D27788: Implement EGL_KHR_partial_update and EGL_EXT_swap_buffers_with_damage
Vlad Zahorodnii <[email protected]>
| Newsgroups | gmane.comp.kde.devel.kwin |
|---|---|
| Message-ID | <[email protected]> |
zzag added a comment. The key point about EGL_KHR_partial_update is that the compositor know what parts of the back buffer are going to be updated. So the compositor could specify the damage region _beforehand_ recording command buffers. If the compositor doesn't have a sophisticated effects pipeline, e.g. weston, then it's pretty trivial to determine the buffer damage, but that's not the case for us, unfortunately. Ideally, a compositing cycle must be split into several phases - pre paint, paint, and maybe post paint. In the pre paint phase, the compositor would do necessary bookkeeping so we know what the damage region for the next frame is. I really like the idea of having a hook that is called when the scene is about to start recording a command buffer for the next frame, but given the current design of the effects pipeline I am starting to think that we cannot support EGL_KHR_partial_update at the moment. The dirty region for the next frame is computed too late and I don't see how we can fix it because effects have the final say in what area of the screen will be repainted. :/ I would love to be proven wrong, though. REPOSITORY R108 KWin REVISION DETAIL https://phabricator.kde.org/D27788 To: apol, #kwin, #plasma:_mobile Cc: mwolff, zzag, davidedmundson, kwin, Orage, cacarry, LeGast00n, The-Feren-OS-Dev, cblack, jraleigh, zachus, fbampaloukas, GB_2, mkulinski, ragreen, jackyalcine, iodelay, crozbo, bwowk, ZrenBot, ngraham, alexeymin, himcesjf, lesliezhai, ali-mohamed, hardening, romangg, jensreuterberg, abetts, sebas, apol, ahiemstra, mart