be ready to be reviewed. @bors-servo try |
Send an IpcSharedMemory in tex_image_2d and tex_sub_image_2d <!-- Reviewable:start --> This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/22225) <!-- Reviewable:end -->
|
|
|
Unhappy tidy, fixed. |
|
r? @jdm |
This avoids a copy in the case of textures coming from HTMLImageElement.
|
Looks good, with that question. Have you checked any perf numbers out of curiosity? Just wondering what the relative speed is between creating an IPC channel vs. a shared memory thing. |
| ); | ||
| let effective_data_type = self | ||
| .extension_manager | ||
| .effective_type(data_type.as_gl_constant()); |
This comment has been minimized.
This comment has been minimized.
emilio
Nov 21, 2018
Member
Why doesn't textSubImage2D need to go through the EM in order to get the effective internal format?
This comment has been minimized.
This comment has been minimized.
nox
Nov 21, 2018
Author
Member
It was like that before, and using it makes everything panic on the GL side.
|
@bors-servo r=emilio |
|
|
Send an IpcSharedMemory in tex_image_2d and tex_sub_image_2d <!-- Reviewable:start --> This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/22225) <!-- Reviewable:end -->
|
|
nox commentedNov 19, 2018
•
edited by SimonSapin
This change is