You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: sycl/doc/extensions/deprecated/sycl_ext_oneapi_group_algorithms.asciidoc
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -146,7 +146,7 @@ Group algorithms are performed collaboratively by the work-items in a group. All
146
146
147
147
Many functions provide at least two overloads: one operating directly on data produced by the work-items in the group, and another operating on a range of data in memory specified by a pair of pointers. If the pointers passed to such a group function are not the same for all work-items in the group, their behavior is undefined. How the elements of a range are processed by the work-items in a group is undefined.
148
148
149
-
Using functions from the group algorithms library inside of a kernel may introduce additional limits on the resources available to user code inside the same kernel (e.g. private memory, work-group local memory). The behavior of these limits is implementation-defined, but must be reflected by calls to kernel querying functions such as +kernel::get_work_group_info+.
149
+
Using functions from the group algorithms library inside of a kernel may introduce additional limits on the resources available to user code inside the same kernel (e.g. private memory, work-group local memory). The behavior of these limits is implementation-defined, but must be reflected by calls to kernel querying functions such as +kernel::get_info+.
150
150
151
151
It is undefined behavior for any of these functions to be invoked within a +parallel_for_work_group+ or +parallel_for_work_item+ context, but this restriction may be lifted in a future version of the proposal.
0 commit comments