Matrix indexing (#1172)

* Added hlsl-intrinsic test folder.
Enabled ceil as works across targets.

* log10 support.

* Fix float % on CPU/CUDA to match HLSL which is fmod (not fremainder).

* Added log10 tests back to scalar-float.slang

* Don't add the ( for $Sx - it's clearer what's going on without it.

* Works on CUDA/CPU. Problem with asint/asuint do not seem to be found.

* Only asuint exists for double.

* Support countbits on CUDA and C++.

* Fix typo in C++ population count.

* First pass at int vector intrinsic tests.

* Swizzle for int.

* Bit cast tests on CUDA.

* Fix warning on gcc.

* Fix bit-cast-double execution on CUDA.

* scalar-int test working on gcc release.

* GetAt working on CUDA/C++

* Split out runtime index into it's own test.

* Removed SetAt, as can use assignment with GetAt.

* Allowing getAt to be used on matrices.

* Don't need [] on matrix type any longer because use getAt.

* Enable clamp on matrix-int.

* Fix matrix-int.slang test - because clamp behavior varied if min and max were say inverted.
Added runtime indexing version of matrix-int.

7 files changed, 143 insertions, 19 deletions

diff --git a/tests/hlsl-intrinsic/matrix-int-runtime-index.slang b/tests/hlsl-intrinsic/matrix-int-runtime-index.slang
new file mode 100644
index 000000000..e3e2733b7
--- /dev/null
+++ b/tests/hlsl-intrinsic/matrix-int-runtime-index.slang
@@ -0,0 +1,37 @@
+// NOTE we can't test on VK/gl because we will output imat3 - and imat3 is not supported by glsl.
+// NOTE we also can't do this test on fxc - as it does not allow runtime non constant indexing.
+
+//TEST(compute):COMPARE_COMPUTE_EX:-cpu -compute 
+//DISABLE_TEST(compute):COMPARE_COMPUTE_EX:-slang -compute
+//TEST(compute):COMPARE_COMPUTE_EX:-slang -compute -dx12 -use-dxil
+//DISABLE_TEST(compute, vulkan):COMPARE_COMPUTE_EX:-vk -compute
+//TEST(compute, vulkan):COMPARE_COMPUTE_EX:-cuda -compute
+
+//TEST_INPUT:ubuffer(data=[0 0 0 0], stride=4):out,name outputBuffer
+RWStructuredBuffer<int> outputBuffer;
+
+int horizontalAdd(int3 v) { return v.x + v.y + v.z; }
+
+[numthreads(4, 1, 1)]
+void computeMain(uint3 dispatchThreadID : SV_DispatchThreadID)
+{
+    int idx = int(dispatchThreadID.x);
+
+    matrix<int, 3, 3> a = { { 0, 1, 2}, {2, 4, 6}, {7, 21, 32}};
+    matrix<int, 3, 3> b = { { 4, 9, -1}, {-2, 4, 2}, {31, -3, 7}};
+
+    // Do a dynamic row swap
+    {
+        matrix<int, 3, 3> c = a;
+    
+        int row0 = idx % 3;
+        int row1 = (idx + 1) % 3; 
+        int row2 = (idx + 2) % 3;
+        
+        a[row0] = c[0];
+        a[row1] = c[1];
+        a[row2] = c[2];
+    }
+ 
+    outputBuffer[idx] = (a[0].x & 15) + ((a[1].x & 15) << 4) + ((a[2].x & 15) << 8);
+}
\ No newline at end of file
diff --git a/tests/hlsl-intrinsic/matrix-int-runtime-index.slang.expected.txt b/tests/hlsl-intrinsic/matrix-int-runtime-index.slang.expected.txt
new file mode 100644
index 000000000..b31b447b8
--- /dev/null
+++ b/tests/hlsl-intrinsic/matrix-int-runtime-index.slang.expected.txt
@@ -0,0 +1,4 @@
+720
+207
+72
+720
diff --git a/tests/hlsl-intrinsic/matrix-int.slang b/tests/hlsl-intrinsic/matrix-int.slang
new file mode 100644
index 000000000..cd69156a7
--- /dev/null
+++ b/tests/hlsl-intrinsic/matrix-int.slang
@@ -0,0 +1,46 @@
+// NOTE we can't test on VK/gl at the moment because we don't support intrinsics over matrices on that target currently
+
+//TEST(compute):COMPARE_COMPUTE_EX:-cpu -compute 
+//TEST(compute):COMPARE_COMPUTE_EX:-slang -compute
+//TEST(compute):COMPARE_COMPUTE_EX:-slang -compute -dx12 -use-dxil
+//DISABLE_TEST(compute, vulkan):COMPARE_COMPUTE_EX:-vk -compute
+//TEST(compute, vulkan):COMPARE_COMPUTE_EX:-cuda -compute
+
+//TEST_INPUT:ubuffer(data=[0 0 0 0], stride=4):out,name outputBuffer
+RWStructuredBuffer<int> outputBuffer;
+
+int horizontalAdd(int3 v) { return v.x + v.y + v.z; }
+
+[numthreads(4, 1, 1)]
+void computeMain(uint3 dispatchThreadID : SV_DispatchThreadID)
+{
+    int idx = int(dispatchThreadID.x);
+
+    matrix<int, 3, 3> a = { { 0, 1, 2}, {2, 4, 6}, {16, 21, 32}};
+    matrix<int, 3, 3> b = { { 4, 9, -1}, {-2, 4, 2}, {31, -3, 7}};
+
+    matrix<int, 3, 3> t = {};
+    
+    t += max(a, b);
+    t += min(a, b);
+    t += abs(a);
+    t += b % 5;
+
+    {
+        int3 low = int3(3 + idx * 1);
+        int3 high = int3(5 + idx * 2);
+       
+        t += clamp(a, matrix<int, 3, 3>(low, low, low), matrix<int, 3, 3>(high, high, high));
+    }
+       
+    // Access rows and elements, both read and write
+   
+    a[0].x ++;
+    a[1][1] --;
+    a[0] = a[1];
+    a[2].y += b[1].x;
+   
+    t += a;
+   
+    outputBuffer[idx] = horizontalAdd(t[0]) + horizontalAdd(t[1]) + horizontalAdd(t[2]);
+}
\ No newline at end of file
diff --git a/tests/hlsl-intrinsic/matrix-int.slang.expected.txt b/tests/hlsl-intrinsic/matrix-int.slang.expected.txt
new file mode 100644
index 000000000..ad9f3b0fb
--- /dev/null
+++ b/tests/hlsl-intrinsic/matrix-int.slang.expected.txt
@@ -0,0 +1,4 @@
+163
+16E
+179
+184
diff --git a/tests/hlsl-intrinsic/vector-int-runtime-index.slang b/tests/hlsl-intrinsic/vector-int-runtime-index.slang
new file mode 100644
index 000000000..bb704ed1a
--- /dev/null
+++ b/tests/hlsl-intrinsic/vector-int-runtime-index.slang
@@ -0,0 +1,46 @@
+//TEST(compute):COMPARE_COMPUTE_EX:-cpu -compute 
+//DISABLE_TEST(compute):COMPARE_COMPUTE_EX:-slang -compute
+//TEST(compute):COMPARE_COMPUTE_EX:-slang -compute -dx12 -use-dxil
+//TEST(compute, vulkan):COMPARE_COMPUTE_EX:-vk -compute
+//TEST(compute, vulkan):COMPARE_COMPUTE_EX:-cuda -compute
+
+//TEST_INPUT:ubuffer(data=[0 0 0 0], stride=4):out,name outputBuffer
+RWStructuredBuffer<int> outputBuffer;
+
+[numthreads(4, 1, 1)]
+void computeMain(uint3 dispatchThreadID : SV_DispatchThreadID)
+{
+    int idx = int(dispatchThreadID.x);
+
+    int3 a = { idx + 10, idx - 9, idx + 3};
+    int3 b = { idx * 2, idx * 3, idx * 10};
+    
+    int3 t = { 0, 0, 0};
+    
+    t += max(a, b);
+    t += min(a, b);
+    t += abs(a);
+    t += b % 5;
+    
+    t += clamp(a, int3(10), int3(23));
+   
+    // Swizzle
+    t += a.zyx;
+    // Swizzle from scalar
+    t += idx.xxx;
+    
+    {   
+        t += int3(a[idx % 3], a[0], b[2]);
+    }
+   
+    // Writes 
+    {   
+        int3 v = int3(b[(idx + 1) % 3], b[(idx + 2) % 3], b[(idx + 3) % 3]);
+        v[1] += v[0];
+        v[idx & 1] += v[2];
+    
+        t += v;
+    }
+   
+    outputBuffer[idx] = t.x + t.y + t.z;
+}
\ No newline at end of file
diff --git a/tests/hlsl-intrinsic/vector-int-runtime-index.slang.expected.txt b/tests/hlsl-intrinsic/vector-int-runtime-index.slang.expected.txt
new file mode 100644
index 000000000..1ff551cc3
--- /dev/null
+++ b/tests/hlsl-intrinsic/vector-int-runtime-index.slang.expected.txt
@@ -0,0 +1,4 @@
+50
+84
+D0
+103
diff --git a/tests/hlsl-intrinsic/vector-int.slang b/tests/hlsl-intrinsic/vector-int.slang
index 9a3f1faaa..d3327cfd0 100644
--- a/tests/hlsl-intrinsic/vector-int.slang
+++ b/tests/hlsl-intrinsic/vector-int.slang
@@ -1,6 +1,6 @@
 //TEST(compute):COMPARE_COMPUTE_EX:-cpu -compute 
 //TEST(compute):COMPARE_COMPUTE_EX:-slang -compute
-//TEST(compute):COMPARE_COMPUTE_EX:-slang -compute -dx12
+//TEST(compute):COMPARE_COMPUTE_EX:-slang -compute -dx12 -use-dxil
 //TEST(compute, vulkan):COMPARE_COMPUTE_EX:-vk -compute
 //TEST(compute, vulkan):COMPARE_COMPUTE_EX:-cuda -compute
 
@@ -28,23 +28,6 @@ void computeMain(uint3 dispatchThreadID : SV_DispatchThreadID)
     t += a.zyx;
     // Swizzle from scalar
     t += idx.xxx;
-    
-#if 0 
-    // TODO(JS): On C++ not all accesses are turned into getAt/setAt, so disable for now.
-    // Reads
-    {   
-        t += int3(a[idx % 3], a[0], b[2]);
-    }
-   
-    // Writes 
-    {   
-        int3 v = int3(b[(idx + 1) % 3], b[(idx + 2) % 3], b[(idx + 3) % 3]);
-        v[1] += v[0];
-        v[idx & 1] += v[2];
-    
-        t += v;
-    }
-#endif   
-   
+       
     outputBuffer[idx] = t.x + t.y + t.z;
 }
\ No newline at end of file