Minor bug fixes in scalar_born.cu and elastic.cu. Untested.

Author: ar4

src/deepwave/elastic.cu

@@ -668,12 +668,9 @@ __global__ void forward_kernel_v(
   int64_t x = blockIdx.x * blockDim.x + threadIdx.x;
   int64_t y = blockIdx.y * blockDim.y + threadIdx.y;
   int64_t shot_idx = blockIdx.z * blockDim.z + threadIdx.z;
-  int64_t z = 0;
 #else /* DW_NDIM == 1 */
   int64_t x = blockIdx.x * blockDim.x + threadIdx.x;
   int64_t shot_idx = blockIdx.y * blockDim.y + threadIdx.y;
-  int64_t z = 0;
-  int64_t y = 0;
 #endif
 
 #if DW_NDIM == 3
@@ -899,12 +896,9 @@ __global__ void forward_kernel_sigma(
   int64_t x = blockIdx.x * blockDim.x + threadIdx.x;
   int64_t y = blockIdx.y * blockDim.y + threadIdx.y;
   int64_t shot_idx = blockIdx.z * blockDim.z + threadIdx.z;
-  int64_t z = 0;
 #else /* DW_NDIM == 1 */
   int64_t x = blockIdx.x * blockDim.x + threadIdx.x;
   int64_t shot_idx = blockIdx.y * blockDim.y + threadIdx.y;
-  int64_t z = 0;
-  int64_t y = 0;
 #endif
 
 #if DW_NDIM == 3
@@ -924,9 +918,9 @@ __global__ void forward_kernel_sigma(
 #if DW_NDIM >= 2
       int64_t const pml_y0h = pml_y0;
       int64_t const pml_y1h = MAX(pml_y0, pml_y1 - 1);
-#endif
       int64_t const pml_x0h = pml_x0;
       int64_t const pml_x1h = MAX(pml_x0, pml_x1 - 1);
+#endif
 
 #if DW_NDIM == 3
       int64_t j = z * ny * nx + y * nx + x;
@@ -1168,12 +1162,9 @@ __global__ void backward_kernel_sigma(
   int64_t x = blockIdx.x * blockDim.x + threadIdx.x;
   int64_t y = blockIdx.y * blockDim.y + threadIdx.y;
   int64_t shot_idx = blockIdx.z * blockDim.z + threadIdx.z;
-  int64_t z = 0;
 #else /* DW_NDIM == 1 */
   int64_t x = blockIdx.x * blockDim.x + threadIdx.x;
   int64_t shot_idx = blockIdx.y * blockDim.y + threadIdx.y;
-  int64_t z = 0;
-  int64_t y = 0;
 #endif
 
 #if DW_NDIM == 3
@@ -1435,12 +1426,9 @@ __global__ void backward_kernel_v(
   int64_t x = blockIdx.x * blockDim.x + threadIdx.x;
   int64_t y = blockIdx.y * blockDim.y + threadIdx.y;
   int64_t shot_idx = blockIdx.z * blockDim.z + threadIdx.z;
-  int64_t z = 0;
 #else /* DW_NDIM == 1 */
   int64_t x = blockIdx.x * blockDim.x + threadIdx.x;
   int64_t shot_idx = blockIdx.y * blockDim.y + threadIdx.y;
-  int64_t z = 0;
-  int64_t y = 0;
 #endif
 
 #if DW_NDIM == 3
@@ -1485,9 +1473,9 @@ __global__ void backward_kernel_v(
 #if DW_NDIM >= 2
       int64_t const pml_y0h = pml_y0;
       int64_t const pml_y1h = MAX(pml_y0, pml_y1 - 1);
-#endif
       int64_t const pml_x0h = pml_x0;
       int64_t const pml_x1h = MAX(pml_x0, pml_x1 - 1);
+#endif
 
 #if DW_NDIM == 3
       int64_t j = z * ny * nx + y * nx + x;
@@ -2506,8 +2494,20 @@ extern "C"
 #if DW_NDIM >= 2
           m_sigmayyy, m_sigmaxyy, m_sigmaxyx,
 #endif
-          m_sigmaxxx, dvxdbuoyancy + store_i, dvxdbuoyancy + store_i,
-          dvxdbuoyancy + store_i, grad_buoyancy_z_shot, grad_buoyancy_y_shot,
+          m_sigmaxxx,
+#if DW_NDIM >= 3
+          dvzdbuoyancy + store_i,
+#endif
+#if DW_NDIM >= 2
+          dvydbuoyancy + store_i,
+#endif
+          dvxdbuoyancy + store_i,
+#if DW_NDIM >= 3
+	  grad_buoyancy_z_shot,
+#endif
+#if DW_NDIM >= 2
+	  grad_buoyancy_y_shot,
+#endif
           grad_buoyancy_x_shot,
 #if DW_NDIM >= 3
           az, azh, bz, bzh,
@@ -2567,8 +2567,20 @@ extern "C"
 #if DW_NDIM >= 2
           m_sigmayyyn, m_sigmaxyyn, m_sigmaxyxn,
 #endif
-          m_sigmaxxxn, dvxdbuoyancy + store_i, dvxdbuoyancy + store_i,
-          dvxdbuoyancy + store_i, grad_buoyancy_z_shot, grad_buoyancy_y_shot,
+          m_sigmaxxxn,
+#if DW_NDIM >= 3
+          dvzdbuoyancy + store_i,
+#endif
+#if DW_NDIM >= 2
+          dvydbuoyancy + store_i,
+#endif
+          dvxdbuoyancy + store_i,
+#if DW_NDIM >= 3
+	  grad_buoyancy_z_shot,
+#endif
+#if DW_NDIM >= 2
+	  grad_buoyancy_y_shot,
+#endif
           grad_buoyancy_x_shot,
 #if DW_NDIM >= 3
           az, azh, bz, bzh,