From 52753b51f1dbf51221d7856a9288aad1ab2d351a Mon Sep 17 00:00:00 2001 From: Deepak Nibade Date: Wed, 7 Oct 2015 16:20:07 +0530 Subject: gpu: nvgpu: create sync_fence only if needed Currently, we create sync_fence (from nvhost_sync_create_fence()) for every submit But not all submits request for a sync_fence. Also, nvhost_sync_create_fence() API takes about 1/3rd of the total submit path. Hence to optimize, we can allocate sync_fence only when user explicitly asks for it using (NVGPU_SUBMIT_GPFIFO_FLAGS_FENCE_GET && NVGPU_SUBMIT_GPFIFO_FLAGS_SYNC_FENCE) Also, in CDE path from gk20a_prepare_compressible_read(), we reuse existing fence stored in "state" and that can result into not returning sync_fence_fd when user asked for it Hence, force allocation of sync_fence when job submission comes from CDE path Bug 200141116 Change-Id: Ia921701bf0e2432d6b8a5e8b7d91160e7f52db1e Signed-off-by: Deepak Nibade Reviewed-on: http://git-master/r/812845 (cherry picked from commit 5fd47015eeed00352cc8473eff969a66c94fee98) Reviewed-on: http://git-master/r/837662 Reviewed-by: Automatic_Commit_Validation_User GVS: Gerrit_Virtual_Submit Reviewed-by: Sachin Nikam --- drivers/gpu/nvgpu/gk20a/channel_sync_gk20a.h | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'drivers/gpu/nvgpu/gk20a/channel_sync_gk20a.h') diff --git a/drivers/gpu/nvgpu/gk20a/channel_sync_gk20a.h b/drivers/gpu/nvgpu/gk20a/channel_sync_gk20a.h index a347cbab..618e1b26 100644 --- a/drivers/gpu/nvgpu/gk20a/channel_sync_gk20a.h +++ b/drivers/gpu/nvgpu/gk20a/channel_sync_gk20a.h @@ -3,7 +3,7 @@ * * GK20A Channel Synchronization Abstraction * - * Copyright (c) 2014, NVIDIA CORPORATION. All rights reserved. + * Copyright (c) 2014-2015, NVIDIA CORPORATION. All rights reserved. * * This program is free software; you can redistribute it and/or modify it * under the terms and conditions of the GNU General Public License, @@ -54,7 +54,8 @@ struct gk20a_channel_sync { */ int (*incr)(struct gk20a_channel_sync *s, struct priv_cmd_entry **entry, - struct gk20a_fence **fence); + struct gk20a_fence **fence, + bool need_sync_fence); /* Increment syncpoint/semaphore, preceded by a wfi. * Returns @@ -76,7 +77,8 @@ struct gk20a_channel_sync { int wait_fence_fd, struct priv_cmd_entry **entry, struct gk20a_fence **fence, - bool wfi); + bool wfi, + bool need_sync_fence); /* Reset the channel syncpoint/semaphore. */ void (*set_min_eq_max)(struct gk20a_channel_sync *s); -- cgit v1.2.2