From patchwork Tue Mar 4 17:38:56 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ben Widawsky X-Patchwork-Id: 3764381 Return-Path: X-Original-To: patchwork-intel-gfx@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork1.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.19.201]) by patchwork1.web.kernel.org (Postfix) with ESMTP id AA6139F35F for ; Tue, 4 Mar 2014 17:39:08 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id AFE6520382 for ; Tue, 4 Mar 2014 17:39:07 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) by mail.kernel.org (Postfix) with ESMTP id 9543520374 for ; Tue, 4 Mar 2014 17:39:06 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 63705FB826; Tue, 4 Mar 2014 09:39:05 -0800 (PST) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from mga09.intel.com (mga09.intel.com [134.134.136.24]) by gabe.freedesktop.org (Postfix) with ESMTP id 62A62FB826 for ; Tue, 4 Mar 2014 09:39:03 -0800 (PST) Received: from orsmga002.jf.intel.com ([10.7.209.21]) by orsmga102.jf.intel.com with ESMTP; 04 Mar 2014 09:34:37 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.97,586,1389772800"; d="scan'208";a="493960006" Received: from ironside.jf.intel.com ([10.7.197.218]) by orsmga002.jf.intel.com with ESMTP; 04 Mar 2014 09:39:02 -0800 From: Ben Widawsky To: Intel GFX Date: Tue, 4 Mar 2014 09:38:56 -0800 Message-Id: <1393954736-1397-1-git-send-email-benjamin.widawsky@intel.com> X-Mailer: git-send-email 1.9.0 Cc: Ben Widawsky , Ben Widawsky Subject: [Intel-gfx] [PATCH] drm/i915/bdw: MU_FLUSH_DW a qword instead of dword X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.13 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: intel-gfx-bounces@lists.freedesktop.org Errors-To: intel-gfx-bounces@lists.freedesktop.org X-Spam-Status: No, score=-4.2 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_MED, RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP The actual post sync op is "Write Immediate Data QWord." It is therefore arguable that we should have always done a qword write. The actual impetus for this patch is our decoder complains when we write a dword and I was trying to eliminate the spurious errors. With this patch, I've noticed a really strange reproducible error turns into a different strange reproducible error - so it does indeed have some effect of some sort. This was also recommended to me by someone that is familiar with the Windows driver. It's based on top of the semaphore series, so it won't be easily applied, I'd guess. Signed-off-by: Ben Widawsky --- drivers/gpu/drm/i915/intel_ringbuffer.c | 94 +++++++++++++++++++++++++-------- 1 file changed, 73 insertions(+), 21 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index 5f7bee8..2f47abb 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -684,7 +684,7 @@ static int gen8_rcs_signal(struct intel_ring_buffer *signaller, static int gen8_xcs_signal(struct intel_ring_buffer *signaller, unsigned int num_dwords) { -#define MBOX_UPDATE_DWORDS 6 +#define MBOX_UPDATE_DWORDS 8 struct drm_device *dev = signaller->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct intel_ring_buffer *waiter; @@ -704,15 +704,18 @@ static int gen8_xcs_signal(struct intel_ring_buffer *signaller, if (gtt_offset == MI_SEMAPHORE_SYNC_INVALID) continue; - intel_ring_emit(signaller, (MI_FLUSH_DW + 1) | + intel_ring_emit(signaller, (MI_FLUSH_DW + 2) | MI_FLUSH_DW_OP_STOREDW); intel_ring_emit(signaller, lower_32_bits(gtt_offset) | MI_FLUSH_DW_USE_GTT); intel_ring_emit(signaller, upper_32_bits(gtt_offset)); intel_ring_emit(signaller, signaller->outstanding_lazy_seqno); + intel_ring_emit(signaller, 0); /* upper dword */ + intel_ring_emit(signaller, MI_SEMAPHORE_SIGNAL | MI_SEMAPHORE_TARGET(waiter->id)); intel_ring_emit(signaller, 0); + intel_ring_emit(signaller, MI_NOOP); } WARN_ON(i != num_rings); @@ -1830,6 +1833,35 @@ static void gen6_bsd_ring_write_tail(struct intel_ring_buffer *ring, _MASKED_BIT_DISABLE(GEN6_BSD_SLEEP_MSG_DISABLE)); } +static int gen8_bsd_ring_flush(struct intel_ring_buffer *ring, + u32 invalidate, u32 flush) +{ + uint32_t cmd; + int ret; + + ret = intel_ring_begin(ring, 6); + if (ret) + return ret; + + cmd = MI_FLUSH_DW + 2; + /* + * Bspec vol 1c.5 - video engine command streamer: + * "If ENABLED, all TLBs will be invalidated once the flush + * operation is complete. This bit is only valid when the + * Post-Sync Operation field is a value of 1h or 3h." + */ + if (invalidate & I915_GEM_GPU_DOMAINS) + cmd |= MI_INVALIDATE_TLB | MI_INVALIDATE_BSD | + MI_FLUSH_DW_STORE_INDEX | MI_FLUSH_DW_OP_STOREDW; + intel_ring_emit(ring, cmd); + intel_ring_emit(ring, I915_GEM_HWS_SCRATCH_ADDR | MI_FLUSH_DW_USE_GTT); + intel_ring_emit(ring, 0); /* upper addr */ + intel_ring_emit(ring, 0); /* value */ + intel_ring_emit(ring, 0); /* value */ + intel_ring_emit(ring, MI_NOOP); + intel_ring_advance(ring); + return 0; +} static int gen6_bsd_ring_flush(struct intel_ring_buffer *ring, u32 invalidate, u32 flush) { @@ -1841,8 +1873,6 @@ static int gen6_bsd_ring_flush(struct intel_ring_buffer *ring, return ret; cmd = MI_FLUSH_DW; - if (INTEL_INFO(ring->dev)->gen >= 8) - cmd += 1; /* * Bspec vol 1c.5 - video engine command streamer: * "If ENABLED, all TLBs will be invalidated once the flush @@ -1854,13 +1884,8 @@ static int gen6_bsd_ring_flush(struct intel_ring_buffer *ring, MI_FLUSH_DW_STORE_INDEX | MI_FLUSH_DW_OP_STOREDW; intel_ring_emit(ring, cmd); intel_ring_emit(ring, I915_GEM_HWS_SCRATCH_ADDR | MI_FLUSH_DW_USE_GTT); - if (INTEL_INFO(ring->dev)->gen >= 8) { - intel_ring_emit(ring, 0); /* upper addr */ - intel_ring_emit(ring, 0); /* value */ - } else { - intel_ring_emit(ring, 0); - intel_ring_emit(ring, MI_NOOP); - } + intel_ring_emit(ring, 0); + intel_ring_emit(ring, MI_NOOP); intel_ring_advance(ring); return 0; } @@ -1931,8 +1956,38 @@ gen6_ring_dispatch_execbuffer(struct intel_ring_buffer *ring, return 0; } -/* Blitter support (SandyBridge+) */ +static int gen8_ring_flush(struct intel_ring_buffer *ring, + u32 invalidate, u32 flush) +{ + uint32_t cmd; + int ret; + + ret = intel_ring_begin(ring, 6); + if (ret) + return ret; + + cmd = MI_FLUSH_DW + 2; + /* + * Bspec vol 1c.3 - blitter engine command streamer: + * "If ENABLED, all TLBs will be invalidated once the flush + * operation is complete. This bit is only valid when the + * Post-Sync Operation field is a value of 1h or 3h." + */ + if (invalidate & I915_GEM_DOMAIN_RENDER) + cmd |= MI_INVALIDATE_TLB | MI_FLUSH_DW_STORE_INDEX | + MI_FLUSH_DW_OP_STOREDW; + intel_ring_emit(ring, cmd); + intel_ring_emit(ring, I915_GEM_HWS_SCRATCH_ADDR | MI_FLUSH_DW_USE_GTT); + intel_ring_emit(ring, 0); /* upper addr */ + intel_ring_emit(ring, 0); /* value */ + intel_ring_emit(ring, 0); /* value */ + intel_ring_emit(ring, MI_NOOP); + intel_ring_advance(ring); + + return 0; +} +/* Blitter support (SandyBridge+) */ static int gen6_ring_flush(struct intel_ring_buffer *ring, u32 invalidate, u32 flush) { @@ -1945,8 +2000,7 @@ static int gen6_ring_flush(struct intel_ring_buffer *ring, return ret; cmd = MI_FLUSH_DW; - if (INTEL_INFO(ring->dev)->gen >= 8) - cmd += 1; + /* * Bspec vol 1c.3 - blitter engine command streamer: * "If ENABLED, all TLBs will be invalidated once the flush @@ -1958,13 +2012,8 @@ static int gen6_ring_flush(struct intel_ring_buffer *ring, MI_FLUSH_DW_OP_STOREDW; intel_ring_emit(ring, cmd); intel_ring_emit(ring, I915_GEM_HWS_SCRATCH_ADDR | MI_FLUSH_DW_USE_GTT); - if (INTEL_INFO(ring->dev)->gen >= 8) { - intel_ring_emit(ring, 0); /* upper addr */ - intel_ring_emit(ring, 0); /* value */ - } else { - intel_ring_emit(ring, 0); - intel_ring_emit(ring, MI_NOOP); - } + intel_ring_emit(ring, 0); + intel_ring_emit(ring, MI_NOOP); intel_ring_advance(ring); if (IS_GEN7(dev) && !invalidate && flush) @@ -2190,6 +2239,7 @@ int intel_init_bsd_ring_buffer(struct drm_device *dev) if (INTEL_INFO(dev)->gen >= 8) { ring->irq_enable_mask = GT_RENDER_USER_INTERRUPT << GEN8_VCS1_IRQ_SHIFT; + ring->flush = gen8_bsd_ring_flush; ring->irq_get = gen8_ring_get_irq; ring->irq_put = gen8_ring_put_irq; ring->dispatch_execbuffer = @@ -2257,6 +2307,7 @@ int intel_init_blt_ring_buffer(struct drm_device *dev) if (INTEL_INFO(dev)->gen >= 8) { ring->irq_enable_mask = GT_RENDER_USER_INTERRUPT << GEN8_BCS_IRQ_SHIFT; + ring->flush = gen8_ring_flush; ring->irq_get = gen8_ring_get_irq; ring->irq_put = gen8_ring_put_irq; ring->dispatch_execbuffer = gen8_ring_dispatch_execbuffer; @@ -2306,6 +2357,7 @@ int intel_init_vebox_ring_buffer(struct drm_device *dev) if (INTEL_INFO(dev)->gen >= 8) { ring->irq_enable_mask = GT_RENDER_USER_INTERRUPT << GEN8_VECS_IRQ_SHIFT; + ring->flush = gen8_ring_flush; ring->irq_get = gen8_ring_get_irq; ring->irq_put = gen8_ring_put_irq; ring->dispatch_execbuffer = gen8_ring_dispatch_execbuffer;