diff --git a/clang/lib/CIR/Lowering/ThroughMLIR/LowerCIRLoopToSCF.cpp b/clang/lib/CIR/Lowering/ThroughMLIR/LowerCIRLoopToSCF.cpp
index 01be972711b1..b7afcb9019f1 100644
--- a/clang/lib/CIR/Lowering/ThroughMLIR/LowerCIRLoopToSCF.cpp
+++ b/clang/lib/CIR/Lowering/ThroughMLIR/LowerCIRLoopToSCF.cpp
@@ -25,7 +25,6 @@
 #include "clang/CIR/Dialect/IR/CIRTypes.h"
 #include "clang/CIR/LowerToMLIR.h"
 #include "llvm/ADT/TypeSwitch.h"
-#include "llvm/IR/Module.h"
 
 using namespace cir;
 using namespace llvm;
@@ -483,6 +482,19 @@ class CIRWhileOpLowering : public mlir::OpConversionPattern<cir::WhileOp> {
       return;
 
     for (auto continueOp : continues) {
+      bool nested = false;
+      // When there is another loop between this WhileOp and the ContinueOp,
+      // we shouldn't change that loop instead.
+      for (mlir::Operation *parent = continueOp->getParentOp();
+           parent != whileOp; parent = parent->getParentOp()) {
+        if (isa<WhileOp>(parent)) {
+          nested = true;
+          break;
+        }
+      }
+      if (nested)
+        continue;
+
       // When the ContinueOp is under an IfOp, a direct replacement of
       // `scf.yield` won't work: the yield would jump out of that IfOp instead.
       // We might need to change the WhileOp itself to achieve the same effect.
diff --git a/clang/lib/CIR/Lowering/ThroughMLIR/LowerCIRToMLIR.cpp b/clang/lib/CIR/Lowering/ThroughMLIR/LowerCIRToMLIR.cpp
index 71752bb19c25..5800773e715b 100644
--- a/clang/lib/CIR/Lowering/ThroughMLIR/LowerCIRToMLIR.cpp
+++ b/clang/lib/CIR/Lowering/ThroughMLIR/LowerCIRToMLIR.cpp
@@ -1552,7 +1552,11 @@ void ConvertCIRToMLIRPass::runOnOperation() {
                          mlir::scf::SCFDialect, mlir::cf::ControlFlowDialect,
                          mlir::math::MathDialect, mlir::vector::VectorDialect,
                          mlir::LLVM::LLVMDialect>();
-  target.addIllegalDialect<cir::CIRDialect>();
+  // We cannot mark cir dialect as illegal before conversion.
+  // The conversion of WhileOp relies on partially preserving operations from
+  // cir dialect, for example the `cir.continue`. If we marked cir as illegal
+  // here, then MLIR would think any remaining `cir.continue` indicates a
+  // failure, which is not what we want.
 
   if (failed(applyPartialConversion(module, target, std::move(patterns))))
     signalPassFailure();
@@ -1616,8 +1620,9 @@ mlir::ModuleOp lowerFromCIRToMLIR(mlir::ModuleOp theModule,
 
   auto result = !mlir::failed(pm.run(theModule));
   if (!result)
-    report_fatal_error(
-        "The pass manager failed to lower CIR to MLIR standard dialects!");
+    theModule.dump(),
+        report_fatal_error(
+            "The pass manager failed to lower CIR to MLIR standard dialects!");
   // Now that we ran all the lowering passes, verify the final output.
   if (theModule.verify().failed())
     report_fatal_error(
diff --git a/clang/test/CIR/Lowering/ThroughMLIR/while-with-continue.cpp b/clang/test/CIR/Lowering/ThroughMLIR/while-with-continue.cpp
index 07ebcfab8ac9..fa13e11b80fb 100644
--- a/clang/test/CIR/Lowering/ThroughMLIR/while-with-continue.cpp
+++ b/clang/test/CIR/Lowering/ThroughMLIR/while-with-continue.cpp
@@ -68,3 +68,38 @@ void while_continue_2() {
   // CHECK:   scf.yield
   // CHECK: }
 }
+
+void while_continue_nested() {
+  int i = 0;
+  while (i < 10) {
+    while (true) {
+      continue;
+      i--;
+    }
+    i++;
+  }
+  // The continue will only work on the inner while.
+
+  // CHECK: scf.while : () -> () {
+  // CHECK:   %[[IV:.+]] = memref.load %alloca[]
+  // CHECK:   %[[TEN:.+]] = arith.constant 10
+  // CHECK:   %[[LT:.+]] = arith.cmpi slt, %[[IV]], %[[TEN]]
+  // CHECK:   scf.condition(%[[LT]])
+  // CHECK: } do {
+  // CHECK:   memref.alloca_scope  {
+  // CHECK:     memref.alloca_scope  {
+  // CHECK:       scf.while : () -> () {
+  // CHECK:         %[[TRUE:.+]] = arith.constant true
+  // CHECK:         scf.condition(%[[TRUE]])
+  // CHECK:       } do {
+  // CHECK:         scf.yield
+  // CHECK:       }
+  // CHECK:     }
+  // CHECK:     %[[IV2:.+]] = memref.load %alloca[]
+  // CHECK:     %[[ONE:.+]] = arith.constant 1
+  // CHECK:     %[[ADD:.+]] = arith.addi %[[IV2]], %[[ONE]]
+  // CHECK:     memref.store %[[ADD]], %alloca[]
+  // CHECK:   }
+  // CHECK:   scf.yield
+  // CHECK: }
+}