tint/test-runner: Split expectations for FXC and DXC

Change tint's `--fxc` flag to take the path of the FXC compiler DLL.
Have tint attempt to validate with both FXC and DXC if `--validate` is
passed.

Fix the 'dirsWithNoPassExpectations' logic which looks like it got
broken with the tint -> dawn merge. It also incorrectly applied
filepath.FromSlash() on windows.

Change-Id: I0f46aa5c21bc48a2abc48402c41f846aff4a8633
Reviewed-on: https://dawn-review.googlesource.com/c/dawn/+/96800
Reviewed-by: Antonio Maiorano <amaiorano@google.com>
Commit-Queue: Ben Clayton <bclayton@chromium.org>

test/tint/expressions/binary/mul/mat2x4-mat4x2/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,7 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const float2x4 a = float2x4(float4(1.0f, 2.0f, 3.0f, 4.0f), float4(5.0f, 6.0f, 7.0f, 8.0f));
+  const float4x2 b = float4x2(float2(-1.0f, -2.0f), float2(-3.0f, -4.0f), float2(-5.0f, -6.0f), float2(-7.0f, -8.0f));
+  const float4x4 r = mul(b, a);
+  return;
+}

test/tint/expressions/binary/mul/mat3x2-vec3/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,18 @@
+cbuffer cbuffer_data : register(b0, space0) {
+  uint4 data[3];
+};
+
+float3x2 tint_symbol_2(uint4 buffer[3], uint offset) {
+  const uint scalar_offset = ((offset + 0u)) / 4;
+  uint4 ubo_load = buffer[scalar_offset / 4];
+  const uint scalar_offset_1 = ((offset + 8u)) / 4;
+  uint4 ubo_load_1 = buffer[scalar_offset_1 / 4];
+  const uint scalar_offset_2 = ((offset + 16u)) / 4;
+  uint4 ubo_load_2 = buffer[scalar_offset_2 / 4];
+  return float3x2(asfloat(((scalar_offset & 2) ? ubo_load.zw : ubo_load.xy)), asfloat(((scalar_offset_1 & 2) ? ubo_load_1.zw : ubo_load_1.xy)), asfloat(((scalar_offset_2 & 2) ? ubo_load_2.zw : ubo_load_2.xy)));
+}
+
+void main() {
+  const float2 x = mul(asfloat(data[2].xyz), tint_symbol_2(data, 0u));
+  return;
+}

test/tint/expressions/binary/mul/mat3x3-mat3x3/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,7 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const float3x3 a = float3x3(float3(1.0f, 2.0f, 3.0f), float3(4.0f, 5.0f, 6.0f), float3(7.0f, 8.0f, 9.0f));
+  const float3x3 b = float3x3(float3(-1.0f, -2.0f, -3.0f), float3(-4.0f, -5.0f, -6.0f), float3(-7.0f, -8.0f, -9.0f));
+  const float3x3 r = mul(b, a);
+  return;
+}

test/tint/expressions/binary/mul/mat3x3-vec3/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,15 @@
+cbuffer cbuffer_data : register(b0, space0) {
+  uint4 data[4];
+};
+
+float3x3 tint_symbol_2(uint4 buffer[4], uint offset) {
+  const uint scalar_offset = ((offset + 0u)) / 4;
+  const uint scalar_offset_1 = ((offset + 16u)) / 4;
+  const uint scalar_offset_2 = ((offset + 32u)) / 4;
+  return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
+}
+
+void main() {
+  const float3 x = mul(asfloat(data[3].xyz), tint_symbol_2(data, 0u));
+  return;
+}

test/tint/expressions/binary/mul/mat4x2-mat2x4/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,7 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const float4x2 a = float4x2(float2(-1.0f, -2.0f), float2(-3.0f, -4.0f), float2(-5.0f, -6.0f), float2(-7.0f, -8.0f));
+  const float2x4 b = float2x4(float4(1.0f, 2.0f, 3.0f, 4.0f), float4(5.0f, 6.0f, 7.0f, 8.0f));
+  const float2x2 r = mul(b, a);
+  return;
+}

test/tint/expressions/binary/mul/scalar-scalar/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,5 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const float r = (1.0f * 2.0f);
+  return;
+}

test/tint/expressions/binary/mul/scalar-scalar/i32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,5 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const int r = (1 * 2);
+  return;
+}

test/tint/expressions/binary/mul/scalar-scalar/u32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,5 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const uint r = (1u * 2u);
+  return;
+}

test/tint/expressions/binary/mul/scalar-vec3/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,7 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const float a = 4.0f;
+  const float3 b = float3(1.0f, 2.0f, 3.0f);
+  const float3 r = (a * b);
+  return;
+}

test/tint/expressions/binary/mul/scalar-vec3/i32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,7 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const int a = 4;
+  const int3 b = int3(1, 2, 3);
+  const int3 r = (a * b);
+  return;
+}

test/tint/expressions/binary/mul/scalar-vec3/u32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,7 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const uint a = 4u;
+  const uint3 b = uint3(1u, 2u, 3u);
+  const uint3 r = (a * b);
+  return;
+}

test/tint/expressions/binary/mul/vec3-mat3x3/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,15 @@
+cbuffer cbuffer_data : register(b0, space0) {
+  uint4 data[4];
+};
+
+float3x3 tint_symbol_3(uint4 buffer[4], uint offset) {
+  const uint scalar_offset = ((offset + 0u)) / 4;
+  const uint scalar_offset_1 = ((offset + 16u)) / 4;
+  const uint scalar_offset_2 = ((offset + 32u)) / 4;
+  return float3x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz));
+}
+
+void main() {
+  const float3 x = mul(tint_symbol_3(data, 0u), asfloat(data[3].xyz));
+  return;
+}

test/tint/expressions/binary/mul/vec3-mat4x3/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,16 @@
+cbuffer cbuffer_data : register(b0, space0) {
+  uint4 data[5];
+};
+
+float4x3 tint_symbol_3(uint4 buffer[5], uint offset) {
+  const uint scalar_offset = ((offset + 0u)) / 4;
+  const uint scalar_offset_1 = ((offset + 16u)) / 4;
+  const uint scalar_offset_2 = ((offset + 32u)) / 4;
+  const uint scalar_offset_3 = ((offset + 48u)) / 4;
+  return float4x3(asfloat(buffer[scalar_offset / 4].xyz), asfloat(buffer[scalar_offset_1 / 4].xyz), asfloat(buffer[scalar_offset_2 / 4].xyz), asfloat(buffer[scalar_offset_3 / 4].xyz));
+}
+
+void main() {
+  const float4 x = mul(tint_symbol_3(data, 0u), asfloat(data[4].xyz));
+  return;
+}

test/tint/expressions/binary/mul/vec3-scalar/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,6 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const float3 a = float3(1.0f, 2.0f, 3.0f);
+  const float3 r = (a * 4.0f);
+  return;
+}

test/tint/expressions/binary/mul/vec3-scalar/i32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,6 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const int3 a = int3(1, 2, 3);
+  const int3 r = (a * 4);
+  return;
+}

test/tint/expressions/binary/mul/vec3-scalar/u32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,6 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const uint3 a = uint3(1u, 2u, 3u);
+  const uint3 r = (a * 4u);
+  return;
+}

test/tint/expressions/binary/mul/vec3-vec3/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,7 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const float3 a = float3(1.0f, 2.0f, 3.0f);
+  const float3 b = float3(4.0f, 5.0f, 6.0f);
+  const float3 r = (a * b);
+  return;
+}

test/tint/expressions/binary/mul/vec3-vec3/i32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,7 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const int3 a = int3(1, 2, 3);
+  const int3 b = int3(4, 5, 6);
+  const int3 r = (a * b);
+  return;
+}

test/tint/expressions/binary/mul/vec3-vec3/u32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,7 @@
+[numthreads(1, 1, 1)]
+void f() {
+  const uint3 a = uint3(1u, 2u, 3u);
+  const uint3 b = uint3(4u, 5u, 6u);
+  const uint3 r = (a * b);
+  return;
+}