Fix FXC compile errors on modulo by zero

Just like for divide, FXC fails with the exact same error when
performing a modulo on a value that FXC determines to be zero. We
address it in the same way as we do for divide.

This also fixes a couple of the vk-gl-cts tests for which I manually
generated expectation files for.

Bug: tint:1083
Change-Id: Ia388bf002112afded542adb791d37e88e35a77ff
Reviewed-on: https://dawn-review.googlesource.com/c/tint/+/74220
Reviewed-by: James Price <jrprice@google.com>
Kokoro: Kokoro <noreply+kokoro@google.com>
Commit-Queue: Antonio Maiorano <amaiorano@google.com>

test/expressions/binary/mod_by_zero/by_constant/scalar-vec3/i32.wgsl

@@ -0,0 +1,6 @@
+[[stage(compute), workgroup_size(1)]]
+fn f() {
+    let a = 4;
+    let b = vec3<i32>(0, 2, 0);
+    let r : vec3<i32> = a % b;
+}

test/expressions/binary/mod_by_zero/by_constant/scalar-vec3/i32.wgsl.expected.hlsl

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

test/expressions/binary/mod_by_zero/by_constant/scalar-vec3/i32.wgsl.expected.msl

@@ -0,0 +1,10 @@
+#include <metal_stdlib>
+
+using namespace metal;
+kernel void f() {
+  int const a = 4;
+  int3 const b = int3(0, 2, 0);
+  int3 const r = (a % b);
+  return;
+}
+

test/expressions/binary/mod_by_zero/by_constant/scalar-vec3/i32.wgsl.expected.spvasm

@@ -0,0 +1,27 @@
+; SPIR-V
+; Version: 1.3
+; Generator: Google Tint Compiler; 0
+; Bound: 16
+; Schema: 0
+               OpCapability Shader
+               OpMemoryModel Logical GLSL450
+               OpEntryPoint GLCompute %f "f"
+               OpExecutionMode %f LocalSize 1 1 1
+               OpName %f "f"
+       %void = OpTypeVoid
+          %1 = OpTypeFunction %void
+        %int = OpTypeInt 32 1
+      %int_4 = OpConstant %int 4
+      %v3int = OpTypeVector %int 3
+      %int_0 = OpConstant %int 0
+      %int_2 = OpConstant %int 2
+         %10 = OpConstantComposite %v3int %int_0 %int_2 %int_0
+%_ptr_Function_v3int = OpTypePointer Function %v3int
+         %14 = OpConstantNull %v3int
+          %f = OpFunction %void None %1
+          %4 = OpLabel
+         %12 = OpVariable %_ptr_Function_v3int Function %14
+         %15 = OpCompositeConstruct %v3int %int_4 %int_4 %int_4
+         %11 = OpSMod %v3int %15 %10
+               OpReturn
+               OpFunctionEnd

test/expressions/binary/mod_by_zero/by_constant/scalar-vec3/i32.wgsl.expected.wgsl

@@ -0,0 +1,6 @@
+[[stage(compute), workgroup_size(1)]]
+fn f() {
+  let a = 4;
+  let b = vec3<i32>(0, 2, 0);
+  let r : vec3<i32> = (a % b);
+}

test/expressions/binary/mod_by_zero/by_constant/scalar-vec3/u32.wgsl

@@ -0,0 +1,6 @@
+[[stage(compute), workgroup_size(1)]]
+fn f() {
+    let a = 4u;
+    let b = vec3<u32>(0u, 2u, 0u);
+    let r : vec3<u32> = a % b;
+}

test/expressions/binary/mod_by_zero/by_constant/scalar-vec3/u32.wgsl.expected.hlsl

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

test/expressions/binary/mod_by_zero/by_constant/scalar-vec3/u32.wgsl.expected.msl

@@ -0,0 +1,10 @@
+#include <metal_stdlib>
+
+using namespace metal;
+kernel void f() {
+  uint const a = 4u;
+  uint3 const b = uint3(0u, 2u, 0u);
+  uint3 const r = (a % b);
+  return;
+}
+

test/expressions/binary/mod_by_zero/by_constant/scalar-vec3/u32.wgsl.expected.spvasm

@@ -0,0 +1,27 @@
+; SPIR-V
+; Version: 1.3
+; Generator: Google Tint Compiler; 0
+; Bound: 16
+; Schema: 0
+               OpCapability Shader
+               OpMemoryModel Logical GLSL450
+               OpEntryPoint GLCompute %f "f"
+               OpExecutionMode %f LocalSize 1 1 1
+               OpName %f "f"
+       %void = OpTypeVoid
+          %1 = OpTypeFunction %void
+       %uint = OpTypeInt 32 0
+     %uint_4 = OpConstant %uint 4
+     %v3uint = OpTypeVector %uint 3
+     %uint_0 = OpConstant %uint 0
+     %uint_2 = OpConstant %uint 2
+         %10 = OpConstantComposite %v3uint %uint_0 %uint_2 %uint_0
+%_ptr_Function_v3uint = OpTypePointer Function %v3uint
+         %14 = OpConstantNull %v3uint
+          %f = OpFunction %void None %1
+          %4 = OpLabel
+         %12 = OpVariable %_ptr_Function_v3uint Function %14
+         %15 = OpCompositeConstruct %v3uint %uint_4 %uint_4 %uint_4
+         %11 = OpUMod %v3uint %15 %10
+               OpReturn
+               OpFunctionEnd

test/expressions/binary/mod_by_zero/by_constant/scalar-vec3/u32.wgsl.expected.wgsl

@@ -0,0 +1,6 @@
+[[stage(compute), workgroup_size(1)]]
+fn f() {
+  let a = 4u;
+  let b = vec3<u32>(0u, 2u, 0u);
+  let r : vec3<u32> = (a % b);
+}