9b136c90de
buffer: Add specific handling for different buffers
2020-07-11 20:52:01 +00:00
480392e40c
buffer: Add buffer type to interface
2020-07-11 20:52:01 +00:00
59cc119d76
main: Refine damage hints
2020-07-11 20:52:01 +00:00
fe590f3940
damage-refinery: Use damage hint
2020-07-11 20:52:01 +00:00
e398dcc235
damage-refinery: Hash directly instead of copying
2020-07-11 20:52:01 +00:00
439d3997d4
damage-refinery: Use negative stride to handle y-inversion
2020-07-11 20:52:01 +00:00
c515a29951
Create hash based damage checker
2020-07-11 20:52:01 +00:00
946ace1760
Fix damage transform
2020-07-11 20:52:01 +00:00
12555bea13
pixman-renderer: Use negative stride for y-inversion
2020-07-11 20:52:01 +00:00
1cafc25655
Transform damage coordinates
2020-07-11 20:52:01 +00:00
6b1dc2e6c3
Add utilities for managing output transforms
2020-07-11 20:52:01 +00:00
507b76dfbd
Create pixman renderer
2020-07-11 20:52:01 +00:00
29f1669d55
buffer: Leave pixman out of buffers
2020-07-11 20:52:01 +00:00
e65cae43c2
Extract pixel format conversion into own file
2020-07-11 20:52:01 +00:00
a6e1ba25ea
Render using pixman
2020-07-11 20:52:00 +00:00
4fb472f611
Create a buffer abstraction
2020-07-11 20:52:00 +00:00
0a139a1ac4
Align with aml API changes
2020-07-06 16:46:30 +00:00
4f1bca55c3
Exit when main selected seat/output goes away
2020-06-12 23:15:22 +00:00
041ac60ef6
renderer: Add more error messages
2020-06-10 22:09:22 +00:00
ff882b7774
keyboard: Ignore shift level on modifiers
2020-05-26 22:19:48 +00:00
ff1223e4f9
keyboard: Fix comparison of signed and unsigned value
2020-05-18 20:38:48 +00:00
9b48290d09
Improve error handling in wayland dispatch
2020-05-10 12:21:10 +00:00
151e916752
renderer: Set GL_PACK_ALIGNMENT to 1
...
Otherwise row stride will multiple of 4, even for damage data.
2020-05-10 12:21:10 +00:00
0761a5b818
damage: Fix uninitialised value
2020-05-10 12:21:10 +00:00
4f5933c07f
dmabuf: Limit rate to 30 FPS
2020-05-03 19:58:50 +00:00
f5453ffe1e
keyboard: Relax matching criteria on shift level
...
Clients behave differently when it comes to sending release events for keys,
so it's hard to rely on any one kind of behaviour.
2020-05-01 17:34:22 +00:00
405268fc58
keyboard: Log keyboard symbol lookup errors
2020-05-01 17:19:14 +00:00
6e889211db
renderer: Run glFinish() before glReadPixels()
2020-04-26 14:15:57 +00:00
8cc8c198d2
renderer: Run glFinish() after rendering dmabuf frame
2020-04-26 14:14:06 +00:00
06a249897b
dmabuf: Show a debug message when frames are held for too long.
2020-04-26 14:14:06 +00:00
04569b01f2
dmabuf: Add dtrace probes
2020-04-26 11:24:29 +00:00
9d6310cb14
dmabuf: Remove rate limiting
2020-04-26 10:39:58 +00:00
6becbacbe5
Align with NeatVNC interface changes
2020-04-12 21:46:48 +00:00
1fc664a014
Remove call to nvnc_set_dimensions()
2020-04-12 16:14:18 +00:00
f9deca5c9c
shm: Add copyright notice
...
Putting a proper license on this to appease package maintainers.
2020-04-12 13:08:42 +00:00
3f62295214
keyboard: rename keymap_len to keymap_size
2020-04-09 15:39:09 +03:00
fa49aca45a
keyboard: check that write finished writing everything
...
(Use loop to write until the end)
2020-04-09 15:38:13 +03:00
ccc582cd58
keyboard: allocate shm with *actually* correct size
...
strlen returns "the number of characters that precede the terminating NUL".
We do need to send the \0 byte, since xkbcommon on the receiving end
*is* expecting a C string. And e.g. wlroots does strlen+1.
2020-04-09 15:37:18 +03:00
14d62d0029
Fix transformations
...
This fixes #29
2020-04-09 11:50:48 +00:00
58a181ccbc
Redesign update/rendering loop
...
Every change to the framebuffer is now copied into a single buffer
inside a single callback. This simplifies things a lot, and might even
perform better.
2020-04-07 23:37:31 +00:00
d610076614
shm: support memfd and SHM_ANON
2020-04-06 02:07:25 +03:00
720b127dee
keyboard: allocate shm with correct size
...
This was working on Linux because shm_open there returns a regular file,
but this is required on FreeBSD where shm_open returns a memory-backed fd.
2020-04-06 00:26:22 +03:00
be401b5e4a
Report error when binding to address fails
...
This fixes #21
2020-04-04 22:38:23 +00:00
6e521a07a0
Merge frames rather than dropping them when encoder is too slow
2020-04-04 21:03:49 +00:00
238c196e6b
Capture a new frame immediately when a client requests a whole frame
2020-04-04 15:01:38 +00:00
ee4917f200
Make sure that frames being written to are not in use by the encoder
2020-04-04 13:14:07 +00:00
8ce312e9eb
renderer: Add dtrace probes
2020-04-02 00:55:18 +00:00
075680994e
screencopy: Add dtrace probes
2020-04-02 00:34:00 +00:00
dc7adf8f3b
renderer: Use ATTR_INDEX_ constants for glEnableVertex*
2020-03-29 14:05:33 +00:00
e85f219aff
Make sure damage buffer is properly aligned
2020-03-29 13:09:28 +00:00
Andri Yngvason
Greg V