diff options
author | Chris Wilson <[email protected]> | 2015-02-26 11:25:18 +0000 |
---|---|---|
committer | Chris Wilson <[email protected]> | 2015-03-18 09:33:33 +0000 |
commit | 8b9bd19021c0efef33d66ae24f8871b826d66e8a (patch) | |
tree | 083c874b30a10d14c4ebfea0d82cc5d7ccf273ed /src/mesa | |
parent | 27bf37ba05b69ebf6f373d1637a26b4839265921 (diff) |
i965: Throttle rendering to an fbo
When rendering to an fbo, even though it may be acting as a winsys
frontbuffer or just generally, we never throttle. However, when rendering
to an fbo, there is no natural frame boundary. Conventionally we use
SwapBuffers and glFinish, but potential callers avoid often glFinish for
being too heavy handed (waiting on all outstanding rendering to complete).
The kernel provides a soft-throttling option for this case that waits for
rendering older than 20ms to be complete (that's a little too lax to be
used for swapbuffers, but is here a useful safety net). The remaining
choice is then either never to throttle, throttle after every draw call,
or at after intermediate user defined point such as glFlush and thus all the
implied flushes. This patch opts for the latter as that is the current
method used for flushing to front buffers.
v2: Defer the throttling from inside the flush to the next
intel_prepare_render() and switch non-fbo frontbuffer throttling over to
use the same lax method. The issuing being that
glFlush()/intel_prepare_read() is just as likely to be called inside a
tight loop and not at "frame" boundaries.
v3: Rename from need_front_throttle to need_flush_throttle to avoid any
ambiguity between front buffer rendering and fbo rendering. (Chad)
v4: Whitespace
Signed-off-by: Chris Wilson <[email protected]>
Cc: Daniel Vetter <[email protected]>
Cc: Kenneth Graunke <[email protected]>
Cc: Ben Widawsky <[email protected]>
Cc: Kristian Høgsberg <[email protected]>
Cc: Chad Versace <[email protected]>
Cc: Ian Romanick <[email protected]>
Reviewed-by: Chad Versace <[email protected]>
Diffstat (limited to 'src/mesa')
-rw-r--r-- | src/mesa/drivers/dri/i965/brw_context.c | 16 | ||||
-rw-r--r-- | src/mesa/drivers/dri/i965/brw_context.h | 14 | ||||
-rw-r--r-- | src/mesa/drivers/dri/i965/intel_screen.c | 8 |
3 files changed, 29 insertions, 9 deletions
diff --git a/src/mesa/drivers/dri/i965/brw_context.c b/src/mesa/drivers/dri/i965/brw_context.c index 0881e489fe2..51203726438 100644 --- a/src/mesa/drivers/dri/i965/brw_context.c +++ b/src/mesa/drivers/dri/i965/brw_context.c @@ -234,8 +234,8 @@ intel_glFlush(struct gl_context *ctx) intel_batchbuffer_flush(brw); intel_flush_front(ctx); - if (brw_is_front_buffer_drawing(ctx->DrawBuffer)) - brw->need_throttle = true; + + brw->need_flush_throttle = true; } static void @@ -1245,12 +1245,20 @@ intel_prepare_render(struct brw_context *brw) * 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 (brw->need_throttle && brw->first_post_swapbuffers_batch) { + if (brw->need_swap_throttle && brw->first_post_swapbuffers_batch) { if (!brw->disable_throttling) drm_intel_bo_wait_rendering(brw->first_post_swapbuffers_batch); drm_intel_bo_unreference(brw->first_post_swapbuffers_batch); brw->first_post_swapbuffers_batch = NULL; - brw->need_throttle = false; + brw->need_swap_throttle = false; + /* Throttling here is more precise than the throttle ioctl, so skip it */ + brw->need_flush_throttle = false; + } + + if (brw->need_flush_throttle) { + __DRIscreen *psp = brw->intelScreen->driScrnPriv; + drmCommandNone(psp->fd, DRM_I915_GEM_THROTTLE); + brw->need_flush_throttle = false; } } diff --git a/src/mesa/drivers/dri/i965/brw_context.h b/src/mesa/drivers/dri/i965/brw_context.h index 682fbe97db0..eebd7cec72e 100644 --- a/src/mesa/drivers/dri/i965/brw_context.h +++ b/src/mesa/drivers/dri/i965/brw_context.h @@ -1031,7 +1031,19 @@ struct brw_context /** Framerate throttling: @{ */ drm_intel_bo *first_post_swapbuffers_batch; - bool need_throttle; + + /* Limit the number of outstanding SwapBuffers by waiting for an earlier + * frame of rendering to complete. This gives a very precise cap to the + * latency between input and output such that rendering never gets more + * than a frame behind the user. (With the caveat that we technically are + * not using the SwapBuffers itself as a barrier but the first batch + * submitted afterwards, which may be immediately prior to the next + * SwapBuffers.) + */ + bool need_swap_throttle; + + /** General throttling, not caught by throttling between SwapBuffers */ + bool need_flush_throttle; /** @} */ GLuint stats_wm; diff --git a/src/mesa/drivers/dri/i965/intel_screen.c b/src/mesa/drivers/dri/i965/intel_screen.c index cea7ddfe67a..3640b675471 100644 --- a/src/mesa/drivers/dri/i965/intel_screen.c +++ b/src/mesa/drivers/dri/i965/intel_screen.c @@ -174,10 +174,10 @@ intel_dri2_flush_with_flags(__DRIcontext *cPriv, if (flags & __DRI2_FLUSH_DRAWABLE) intel_resolve_for_dri2_flush(brw, dPriv); - if (reason == __DRI2_THROTTLE_SWAPBUFFER || - reason == __DRI2_THROTTLE_FLUSHFRONT) { - brw->need_throttle = true; - } + if (reason == __DRI2_THROTTLE_SWAPBUFFER) + brw->need_swap_throttle = true; + if (reason == __DRI2_THROTTLE_FLUSHFRONT) + brw->need_flush_throttle = true; intel_batchbuffer_flush(brw); |