[Hopper][WS] Fix use-after-free due to iterator invalidation (#7796)

Getting a crash internally when running `09-persistent-matmul.py`
tutorial, and ASAN reports the following:

```
==7854==ERROR: AddressSanitizer: heap-use-after-free on address 0x7c884c02e800 at pc 0x557f344112d9 bp 0x7b35908a1840 sp 0x7b35908a1838
READ of size 8 at 0x7c884c02e800 thread T1128
    #0 0x557f344112d8 in getNextOperandUsingThisValue third_party/llvm/llvm-project/mlir/include/mlir/IR/UseDefLists.h:43:58
    #1 0x557f344112d8 in operator++ third_party/llvm/llvm-project/mlir/include/mlir/IR/UseDefLists.h:322:39
    #2 0x557f344112d8 in mlir::ResultRange::UseIterator::operator++() third_party/llvm/llvm-project/mlir/lib/IR/OperationSupport.cpp:613:5
    #3 0x557f2ab70625 in mlir::lowerTokenOperations(mlir::Operation*, int, int) third_party/triton/third_party/nvidia/hopper/lib/Transforms/WarpSpecialization/WSLowerToken.cpp:269:27
    #4 0x557f2ab70de8 in mlir::doTokenLowering(mlir::triton::FuncOp&, unsigned int) third_party/triton/third_party/nvidia/hopper/lib/Transforms/WarpSpecialization/WSLowerToken.cpp:321:3
    #5 0x557f2ab2d018 in mlir::NVGPUWarpSpecializationPass::runOnFuncOp(mlir::triton::FuncOp) third_party/triton/third_party/nvidia/hopper/lib/Transforms/WarpSpecialization.cpp:99:5
    #6 0x557f2ab2c5d6 in operator() third_party/triton/third_party/nvidia/hopper/lib/Transforms/WarpSpecialization.cpp:108:55
    #7 0x557f2ab2c5d6 in operator() third_party/llvm/llvm-project/mlir/include/mlir/IR/Visitors.h:304:7
    #8 0x557f2ab2c5d6 in void llvm::function_ref<void (mlir::Operation*)>::callback_fn<std::__u::enable_if<!llvm::is_one_of<mlir::triton::FuncOp, mlir::Operation*, mlir::Region*, mlir::Block*>::value && std::is_same<void, void>::value, void>::type mlir::detail::walk<(mlir::WalkOrder)1, mlir::ForwardIterator, mlir::NVGPUWarpSpecializationPass::runOnOperation()::'lambda'(mlir::triton::FuncOp), mlir::triton::FuncOp, void>(mlir::Operation*, mlir::NVGPUWarpSpecializationPass::runOnOperation()::'lambda'(mlir::triton::FuncOp)&&)::'lambda'(mlir::Operation*)>(long, mlir::Operation*) third_party/llvm/llvm-project/llvm/include/llvm/ADT/STLFunctionalExtras.h:46:12
    #9 0x557f2820ce45 in operator() third_party/llvm/llvm-project/llvm/include/llvm/ADT/STLFunctionalExtras.h:69:12
    #10 0x557f2820ce45 in void mlir::detail::walk<mlir::ForwardIterator>(mlir::Operation*, llvm::function_ref<void (mlir::Operation*)>, mlir::WalkOrder) third_party/llvm/llvm-project/mlir/include/mlir/IR/Visitors.h:152:5
    #11 0x557f2820ce2c in void mlir::detail::walk<mlir::ForwardIterator>(mlir::Operation*, llvm::function_ref<void (mlir::Operation*)>, mlir::WalkOrder) third_party/llvm/llvm-project/mlir/include/mlir/IR/Visitors.h:147:9
    #12 0x557f2ab2c0c9 in walk<(mlir::WalkOrder)1, mlir::ForwardIterator, (lambda at third_party/triton/third_party/nvidia/hopper/lib/Transforms/WarpSpecialization.cpp:108:26), mlir::triton::FuncOp, void> third_party/llvm/llvm-project/mlir/include/mlir/IR/Visitors.h:306:10
    #13 0x557f2ab2c0c9 in walk<(mlir::WalkOrder)1, mlir::ForwardIterator, (lambda at third_party/triton/third_party/nvidia/hopper/lib/Transforms/WarpSpecialization.cpp:108:26), void> third_party/llvm/llvm-project/mlir/include/mlir/IR/Operation.h:798:12
    #14 0x557f2ab2c0c9 in mlir::NVGPUWarpSpecializationPass::runOnOperation() third_party/triton/third_party/nvidia/hopper/lib/Transforms/WarpSpecialization.cpp:108:21
...
```

The problem seems to be that we are iterating through uses, and then
removing some of them inside the loop, which invalidates the iterator.

Author: gflegar

third_party/nvidia/hopper/lib/Transforms/WarpSpecialization/WSLowerToken.cpp

@@ -13,6 +13,7 @@
 #include "triton/Dialect/TritonGPU/IR/Dialect.h"
 #include "triton/Dialect/TritonGPU/Transforms/Utility.h"
 #include "triton/Dialect/TritonNvidiaGPU/IR/Dialect.h"
+#include "llvm/ADT/STLExtras.h"
 
 namespace tt = mlir::triton;
 namespace ttg = mlir::triton::gpu;
@@ -252,7 +253,7 @@ void lowerTokenOperations(Operation *parentOp, int numCTAs,
     if (auto tokenOp = dyn_cast<ttnvws::CreateTokenOp>(op)) {
       // Check to see if it is used by warpSpec. If yes, eraseOperand and
       // eraseArgument.
-      for (OpOperand &use : tokenOp->getUses()) {
+      for (OpOperand &use : llvm::make_early_inc_range(tokenOp->getUses())) {
         Operation *user = use.getOwner();
         if (auto wsOp = dyn_cast<ttg::WarpSpecializeOp>(user)) {
           unsigned opndNum = use.getOperandNumber();