Add a test for scatter add. (#238)

* Add a test for scatter add (segfaults on gpus for now).

* Bugfix for the scatter add cuda kernel.

candle-core/src/cuda_backend.rs

@@ -846,7 +846,7 @@ impl<'a> Map2InPlace for ScatterAdd<'a> {
     fn f<T: DeviceRepr + WithDType + ValidAsZeroBits>(
         &self,
         dst: &mut CudaSlice<T>,
-        _dst_shape: &Shape,
+        dst_shape: &Shape,
         src: &CudaSlice<T>,
         src_l: &Layout,
         dev: &CudaDevice,
@@ -874,11 +874,11 @@ impl<'a> Map2InPlace for ScatterAdd<'a> {
         let left_sz: usize = src_l.dims()[..dim].iter().product();
         let right_sz: usize = src_l.dims()[dim + 1..].iter().product();
         let src_dim_sz = src_l.dims()[dim];
-        let ids_dim_sz = ids_l.dims()[dim];
+        let dst_dim_sz = dst_shape.dims()[dim];
         let cfg = LaunchConfig::for_num_elems((left_sz * right_sz) as u32);
         let func = dev.get_or_load_func(&kernel_name::<T>(name), kernels::INDEXING)?;
         // SAFETY: Set later by running the kernel.
-        let params = (ids, &src, dst, left_sz, src_dim_sz, ids_dim_sz, right_sz);
+        let params = (ids, &src, dst, left_sz, src_dim_sz, dst_dim_sz, right_sz);
         // SAFETY: ffi.
         unsafe { func.launch(cfg, params) }.w()?;
         Ok(())

candle-core/tests/tensor_tests.rs

@@ -389,6 +389,46 @@ fn index_add(device: &Device) -> Result<()> {
     Ok(())
 }
 
+fn scatter_add(device: &Device) -> Result<()> {
+    let t = Tensor::arange(0f32, 12f32, device)?.reshape((4, 3))?;
+    assert_eq!(
+        t.to_vec2::<f32>()?,
+        &[
+            [0.0, 1.0, 2.0],
+            [3.0, 4.0, 5.0],
+            [6.0, 7.0, 8.0],
+            [9.0, 10.0, 11.0]
+        ]
+    );
+    let ids = Tensor::new(&[[0u32, 1, 2], [3, 4, 0], [3, 3, 1], [2, 0, 4]], device)?;
+    let init = Tensor::ones((4, 5), DType::F32, device)?;
+    let hs = init.scatter_add(&ids, &t, 1)?;
+    assert_eq!(
+        hs.to_vec2::<f32>()?,
+        &[
+            [1.0, 2.0, 3.0, 1.0, 1.0],
+            [6.0, 1.0, 1.0, 4.0, 5.0],
+            [1.0, 9.0, 1.0, 14.0, 1.0],
+            [11.0, 1.0, 10.0, 1.0, 12.0]
+        ]
+    );
+
+    let init = Tensor::ones((6, 3), DType::F32, device)?;
+    let hs = init.scatter_add(&ids, &t, 0)?;
+    assert_eq!(
+        hs.to_vec2::<f32>()?,
+        &[
+            [1.0, 11.0, 6.0],
+            [1.0, 2.0, 9.0],
+            [10.0, 1.0, 3.0],
+            [10.0, 8.0, 1.0],
+            [1.0, 5.0, 12.0],
+            [1.0, 1.0, 1.0]
+        ]
+    );
+    Ok(())
+}
+
 fn gather(device: &Device) -> Result<()> {
     let ids = Tensor::new(&[[0u32], [2u32], [1u32], [0u32]], device)?;
     let t = Tensor::arange(0f32, 12f32, device)?.reshape((4, 3))?;
@@ -588,3 +628,4 @@ test_device!(broadcasting, broadcasting_cpu, broadcasting_gpu);
 test_device!(index_select, index_select_cpu, index_select_gpu);
 test_device!(index_add, index_add_cpu, index_add_gpu);
 test_device!(gather, gather_cpu, gather_gpu);
+test_device!(scatter_add, scatter_add_cpu, scatter_add_gpu);

candle-kernels/src/indexing.cu

@@ -144,7 +144,6 @@ extern "C" __global__ void FN_NAME(  \
 template<typename T, typename I>
 __device__ void scatter_add(
     const I *ids,
-    const size_t ids_dim_size,
     const T *inp,
     T *out,
     const size_t left_size,
@@ -156,8 +155,8 @@ __device__ void scatter_add(
       for (unsigned int i = blockIdx.x * blockDim.x + threadIdx.x; i < numel; i += blockDim.x * gridDim.x) {
           const size_t pre = i / right_size;
           const size_t post = i % right_size;
-          for (unsigned int j = 0; j < ids_dim_size; ++j) {
-              const size_t src_i = (pre * ids_dim_size + j) * right_size + post;
+          for (unsigned int j = 0; j < src_dim_size; ++j) {
+              const size_t src_i = (pre * src_dim_size + j) * right_size + post;
               const size_t idx = ids[src_i];
               const size_t dst_i = (pre * dst_dim_size + idx) * right_size + post;
               out[dst_i] += inp[src_i];
@@ -168,14 +167,13 @@ __device__ void scatter_add(
 #define SA_OP(TYPENAME, INDEX_TYPENAME, FN_NAME) \
 extern "C" __global__ void FN_NAME(  \
     const INDEX_TYPENAME *ids, \
-    const size_t ids_dim_size, \
     const TYPENAME *inp, \
     TYPENAME *out, \
     const size_t left_size, \
     const size_t src_dim_size, \
     const size_t dst_dim_size, \
     const size_t right_size \
-) { scatter_add(ids, ids_dim_size, inp, out, left_size, src_dim_size, dst_dim_size, right_size); } \
+) { scatter_add(ids, inp, out, left_size, src_dim_size, dst_dim_size, right_size); } \
 
 
 #if __CUDA_ARCH__ >= 800