ggml: skip cublasGemmBatchedEx during graph reservation

cublasGemmBatchedEx fails during graph capture when pool allocations
return fake pointers. This is triggered when NUM_PARALLEL is greater
than 1 for models like gemma4 that use batched matmuls. Skip it
during reservation since the memory tracking is already handled by
the pool allocations.

Fixes #15249

llama/patches/0020-ggml-No-alloc-mode.patch

@@ -11,9 +11,9 @@ must be recreated with no-alloc set to false before loading data.
  ggml/include/ggml-backend.h     |   1 +
  ggml/src/ggml-backend-impl.h    |  16 +++
  ggml/src/ggml-backend.cpp       |  75 ++++++++++-
- ggml/src/ggml-cuda/common.cuh   |  62 ++++++++-
+ ggml/src/ggml-cuda/common.cuh   |  83 +++++++++++-
  ggml/src/ggml-cuda/ggml-cuda.cu | 224 ++++++++++++++++++++++++++------
- 5 files changed, 333 insertions(+), 45 deletions(-)
+ 5 files changed, 354 insertions(+), 45 deletions(-)
 
 diff --git a/ggml/include/ggml-backend.h b/ggml/include/ggml-backend.h
 index 93c95602d..dbbb61d9c 100644
@@ -229,7 +229,7 @@ diff --git a/ggml/src/ggml-cuda/common.cuh b/ggml/src/ggml-cuda/common.cuh
 index 9fcb2f9fd..e800ee8f6 100644
 --- a/ggml/src/ggml-cuda/common.cuh
 +++ b/ggml/src/ggml-cuda/common.cuh
-@@ -37,6 +37,41 @@
+@@ -37,6 +37,62 @@
  #include "vendors/cuda.h"
  #endif // defined(GGML_USE_HIP)
  
@@ -261,12 +261,33 @@ index 9fcb2f9fd..e800ee8f6 100644
 +    }
 +}
 +
++static cublasStatus_t cublasGemmBatchedExReserve(
++        cublasHandle_t handle, cublasOperation_t transa, cublasOperation_t transb,
++        int m, int n, int k,
++        const void *alpha,
++        const void *const Aarray[], cudaDataType_t Atype, int lda,
++        const void *const Barray[], cudaDataType_t Btype, int ldb,
++        const void *beta,
++        void *const Carray[], cudaDataType_t Ctype, int ldc,
++        int batchCount,
++        cublasComputeType_t computeType, cublasGemmAlgo_t algo) {
++    if (!reserving_graph) {
++        return cublasGemmBatchedEx(handle, transa, transb, m, n, k,
++            alpha, Aarray, Atype, lda, Barray, Btype, ldb,
++            beta, Carray, Ctype, ldc, batchCount, computeType, algo);
++    } else {
++        return CUBLAS_STATUS_SUCCESS;
++    }
++}
++
 +#undef cudaMemcpyAsync
 +#define cudaMemcpyAsync cudaMemcpyAsyncReserve
 +#undef cudaMemcpy2DAsync
 +#define cudaMemcpy2DAsync cudaMemcpy2DAsyncReserve
 +#undef cudaMemsetAsync
 +#define cudaMemsetAsync cudaMemsetAsyncReserve
++#undef cublasGemmBatchedEx
++#define cublasGemmBatchedEx cublasGemmBatchedExReserve
 +
  #define STRINGIZE_IMPL(...) #__VA_ARGS__
  #define STRINGIZE(...) STRINGIZE_IMPL(__VA_ARGS__)

ml/backend/ggml/ggml/src/ggml-cuda/common.cuh

@@ -65,12 +65,33 @@ static cudaError_t cudaMemsetAsyncReserve ( void* devPtr, int value, size_t coun
     }
 }
 
+static cublasStatus_t cublasGemmBatchedExReserve(
+        cublasHandle_t handle, cublasOperation_t transa, cublasOperation_t transb,
+        int m, int n, int k,
+        const void *alpha,
+        const void *const Aarray[], cudaDataType_t Atype, int lda,
+        const void *const Barray[], cudaDataType_t Btype, int ldb,
+        const void *beta,
+        void *const Carray[], cudaDataType_t Ctype, int ldc,
+        int batchCount,
+        cublasComputeType_t computeType, cublasGemmAlgo_t algo) {
+    if (!reserving_graph) {
+        return cublasGemmBatchedEx(handle, transa, transb, m, n, k,
+            alpha, Aarray, Atype, lda, Barray, Btype, ldb,
+            beta, Carray, Ctype, ldc, batchCount, computeType, algo);
+    } else {
+        return CUBLAS_STATUS_SUCCESS;
+    }
+}
+
 #undef cudaMemcpyAsync
 #define cudaMemcpyAsync cudaMemcpyAsyncReserve
 #undef cudaMemcpy2DAsync
 #define cudaMemcpy2DAsync cudaMemcpy2DAsyncReserve
 #undef cudaMemsetAsync
 #define cudaMemsetAsync cudaMemsetAsyncReserve
+#undef cublasGemmBatchedEx
+#define cublasGemmBatchedEx cublasGemmBatchedExReserve
 
 #define STRINGIZE_IMPL(...) #__VA_ARGS__
 #define STRINGIZE(...) STRINGIZE_IMPL(__VA_ARGS__)