gallium: add global buffer memory barrier bit

Currently radeonsi synchronizes after every dispatch and Clover
does nothing to synchronize. This is overzealous, especially with
GL compute, so add a barrier for global buffers.

Signed-off-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Dave Airlie <airlied@redhat.com>

diff --git a/src/gallium/include/pipe/p_defines.h b/src/gallium/include/pipe/p_defines.h
index 8257b4a..6f30f9e 100644 (file)
--- a/src/gallium/include/pipe/p_defines.h
+++ b/src/gallium/include/pipe/p_defines.h
@@ -368,6 +368,7 @@ enum pipe_flush_flags
 #define PIPE_BARRIER_IMAGE             (1 << 8)
 #define PIPE_BARRIER_FRAMEBUFFER       (1 << 9)
 #define PIPE_BARRIER_STREAMOUT_BUFFER  (1 << 10)
+#define PIPE_BARRIER_GLOBAL_BUFFER     (1 << 11)
 
 /**
  * Resource binding flags -- state tracker must specify in advance all

diff --git a/src/gallium/state_trackers/clover/core/kernel.cpp b/src/gallium/state_trackers/clover/core/kernel.cpp
index bce3b52..266d50e 100644 (file)
--- a/src/gallium/state_trackers/clover/core/kernel.cpp
+++ b/src/gallium/state_trackers/clover/core/kernel.cpp
@@ -89,6 +89,8 @@ kernel::launch(command_queue &q,
                              exec.sviews.size(), NULL);
    q.pipe->bind_sampler_states(q.pipe, PIPE_SHADER_COMPUTE, 0,
                                exec.samplers.size(), NULL);
+
+   q.pipe->memory_barrier(q.pipe, PIPE_BARRIER_GLOBAL_BUFFER);
    exec.unbind();
 }