From 2b52fbd24a63e6d43081b4db0913b6e9cca8e400 Mon Sep 17 00:00:00 2001
From: Chao Liu <lc.roy86@gmail.com>
Date: Wed, 16 Jan 2019 12:58:44 -0600
Subject: [PATCH] bug fix

---
 .../gridwise_implicit_gemm_convolution_nchw_kcsr.cuh   | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)

diff --git a/src/include/gridwise_implicit_gemm_convolution_nchw_kcsr.cuh b/src/include/gridwise_implicit_gemm_convolution_nchw_kcsr.cuh
index 662e839b52..67becb5405 100644
--- a/src/include/gridwise_implicit_gemm_convolution_nchw_kcsr.cuh
+++ b/src/include/gridwise_implicit_gemm_convolution_nchw_kcsr.cuh
@@ -94,7 +94,7 @@ __global__ void gridwise_implicit_gemm_convolution_nchw_kcsr(InGlobalDesc,
     constexpr auto out_hkwn_thread_desc =
         make_ConstantTensorDescriptor(Sequence<HoPerThread, KPerThread, WoPerThread, NPerThread>{});
 
-#if 1
+#if 0
     if(get_thread_local_1d_id() == 0 && get_block_1d_id() == 0)
     {
         print_ConstantTensorDescriptor(in_nchw_block_desc, "in_nchw_block_desc");
@@ -156,7 +156,6 @@ __global__ void gridwise_implicit_gemm_convolution_nchw_kcsr(InGlobalDesc,
     for(unsigned c_block_data_begin = 0; c_block_data_begin < in_nchw_global_desc.GetLength(I1);
         c_block_data_begin += CPerBlock, __syncthreads())
     {
-#if 1
         // input: global mem to LDS,
         //   convert 4d-tensor in[N,C,Hi,Wi] to matrix in_matrix[C,Hi*Wi*N]
         blockwise_4d_tensor_copy_reorder_by_get_dst_from_src<BlockSize>(
@@ -169,9 +168,7 @@ __global__ void gridwise_implicit_gemm_convolution_nchw_kcsr(InGlobalDesc,
             p_in_block,
             in_nchw_block_desc.GetLengths(),
             reorder_chwn_from_nchw);
-#endif
 
-#if 1
         // weight: global mem to LDS,
         blockwise_4d_tensor_copy_reorder_by_get_dst_from_src<BlockSize>(
             wei_kcsr_global_desc,
@@ -181,11 +178,9 @@ __global__ void gridwise_implicit_gemm_convolution_nchw_kcsr(InGlobalDesc,
             p_wei_block,
             wei_kcsr_block_desc.GetLengths(),
             reorder_srck_from_kcsr);
-#endif
 
         __syncthreads();
 
-#if 1
         // a series of batched GEMM
         for(unsigned s = 0; s < S; ++s)
         {
@@ -194,12 +189,11 @@ __global__ void gridwise_implicit_gemm_convolution_nchw_kcsr(InGlobalDesc,
                 auto f_accum = [](auto& c, const auto&& ab) { c += ab; };
 
                 blockwise_batch_gemm.run(p_wei_block + wei_srck_block_desc.Get1dIndex(s, r, 0, 0),
-                                         p_in_block + in_chwn_block_desc.Get1dIndex(0, 0, r, 0),
+                                         p_in_block + in_chwn_block_desc.Get1dIndex(0, s, r, 0),
                                          p_out_thread,
                                          f_accum);
             }
         }
-#endif
     }
 
     const auto matrix_c_index =