[llvm] [HLSL] Diagnose overlapping resource bindings (PR #140982)

[Justin Bogner via llvm-commits]

================
@@ -50,15 +51,55 @@ static void reportInvalidDirection(Module &M, DXILResourceMap &DRM) {
   }
 }
 
-} // namespace
+static void reportOverlappingError(Module &M, ResourceInfo R1,
+                                   ResourceInfo R2) {
+  SmallString<64> Message;
+  raw_svector_ostream OS(Message);
+  OS << "resource " << R1.getName() << " at register "
+     << R1.getBinding().LowerBound << " overlaps with resource " << R2.getName()
+     << " at register " << R2.getBinding().LowerBound << ", space "
+     << R2.getBinding().Space;
+  M.getContext().diagnose(DiagnosticInfoGeneric(Message));
+}
 
-PreservedAnalyses
-DXILPostOptimizationValidation::run(Module &M, ModuleAnalysisManager &MAM) {
-  DXILResourceMap &DRM = MAM.getResult<DXILResourceAnalysis>(M);
+static void reportOverlappingBinding(Module &M, DXILResourceMap &DRM) {
+  if (DRM.empty())
+    return;
 
+  for (auto ResList :
+       {DRM.srvs(), DRM.uavs(), DRM.cbuffers(), DRM.samplers()}) {
+    if (ResList.empty())
+      continue;
+    const ResourceInfo *PrevRI = &*ResList.begin();
+    for (auto *I = ResList.begin() + 1; I != ResList.end(); ++I) {
+      const ResourceInfo *RI = &*I;
+      if (PrevRI->getBinding().overlapsWith(RI->getBinding())) {
+        reportOverlappingError(M, *PrevRI, *RI);
+        continue;
----------------
bogner wrote:

I'm not sure there's a "correct" answer here, but choosing to continue here makes one case better at the cost of another. Just to confirm the behaviour:
- If we have ranges 1-5, 2-3, and 2-5, we'll refer to the first range instead of the second in the error about the third. This seems fairly neutral / fine either way
- If we have ranges 1-5, 2-3, and 4-5, we'll report an error about the third range, where we wouldn't if we did this the other way
- If we have ranges 1-5, 2-7, and 6-8, we won't report an error about the third range, where we would if we did this the other way

Is this what we want? I suppose it's probably reasonable. Could you make sure there's a test that demonstrates this choice is intentional please?

https://github.com/llvm/llvm-project/pull/140982