Publish NVIDIA update to NV_gpu_multicast not pushed t (#380)
* Publish NVIDIA update to NV_gpu_multicast not pushed t
clarify that the interactions with uniform APIs only apply to
EXT_bindable_uniform (not ARB_uniform_buffer_object).
optionally allow MulticastCopyBufferSubDataNV with buffers lacking
per-GPU storage
* Also fix a non-ASCII quote character.
diff --git a/extensions/NV/NV_gpu_multicast.txt b/extensions/NV/NV_gpu_multicast.txt
index 1fc2e7e..0278792 100644
--- a/extensions/NV/NV_gpu_multicast.txt
+++ b/extensions/NV/NV_gpu_multicast.txt
@@ -25,8 +25,8 @@
Version
- Last Modified Date: January 3, 2019
- Revision: 6
+ Last Modified Date: April 2, 2019
+ Revision: 7
Number
@@ -45,6 +45,8 @@
This extension requires EXT_direct_state_access.
+ This extension interacts with EXT_bindable_uniform
+
Overview
This extension enables novel multi-GPU rendering techniques by providing application control
@@ -201,7 +203,6 @@
The following rules apply to buffer objects with per-GPU storage:
When mapped updates apply to all GPUs (only WRITE_ONLY access is supported).
- When bound to UNIFORM_BUFFER, client uniform updates apply to all GPUs.
When used as the write buffer for CopyBufferSubData or CopyNamedBufferSubData, writes apply
to all GPUs.
@@ -212,8 +213,6 @@
An INVALID_VALUE error is generated if BufferStorage/NamedBufferStorageEXT is called with
PER_GPU_STORAGE_BIT_NV set with MAP_READ_BIT or SPARSE_STORAGE_BIT_ARB.
- An INVALID_OPERATION is generated if a buffer with PER_GPU_STORAGE_BIT_NV is bound to
- UNIFORM_BUFFER and GetUniformfv, GetUniformiv, GetUniformuiv or GetUniformdv is called.
To modify buffer object data on one or more GPUs, the client may use the command
@@ -246,10 +245,28 @@
uint readBuffer, uint writeBuffer,
intptr readOffset, intptr writeOffset, sizeiptr size);
- This command operates similarly to CopyNamedBufferSubData, with the exception that it operates
- on per-GPU instances of the buffer object. The read GPU index is specified by <readGpu> and
- the set of write GPUs is specified by the mask in <writeGpuMask>. The following errors apply
- to MulticastCopyBufferSubDataNV:
+ This command operates similarly to CopyNamedBufferSubData, while adding control over the
+ source and destination GPU(s). The read GPU index is specified by <readGpu> and
+ the set of write GPUs is specified by the mask in <writeGpuMask>.
+
+ Implementations may also support this command with buffers not created with
+ PER_GPU_STORAGE_BIT_NV. This support can be determined with one test copy with an error check
+ (see error discussion below). Note that a buffer created without PER_GPU_STORAGE_BIT_NV is
+ considered to have undefined storage and the behavior of the command depends on the storage
+ type (per-GPU or GPU-shared) currently used for <writeBuffer>. If <writeBuffer> is using
+ GPU-shared storage, the normal error checks apply but the command behaves as if <writeGpuMask>
+ includes all GPUs. If <writeBuffer> is using per-GPU storage, the command behaves as if
+ PER_GPU_STORAGE_BIT_NV were set, however performance may be reduced.
+
+ This following error may apply to MulticastCopyBufferSubDataNV on some implementations and not
+ on others. In earlier revisions of this extension the error was required, therefore
+ applications should perform a test copy using buffers without PER_GPU_STORAGE_BIT_NV before
+ relying on that functionality:
+
+ An INVALID_OPERATION error is generated if the value of BUFFER_STORAGE_FLAGS for <readBuffer>
+ or <writeBuffer> does not have PER_GPU_STORAGE_BIT_NV set.
+
+ The following errors apply to MulticastCopyBufferSubDataNV:
An INVALID_OPERATION error is generated if <readBuffer> or <writeBuffer> is not the name of an
existing buffer object.
@@ -259,8 +276,6 @@
An INVALID_OPERATION error is generated if either the source or destination buffer objects is
mapped, unless they were mapped with MAP_PERSISTENT_BIT set in the Map*BufferRange access
flags.
- An INVALID_OPERATION error is generated if the value of BUFFER_STORAGE_FLAGS for <readBuffer>
- or <writeBuffer> does not have PER_GPU_STORAGE_BIT_NV set.
An INVALID_VALUE error is generated if <readGpu> is greater than or equal to
MULTICAST_GPUS_NV.
An INVALID_OPERATION error is generated if <writeGpuMask> is zero. An INVALID_VALUE error is
@@ -574,6 +589,15 @@
If ARB_sparse_buffer is not supported, any reference to SPARSE_STORAGE_BIT_ARB should be
removed.
+Interactions with EXT_bindable_uniform
+
+ When using the functionality of EXT_bindable_uniform and a per-GPU storage buffer is bound
+ to a bindable location in a program object, client uniform updates apply to all GPUs.
+
+ An INVALID_OPERATION is generated if a buffer with PER_GPU_STORAGE_BIT_NV is bound to a
+ program object's bindable location and GetUniformfv, GetUniformiv, GetUniformuiv or
+ GetUniformdv is called.
+
Errors
Relaxation of INVALID_ENUM errors
@@ -776,6 +800,10 @@
Rev. Date Author Changes
---- -------- -------- -----------------------------------------------
+ 7 04/02/19 jschnarr clarify that the interactions with uniform APIs only apply to
+ EXT_bindable_uniform (not ARB_uniform_buffer_object).
+ optionally allow MulticastCopyBufferSubDataNV with buffers lacking
+ per-GPU storage
6 01/03/19 jschnarr reduce MULTICAST_GPUS_NV minimum to 1
clarify that MULTICAST_GPUS_NV is constant for a context
5 10/07/16 jschnarr trivial typo fix
