Commit Graph

476 Commits

Author SHA1 Message Date
bunnei
34b2c60f95
Merge pull request #2823 from ReinUsesLisp/shr-clamp
shader/shift: Implement SHR wrapped and clamped variants
2019-09-10 11:56:17 -04:00
ReinUsesLisp
1f43e5296f gl_shader_decompiler: Keep track of written images and mark them as modified 2019-09-05 23:26:05 -03:00
ReinUsesLisp
3a450c1395 kepler_compute: Implement texture queries 2019-09-05 20:35:51 -03:00
ReinUsesLisp
4de04eba39 shader_ir: Implement LD_S
Loads from shared memory.
2019-09-05 01:38:37 -03:00
ReinUsesLisp
f17415d431 shader_ir: Implement ST_S
This instruction writes to a memory buffer shared with threads within
the same work group. It is known as "shared" memory in GLSL.
2019-09-05 01:38:37 -03:00
ReinUsesLisp
77ef4fa907 shader/shift: Implement SHR wrapped and clamped variants
Nvidia defaults to wrapped shifts, but this is undefined behaviour on
OpenGL's spec. Explicitly mask/clamp according to what the guest shader
requires.
2019-09-04 01:55:24 -03:00
ReinUsesLisp
dfae2d141a half_set_predicate: Fix predicate assignments 2019-09-04 01:54:23 -03:00
bunnei
81fbc5370d
Merge pull request #2812 from ReinUsesLisp/f2i-selector
shader_ir/conversion: Implement F2I and F2F F16 selector
2019-09-03 22:35:33 -04:00
bunnei
d4f33b822b
Merge pull request #2811 from ReinUsesLisp/fsetp-fix
float_set_predicate: Add missing negation bit for the second operand
2019-09-03 22:34:34 -04:00
Rodrigo Locatti
4d4f9cc104 video_core: Silent miscellaneous warnings (#2820)
* texture_cache/surface_params: Remove unused local variable

* rasterizer_interface: Add missing documentation commentary

* maxwell_dma: Remove unused rasterizer reference

* video_core/gpu: Sort member declaration order to silent -Wreorder warning

* fermi_2d: Remove unused MemoryManager reference

* video_core: Silent unused variable warnings

* buffer_cache: Silent -Wreorder warnings

* kepler_memory: Remove unused MemoryManager reference

* gl_texture_cache: Add missing override

* buffer_cache: Add missing include

* shader/decode: Remove unused variables
2019-08-30 14:08:00 -04:00
bunnei
f8cc5668f8
Merge pull request #2758 from ReinUsesLisp/packed-tid
shader/decode: Implement S2R Tic
2019-08-29 12:58:43 -04:00
ReinUsesLisp
e3534700d7 shader_ir/conversion: Split int and float selector and implement F2F H1 2019-08-28 16:09:33 -03:00
ReinUsesLisp
b13fbc25b8 shader_ir/conversion: Implement F2I F16 Ra.H1 2019-08-27 23:40:40 -03:00
ReinUsesLisp
6207751b00 float_set_predicate: Add missing negation bit for the second operand 2019-08-27 21:57:43 -03:00
ReinUsesLisp
4e35177e23 shader_ir: Implement VOTE
Implement VOTE using Nvidia's intrinsics. Documentation about these can
be found here
https://developer.nvidia.com/reading-between-threads-shader-intrinsics

Instead of using portable ARB instructions I opted to use Nvidia
intrinsics because these are the closest we have to how Tegra X1
hardware renders.

To stub VOTE on non-Nvidia drivers (including nouveau) this commit
simulates a GPU with a warp size of one, returning what is meaningful
for the instruction being emulated:

* anyThreadNV(value) -> value
* allThreadsNV(value) -> value
* allThreadsEqualNV(value) -> true

ballotARB, also known as "uint64_t(activeThreadsNV())", emits

VOTE.ANY Rd, PT, PT;

on nouveau's compiler. This doesn't match exactly to Nvidia's code

VOTE.ALL Rd, PT, PT;

Which is emulated with activeThreadsNV() by this commit. In theory this
shouldn't really matter since .ANY, .ALL and .EQ affect the predicates
(set to PT on those cases) and not the registers.
2019-08-21 14:50:38 -03:00
bunnei
dfdd20142e
Merge pull request #2777 from ReinUsesLisp/hsetp2-fe3h-fix
half_set_predicate: Fix HSETP2_C constant buffer offset
2019-08-21 10:29:17 -04:00
bunnei
cedc1aab4a
Merge pull request #2753 from FernandoS27/float-convert
Shader_Ir: Implement F16 Variants of F2F, F2I, I2F.
2019-08-21 10:27:57 -04:00
bunnei
ca61e298b3
Merge pull request #2778 from ReinUsesLisp/nop
shader_ir: Implement NOP
2019-08-18 08:51:34 -04:00
ReinUsesLisp
2ff8044806 shader_ir: Implement NOP 2019-08-04 03:02:55 -03:00
ReinUsesLisp
ec0da3ef64 half_set_predicate: Fix HSETP2_C constant buffer offset 2019-08-04 02:50:55 -03:00
ReinUsesLisp
77f1a676a1 decode/half_set_predicate: Fix predicates 2019-07-26 00:12:38 -03:00
bunnei
b0ff3179ef
Merge pull request #2739 from lioncash/cflow
video_core/control_flow: Minor changes/warning cleanup
2019-07-25 13:04:56 -04:00
bunnei
4d26550f5f
Merge pull request #2737 from FernandoS27/track-fix
Shader_Ir: Correct tracking to track from right to left
2019-07-25 12:41:52 -04:00
bunnei
31e8a61527
Merge pull request #2743 from FernandoS27/surpress-assert
Downgrade and suppress a series of GPU asserts and debug messages.
2019-07-25 12:34:36 -04:00
ReinUsesLisp
104641db07 shader/decode: Implement S2R Tic 2019-07-22 16:16:10 -03:00
Fernando Sahmkow
11f4e739bd Shader_Ir: Implement F16 Variants of F2F, F2I, I2F.
This commit takes care of implementing the F16 Variants of the 
conversion instructions and makes sure conversions are done.
2019-07-20 17:38:25 -04:00
Fernando Sahmkow
1158777737 Shader_Ir: Change Debug Asserts for Log Warnings 2019-07-19 22:15:34 -04:00
ReinUsesLisp
45c162444d shader/half_set_predicate: Fix HSETP2 implementation 2019-07-19 22:21:22 -03:00
ReinUsesLisp
6c4985edc9 shader/half_set_predicate: Implement missing HSETP2 variants 2019-07-19 22:20:47 -03:00
Lioncash
c1c89411da video_core/control_flow: Provide operator!= for types with operator==
Provides operational symmetry for the respective structures.
2019-07-18 21:03:31 -04:00
Lioncash
1780e0e3d0 video_core/control_flow: Prevent sign conversion in TryGetBlock()
The return value is a u32, not an s32, so this would result in an
implicit signedness conversion.
2019-07-18 21:03:31 -04:00
Lioncash
a162a844d2 video_core/control_flow: Remove unnecessary BlockStack copy constructor
This is the default behavior of the copy constructor, so it doesn't need
to be specified.

While we're at it we can make the other non-default constructor
explicit.
2019-07-18 21:03:30 -04:00
Lioncash
56bc11d952 video_core/control_flow: Use std::move where applicable
Results in less work being done where avoidable.
2019-07-18 21:03:30 -04:00
Lioncash
e7b39f47f8 video_core/control_flow: Use the prefix variant of operator++ for iterators
Same thing, but potentially allows a standard library implementation to
pick a more efficient codepath.
2019-07-18 21:03:30 -04:00
Lioncash
6885e7e7ec video_core/control_flow: Use empty() member function for checking emptiness
It's what it's there for.
2019-07-18 21:03:30 -04:00
Lioncash
45fa12a05c video_core: Resolve -Wreorder warnings
Ensures that the constructor members are always initialized in the order
that they're declared in.
2019-07-18 21:03:30 -04:00
Lioncash
47df844338 video_core/control_flow: Make program_size for ScanFlow() a std::size_t
Prevents a truncation warning from occurring with MSVC. Also the
internal data structures already treat it as a size_t, so this is just a
discrepancy in the interface.
2019-07-18 21:03:29 -04:00
Lioncash
3df9558593 video_core/control_flow: Place all internally linked types/functions within an anonymous namespace
Previously, quite a few functions were being linked with external
linkage.
2019-07-18 21:03:29 -04:00
Lioncash
1109db86b7 video_core/shader/decode: Prevent sign-conversion warnings
Makes it explicit that the conversions here are intentional.
2019-07-18 21:03:29 -04:00
bunnei
63bda67a34
Merge pull request #2738 from lioncash/shader-ir
shader-ir: Minor cleanup-related changes
2019-07-18 13:52:01 -04:00
Fernando Sahmkow
5a06e33859 Shader_Ir: correct clang format 2019-07-18 10:09:26 -04:00
Fernando Sahmkow
0b65e9335e Shader_Ir: Downgrade precision and rounding asserts to debug asserts.
This commit reduces the sevirity of asserts for FP precision and 
rounding as this are well known and have little to no consequences in 
gpu's accuracy.
2019-07-18 08:17:19 -04:00
Fernando Sahmkow
223a535f3f
Merge pull request #2740 from lioncash/bra
shader/decode/other: Correct branch indirect argument within BRA handling
2019-07-17 14:25:08 -04:00
Lioncash
bebbdc2067 shader_ir: std::move Node instance where applicable
These are std::shared_ptr instances underneath the hood, which means
copying them isn't as cheap as a regular pointer. Particularly so on
weakly-ordered systems.

This avoids atomic reference count increments and decrements where they
aren't necessary for the core set of operations.
2019-07-16 19:49:23 -04:00
Lioncash
60926ac16b shader_ir: Rename Get/SetTemporal to Get/SetTemporary
This is more accurate in terms of describing what the functions are
actually doing. Temporal relates to time, not the setting of a temporary
itself.
2019-07-16 19:47:43 -04:00
Lioncash
44d87ff641 shader_ir: Remove unused includes
Removes unnecessary header dependencies.
2019-07-16 19:47:42 -04:00
Fernando Sahmkow
d614193e49 Shader_Ir: Correct tracking to track from right to left 2019-07-16 15:06:59 -04:00
Fernando Sahmkow
b56e7f870a
Merge pull request #2565 from ReinUsesLisp/track-indirect
shader/track: Track indirect buffers
2019-07-16 14:58:35 -04:00
Lioncash
e2d7dda166 shader/decode/other: Correct branch indirect argument within BRA handling
This appears to have been a copy/paste error introduced within
8a6fc529a9
2019-07-16 12:20:45 -04:00
Fernando Sahmkow
1bdb59fc6e
Merge pull request #2695 from ReinUsesLisp/layer-viewport
gl_shader_decompiler: Implement gl_ViewportIndex and gl_Layer in vertex shaders
2019-07-15 16:28:07 -04:00
ReinUsesLisp
afa8096df5 shader: Allow tracking of indirect buffers without variable offset
While changing this code, simplify tracking code to allow returning
the base address node, this way callers don't have to manually rebuild
it on each invocation.
2019-07-14 22:36:44 -03:00
Fernando Sahmkow
0ec9da2f9f
Merge pull request #2692 from ReinUsesLisp/tlds-f16
shader/texture: Add F16 support for TLDS
2019-07-14 08:44:38 -04:00
Fernando Sahmkow
f2549739d1 shader_ir: Add comments on missing instruction.
Also shows Nvidia's address space on comments.
2019-07-09 17:15:45 -04:00
Fernando Sahmkow
2de7649311 shader_ir: limit explorastion to best known program size. 2019-07-09 08:14:43 -04:00
Fernando Sahmkow
e7c6045a03 control_flow: Correct block breaking algorithm. 2019-07-09 08:14:43 -04:00
Fernando Sahmkow
dc4a93594c control_flow: Assert shaders bigger than limit. 2019-07-09 08:14:42 -04:00
Fernando Sahmkow
e7a88f0ab3 control_flow: Address feedback. 2019-07-09 08:14:42 -04:00
Fernando Sahmkow
34357b110c shader_ir: Correct parsing of scheduling instructions and correct sizing 2019-07-09 08:14:41 -04:00
Fernando Sahmkow
cfb3db1a32 shader_ir: Correct max sizing 2019-07-09 08:14:40 -04:00
Fernando Sahmkow
d45fed3030 shader_ir: Remove unnecessary constructors and use optional for ScanFlow result 2019-07-09 08:14:40 -04:00
Fernando Sahmkow
01b21ee1e8 shader_ir: Corrections, documenting and asserting control_flow 2019-07-09 08:14:39 -04:00
Fernando Sahmkow
d5533b440c shader_ir: Unify blocks in decompiled shaders. 2019-07-09 08:14:39 -04:00
Fernando Sahmkow
926b80102f shader_ir: Decompile Flow Stack 2019-07-09 08:14:38 -04:00
Fernando Sahmkow
459fce3a8f shader_ir: propagate shader size to the IR 2019-07-09 08:14:37 -04:00
Fernando Sahmkow
8a6fc529a9 shader_ir: Implement BRX & BRA.CC 2019-07-09 08:14:37 -04:00
Fernando Sahmkow
c218ae4b02 shader_ir: Remove the old scanner. 2019-07-09 08:14:36 -04:00
Fernando Sahmkow
8af6e6a052 shader_ir: Implement a new shader scanner 2019-07-09 08:14:36 -04:00
ReinUsesLisp
c9d886c84e gl_shader_decompiler: Implement gl_ViewportIndex and gl_Layer in vertex shaders
This commit implements gl_ViewportIndex and gl_Layer in vertex and
geometry shaders. In the case it's used in a vertex shader, it requires
ARB_shader_viewport_layer_array. This extension is available on AMD and
Nvidia devices (mesa and proprietary drivers), but not available on
Intel on any platform. At the moment of writing this description I don't
know if this is a hardware limitation or a driver limitation.

In the case that ARB_shader_viewport_layer_array is not available,
writes to these registers on a vertex shader are ignored, with the
appropriate logging.
2019-07-07 20:42:55 -03:00
Tobias
be020f7621
Delete decode_integer_set.cpp 2019-07-07 21:40:33 +02:00
ReinUsesLisp
d0966b9f7c shader/texture: Add F16 support for TLDS 2019-07-07 16:05:56 -03:00
ReinUsesLisp
10a83653ee decode/texture: Address feedback 2019-06-24 02:05:05 -03:00
Fernando Sahmkow
d1812316e1 texture_cache: Style and Corrections 2019-06-20 21:24:47 -04:00
Fernando Sahmkow
b7de31ac97 shader_ir: Fix image copy rebase issues 2019-06-20 21:38:34 -03:00
ReinUsesLisp
9097301d92 shader: Implement bindless images 2019-06-20 21:38:33 -03:00
ReinUsesLisp
06c4ce8645 shader: Decode SUST and implement backing image functionality 2019-06-20 21:38:33 -03:00
ReinUsesLisp
4e81fc8296 shader: Implement texture buffers 2019-06-20 21:36:12 -03:00
ReinUsesLisp
fe8e6618f2 shader: Split SSY and PBK stack
Hardware testing revealed that SSY and PBK push to a different stack,
allowing code like this:

        SSY label1;
        PBK label2;
        SYNC;
label1: PBK;
label2: EXIT;
2019-06-07 02:18:27 -03:00
ReinUsesLisp
769a50661a shader/node: Minor changes
Reflect std::shared_ptr nature of Node on initializers and remove
constant members in nodes.

Add some commentaries.
2019-06-06 20:03:33 -03:00
ReinUsesLisp
e1b3be7ced shader: Move Node declarations out of the shader IR header
Analysis passes do not have a good reason to depend on shader_ir.h to
work on top of nodes. This splits node-related declarations to their own
file and leaves the IR in shader_ir.h
2019-06-06 20:02:37 -03:00
ReinUsesLisp
bf4dfb3ad4 shader: Use shared_ptr to store nodes and move initialization to file
Instead of having a vector of unique_ptr stored in a vector and
returning star pointers to this, use shared_ptr. While changing
initialization code, move it to a separate file when possible.

This is a first step to allow code analysis and node generation beyond
the ShaderIR class.
2019-06-05 20:41:52 -03:00
bunnei
e3608578e4
Merge pull request #2446 from ReinUsesLisp/tid
shader: Implement S2R Tid{XYZ} and CtaId{XYZ}
2019-05-29 12:21:17 -04:00
bunnei
1a2d90ab09
Merge pull request #2485 from ReinUsesLisp/generic-memory
shader/memory: Implement generic memory stores and loads (ST and LD)
2019-05-24 18:24:26 -04:00
Lioncash
b6dcb1ae4d shader/shader_ir: Make Comment() take a std::string by value
This allows for forming comment nodes without making unnecessary copies
of the std::string instance.

e.g. previously:

Comment(fmt::format("Base address is c[0x{:x}][0x{:x}]",
        cbuf->GetIndex(), cbuf_offset));

Would result in a copy of the string being created, as CommentNode()
takes a std::string by value (a const ref passed to a value parameter
results in a copy).

Now, only one instance of the string is ever moved around. (fmt::format
returns a std::string, and since it's returned from a function by value,
this is a prvalue (which can be treated like an rvalue), so it's moved
into Comment's string parameter), we then move it into the CommentNode
constructor, which then moves the string into its member variable).
2019-05-23 03:01:55 -03:00
Lioncash
228e58d0a5 shader/decode/*: Add missing newline to files lacking them
Keeps the shader code file endings consistent.
2019-05-23 02:55:52 -03:00
Lioncash
87b4c1ac5e shader/decode/*: Eliminate indirect inclusions
Amends cases where we were using things that were indirectly being
satisfied through other headers. This way, if those headers change and
eliminate dependencies on other headers in the future, we don't have
cascading compilation errors.
2019-05-23 02:55:52 -03:00
Lioncash
195b54602f shader/decode/memory: Remove left in debug pragma 2019-05-22 17:08:50 -04:00
ReinUsesLisp
75e7b45d69 shader/memory: Implement ST (generic memory) 2019-05-20 22:41:53 -03:00
ReinUsesLisp
f78ef617b6 shader/memory: Implement LD (generic memory) 2019-05-20 22:38:59 -03:00
ReinUsesLisp
9c3461604c shader: Implement S2R Tid{XYZ} and CtaId{XYZ} 2019-05-20 16:36:49 -03:00
bunnei
d49efbfb4a
Merge pull request #2441 from ReinUsesLisp/al2p
shader: Implement AL2P and ALD.PHYS
2019-05-19 14:02:58 -04:00
Lioncash
e310d943b8 shader/shader_ir: Remove unnecessary inline specifiers
constexpr internally links by default, so the inline specifier is
unnecessary.
2019-05-19 08:23:15 -04:00
Lioncash
212b148923 shader/shader_ir: Simplify constructors for OperationNode
Many of these constructors don't even need to be templated. The only
ones that need to be templated are the ones that actually make use of
the parameter pack.

Even then, since std::vector accepts an initializer list, we can supply
the parameter pack directly to it instead of creating our own copy of
the list, then copying it again into the std::vector.
2019-05-19 08:23:14 -04:00
Lioncash
81e7e63080 shader/shader_ir: Remove unnecessary template parameter packs from Operation() overloads where applicable
These overloads don't actually make use of the parameter pack, so they
can be turned into regular non-template function overloads.
2019-05-19 08:23:14 -04:00
Lioncash
e09ee0ff23 shader/shader_ir: Mark tracking functions as const member functions
These don't actually modify instance state, so they can be marked as
const member functions
2019-05-19 08:23:09 -04:00
Lioncash
ce04ab38bb shader/shader_ir: Place implementations of constructor and destructor in cpp file
Given the class contains quite a lot of non-trivial types, place the
constructor and destructor within the cpp file to avoid inlining
construction and destruction code everywhere the class is used.
2019-05-19 04:02:02 -04:00
Lioncash
e43ba3acd4 video_core/shader/decode/texture: Remove unused variable from GetTld4Code() 2019-05-09 18:49:56 -04:00
Lioncash
9e15193ef8
shader/decode/texture: Remove unused variable
This isn't used anywhere, so we can get rid of it.
2019-05-04 02:10:38 -04:00
ReinUsesLisp
d4df803b2b shader_ir/other: Implement IPA.IDX 2019-05-02 21:46:37 -03:00
ReinUsesLisp
28bffb1ffa shader_ir/memory: Assert on non-32 bits ALD.PHYS 2019-05-02 21:46:25 -03:00
ReinUsesLisp
fe700e1856 shader: Add physical attributes commentaries 2019-05-02 21:46:25 -03:00