Commit Graph

425 Commits

Author SHA1 Message Date
Rex Xu
286ca432cf SPV: Memory qualifiers should decorate top-level block members 2017-07-27 14:33:16 +08:00
John Kessenich
d2d3a14237 SPV: Update to latest 1.0 headers, removing redundancies in GLSL.ext.AMD.h. 2017-07-25 21:03:29 -06:00
LoopDawg
ef94b1a5ca Fix dref explicit LOD form of sample with cube texture arrays
The dref parameter was being used as the LOD.  Now it it's properly the dref.
2017-07-24 18:45:37 -06:00
John Kessenich
961cd35b73 SPV: Fix #995: Include memory decorations on parameters. 2017-07-18 03:07:43 -06:00
John Kessenich
fad6297206 SPV: Non-functional: support lists of decorations per parameter. 2017-07-18 02:46:59 -06:00
John Kessenich
735d7e56a6 Address part A of #982: $Global will use std140 instead of HLSL offsets.
From comment about this:
Adjust alignment for HLSL rules
TODO: make this consistent in early phases of code: adjusting this late means inconsistencies with earlier code, which for reflection is an issue.
Until reflection is brought in sync with these adjustments, don't apply to $Global,
which is the most likely to rely on reflection, and least likely to rely
implicit layouts.
2017-07-13 11:39:16 -06:00
Rex Xu
57e65929e4 HLSL: Translate directive [flatten] and [branch] to SPV control mask. 2017-07-06 11:31:33 +08:00
John Kessenich
9645f78293 Merge pull request #965 from chaoc/spv-khr-post-depth-coverage
Implement SPV_KHR_post_depth_coverage
2017-07-05 14:48:19 -06:00
chaoc
c120452754 Implement SPV_KHR_post_depth_coverage
Added support for both extension GL_ARB_post_depth_coverage and GL_EXT_post_depth_coverage.
2017-07-05 12:27:15 -07:00
John Kessenich
17b5f9175d Merge pull request #969 from d3x0r/patch-1
Add option to skip installation
2017-07-04 11:00:42 -06:00
Rohith Chandran
6206091e63 Fix CMake scripts: The set_property script can be used to set only a single property, so now setting the POSITION_INDEPENDENT_CODE property correctly. 2017-07-04 10:53:45 -04:00
d3x0r
4e53d9052d Update CMakeLists.txt 2017-07-04 06:00:35 -07:00
Rex Xu
37cdceed41 Implement extension GL_ARB_shader_stencil_export 2017-06-29 17:50:46 +08:00
John Kessenich
e2ff404f6b Merge pull request #917 from KhronosGroup/remove-redundant-locations
Replace #422: Remove the redundant location setting in AST->SPIR-V.
2017-06-15 12:11:37 -06:00
Rex Xu
301a2bc8a7 SPV: Fix an typo of SPV_AMD_texture_gather_bias_lod 2017-06-14 23:09:39 +08:00
Rex Xu
cabbb788b4 Implement extension GL_AMD_gpu_shader_int16
- Add int16 types (int16_t, uint16_t, i16vec, u16vec).
- Add int16 support to GLSL operators.
- Add int16 type conversions (to int16, from int16).
- Add int16 built-in functions.
2017-06-09 17:11:23 +08:00
Rex Xu
225e0fcadd Implement the extension GL_AMD_texture_gather_bias_lod 2017-06-05 16:41:06 +08:00
John Kessenich
7cdf3fc3c6 Replace #422: Remove the redundant location setting in AST->SPIR-V.
This was redundant in two ways:
1) it replicated algorithms owned in the front end, and
2) it sometimes left location information on both a block and its members.
2017-06-04 13:22:39 -06:00
John Kessenich
e485c7af58 SPV: Debug output: Include OpLine information for execution path.
Note that declaratives are not handled, only procedurals.
2017-05-31 18:50:53 -06:00
John Kessenich
121853f4df SPV: Add OpSource shader source code and file name. 2017-05-31 17:14:15 -06:00
John Kessenich
d6af18f621 Merge pull request #901 from LoopDawg/imat-construct
HLSL: Add imat, umat, and bmat constructors
2017-05-24 23:12:58 -06:00
John Kessenich
84cc15f0d0 HLSL: Fix #903: Don't short-circuit && or ||. 2017-05-24 16:47:32 -06:00
LoopDawg
174ccb8f1d HLSL: Add imat, umat, and bmat constructors
Fixes #894
2017-05-20 21:54:16 -06:00
John Kessenich
7b8c386c5d More non-determinism fixed. 2017-05-19 23:44:51 -06:00
John Kessenich
b6cabc4f24 Fix C++ portability, non-deterministic order of argument evaluation. 2017-05-19 23:29:50 -06:00
John Kessenich
80f92a190a Fix #857: Convert uniform int to local bool for struct alias assignment.
This was done for one direction, but not both directions, so this commit
picks up the other direction.
2017-05-19 23:00:13 -06:00
John Kessenich
0d2b4713c5 HLSL: Don't do logical short-circuits when the operands are bool-vectors.
This seems a bit ill-defined, and was generating code that made OpPhi of two
operands that were Boolean vectors result in a scalar bool.
2017-05-19 20:19:00 -06:00
LoopDawg
65c2eed65d Remapper: handle embedded opcode in OpSpecConstantOp
OpSpecConstantOp contains an embedded opcode which is given as a literal
argument to the OpSpecConstantOp.  The subsequent arguments are as the
embedded op would expect, which may be a mixture of IDs and literals.  This
adds support for that to the remapper binary parser.  Upon seeing such an
embedded op, the parser flips over to parsing the argument list as
appropriate for that opcode.

Fixes #882.
2017-05-18 16:13:04 -06:00
Lei Zhang
2840f63fbc Use unordered_set in ReadableOrderTraverser.
Previously we are using unordered_map to map to bool, which can
be simplified with unordered_set.
2017-05-11 20:48:34 -04:00
Aras Pranckevicius
23d3c712bc [lumped builds] Add include guards (#pragma once) to header files that did not have any.
The codebase seems to use both “#pragma once“ approach, and “#ifndef / #define” approach,
so I picked pragma once as that one is less typing & less brittle.

When glslang is built with some other build system and lumped/unity builds are used,
without the include guards some headers would get included multiple times, leading to duplicate
declaration errors.
2017-05-10 16:58:38 +03:00
John Kessenich
a5c5fb6118 SPV: Emit extension for using StorageClassStorageBuffer. 2017-05-05 05:09:58 -06:00
steve-lunarg
f1709e7146 HLSL: implement [unroll] and [loop] attributes
This adds infrastructure suitable for any front end to create SPIR-V loop
control flags.  The only current front end doing so is HLSL.

[unroll] turns into spv::LoopControlUnrollMask
[loop] turns into spv::LoopControlDontUnrollMask
no specification means spv::LoopControlMaskNone
2017-05-03 13:44:40 -06:00
David Seifert
22afc38b55 Modernise CMake #2
This reverts commit cfc69d95af.
* Change CMAKE_INSTALL_PREFIX default on Windows in order
  to prevent permission denied errors when trying to install
  to "Program Files".
2017-04-29 11:23:15 +02:00
John Kessenich
cfc69d95af Revert "Modernise CMake" 2017-04-28 22:04:24 -06:00
David Seifert
5a5699bdbd Modernise CMake
* Use `GNUInstallDirs` in order to respect GNU conventions.
  This is especially important for multi-arch/multi-lib setups.
* Specify position independent mode building properly, without
  using the historic hack of adding `-fPIC` as a definition.
  This makes the build system more portable.
* Only detect C++ (and not C) to slightly speed up configuring.
* Specify C++11 mode using modern CMake idioms.
* Fix some whitespace issues.
2017-04-28 22:46:52 +02:00
John Kessenich
670271890d SPV: Support test #pragma for generating the StorageBuffer storage class.
Longer term, this storage class should be generated based on the mode
of compilation.
2017-04-19 18:34:49 -06:00
John Kessenich
d1141843c0 SPV Non-functional: fix ordering causing problems with another branch. 2017-04-13 17:08:11 -06:00
John Kessenich
ae79697db1 Merge pull request #827 from amdrexu/feature4
Implement the extension SPV_KHR_16bit_storage
2017-04-08 10:03:45 -06:00
John Kessenich
a09eefd8aa Non-functional: Fix typos in comments. 2017-04-07 15:40:30 -06:00
John Kessenich
6fa17641b5 HLSL: Emit the OpSource HLSL instruction for HLSL, using new headers. 2017-04-07 15:40:01 -06:00
Rex Xu
f89ad98013 Implement the extension SPV_KHR_16bit_storage 2017-04-07 23:22:33 +08:00
John Kessenich
38a425355b SPV: Move to latest public headers. 2017-04-06 17:21:58 -06:00
John Kessenich
4f1403ed1b SPV: Fix #807: use --hlsl-offsets to allow hlsl-style offsets in a buffer.
Corresponds to the EShMsgHlslOffsets flag in messages.
Works for both GLSL and HLSL.
2017-04-05 17:38:20 -06:00
steve-lunarg
e741249b72 HLSL: pass tessellation execution modes through to SPIR-V
The SPIR-V generator had assumed tessellation modes such as
primitive type and vertex order would only appear in tess eval
(domain) shaders.  SPIR-V allows either, and HLSL allows and
possibly requires them to be in the hull shader.

This change:

1. Passes them through for either tessellation stage, and,

2. Does not set up defaults in the domain stage for HLSl compilation,
to avoid conflicting definitions.
2017-03-31 11:47:18 -06:00
John Kessenich
e434ad923e Fix #809: smear scalar condition in OpSelect for selecting vector operands. 2017-03-30 10:16:22 -06:00
John Kessenich
714e58b2fc Merge pull request #801 from amdrexu/bugfix
HLSL: Fix an issue of frexp().
2017-03-29 10:04:30 -06:00
Rex Xu
bcf291a7ba Don't emit NV-specific interface members if NV extensions are disabled. 2017-03-29 23:01:36 +08:00
Rex Xu
470026f9d7 HLSL: Fix an issue of frexp().
The "exp" parameter is floating-point type in HLSL intrinsic while it is
integer type in GLSL built-in function.
2017-03-29 17:12:40 +08:00
John Kessenich
aa3c64c214 Fix #800 (mostly): set of Linux warnings. 2017-03-28 09:52:38 -06:00
John Kessenich
0e7378446b Fix #790: Don't emit NV-specific interface members unless enabled by extension. 2017-03-24 18:40:40 -06:00