|
@@ -813,7 +813,7 @@ various targets.
|
|
|
@item @emph{Fields}:
|
|
@item @emph{Fields}:
|
|
|
@table @asis
|
|
@table @asis
|
|
|
@item @code{where}:
|
|
@item @code{where}:
|
|
|
-Indicates which types of processing units are able to execute that codelet.
|
|
|
|
|
|
|
+Indicates which types of processing units are able to execute the codelet.
|
|
|
@code{STARPU_CPU|STARPU_CUDA} for instance indicates that the codelet is
|
|
@code{STARPU_CPU|STARPU_CUDA} for instance indicates that the codelet is
|
|
|
implemented for both CPU cores and CUDA devices while @code{STARPU_GORDON}
|
|
implemented for both CPU cores and CUDA devices while @code{STARPU_GORDON}
|
|
|
indicates that it is only available on Cell SPUs.
|
|
indicates that it is only available on Cell SPUs.
|
|
@@ -892,17 +892,17 @@ argument.
|
|
|
In the case of the Cell processor, the @code{cl_arg} pointer is not directly
|
|
In the case of the Cell processor, the @code{cl_arg} pointer is not directly
|
|
|
given to the SPU function. A buffer of size @code{cl_arg_size} is allocated on
|
|
given to the SPU function. A buffer of size @code{cl_arg_size} is allocated on
|
|
|
the SPU. This buffer is then filled with the @code{cl_arg_size} bytes starting
|
|
the SPU. This buffer is then filled with the @code{cl_arg_size} bytes starting
|
|
|
-at address @code{cl_arg}. In that case, the argument given to the SPU codelet
|
|
|
|
|
|
|
+at address @code{cl_arg}. In this case, the argument given to the SPU codelet
|
|
|
is therefore not the @code{cl_arg} pointer, but the address of the buffer in
|
|
is therefore not the @code{cl_arg} pointer, but the address of the buffer in
|
|
|
local store (LS) instead. This field is ignored for CPU, CUDA and OpenCL
|
|
local store (LS) instead. This field is ignored for CPU, CUDA and OpenCL
|
|
|
codelets.
|
|
codelets.
|
|
|
|
|
|
|
|
@item @code{callback_func} (optional) (default = @code{NULL}):
|
|
@item @code{callback_func} (optional) (default = @code{NULL}):
|
|
|
This is a function pointer of prototype @code{void (*f)(void *)} which
|
|
This is a function pointer of prototype @code{void (*f)(void *)} which
|
|
|
-specifies a possible callback. If that pointer is non-null, the callback
|
|
|
|
|
|
|
+specifies a possible callback. If this pointer is non-null, the callback
|
|
|
function is executed @emph{on the host} after the execution of the task. The
|
|
function is executed @emph{on the host} after the execution of the task. The
|
|
|
callback is passed the value contained in the @code{callback_arg} field. No
|
|
callback is passed the value contained in the @code{callback_arg} field. No
|
|
|
-callback is executed if that field is null.
|
|
|
|
|
|
|
+callback is executed if the field is null.
|
|
|
|
|
|
|
|
@item @code{callback_arg} (optional) (default = @code{NULL}):
|
|
@item @code{callback_arg} (optional) (default = @code{NULL}):
|
|
|
This is the pointer passed to the callback function. This field is ignored if
|
|
This is the pointer passed to the callback function. This field is ignored if
|
|
@@ -943,7 +943,7 @@ returned by @code{starpu_worker_get_id}). This field is ignored if
|
|
|
@item @code{detach} (optional) (default = 1):
|
|
@item @code{detach} (optional) (default = 1):
|
|
|
If this flag is set, it is not possible to synchronize with the task
|
|
If this flag is set, it is not possible to synchronize with the task
|
|
|
by the means of @code{starpu_task_wait} later on. Internal data structures
|
|
by the means of @code{starpu_task_wait} later on. Internal data structures
|
|
|
-are only guaranteed to be freed once @code{starpu_task_wait} is called if that
|
|
|
|
|
|
|
+are only guaranteed to be freed once @code{starpu_task_wait} is called if the
|
|
|
flag is not set.
|
|
flag is not set.
|
|
|
|
|
|
|
|
@item @code{destroy} (optional) (default = 1):
|
|
@item @code{destroy} (optional) (default = 1):
|
|
@@ -1043,7 +1043,7 @@ asynchronous tasks by the means of tags, using the @code{starpu_tag_wait}
|
|
|
function for instance.
|
|
function for instance.
|
|
|
@item @emph{Return value}:
|
|
@item @emph{Return value}:
|
|
|
In case of success, this function returns 0, a return value of @code{-ENODEV}
|
|
In case of success, this function returns 0, a return value of @code{-ENODEV}
|
|
|
-means that there is no worker able to process that task (e.g. there is no GPU
|
|
|
|
|
|
|
+means that there is no worker able to process this task (e.g. there is no GPU
|
|
|
available and this task is only implemented for CUDA devices).
|
|
available and this task is only implemented for CUDA devices).
|
|
|
@item @emph{Prototype}:
|
|
@item @emph{Prototype}:
|
|
|
@code{int starpu_task_submit(struct starpu_task *task);}
|
|
@code{int starpu_task_submit(struct starpu_task *task);}
|
|
@@ -1086,7 +1086,7 @@ dependencies between tasks by the means of those tags. To do so, fill the
|
|
|
@code{tag_id} field of the @code{starpu_task} structure with a tag number (can
|
|
@code{tag_id} field of the @code{starpu_task} structure with a tag number (can
|
|
|
be arbitrary) and set the @code{use_tag} field to 1.
|
|
be arbitrary) and set the @code{use_tag} field to 1.
|
|
|
|
|
|
|
|
-If @code{starpu_tag_declare_deps} is called with that tag number, the task will
|
|
|
|
|
|
|
+If @code{starpu_tag_declare_deps} is called with this tag number, the task will
|
|
|
not be started until the tasks which holds the declared dependency tags are
|
|
not be started until the tasks which holds the declared dependency tags are
|
|
|
completed.
|
|
completed.
|
|
|
@end table
|
|
@end table
|
|
@@ -1178,8 +1178,8 @@ terminated.
|
|
|
@table @asis
|
|
@table @asis
|
|
|
@item @emph{Description}:
|
|
@item @emph{Description}:
|
|
|
This function releases the resources associated to tag @code{id}. It can be
|
|
This function releases the resources associated to tag @code{id}. It can be
|
|
|
-called once the corresponding task has been executed and when there is no tag
|
|
|
|
|
-that depend on that one anymore.
|
|
|
|
|
|
|
+called once the corresponding task has been executed and when there is
|
|
|
|
|
+no other tag that depend on this tag anymore.
|
|
|
@item @emph{Prototype}:
|
|
@item @emph{Prototype}:
|
|
|
@code{void starpu_tag_remove(starpu_tag_t id);}
|
|
@code{void starpu_tag_remove(starpu_tag_t id);}
|
|
|
@end table
|
|
@end table
|
|
@@ -1550,7 +1550,7 @@ different interfaces describing each buffer. In the case of the @b{vector
|
|
|
interface}, the location of the vector (resp. its length) is accessible in the
|
|
interface}, the location of the vector (resp. its length) is accessible in the
|
|
|
@code{ptr} (resp. @code{nx}) of this array. Since the vector is accessed in a
|
|
@code{ptr} (resp. @code{nx}) of this array. Since the vector is accessed in a
|
|
|
read-write fashion, any modification will automatically affect future accesses
|
|
read-write fashion, any modification will automatically affect future accesses
|
|
|
-to that vector made by other tasks.
|
|
|
|
|
|
|
+to this vector made by other tasks.
|
|
|
|
|
|
|
|
@node Scaling a Vector (hybrid)
|
|
@node Scaling a Vector (hybrid)
|
|
|
@section Vector Scaling on an Hybrid CPU/GPU Machine
|
|
@section Vector Scaling on an Hybrid CPU/GPU Machine
|
Nathalie Furmento