Commit Graph

64 Commits

Author SHA1 Message Date
Hajime Hoshi
5147bbde9d restorable: Add MakeVolatile 2019-02-13 00:39:54 +09:00
Hajime Hoshi
2b0dff197e restorable: Fix comments (#793) 2019-01-22 23:33:56 +09:00
Hajime Hoshi
158f4fb3ed restorable: Bug fix: Reading pixels from a volatile image might fail
Fixes #793
2019-01-22 03:35:54 +09:00
Hajime Hoshi
107dfe5074 graphics: Improve Set speed
On GopherJS, copying a struct is very heavy. This change avoids
copying (color) structs when possible.
2019-01-14 04:25:43 +09:00
Hajime Hoshi
b34834a895 graphicscommand: Explicitly forbide ReplacePixels for a part after DrawImage 2019-01-06 05:15:27 +09:00
Hajime Hoshi
1cfd97cde0 restorable: Clearing is not needed after NewImage 2019-01-06 03:40:59 +09:00
Hajime Hoshi
62cdb93105 restorable: Improve tests 2018-12-26 21:14:29 +09:00
Hajime Hoshi
fbf7007056 restorable: Pixels() should return immediately when it doesn't have to access GPU
Fixes #763
2018-12-26 20:25:40 +09:00
Hajime Hoshi
f1582c2d73 graphics: Add Address representing a sampler address mode
Fixes #761
2018-12-24 17:29:31 +09:00
Hajime Hoshi
b2b09ccec0 restorable: Avoid memory allocating when an entire image is cleared 2018-10-30 11:08:38 +09:00
Hajime Hoshi
ce1c616f69 Merge graphics and graphicsutil 2018-10-28 23:03:06 +09:00
Hajime Hoshi
ff54d3b681 Move CompositeMode to graphics package 2018-10-28 21:49:47 +09:00
Hajime Hoshi
06f2052817 Add graphics package and move Filter to graphics 2018-10-28 20:25:52 +09:00
Hajime Hoshi
2da5192510 Rename graphics -> graphicscommand 2018-10-28 20:10:05 +09:00
Hajime Hoshi
690c3cf981 graphics: Embed 'scale' part of the color matrix
If the color matrix includes only 'scale' part, they are embedded
into vertices in order to reduce draw calls.

Fixes #662
2018-08-10 23:20:55 +09:00
Hajime Hoshi
77e905eb25 Revert "graphics: Embed color matrices values to vertices (#655)"
This reverts commit 5a9235aaf9.
2018-08-06 01:44:31 +09:00
Hajime Hoshi
5a9235aaf9
graphics: Embed color matrices values to vertices (#655)
Fixes #531
2018-07-26 11:37:27 +09:00
Hajime Hoshi
5249afc06b graphicsutil: Add QuadIndices 2018-07-21 03:40:37 +09:00
Hajime Hoshi
6c8b7f8e9c graphics: Refactoring: Reduce error propagations 2018-07-12 03:08:50 +09:00
Hajime Hoshi
840c4d24df graphicsutil: Accept matrix parameters directly for performance 2018-06-19 00:19:32 +09:00
Hajime Hoshi
ac24a8382f graphicsutil: Extract matrix applying for performance 2018-06-17 19:45:09 +09:00
Hajime Hoshi
95561bbf6b affine: Use float32 values for GeoM
float32 was slow on GopherJS, but not slow on Wasm.
2018-06-17 15:38:10 +09:00
Hajime Hoshi
4c2fc30311 graphicsutil: Don't allow nil GeoM 2018-06-17 01:48:56 +09:00
Hajime Hoshi
b4ace53596 affine: Remove GeoM to avoid GeoM allocations on the heap 2018-06-17 00:59:50 +09:00
Hajime Hoshi
3ee76ffd8e graphicsutil: Move vertices.go to graphicsutil 2018-06-10 23:08:22 +09:00
Hajime Hoshi
9745ccee47 restorable: Make (*Image).DrawImage accept vertices and indices 2018-06-10 22:59:00 +09:00
Hajime Hoshi
45afb6db67 Reland: restorable: Merge Clear to ReplacePixel 2018-04-29 13:40:18 +09:00
Hajime Hoshi
1cf4f30541 Revert "restorable: Merge Clear to ReplacePixels"
This reverts commit fb641d88cd.

Reason: When restoring the image where ReplacePixels is called,
dummyImage might not be restored since there is no record of
relationships between this image and dummyImage. Now pixels is not
nil when reverting by chance, but this would cause problems in the
future.
2018-04-25 22:31:48 +09:00
Hajime Hoshi
fb641d88cd restorable: Merge Clear to ReplacePixels
This can avoid unnecessary stale images that requires loading
pixels from GPU.
2018-04-25 01:30:57 +09:00
Hajime Hoshi
9902497e3d test: Use flock to make tests exclusive
Fixes #575
2018-04-07 00:59:52 +09:00
Hajime Hoshi
70f16daafd restorable: Fix tests (wrong assumption was used) 2018-04-06 02:15:49 +09:00
Hajime Hoshi
e2632e358e restorable: Add TestDoubleResolve 2018-03-25 21:09:26 +09:00
Hajime Hoshi
fc47d2843a restorable: Add TestRestoreChain2 2018-03-25 19:43:32 +09:00
Hajime Hoshi
8149ace52c restorable: Add TestDispose 2018-03-25 00:50:11 +09:00
Hajime Hoshi
be4e3ab801 restorable: Bug fix: partial ReplacePixels broke the image state (#562) 2018-03-20 02:14:00 +09:00
Hajime Hoshi
1068156215 restorable: Add restoring case to TestReplacePixels 2018-03-01 02:43:59 +09:00
Hajime Hoshi
127f6c09c8 graphics: Specify rect to glTexSubImage2D 2018-03-01 01:58:09 +09:00
Hajime Hoshi
2f1908b05c restorable: Use nil instead of empty affine.ColorM 2018-02-28 02:14:10 +09:00
Hajime Hoshi
adbd0db765 restorable: Accept source rect and geom instead of vertices 2018-02-27 01:35:30 +09:00
Hajime Hoshi
85b133dad0 graphics: Remove NewImageFromImage command 2018-02-25 23:22:49 +09:00
Hajime Hoshi
fe7aae32a0 restorable: Remove graphics.Image.Fill usage 2018-02-24 16:18:18 +09:00
Hajime Hoshi
2ba89591db restorable: Remove Fill 2018-02-24 15:14:53 +09:00
Hajime Hoshi
6ef4bbde2d graphics: Add FilterDefault; Make DrawImageOptions specify Filter (#453) 2018-02-14 02:59:44 +09:00
Hajime Hoshi
595e654123 graphics: Bug fix: pixels read from GPU might include errors (#492) 2018-02-03 16:43:42 +09:00
Hajime Hoshi
65b71ad79d restorabel: uint8 -> byte 2018-02-03 16:36:43 +09:00
seebs
0b7ba8e573 Do GeoM projection on CPU (#477)
Handling GeoM projection on CPU may seem like a weird choice, given
how fast GPU is, but it pays off:

* You only have to do a very small subset of the actual matrix
  multiply.
* You don't have to construct a matrix in the vertex shader.
* Six fewer float32 values per vertex.
* You do still have to do the matrix computation for each vertex,
  though.

Signed-off-by: Seebs <seebs@seebs.net>
2018-01-14 16:01:55 +09:00
Hajime Hoshi
723d153800 Reland(2): graphics: Appropriate rendering of edges on linear filter 2017-12-12 00:03:01 +09:00
Hajime Hoshi
362d1c417f Revert 'graphics: Appropriate rendering of edges on linear filter (Reland)' (#458) 2017-12-11 20:30:54 +09:00
Hajime Hoshi
1152439e65 graphics: Appropriate rendering of edges on linear filter (Reland)
Fixes #456
2017-12-07 02:26:16 +09:00
Hajime Hoshi
8091aa5190 Revert 'graphics: Appropriate rendering of edges on linear filter' #456 2017-12-06 23:40:29 +09:00