31ccfffd38
Common: Add alignment header
...
Gets rid of duplicated alignment code.
2016-12-06 20:33:53 +01:00
7192789c11
Merge pull request #4486 from stenzek/gl-minimize-assert
...
OGL: Fix assertion when minimizing window
2016-12-05 12:46:29 +01:00
bff394ec8a
Merge pull request #4489 from stenzek/vulkan-minor-fixes
...
Vulkan: Minor fixes
2016-12-05 12:19:49 +10:00
8f97a24904
Vulkan: Enable GS instancing
2016-12-04 20:53:49 +10:00
9736198c3b
Vulkan: Use explicit barriers instead of dependancies
...
At least on NV, some of these don't seem to have the intended effect. One
known instance of this is in texture conversion.
2016-12-04 20:10:13 +10:00
a475792163
Vulkan: Fix incorrect logic in readback preemption
...
This could have been causing a large number of command buffer
submissions per frame, depending on when the readbacks occured.
2016-12-04 20:10:13 +10:00
58978c1440
Vulkan: Faster path for decoding XFB data
...
Using a texel buffer as the copy destination removes the need to copy to
an intermediate texture first.
2016-12-04 20:10:13 +10:00
804cd0ff03
Vulkan: Move XFB encoding/decoding to TextureConverter
2016-12-04 20:10:13 +10:00
d67463e0a7
Vulkan: Simplify palette texture conversion
2016-12-04 20:10:13 +10:00
add638538b
Vulkan: Combine PaletteTextureConverter and TextureEncoder classes
2016-12-04 20:10:13 +10:00
804af42ccc
Vulkan: Support binding texel buffers in UtilityShaderDraw
2016-12-04 20:10:13 +10:00
e241ec6666
Vulkan: Implement StagingTexture2D on top of StagingBuffer
...
Greatly simplifies things, and we weren't using the linear texture
implementation anyway.
2016-12-04 20:10:13 +10:00
4bc0e14995
Vulkan: Use an enumeration to index pipeline layouts
2016-12-04 20:10:13 +10:00
cd3481fbc7
Vulkan: Differentiate between descriptor set layouts and bind points
...
This also moves the pipeline and descriptor set layouts used for texture
conversion (texel buffers) to ObjectCache, and shares a binding location
with the SSBO set.
2016-12-04 20:10:13 +10:00
4e9018049d
Vulkan: Support logging debug reports without enabling validation layers
...
There is a caveat, Host GPU must be checked prior to starting the game, as
we can't enable the extension at runtime without recreating the instance.
2016-12-04 19:55:12 +10:00
1cfb0a1185
Vulkan: Fix deadlock in some resize scenarios
...
Only have experienced this on a few occasions when using the anv driver.
2016-12-04 19:12:33 +10:00
2ef884893a
Vulkan: Fix map error when texture dumping is enabled
2016-12-04 00:23:34 +10:00
4c860ddb70
Vulkan: Fix typo in PopulateBackendInfoMultisampleModes
2016-12-04 00:23:30 +10:00
730ec3809a
Vulkan: Remove unused variable
2016-12-04 00:23:25 +10:00
107e8c8ee7
OGL: Fix assertion when minimizing window
...
Also fixes a potential divide by zero in Renderer::RenderText.
2016-12-03 20:35:14 +10:00
7b1966a680
Vulkan: Fix assertion triggering when geometry shaders are unsupported
...
Happened when loading the pipeline UID cache.
2016-12-02 17:37:10 +10:00
ca691a9d95
Vulkan: Allow re-use of uniform buffers when doing per-stage uploads
...
This is safe now because we invalidate the pointers after submitting a
command buffer.
2016-11-30 23:34:46 +10:00
6a4eba1153
Vulkan: Replace explicit command buffer submits with wrapper function
...
Should we ever introduce anything else that has to be done when a command
buffer is executed (e.g. invalidating constants from previous commit), we
don't have to update all the callers.
2016-11-30 23:14:36 +10:00
3adeacb78d
Vulkan: Fix case where uniforms could be overwritten
...
If a draw caused a command buffer submission, the current uniform storage
should not be used for the new command buffer.
2016-11-30 23:14:35 +10:00
4c11735bd5
Vulkan: Fix case where a draw's vertices could be overwritten
...
This could happen because the vertex memory was already committed, if a
uniform buffer allocation failed and caused a command buffer to be
executed, it would be associated with the previous command buffer rather
than the buffer containing the draw that consumed these vertices.
2016-11-30 22:45:00 +10:00
37550501cc
Vulkan: Fix incorrect handling of buffer wrap-around in StreamBuffer
...
This was happening when a fence wait happened mid-frame. The data written
between the fence being queued and the allocation occuring was incorrectly
assumed to be consumed by the GPU.
2016-11-30 22:44:52 +10:00
0212741574
Merge pull request #4436 from stenzek/vulkan-full-ir-framedump
...
VideoBackends: Internal resolution frame dumping
2016-11-28 22:05:16 +10:00
b81dee8b9a
OGL: Support full-resolution frame dumping
2016-11-28 21:54:56 +10:00
93221e7f48
OGL: Move frame rendering procedures to seperate methods
2016-11-28 21:54:56 +10:00
1c1a686f63
Vulkan: Support full resolution frame dumping
2016-11-28 21:54:55 +10:00
8d48319414
Vulkan: Validate the pipeline cache before using it
...
This ensures that if a user changes adapters or vendors we're not passing
invalid data to the driver.
2016-11-28 21:21:55 +10:00
9604b336c8
Vulkan: Don't destroy the device's pipeline cache on MSAA mode change
...
The user could switch back again, and this would mean this data would be
lost. Disk space is cheap, and it's not going to be much.
2016-11-28 21:21:54 +10:00
aac66a1b61
Vulkan: Implement a pipeline UID cache
...
This stores enough information to recreate the pipeline, including the
shader UIDs, blend/depth/rasterization state, primitive and vertex format.
2016-11-28 21:21:53 +10:00
6db0ee9561
VideoCommon: Remove backbuffer size parameters from methods
...
We have the s_backbuffer_{width,height} fields to represent this, so
there's no point in passing them as parameters every time.
2016-11-28 20:14:59 +10:00
a0a62c0f46
VideoConfig: Add option for full-resolution frame dumping
2016-11-28 20:14:59 +10:00
e66373b7ff
Merge pull request #4456 from stenzek/vulkan-framedump-fix-2
...
Vulkan: Fix black borders in frame dumps
2016-11-28 20:03:08 +10:00
72e3f1ecec
Remove unnecessary ConfigManager includes
...
Making changes to ConfigManager.h has always been a pain, because
it means rebuilding half of Dolphin, since a lot of files depend on
and include this header.
However, it turns out some includes are unnecessary. This commit
removes ConfigManager includes from files which don't contain
SConfig or GPUDeterminismMode or GPU_DETERMINISM (which means the
ConfigManager include is not used).
(I've also had to get rid of some indirect includes.)
2016-11-27 22:38:38 +01:00
d6d3341183
D3D: Fix strided XFB copies
2016-11-19 20:29:51 +10:00
6b88a854a7
Vulkan: Handle strided XFB copies
...
Where src_rect.width * 2 != dst_stride.
2016-11-19 20:29:47 +10:00
725ef4c5dc
Vulkan: Pass target_rect to framebuffer draw methods
...
Fixes the black borders in frame dumps when the window was not sized to
the framebuffer aspect ratio.
2016-11-18 22:55:22 +10:00
3c92b35422
Vulkan: Use multiple command pools, one per frame
...
Instead of resetting two command buffers, now we only have to call
vkResetCommandPool once at the start of a frame.
NV's recommends using one pool per frame/thread. May offer a very small
boost in performance on some systems.
2016-11-15 00:40:15 +10:00
bd67adb362
Vulkan: Use correct sample count for EFB pokes with MSAA enabled
2016-11-14 20:24:16 +10:00
89176fe2ab
Vulkan: Fix crash on EFB poke
2016-11-14 20:19:59 +10:00
99de9fbe33
Merge pull request #4443 from Armada651/exclusive-ui
...
D3D: Move exclusive mode switching to UI thread.
2016-11-14 01:45:07 +01:00
7e35a47b51
Cosmetics.
2016-11-13 22:17:40 +01:00
bc98ec77be
Merge pull request #4441 from stenzek/vulkan-max-image-count
...
Vulkan: Handle maxImageCount of zero when creating swap chain
2016-11-13 14:12:32 +01:00
c723532f0d
Merge pull request #4430 from stenzek/vulkan-no-relaxed-vsync
...
Vulkan: Don't use FIFO_RELAXED present mode for vsync.
2016-11-13 14:10:58 +01:00
d7cf5e28b6
Frame: Use PauseAndLock when switching fullscreen modes.
...
This allows us to regain exclusive mode directly from OnActive().
2016-11-11 20:36:10 +01:00
9909babe2c
D3DBase: Create the swapchain in fullscreen mode if enabled.
2016-11-11 20:36:10 +01:00
0a194f8a3e
VideoConfig: Remove fullscreen flags.
...
These weren't actually settings, they were used as a bad way to communicate with the GPU thread.
2016-11-11 20:36:09 +01:00
Léo Lam