From f50dcc16e63cc8984da8194d54abc89921e5071c Mon Sep 17 00:00:00 2001
From: Andreas Kloeckner <inform@tiker.net>
Date: Mon, 4 Jul 2011 09:38:41 -0400
Subject: [PATCH] Document g_times_l.
---
doc/source/misc.rst | 6 ++++++
doc/source/runtime.rst | 21 +++++++++++++++++++--
2 files changed, 25 insertions(+), 2 deletions(-)
diff --git a/doc/source/misc.rst b/doc/source/misc.rst
index 326ecc0b..954da015 100644
--- a/doc/source/misc.rst
+++ b/doc/source/misc.rst
@@ -73,6 +73,11 @@ Version 2011.2
This version is currently under development. You can get snapshots from
PyOpenCL's git version control.
+Version 2011.1.2
+----------------
+
+* More bug fixes.
+
Version 2011.1.1
----------------
@@ -108,6 +113,7 @@ Version 2011.1
* Add :func:`pyopencl.enqueue_copy`. Deprecate all other transfer functions.
* Add support for numerous extensions, among them device fission.
* Add a compiler cache.
+* Add the 'g_times_l' keyword arg to kernel execution.
Version 0.92
------------
diff --git a/doc/source/runtime.rst b/doc/source/runtime.rst
index 538ae09d..88e72140 100644
--- a/doc/source/runtime.rst
+++ b/doc/source/runtime.rst
@@ -782,7 +782,7 @@ Programs and Kernels
prg.kernel(queue, n_globals, None, args)
- .. method:: __call__(queue, global_size, local_size, *args, global_offset=None, wait_for=None)
+ .. method:: __call__(queue, global_size, local_size, *args, global_offset=None, wait_for=None, g_times_l=False)
Use :func:`enqueue_nd_range_kernel` to enqueue a kernel execution, after using
:meth:`set_args` to set each argument in turn. See the documentation for
@@ -791,6 +791,11 @@ Programs and Kernels
*None* may be passed for local_size.
+ If *g_times_l* is specified, the global size will be multiplied by the
+ local size. (which makes the behavior more like Nvidia CUDA) In this case,
+ *global_size* and *local_size* also do not have to have the same number
+ of dimensions.
+
.. versionchanged:: 0.92
*local_size* was promoted to third positional argument from being a
keyword argument. The old keyword argument usage will continue to
@@ -802,6 +807,9 @@ Programs and Kernels
its treatment in the wrapper had a bug (now fixed) that prevented
it from working.
+ .. versionchanged:: 2011.1
+ Added the *g_times_l* keyword arg.
+
|comparable|
.. class:: LocalMemory(size)
@@ -814,9 +822,18 @@ Programs and Kernels
The size of local buffer in bytes to be provided.
-.. function:: enqueue_nd_range_kernel(queue, kernel, global_work_size, local_work_size, global_work_offset=None, wait_for=None)
+.. function:: enqueue_nd_range_kernel(queue, kernel, global_work_size, local_work_size, global_work_offset=None, wait_for=None, g_times_l=True)
|std-enqueue-blurb|
+
+ If *g_times_l* is specified, the global size will be multiplied by the
+ local size. (which makes the behavior more like Nvidia CUDA) In this case,
+ *global_size* and *local_size* also do not have to have the same number
+ of dimensions.
+
+ .. versionchanged:: 2011.1
+ Added the *g_times_l* keyword arg.
+
.. function:: enqueue_task(queue, kernel, wait_for=None)
--
GitLab