tabi.katalin
d2cdca45a5
Fix bad alloc error when creating window surface
...
The screen doesn't refresh when changing between captures during replay on remote Android device because it fails to display the splash screen and the new preview window.
2018-05-24 10:45:46 +01:00
baldurk
9e3a454a18
Unwrap resources passed to IDXGIDevice::Reclaim/OfferResources*
2018-05-23 22:29:05 +01:00
baldurk
6cfdd09774
Allow specifying a NULL bytecode with a non-0 bytecode length
...
* Apparently D3D12 warns about this but still creates the PSO
2018-05-23 19:29:10 +01:00
baldurk
ebe7ffca80
Don't divide by zero when picking if a vertexbuffer has a stride of 0
2018-05-23 17:39:53 +01:00
baldurk
b3bc00738d
Don't enumerate window titles when we just need a list of PIDs
2018-05-23 17:28:56 +01:00
baldurk
ab9a212b8d
Update qprocessinfo to 0d591f19f17805d65ef4880c3fc51021523003df
2018-05-23 17:28:42 +01:00
baldurk
0a041601f8
Update toolwindowmanger to d602babc98639c5d8206f1667f95062b2f3709fc
2018-05-23 17:22:18 +01:00
baldurk
df6a95edf0
For backwards compatibility, bind fake backbuffer FBO instead of 0
...
* In newer captures we should always serialise an actual correct ID for the
backbuffer, but in older captures an empty resource ID will be serialised, so
we should pick this up and bind the backbuffer.
2018-05-23 17:08:39 +01:00
baldurk
6e3a173c7f
Fetch ordinal names if we emergency-fetch module in GetProcAddress
2018-05-23 16:08:44 +01:00
baldurk
851a88133a
Remove the 'save all initials' option. Behave as if it's always enabled
...
* This option has always been a mixed bag - when originally written captures
weren't compressed at all so saving the cost of a initial contents on a
gbuffer would have a significant savings.
* Now with compression the savings are lesser, and it's a source of
bugs/confusion for the case where either a bug is caused by leaking data from
the previous frame or worse still the contents are discarded incorrectly.
* D3D11 will now behave as the other APIs will - saving initial contents
whenever needed even if they seem like they might not be used.
2018-05-23 16:08:44 +01:00
baldurk
54ce79d8ae
Don't change an already sized internalformat. Closes #990
2018-05-23 16:08:44 +01:00
baldurk
921d5b7266
Fall back to using PLT hooking on android if interceptor-lib fails
...
* This started happening recently with eglSwapBuffers for unknown reasons.
2018-05-23 11:22:06 +01:00
baldurk
474491c991
Fix missing close-bracket on vkCmdDispatchIndirect name
2018-05-18 09:31:52 +01:00
baldurk
d90d1d81e6
Handle empty/no value defines in command-line HLSL definitions
2018-05-17 15:04:40 +01:00
baldurk
85b35cd352
Set usage panel in resource inspector to auto-resize columns to fit data
2018-05-17 15:04:40 +01:00
baldurk
f8ced4c0ee
Update SPDB handling to use cvinfo, and store more/better data
...
* This gives us local ranges accurately, as well as fixes bugs with not
getting inline sites. Closes #969
2018-05-17 15:04:40 +01:00
baldurk
e4a15d0ecb
Don't discard short lines (particularly closing braces)
2018-05-17 15:04:26 +01:00
baldurk
e4dcfb7c28
ReadVarLenUInt is available as CodeViewInfo::CVUncompressData
2018-05-17 15:04:26 +01:00
baldurk
941c307e86
Minor changes to cvinfo.h to compile cleanly
2018-05-17 15:04:26 +01:00
baldurk
80724f7130
Add cvconst.h / cvinfo.h from microsoft-pdb repository for SPDB decoding
...
* This will be used to replace/improve the old reverse-engineered decoding.
2018-05-17 15:04:26 +01:00
baldurk
df0246ab68
Remap SPDB line-numbers after sorting filenames
2018-05-17 15:04:23 +01:00
baldurk
7b686d0105
Split SPDB/SDBG processing out into separate files
2018-05-17 01:36:06 +01:00
baldurk
bf09eb59a3
Update DevDriverTools repo to 21c93b7230c4188
2018-05-17 01:36:06 +01:00
baldurk
72eb734964
Correctly pass selected primitive to DebugPixel from pixel history
2018-05-17 01:35:26 +01:00
baldurk
1c75a63176
Handle NULL descriptor structs used to init CPU-side descriptors
2018-05-15 19:44:17 +01:00
baldurk
5c143acccf
Don't underestimate chunk size for empty descriptor heaps
2018-05-15 19:44:17 +01:00
baldurk
d8a64c4172
Fix compilation on GCC 8. Closes #984
2018-05-15 19:44:17 +01:00
baldurk
8c5e8081c5
Fix typo in documentation of ShaderVariable::members
2018-05-15 19:44:17 +01:00
baldurk
a31a5a9cfb
Provide reasonable default values for TextureDisplay struct
2018-05-15 19:44:17 +01:00
baldurk
afb7ea5dab
Improve python wrapping of WindowingData
...
* If we make WindowingData an empty struct not an opaque one, it won't leak memory
* Similarly we need typemaps to allow python to pass plain ints and have them cast to pointers
2018-05-15 19:44:17 +01:00
baldurk
f0d5bc8f2e
Add pass-through unwrapping support for nvenc
...
* nvenc takes a D3D11 object pointer, which must be unwrapped or else it will crash.
2018-05-15 13:02:39 +01:00
baldurk
757671b97f
Add custom right-click menu to texture tabs in texture viewer
2018-05-15 11:27:34 +01:00
baldurk
f12b8b5ee1
Check resource name cache ID to avoid unnecessary model resets
...
* Prevents annoying focus loss when selecting resources when no name has actually changed
2018-05-15 11:27:12 +01:00
baldurk
7eb19ab13f
Add better structured data helper functions to expose to python
2018-05-15 10:44:26 +01:00
baldurk
a35f4bbeae
Add a temporary hack to store the current imageLayout of a descriptor
2018-05-14 18:15:42 +01:00
baldurk
1100d8c0d3
Allow fetching data directly from VkDeviceMemory as well as VkBuffers
2018-05-14 18:15:10 +01:00
baldurk
3861e91d6e
Mark D3D12 pure heaps used for placement as memory, not pools
2018-05-14 16:48:06 +01:00
baldurk
913cbf370b
Make sure to grab a new command buffer each iteration in GetBufferData
2018-05-14 16:47:50 +01:00
baldurk
f698bac6bf
Add hack to avoid breaking VkMarkerRegion with multiple vulkan instances
...
* This can happen when converting to structured data. We let the first
WrappedVulkan 'claim' the static pointer, and ignore others. It's not
perfect but avoids the need to refactor and pass in handle everywhere.
2018-05-14 16:43:57 +01:00
baldurk
8a05e29e86
Fix QRenderDoc project filters - some files weren't included properly
2018-05-14 16:42:34 +01:00
baldurk
b213d54c1d
Account for added Divisor column when setting up vulkan pipeline state
2018-05-14 16:41:25 +01:00
baldurk
5c26270de0
Implement __hash__() for ResourceId
2018-05-14 16:40:58 +01:00
baldurk
f1dcb785a1
Fix implementation of array_setsubscript when deleting a range
2018-05-14 16:40:45 +01:00
baldurk
0cc5ed6145
Handle dstArrayElement when updating current bindings from initial state
2018-05-13 10:13:49 +01:00
baldurk
42aefbaf0e
Fix off-by-one error with gathering array descriptor writes
2018-05-13 10:13:49 +01:00
baldurk
a6ae880b64
Fix a missing else causing doubles to get the wrong expanded format
2018-05-13 10:13:49 +01:00
baldurk
5c40914c32
Set disassembly view to NULL when it's deleted
2018-05-13 10:13:49 +01:00
Hans-Kristian Arntzen
5bf0ae3624
Implement RGB10A2SNorm conversion for postvs.
2018-05-12 21:03:01 +01:00
baldurk
cf508427da
Remove assert that aliased functions will return same pointer
...
* The validation layers break this, although that's a very worrying
trend if KHR and core-promoted functions are allowed to be different
and potentially have different implementations.
2018-05-12 12:58:25 +01:00
baldurk
dc8d4f63d5
Don't discard array descriptor update because some elements are invalid
...
* When checking for validity, some array elements might be valid while
others might not be. This is still OK if the application knew that
only certain elements would be accessed by the given shaders - so we
should still update the others.
2018-05-12 12:57:40 +01:00