diff --git a/CHANGE_LOG.TXT b/CHANGE_LOG.TXT
index e93963cc42..63d50ab5b4 100644
--- a/CHANGE_LOG.TXT
+++ b/CHANGE_LOG.TXT
@@ -1,3 +1,9 @@
+1.5.1 12/28/2015
+ - Bug fixes:
+ - Fix for incorrect DeviceRadixSort output for some small problems on
+ Maxwell SM52 architectures
+ - Fix for macro redefinition warnings when compiling with Thrust sort
+
//-----------------------------------------------------------------------------
1.5.0 12/14/2015
diff --git a/README.md b/README.md
index bdc366d77b..273bee6672 100644
--- a/README.md
+++ b/README.md
@@ -1,7 +1,7 @@
About CUB
-Current release: v1.5.0 (12/13/2015)
+Current release: v1.5.1 (12/28/2015)
We recommend the [CUB Project Website](http://nvlabs.github.com/cub) and the [cub-users discussion forum](http://groups.google.com/group/cub-users) for further information and examples.
diff --git a/cub/device/dispatch/dispatch_reduce.cuh b/cub/device/dispatch/dispatch_reduce.cuh
index 1e40cb2960..3539b7d822 100644
--- a/cub/device/dispatch/dispatch_reduce.cuh
+++ b/cub/device/dispatch/dispatch_reduce.cuh
@@ -112,7 +112,7 @@ template <
typename OffsetT, ///< Signed integer type for global offsets
typename ReductionOpT, ///< Binary reduction functor type having member T operator()(const T &a, const T &b)
typename T> ///< Data element type that is convertible to the \p value type of \p InputIteratorT
-__launch_bounds__ (int(ChainedPolicyT::ActivePolicy::ReducePolicy::BLOCK_THREADS), 1)
+__launch_bounds__ (int(ChainedPolicyT::ActivePolicy::SingleTilePolicy::BLOCK_THREADS), 1)
__global__ void DeviceReduceSingleTileKernel(
InputIteratorT d_in, ///< [in] Pointer to the input sequence of data items
OutputIteratorT d_out, ///< [out] Pointer to the output aggregate
@@ -122,7 +122,7 @@ __global__ void DeviceReduceSingleTileKernel(
{
// Thread block type for reducing input tiles
typedef AgentReduce<
- typename ChainedPolicyT::ActivePolicy::ReducePolicy,
+ typename ChainedPolicyT::ActivePolicy::SingleTilePolicy,
InputIteratorT,
OffsetT,
ReductionOpT>
diff --git a/docs/download_cub.html b/docs/download_cub.html
index a9f8ae9530..e900a57c53 100644
--- a/docs/download_cub.html
+++ b/docs/download_cub.html
@@ -37,14 +37,14 @@
If your download doesn't start in 3s:
-
-Download CUB!
+
+Download CUB!
diff --git a/docs/html/CHANGE_LOG.TXT b/docs/html/CHANGE_LOG.TXT
index 96973a5e90..63d50ab5b4 100644
--- a/docs/html/CHANGE_LOG.TXT
+++ b/docs/html/CHANGE_LOG.TXT
@@ -1,3 +1,9 @@
+1.5.1 12/28/2015
+ - Bug fixes:
+ - Fix for incorrect DeviceRadixSort output for some small problems on
+ Maxwell SM52 architectures
+ - Fix for macro redefinition warnings when compiling with Thrust sort
+
//-----------------------------------------------------------------------------
1.5.0 12/14/2015
@@ -12,6 +18,14 @@
errors)
- Fix for Git Issue 26 (CUDA error: misaligned address after
cub::DeviceRadixSort::SortKeys())
+ - Fix for incorrect/crash on 0-length problems, e.g., Git Issue 25 (Floating
+ point exception (core dumped) during cub::DeviceRadixSort::SortKeys)
+ - Fix for CUDA 7.5 issues on SM 5.2 with SHFL-based warp-scan and warp-reduction
+ on non-primitive data types (e.g., user-defined structs)
+ - Fix for small radix sorting problems where 0 temporary bytes were
+ required and users code was invoking malloc(0) on some systems where
+ that returns NULL. (Impl assumed was asking for size again and was not
+ running the sort.)
//-----------------------------------------------------------------------------
diff --git a/docs/html/annotated.html b/docs/html/annotated.html
index 2398acf04e..4a582f21f1 100644
--- a/docs/html/annotated.html
+++ b/docs/html/annotated.html
@@ -240,7 +240,7 @@