Attachment 35744 Details for Bug 28072 – Throttle after doing copyregion/swapbuffers round trip

[patch] Throttle after doing copyregion/swapbuffers round trip

0001-intel-Throttle-after-doing-copyregion-swapbuffers-ro.patch (text/plain), 6.01 KB, created by Kristian Høgsberg on 2010-05-18 20:03:35 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Kristian Høgsberg

Created: 2010-05-18 20:03:35 UTC

Size: 6.01 KB

patch

obsolete

>From b7ef0e9f62e3d324bfd5306099638bc580f3ce43 Mon Sep 17 00:00:00 2001
>From: =?UTF-8?q?Kristian=20H=C3=B8gsberg?= <krh@bitplanet.net>
>Date: Tue, 18 May 2010 21:50:44 -0400
>Subject: [PATCH] intel: Throttle after doing copyregion/swapbuffers round trip
>
>Before we would throttle in the flush callback prior to round-tripping
>to the server to do copyregion or swapbuffer.  Now, instead just note
>that we need to throttle and do it in intel_prepare_render(), which
>will be called after receiving the response from the server but before
>we start rendering the next frame.  Even if the server also throttles
>us in swapbuffer, this just makes the throttling a no-op when we hit
>intel_prepare_render().  With that we can drop the
>using_dri2_swapbuffers hack and just always throttle.
>---
> src/mesa/drivers/dri/intel/intel_batchbuffer.c |    3 +-
> src/mesa/drivers/dri/intel/intel_context.c     |   43 +++++++++++------------
> src/mesa/drivers/dri/intel/intel_context.h     |    2 +-
> src/mesa/drivers/dri/intel/intel_screen.c      |   12 ++-----
> 4 files changed, 26 insertions(+), 34 deletions(-)
>
>diff --git a/src/mesa/drivers/dri/intel/intel_batchbuffer.c b/src/mesa/drivers/dri/intel/intel_batchbuffer.c
>index 9768b0d..ca8e344 100644
>--- a/src/mesa/drivers/dri/intel/intel_batchbuffer.c
>+++ b/src/mesa/drivers/dri/intel/intel_batchbuffer.c
>@@ -130,8 +130,7 @@ _intel_batchbuffer_flush(struct intel_batchbuffer *batch, const char *file,
>    struct intel_context *intel = batch->intel;
>    GLuint used = batch->ptr - batch->map;
> 
>-   if (!intel->using_dri2_swapbuffers &&
>-       intel->first_post_swapbuffers_batch == NULL) {
>+   if (intel->first_post_swapbuffers_batch == NULL) {
>       intel->first_post_swapbuffers_batch = intel->batch->buf;
>       drm_intel_bo_reference(intel->first_post_swapbuffers_batch);
>    }
>diff --git a/src/mesa/drivers/dri/intel/intel_context.c b/src/mesa/drivers/dri/intel/intel_context.c
>index a94f688..bf5323e 100644
>--- a/src/mesa/drivers/dri/intel/intel_context.c
>+++ b/src/mesa/drivers/dri/intel/intel_context.c
>@@ -440,6 +440,25 @@ intel_prepare_render(struct intel_context *intel)
>     */
>    if (intel->is_front_buffer_rendering)
>       intel->front_buffer_dirty = GL_TRUE;
>+
>+   /* We're using glFlush as an indicator that a frame is done, which is
>+    * what DRI2 does before calling SwapBuffers (and means we should catch
>+    * people doing front-buffer rendering, as well)..
>+    *
>+    * Wait for the swapbuffers before the one we just emitted, so we don't
>+    * get too many swaps outstanding for apps that are GPU-heavy but not
>+    * CPU-heavy.
>+    *
>+    * Unfortunately, we don't have a handle to the batch containing the swap,
>+    * and getting our hands on that doesn't seem worth it, so we just us the
>+    * first batch we emitted after the last swap.
>+    */
>+   if (intel->need_throttle && intel->first_post_swapbuffers_batch) {
>+      drm_intel_bo_wait_rendering(intel->first_post_swapbuffers_batch);
>+      drm_intel_bo_unreference(intel->first_post_swapbuffers_batch);
>+      intel->first_post_swapbuffers_batch = NULL;
>+      intel->need_throttle = GL_FALSE;
>+   }
> }
> 
> static void
>@@ -451,8 +470,7 @@ intel_viewport(GLcontext *ctx, GLint x, GLint y, GLsizei w, GLsizei h)
>     if (intel->saved_viewport)
> 	intel->saved_viewport(ctx, x, y, w, h);
> 
>-    if (!intel->using_dri2_swapbuffers &&
>-	!intel->meta.internal_viewport_call && ctx->DrawBuffer->Name == 0) {
>+    if (!intel->meta.internal_viewport_call && ctx->DrawBuffer->Name == 0) {
>        dri2InvalidateDrawable(driContext->driDrawablePriv);
>        dri2InvalidateDrawable(driContext->driReadablePriv);
>     }
>@@ -530,27 +548,8 @@ intel_glFlush(GLcontext *ctx)
>    struct intel_context *intel = intel_context(ctx);
> 
>    intel_flush(ctx);
>-
>    intel_flush_front(ctx);
>-
>-   /* We're using glFlush as an indicator that a frame is done, which is
>-    * what DRI2 does before calling SwapBuffers (and means we should catch
>-    * people doing front-buffer rendering, as well)..
>-    *
>-    * Wait for the swapbuffers before the one we just emitted, so we don't
>-    * get too many swaps outstanding for apps that are GPU-heavy but not
>-    * CPU-heavy.
>-    *
>-    * Unfortunately, we don't have a handle to the batch containing the swap,
>-    * and getting our hands on that doesn't seem worth it, so we just us the
>-    * first batch we emitted after the last swap.
>-    */
>-   if (!intel->using_dri2_swapbuffers &&
>-       intel->first_post_swapbuffers_batch != NULL) {
>-      drm_intel_bo_wait_rendering(intel->first_post_swapbuffers_batch);
>-      drm_intel_bo_unreference(intel->first_post_swapbuffers_batch);
>-      intel->first_post_swapbuffers_batch = NULL;
>-   }
>+   intel->need_throttle = GL_TRUE;
> }
> 
> void
>diff --git a/src/mesa/drivers/dri/intel/intel_context.h b/src/mesa/drivers/dri/intel/intel_context.h
>index dae5896..04d5fc9 100644
>--- a/src/mesa/drivers/dri/intel/intel_context.h
>+++ b/src/mesa/drivers/dri/intel/intel_context.h
>@@ -150,8 +150,8 @@ struct intel_context
> 
>    struct intel_batchbuffer *batch;
>    drm_intel_bo *first_post_swapbuffers_batch;
>+   GLboolean need_throttle;
>    GLboolean no_batch_wrap;
>-   GLboolean using_dri2_swapbuffers;
> 
>    struct
>    {
>diff --git a/src/mesa/drivers/dri/intel/intel_screen.c b/src/mesa/drivers/dri/intel/intel_screen.c
>index 3aed253..36f5ada 100644
>--- a/src/mesa/drivers/dri/intel/intel_screen.c
>+++ b/src/mesa/drivers/dri/intel/intel_screen.c
>@@ -110,23 +110,17 @@ intelDRI2Flush(__DRIdrawable *drawable)
>    if (intel->gen < 4)
>       INTEL_FIREVERTICES(intel);
> 
>+   intel->need_throttle = GL_TRUE;
>+
>    if (intel->batch->map != intel->batch->ptr)
>       intel_batchbuffer_flush(intel->batch);
>-}
>-
>-static void
>-intelDRI2Invalidate(__DRIdrawable *drawable)
>-{
>-   struct intel_context *intel = drawable->driContextPriv->driverPrivate;
> 
>-   intel->using_dri2_swapbuffers = GL_TRUE;
>-   dri2InvalidateDrawable(drawable);
> }
> 
> static const struct __DRI2flushExtensionRec intelFlushExtension = {
>     { __DRI2_FLUSH, __DRI2_FLUSH_VERSION },
>     intelDRI2Flush,
>-    intelDRI2Invalidate,
>+    dri2InvalidateDrawable,
> };
> 
> static __DRIimage *
>-- 
>1.7.0.1
>

Actions: View | Splinter Review

Attachments on bug 28072: 35644 | 35744

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.