| Age | Commit message (Collapse) | Author |
|---|
|
Enables all tests/metal/ tests that can be easily enabled.
These tests were not originally designed as render tests; they are
generally being enabled for pipecleaning purposes, and will not be
rigorously testing the corresponding funcitonality.
Where they cannot be enabled as render tests, and a metallib test wasn't
already enabled, a metallib test was enabled instead (where possible).
Fixes #7892
|
|
Close #8090.
When we do type coerce, we use a cache to store the conversion cost
of different type. The key of the cache is defined by
struct BasicTypeKey
{
uint32_t baseType : 8;
uint32_t dim1 : 4;
uint32_t dim2 : 4;
...
}
where dim1 and dim2 is used for dimension of vector and matrix.
However the dim is only 4 bits, so `vector<int, 16>` will have the same
key as `int`, which is wrong.
Fix the issue by extending it to 8 bit.
Also to make the hash key still within 32 bits, we adjust baseType to 5 bits,
and knownConstantBitCount to 6 bits.
---------
Co-authored-by: kaizhangNV <kazhang@nvidia.com>
|
|
|
|
* Metal: `Interlocked` (atomic) member function support for buffers
fixes: #4654
fixes: #4481
1. Add `Interlocked` (atomic) member function support for buffers to Metal
2. Fix `__getEquivalentStructuredBuffer` so it works with CPP/Metal targets
* add `CompareStore` support
* legalize RWByteAddressBuffer to fully replace StructuredBuffer
* destroy replaced byte-addr buffer
* cleanup as per review and add comment to explain why certain code exists
* fix flow of byte-address-buffer replacement
* toggle on option to translate byteAddrBuffer to StructuredBuffer
* cleanup unused buffers
* add treatGetEquivalentStructuredBufferAsGetThis flag to treat getEquivStructuredBuffer as a byteAddressBuffer
* comment to explain `treatGetEquivalentStructuredBufferAsGetThis`
---------
Co-authored-by: Yong He <yonghe@outlook.com>
|
|
* Implement non member function atomic texture support texture_buffer and texture1d
Fixes: #4538
Related to: #4291, fixes `tests/compute/atomics-buffer.slang`
Texture objects cannot use `__getMetalAtomicRef` to cast objects into atomic value type. [Texture objects mandate use of member functions](https://developer.apple.com/metal/Metal-Shading-Language-Specification.pdf#Texture%20Functions). The implementation is as follows:
* We can detect texture object usage through checking for an `IRImageSubscript` Operation. `__isTextureAccess()` was added to evaluate if we have an `IRImageSubscript` operation at compile time (before `static_assert`). `__isTextureAccess()` only checks if we are targeting Metal.
* We have all parameter data needed to call a texture atomic function embedded inside `IRImageSubscript`. `__extractTextureFromTextureAccess()` and `__extractCoordFromTextureAccess()` was added to extract this data for use with Metal atomics.
Note:
* Metal documentation has various incorrect details (function names)
* Since we currently hardcode metal versions for compiling, the Metal compiler version was changed to target `Metal 3.1` (`slang-gcc-compiler-util.cpp`)
* textures do not permit atomic float operations
* add fallthrough attribute + fix bug with 'exchange instead of xor' + fix warning bug
* incorrect function name fix
* missing filecheck
* disable atomics-buffer.slang compute test since GFX issue causing it to fail
* Array support for metal interlockedAtomic and proper verification of texture with interlockedAtomic functions
* Array support for metal interlockedAtomic
* proper verification of texture with interlockedAtomic functions
note: had to seperate many functions to allow forceInlining to run
* missing getOperand(0)
* push atomic fix for metal
* fix atomic syntax for metal and hlsl emitting extra brackets (breaks tests)
* test changes and meta changes
1. max is 8 rw textures with metal because Metal has this limit. Split up tests to not hit this limit
2. added back `[0]`...,`T` to test since this legalizes metal atomic intrinsic
* macro'ify some of the atomic code
1. addresses review
2. makes code easier to modify in the future (rather than sifting through 1000 lines we can just look at ~10-30
* fix test 'check'
* missing float support due to macro
* add functions macro generates, `InternalAtomicOperationInfo`
---------
Co-authored-by: Yong He <yonghe@outlook.com>
|
