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/splat/call/bool.wgsl.expected.fxc.hlsl

@@ -0,0 +1,17 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+bool get_bool() {
+  return true;
+}
+
+void f() {
+  const bool tint_symbol = get_bool();
+  bool2 v2 = bool2((tint_symbol).xx);
+  const bool tint_symbol_1 = get_bool();
+  bool3 v3 = bool3((tint_symbol_1).xxx);
+  const bool tint_symbol_2 = get_bool();
+  bool4 v4 = bool4((tint_symbol_2).xxxx);
+}

test/tint/expressions/splat/call/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,17 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+float get_f32() {
+  return 1.0f;
+}
+
+void f() {
+  const float tint_symbol = get_f32();
+  float2 v2 = float2((tint_symbol).xx);
+  const float tint_symbol_1 = get_f32();
+  float3 v3 = float3((tint_symbol_1).xxx);
+  const float tint_symbol_2 = get_f32();
+  float4 v4 = float4((tint_symbol_2).xxxx);
+}

test/tint/expressions/splat/call/i32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,17 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+int get_i32() {
+  return 1;
+}
+
+void f() {
+  const int tint_symbol = get_i32();
+  int2 v2 = int2((tint_symbol).xx);
+  const int tint_symbol_1 = get_i32();
+  int3 v3 = int3((tint_symbol_1).xxx);
+  const int tint_symbol_2 = get_i32();
+  int4 v4 = int4((tint_symbol_2).xxxx);
+}

test/tint/expressions/splat/call/u32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,17 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+uint get_u32() {
+  return 1u;
+}
+
+void f() {
+  const uint tint_symbol = get_u32();
+  uint2 v2 = uint2((tint_symbol).xx);
+  const uint tint_symbol_1 = get_u32();
+  uint3 v3 = uint3((tint_symbol_1).xxx);
+  const uint tint_symbol_2 = get_u32();
+  uint4 v4 = uint4((tint_symbol_2).xxxx);
+}

test/tint/expressions/splat/expression/bool.wgsl.expected.fxc.hlsl

@@ -0,0 +1,22 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  bool tint_tmp = true;
+  if (!tint_tmp) {
+    tint_tmp = false;
+  }
+  bool2 v2 = bool2(((tint_tmp)).xx);
+  bool tint_tmp_1 = true;
+  if (!tint_tmp_1) {
+    tint_tmp_1 = false;
+  }
+  bool3 v3 = bool3(((tint_tmp_1)).xxx);
+  bool tint_tmp_2 = true;
+  if (!tint_tmp_2) {
+    tint_tmp_2 = false;
+  }
+  bool4 v4 = bool4(((tint_tmp_2)).xxxx);
+}

test/tint/expressions/splat/expression/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,10 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  float2 v2 = float2(((1.0f + 2.0f)).xx);
+  float3 v3 = float3(((1.0f + 2.0f)).xxx);
+  float4 v4 = float4(((1.0f + 2.0f)).xxxx);
+}

test/tint/expressions/splat/expression/i32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,10 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  int2 v2 = int2(((1 + 2)).xx);
+  int3 v3 = int3(((1 + 2)).xxx);
+  int4 v4 = int4(((1 + 2)).xxxx);
+}

test/tint/expressions/splat/expression/u32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,10 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  uint2 v2 = uint2(((1u + 2u)).xx);
+  uint3 v3 = uint3(((1u + 2u)).xxx);
+  uint4 v4 = uint4(((1u + 2u)).xxxx);
+}

test/tint/expressions/splat/immediate/bool.wgsl.expected.fxc.hlsl

@@ -0,0 +1,10 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  bool2 v2 = (true).xx;
+  bool3 v3 = (true).xxx;
+  bool4 v4 = (true).xxxx;
+}

test/tint/expressions/splat/immediate/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,10 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  float2 v2 = (1.0f).xx;
+  float3 v3 = (1.0f).xxx;
+  float4 v4 = (1.0f).xxxx;
+}

test/tint/expressions/splat/immediate/i32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,10 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  int2 v2 = (1).xx;
+  int3 v3 = (1).xxx;
+  int4 v4 = (1).xxxx;
+}

test/tint/expressions/splat/immediate/u32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,10 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  uint2 v2 = (1u).xx;
+  uint3 v3 = (1u).xxx;
+  uint4 v4 = (1u).xxxx;
+}

test/tint/expressions/splat/var/bool.wgsl.expected.fxc.hlsl

@@ -0,0 +1,15 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  bool tint_tmp = true;
+  if (!tint_tmp) {
+    tint_tmp = false;
+  }
+  bool v = (tint_tmp);
+  bool2 v2 = bool2((v).xx);
+  bool3 v3 = bool3((v).xxx);
+  bool4 v4 = bool4((v).xxxx);
+}

test/tint/expressions/splat/var/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,11 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  float v = (1.0f + 2.0f);
+  float2 v2 = float2((v).xx);
+  float3 v3 = float3((v).xxx);
+  float4 v4 = float4((v).xxxx);
+}

test/tint/expressions/splat/var/i32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,11 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  int v = (1 + 2);
+  int2 v2 = int2((v).xx);
+  int3 v3 = int3((v).xxx);
+  int4 v4 = int4((v).xxxx);
+}

test/tint/expressions/splat/var/u32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,11 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  uint v = (1u + 2u);
+  uint2 v2 = uint2((v).xx);
+  uint3 v3 = uint3((v).xxx);
+  uint4 v4 = uint4((v).xxxx);
+}

test/tint/expressions/splat/with_swizzle/f32.wgsl.expected.fxc.hlsl

@@ -0,0 +1,10 @@
+[numthreads(1, 1, 1)]
+void unused_entry_point() {
+  return;
+}
+
+void f() {
+  float a = 1.0f;
+  float b = 1.0f;
+  float c = 1.0f;
+}