From patchwork Wed Feb 26 09:42:58 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405731 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 2D672138D for ; Wed, 26 Feb 2020 09:43:37 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 161A024673 for ; Wed, 26 Feb 2020 09:43:37 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 161A024673 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 75CA26E405; Wed, 26 Feb 2020 09:43:34 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id C8A786E3EF for ; Wed, 26 Feb 2020 09:43:25 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354697-1500050 for multiple; Wed, 26 Feb 2020 09:43:14 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:42:58 +0000 Message-Id: <20200226094314.1500667-1-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 01/17] drm/i915/gt: Reset queue_priority_hint after wedging X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" An odd and highly unlikely path caught us out. On delayed submission (due to an asynchronous reset handler), we poked the priority_hint and kicked the tasklet. However, we had already marked the device as wedged and swapped out the tasklet for a no-op. The result was that we never cleared the priority hint and became upset when we later checked. <0> [574.303565] i915_sel-6278 2.... 481822445us : __i915_subtests: Running intel_execlists_live_selftests/live_error_interrupt <0> [574.303565] i915_sel-6278 2.... 481822472us : __engine_unpark: 0000:00:02.0 rcs0: <0> [574.303565] i915_sel-6278 2.... 481822491us : __gt_unpark: 0000:00:02.0 <0> [574.303565] i915_sel-6278 2.... 481823220us : execlists_context_reset: 0000:00:02.0 rcs0: context:f4ee reset <0> [574.303565] i915_sel-6278 2.... 481824830us : __intel_context_active: 0000:00:02.0 rcs0: context:f51b active <0> [574.303565] i915_sel-6278 2.... 481825258us : __intel_context_do_pin: 0000:00:02.0 rcs0: context:f51b pin ring:{start:00006000, head:0000, tail:0000} <0> [574.303565] i915_sel-6278 2.... 481825311us : __i915_request_commit: 0000:00:02.0 rcs0: fence f51b:2, current 0 <0> [574.303565] i915_sel-6278 2d..1 481825347us : __i915_request_submit: 0000:00:02.0 rcs0: fence f51b:2, current 0 <0> [574.303565] i915_sel-6278 2d..1 481825363us : trace_ports: 0000:00:02.0 rcs0: submit { f51b:2, 0:0 } <0> [574.303565] i915_sel-6278 2.... 481826809us : __intel_context_active: 0000:00:02.0 rcs0: context:f51c active <0> [574.303565] -0 7d.h2 481827326us : cs_irq_handler: 0000:00:02.0 rcs0: CS error: 1 <0> [574.303565] -0 7..s1 481827377us : process_csb: 0000:00:02.0 rcs0: cs-irq head=3, tail=4 <0> [574.303565] -0 7..s1 481827379us : process_csb: 0000:00:02.0 rcs0: csb[4]: status=0x10000001:0x00000000 <0> [574.305593] -0 7..s1 481827385us : trace_ports: 0000:00:02.0 rcs0: promote { f51b:2*, 0:0 } <0> [574.305611] -0 7..s1 481828179us : execlists_reset: 0000:00:02.0 rcs0: reset for CS error <0> [574.305611] i915_sel-6278 2.... 481828284us : __intel_context_do_pin: 0000:00:02.0 rcs0: context:f51c pin ring:{start:00007000, head:0000, tail:0000} <0> [574.305611] i915_sel-6278 2.... 481828345us : __i915_request_commit: 0000:00:02.0 rcs0: fence f51c:2, current 0 <0> [574.305611] -0 7dNs2 481847823us : __i915_request_unsubmit: 0000:00:02.0 rcs0: fence f51b:2, current 1 <0> [574.305611] -0 7dNs2 481847857us : execlists_hold: 0000:00:02.0 rcs0: fence f51b:2, current 1 on hold <0> [574.305611] -0 7.Ns1 481847863us : intel_engine_reset: 0000:00:02.0 rcs0: flags=4 <0> [574.305611] -0 7.Ns1 481847945us : execlists_reset_prepare: 0000:00:02.0 rcs0: depth<-1 <0> [574.305611] -0 7.Ns1 481847946us : intel_engine_stop_cs: 0000:00:02.0 rcs0: <0> [574.305611] -0 7.Ns1 538584284us : intel_engine_stop_cs: 0000:00:02.0 rcs0: timed out on STOP_RING -> IDLE <0> [574.305611] -0 7.Ns1 538584347us : __intel_gt_reset: 0000:00:02.0 engine_mask=1 <0> [574.305611] -0 7.Ns1 538584406us : execlists_reset_rewind: 0000:00:02.0 rcs0: <0> [574.305611] -0 7dNs2 538585050us : __i915_request_reset: 0000:00:02.0 rcs0: fence f51b:2, current 1 guilty? yes <0> [574.305611] -0 7dNs2 538585063us : __execlists_reset: 0000:00:02.0 rcs0: replay {head:0000, tail:0068} <0> [574.306565] -0 7.Ns1 538588457us : intel_engine_cancel_stop_cs: 0000:00:02.0 rcs0: <0> [574.306565] -0 7dNs2 538588462us : __i915_request_submit: 0000:00:02.0 rcs0: fence f51c:2, current 0 <0> [574.306565] -0 7dNs2 538588471us : trace_ports: 0000:00:02.0 rcs0: submit { f51c:2, 0:0 } <0> [574.306565] -0 7.Ns1 538588474us : execlists_reset_finish: 0000:00:02.0 rcs0: depth->1 <0> [574.306565] kworker/-202 2.... 538588755us : i915_request_retire: 0000:00:02.0 rcs0: fence f51c:2, current 2 <0> [574.306565] ksoftirq-46 7..s. 538588773us : process_csb: 0000:00:02.0 rcs0: cs-irq head=11, tail=1 <0> [574.306565] ksoftirq-46 7..s. 538588774us : process_csb: 0000:00:02.0 rcs0: csb[0]: status=0x10000001:0x00000000 <0> [574.306565] ksoftirq-46 7..s. 538588776us : trace_ports: 0000:00:02.0 rcs0: promote { f51c:2!, 0:0 } <0> [574.306565] ksoftirq-46 7..s. 538588778us : process_csb: 0000:00:02.0 rcs0: csb[1]: status=0x10000018:0x00000020 <0> [574.306565] ksoftirq-46 7..s. 538588779us : trace_ports: 0000:00:02.0 rcs0: completed { f51c:2!, 0:0 } <0> [574.306565] kworker/-202 2.... 538588826us : intel_context_unpin: 0000:00:02.0 rcs0: context:f51c unpin <0> [574.306565] i915_sel-6278 6.... 538589663us : __intel_gt_set_wedged.part.32: 0000:00:02.0 start <0> [574.306565] i915_sel-6278 6.... 538589667us : execlists_reset_prepare: 0000:00:02.0 rcs0: depth<-0 <0> [574.306565] i915_sel-6278 6.... 538589710us : intel_engine_stop_cs: 0000:00:02.0 rcs0: <0> [574.306565] i915_sel-6278 6.... 538589732us : execlists_reset_prepare: 0000:00:02.0 bcs0: depth<-0 <0> [574.307591] i915_sel-6278 6.... 538589733us : intel_engine_stop_cs: 0000:00:02.0 bcs0: <0> [574.307591] i915_sel-6278 6.... 538589757us : execlists_reset_prepare: 0000:00:02.0 vcs0: depth<-0 <0> [574.307591] i915_sel-6278 6.... 538589758us : intel_engine_stop_cs: 0000:00:02.0 vcs0: <0> [574.307591] i915_sel-6278 6.... 538589771us : execlists_reset_prepare: 0000:00:02.0 vcs1: depth<-0 <0> [574.307591] i915_sel-6278 6.... 538589772us : intel_engine_stop_cs: 0000:00:02.0 vcs1: <0> [574.307591] i915_sel-6278 6.... 538589778us : execlists_reset_prepare: 0000:00:02.0 vecs0: depth<-0 <0> [574.307591] i915_sel-6278 6.... 538589780us : intel_engine_stop_cs: 0000:00:02.0 vecs0: <0> [574.307591] i915_sel-6278 6.... 538589786us : __intel_gt_reset: 0000:00:02.0 engine_mask=ff <0> [574.307591] i915_sel-6278 6.... 538591175us : execlists_reset_cancel: 0000:00:02.0 rcs0: <0> [574.307591] i915_sel-6278 6.... 538591970us : execlists_reset_cancel: 0000:00:02.0 bcs0: <0> [574.307591] i915_sel-6278 6.... 538591982us : execlists_reset_cancel: 0000:00:02.0 vcs0: <0> [574.307591] i915_sel-6278 6.... 538591996us : execlists_reset_cancel: 0000:00:02.0 vcs1: <0> [574.307591] i915_sel-6278 6.... 538592759us : execlists_reset_cancel: 0000:00:02.0 vecs0: <0> [574.307591] i915_sel-6278 6.... 538592977us : execlists_reset_finish: 0000:00:02.0 rcs0: depth->0 <0> [574.307591] i915_sel-6278 6.N.. 538592996us : execlists_reset_finish: 0000:00:02.0 bcs0: depth->0 <0> [574.307591] i915_sel-6278 6.N.. 538593023us : execlists_reset_finish: 0000:00:02.0 vcs0: depth->0 <0> [574.307591] i915_sel-6278 6.N.. 538593037us : execlists_reset_finish: 0000:00:02.0 vcs1: depth->0 <0> [574.307591] i915_sel-6278 6.N.. 538593051us : execlists_reset_finish: 0000:00:02.0 vecs0: depth->0 <0> [574.307591] i915_sel-6278 6.... 538593407us : __intel_gt_set_wedged.part.32: 0000:00:02.0 end <0> [574.307591] kworker/-210 7d..1 551958381us : execlists_unhold: 0000:00:02.0 rcs0: fence f51b:2, current 2 hold release <0> [574.307591] i915_sel-6278 0.... 559490788us : i915_request_retire: 0000:00:02.0 rcs0: fence f51b:2, current 2 <0> [574.307591] i915_sel-6278 0.... 559490793us : intel_context_unpin: 0000:00:02.0 rcs0: context:f51b unpin <0> [574.307591] i915_sel-6278 0.... 559490798us : __engine_park: 0000:00:02.0 rcs0: parked <0> [574.307591] i915_sel-6278 0.... 559490982us : __intel_context_retire: 0000:00:02.0 rcs0: context:f51c retire runtime: { total:30004ns, avg:30004ns } <0> [574.307591] i915_sel-6278 0.... 559491372us : __engine_park: __engine_park:261 GEM_BUG_ON(engine->execlists.queue_priority_hint != (-((int)(~0U >> 1)) - 1)) Signed-off-by: Chris Wilson --- drivers/gpu/drm/i915/gt/intel_lrc.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 39b0125b7143..35c5cf786726 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -3724,7 +3724,10 @@ static void execlists_reset_rewind(struct intel_engine_cs *engine, bool stalled) static void nop_submission_tasklet(unsigned long data) { + struct intel_engine_cs * const engine = (struct intel_engine_cs *)data; + /* The driver is wedged; don't process any more events. */ + WRITE_ONCE(engine->execlists.queue_priority_hint, INT_MIN); } static void execlists_reset_cancel(struct intel_engine_cs *engine) From patchwork Wed Feb 26 09:42:59 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405749 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id B02F5138D for ; Wed, 26 Feb 2020 09:43:52 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 9924224673 for ; Wed, 26 Feb 2020 09:43:52 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 9924224673 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id A2C776E423; Wed, 26 Feb 2020 09:43:49 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id B9E276E408 for ; Wed, 26 Feb 2020 09:43:37 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354698-1500050 for multiple; Wed, 26 Feb 2020 09:43:15 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:42:59 +0000 Message-Id: <20200226094314.1500667-2-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 02/17] drm/i915/perf: Mark up the racy use of perf->exclusive_stream X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" Inside the general i915_oa_init_reg_state() we avoid using the perf->mutex. However, we rely on perf->exclusive_stream being valid to access at that point, and for that we have to control the race with disabling perf. This relies on the disabling being a heavy barrier that inspects all active contexts, after marking the perf->exclusive_stream as not available. This should ensure that there are no more concurrent accesses to the perf->exclusive_stream as we destroy it. Mark up the races around the perf->exclusive_stream so that they stand out much more. (And hopefully we will be running kcsan to start validating that the only races we have are carefully controlled.) Signed-off-by: Chris Wilson Cc: Lionel Landwerlin --- drivers/gpu/drm/i915/i915_perf.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c index e34c79df6ebc..0838a12e2dc5 100644 --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -1405,8 +1405,10 @@ static void i915_oa_stream_destroy(struct i915_perf_stream *stream) /* * Unset exclusive_stream first, it will be checked while disabling * the metric set on gen8+. + * + * See i915_oa_init_reg_state() and lrc_configure_all_contexts() */ - perf->exclusive_stream = NULL; + WRITE_ONCE(perf->exclusive_stream, NULL); perf->ops.disable_metric_set(stream); free_oa_buffer(stream); @@ -2847,7 +2849,7 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, goto err_oa_buf_alloc; stream->ops = &i915_oa_stream_ops; - perf->exclusive_stream = stream; + WRITE_ONCE(perf->exclusive_stream, stream); ret = perf->ops.enable_metric_set(stream); if (ret) { @@ -2867,7 +2869,7 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, return 0; err_enable: - perf->exclusive_stream = NULL; + WRITE_ONCE(perf->exclusive_stream, NULL); perf->ops.disable_metric_set(stream); free_oa_buffer(stream); @@ -2893,12 +2895,11 @@ void i915_oa_init_reg_state(const struct intel_context *ce, { struct i915_perf_stream *stream; - /* perf.exclusive_stream serialised by lrc_configure_all_contexts() */ - if (engine->class != RENDER_CLASS) return; - stream = engine->i915->perf.exclusive_stream; + /* perf.exclusive_stream serialised by lrc_configure_all_contexts() */ + stream = READ_ONCE(engine->i915->perf.exclusive_stream); /* * For gen12, only CTX_R_PWR_CLK_STATE needs update, but the caller * is already doing that, so nothing to be done for gen12 here. From patchwork Wed Feb 26 09:43:00 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405747 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id D6C161580 for ; Wed, 26 Feb 2020 09:43:51 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id BFC2A24673 for ; Wed, 26 Feb 2020 09:43:51 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org BFC2A24673 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 54D4F6E41D; Wed, 26 Feb 2020 09:43:49 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id B80F26E405 for ; Wed, 26 Feb 2020 09:43:26 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354699-1500050 for multiple; Wed, 26 Feb 2020 09:43:15 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:00 +0000 Message-Id: <20200226094314.1500667-3-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 03/17] drm/i915: Manually acquire engine-wakeref around use of kernel_context X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" The engine->kernel_context is a special case for request emission. Since it is used as the barrier within the engine's wakeref, we must acquire the wakeref before submitting a request to the kernel_context. Reported-by: Lionel Landwerlin Signed-off-by: Chris Wilson Cc: Lionel Landwerlin --- drivers/gpu/drm/i915/i915_perf.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c index 0838a12e2dc5..2334c45f1d08 100644 --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -2196,7 +2196,9 @@ static int gen8_modify_self(struct intel_context *ce, struct i915_request *rq; int err; + intel_engine_pm_get(ce->engine); rq = i915_request_create(ce); + intel_engine_pm_put(ce->engine); if (IS_ERR(rq)) return PTR_ERR(rq); From patchwork Wed Feb 26 09:43:01 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405725 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 8D831138D for ; Wed, 26 Feb 2020 09:43:30 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 75E7A24673 for ; Wed, 26 Feb 2020 09:43:30 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 75E7A24673 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 781256E3F9; Wed, 26 Feb 2020 09:43:27 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id C95B96E3F9 for ; Wed, 26 Feb 2020 09:43:25 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354700-1500050 for multiple; Wed, 26 Feb 2020 09:43:15 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:01 +0000 Message-Id: <20200226094314.1500667-4-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 04/17] drm/i915/gt: Pull marking vm as closed underneath the vm->mutex X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" Pull the final atomic_dec of vm->open (marking the vm as closed) underneath the same vm->mutex as used to close it. This is required to correctly serialise with attempting to reuse the vma as the vm is closed by a second thread. References: 00de702c6c6f ("drm/i915: Check that the vma hasn't been closed before we insert it") Signed-off-by: Chris Wilson --- drivers/gpu/drm/i915/gt/intel_gtt.c | 5 ++++- drivers/gpu/drm/i915/gt/intel_gtt.h | 3 +-- 2 files changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_gtt.c b/drivers/gpu/drm/i915/gt/intel_gtt.c index bb9a6e638175..dfb1be050cca 100644 --- a/drivers/gpu/drm/i915/gt/intel_gtt.c +++ b/drivers/gpu/drm/i915/gt/intel_gtt.c @@ -171,7 +171,9 @@ void __i915_vm_close(struct i915_address_space *vm) { struct i915_vma *vma, *vn; - mutex_lock(&vm->mutex); + if (!atomic_dec_and_mutex_lock(&vm->open, &vm->mutex)) + return; + list_for_each_entry_safe(vma, vn, &vm->bound_list, vm_link) { struct drm_i915_gem_object *obj = vma->obj; @@ -186,6 +188,7 @@ void __i915_vm_close(struct i915_address_space *vm) i915_gem_object_put(obj); } GEM_BUG_ON(!list_empty(&vm->bound_list)); + mutex_unlock(&vm->mutex); } diff --git a/drivers/gpu/drm/i915/gt/intel_gtt.h b/drivers/gpu/drm/i915/gt/intel_gtt.h index 23004445806a..eac38c682ef4 100644 --- a/drivers/gpu/drm/i915/gt/intel_gtt.h +++ b/drivers/gpu/drm/i915/gt/intel_gtt.h @@ -429,8 +429,7 @@ static inline void i915_vm_close(struct i915_address_space *vm) { GEM_BUG_ON(!atomic_read(&vm->open)); - if (atomic_dec_and_test(&vm->open)) - __i915_vm_close(vm); + __i915_vm_close(vm); i915_vm_put(vm); } From patchwork Wed Feb 26 09:43:02 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405723 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id C82FC138D for ; Wed, 26 Feb 2020 09:43:28 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id B04A324672 for ; Wed, 26 Feb 2020 09:43:28 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org B04A324672 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 48B566E3EF; Wed, 26 Feb 2020 09:43:27 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id C93D96E3F2 for ; Wed, 26 Feb 2020 09:43:25 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354701-1500050 for multiple; Wed, 26 Feb 2020 09:43:15 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:02 +0000 Message-Id: <20200226094314.1500667-5-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 05/17] drm/i915: Protect i915_request_await_start from early waits X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" We need to be extremely careful inside i915_request_await_start() as it needs to walk the list of requests in the foreign timeline with very little protection. As we hold our own timeline mutex, we can not nest inside the signaler's timeline mutex, so all that remains is our RCU protection. However, to be safe we need to tell the compiler that we may be traversing the list only under RCU protection, and furthermore we need to start declaring requests as elements of the timeline from their construction. Fixes: 9ddc8ec027a3 ("drm/i915: Eliminate the trylock for awaiting an earlier request") Fixes: 6a79d848403d ("drm/i915: Lock signaler timeline while navigating") Signed-off-by: Chris Wilson --- drivers/gpu/drm/i915/i915_request.c | 41 ++++++++++++++++++++--------- 1 file changed, 28 insertions(+), 13 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_request.c b/drivers/gpu/drm/i915/i915_request.c index d53af93b919b..e5a55801f753 100644 --- a/drivers/gpu/drm/i915/i915_request.c +++ b/drivers/gpu/drm/i915/i915_request.c @@ -290,7 +290,7 @@ bool i915_request_retire(struct i915_request *rq) spin_unlock_irq(&rq->lock); remove_from_client(rq); - list_del(&rq->link); + list_del_rcu(&rq->link); intel_context_exit(rq->context); intel_context_unpin(rq->context); @@ -736,6 +736,8 @@ __i915_request_create(struct intel_context *ce, gfp_t gfp) rq->infix = rq->ring->emit; /* end of header; start of user payload */ intel_context_mark_active(ce); + list_add_tail_rcu(&rq->link, &tl->requests); + return rq; err_unwind: @@ -792,13 +794,23 @@ i915_request_await_start(struct i915_request *rq, struct i915_request *signal) GEM_BUG_ON(i915_request_timeline(rq) == rcu_access_pointer(signal->timeline)); + if (i915_request_started(signal)) + return 0; + fence = NULL; rcu_read_lock(); spin_lock_irq(&signal->lock); - if (!i915_request_started(signal) && - !list_is_first(&signal->link, - &rcu_dereference(signal->timeline)->requests)) { - struct i915_request *prev = list_prev_entry(signal, link); + do { + struct list_head *pos = READ_ONCE(signal->link.prev); + struct i915_request *prev; + + /* Confirm signal has not been retired, the link is valid */ + if (unlikely(i915_request_started(signal))) + break; + + /* Is signal the earliest request on its timeline? */ + if (pos == &rcu_dereference(signal->timeline)->requests) + break; /* * Peek at the request before us in the timeline. That @@ -806,13 +818,18 @@ i915_request_await_start(struct i915_request *rq, struct i915_request *signal) * after acquiring a reference to it, confirm that it is * still part of the signaler's timeline. */ - if (i915_request_get_rcu(prev)) { - if (list_next_entry(prev, link) == signal) - fence = &prev->fence; - else - i915_request_put(prev); + prev = list_entry(pos, typeof(*prev), link); + if (!i915_request_get_rcu(prev)) + break; + + /* After the strong barrier, confirm prev is still attached */ + if (unlikely(READ_ONCE(prev->link.next) != &signal->link)) { + i915_request_put(prev); + break; } - } + + fence = &prev->fence; + } while (0); spin_unlock_irq(&signal->lock); rcu_read_unlock(); if (!fence) @@ -1253,8 +1270,6 @@ __i915_request_add_to_timeline(struct i915_request *rq) 0); } - list_add_tail(&rq->link, &timeline->requests); - /* * Make sure that no request gazumped us - if it was allocated after * our i915_request_alloc() and called __i915_request_add() before From patchwork Wed Feb 26 09:43:03 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405737 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 39122138D for ; Wed, 26 Feb 2020 09:43:40 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 21AAA24673 for ; Wed, 26 Feb 2020 09:43:40 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 21AAA24673 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 68B5E6E41F; Wed, 26 Feb 2020 09:43:35 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id C9FF26E3FC for ; Wed, 26 Feb 2020 09:43:25 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354702-1500050 for multiple; Wed, 26 Feb 2020 09:43:15 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:03 +0000 Message-Id: <20200226094314.1500667-6-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 06/17] drm/i915/selftests: Verify LRC isolation X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" Record the LRC registers before/after a preemption event to ensure that the first context sees nothing from the second client; at least in the normal per-context register state. References: https://gitlab.freedesktop.org/drm/intel/issues/1233 Signed-off-by: Chris Wilson --- drivers/gpu/drm/i915/gt/selftest_lrc.c | 545 +++++++++++++++++++++++++ 1 file changed, 545 insertions(+) diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c b/drivers/gpu/drm/i915/gt/selftest_lrc.c index febd608c23a7..810f7857ad26 100644 --- a/drivers/gpu/drm/i915/gt/selftest_lrc.c +++ b/drivers/gpu/drm/i915/gt/selftest_lrc.c @@ -4748,6 +4748,550 @@ static int live_lrc_timestamp(void *arg) return 0; } +static struct i915_vma * +create_user_vma(struct i915_address_space *vm, unsigned long size) +{ + struct drm_i915_gem_object *obj; + struct i915_vma *vma; + int err; + + obj = i915_gem_object_create_internal(vm->i915, size); + if (IS_ERR(obj)) + return ERR_CAST(obj); + + vma = i915_vma_instance(obj, vm, NULL); + if (IS_ERR(vma)) { + i915_gem_object_put(obj); + return vma; + } + + err = i915_vma_pin(vma, 0, 0, PIN_USER); + if (err) { + i915_gem_object_put(obj); + return ERR_PTR(err); + } + + return vma; +} + +static struct i915_vma * +store_context(struct intel_context *ce, struct i915_vma *scratch) +{ + struct i915_vma *batch; + u32 dw, x, *cs, *hw; + + batch = create_user_vma(ce->vm, SZ_64K); + if (IS_ERR(batch)) + return ERR_CAST(batch); + + cs = i915_gem_object_pin_map(batch->obj, I915_MAP_WC); + if (IS_ERR(cs)) { + i915_vma_put(batch); + return ERR_CAST(cs); + } + + x = 0; + dw = 0; + hw = ce->engine->pinned_default_state; + hw += LRC_STATE_PN * PAGE_SIZE / sizeof(*hw); + do { + u32 lri = hw[dw]; + + if (lri == 0) { + dw++; + continue; + } + + if ((lri & GENMASK(31, 23)) != MI_INSTR(0x22, 0)) { + lri &= 0x7f; + dw += lri + 2; + continue; + } + + lri &= 0x7f; + lri++; + dw++; + + while (lri) { + *cs++ = MI_STORE_REGISTER_MEM_GEN8; + *cs++ = hw[dw]; + *cs++ = lower_32_bits(scratch->node.start + x); + *cs++ = upper_32_bits(scratch->node.start + x); + + dw += 2; + lri -= 2; + x += 4; + } + } while (dw < PAGE_SIZE / sizeof(u32) && + (hw[dw] & ~BIT(0)) != MI_BATCH_BUFFER_END); + + *cs++ = MI_BATCH_BUFFER_END; + + i915_gem_object_flush_map(batch->obj); + i915_gem_object_unpin_map(batch->obj); + + return batch; +} + +static int move_to_active(struct i915_request *rq, + struct i915_vma *vma, + unsigned int flags) +{ + int err; + + i915_vma_lock(vma); + err = i915_request_await_object(rq, vma->obj, flags); + if (!err) + err = i915_vma_move_to_active(vma, rq, flags); + i915_vma_unlock(vma); + + return err; +} + +static struct i915_request * +record_registers(struct intel_context *ce, + struct i915_vma *before, + struct i915_vma *after, + u32 *sema) +{ + struct i915_vma *b_before, *b_after; + struct i915_request *rq; + u32 *cs; + int err; + + b_before = store_context(ce, before); + if (IS_ERR(b_before)) + return ERR_CAST(b_before); + + b_after = store_context(ce, after); + if (IS_ERR(b_after)) { + rq = ERR_CAST(b_after); + goto err_before; + } + + rq = intel_context_create_request(ce); + if (IS_ERR(rq)) + goto err_after; + + err = move_to_active(rq, before, EXEC_OBJECT_WRITE); + if (err) + goto err_rq; + + err = move_to_active(rq, b_before, 0); + if (err) + goto err_rq; + + err = move_to_active(rq, after, EXEC_OBJECT_WRITE); + if (err) + goto err_rq; + + err = move_to_active(rq, b_after, 0); + if (err) + goto err_rq; + + cs = intel_ring_begin(rq, 14); + if (IS_ERR(cs)) { + err = PTR_ERR(cs); + goto err_rq; + } + + *cs++ = MI_ARB_ON_OFF | MI_ARB_DISABLE; + *cs++ = MI_BATCH_BUFFER_START_GEN8 | BIT(8); + *cs++ = lower_32_bits(b_before->node.start); + *cs++ = upper_32_bits(b_before->node.start); + + *cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE; + *cs++ = MI_SEMAPHORE_WAIT | + MI_SEMAPHORE_GLOBAL_GTT | + MI_SEMAPHORE_POLL | + MI_SEMAPHORE_SAD_NEQ_SDD; + *cs++ = 0; + *cs++ = i915_ggtt_offset(ce->engine->status_page.vma) + + offset_in_page(sema); + *cs++ = 0; + *cs++ = MI_NOOP; + + *cs++ = MI_ARB_ON_OFF | MI_ARB_DISABLE; + *cs++ = MI_BATCH_BUFFER_START_GEN8 | BIT(8); + *cs++ = lower_32_bits(b_after->node.start); + *cs++ = upper_32_bits(b_after->node.start); + + intel_ring_advance(rq, cs); + + WRITE_ONCE(*sema, 0); + i915_request_get(rq); + i915_request_add(rq); +err_after: + i915_vma_put(b_after); +err_before: + i915_vma_put(b_before); + return rq; + +err_rq: + i915_request_add(rq); + rq = ERR_PTR(err); + goto err_after; +} + +static struct i915_vma *load_context(struct intel_context *ce, u32 poison) +{ + struct i915_vma *batch; + u32 dw, *cs, *hw; + + batch = create_user_vma(ce->vm, SZ_64K); + if (IS_ERR(batch)) + return ERR_CAST(batch); + + cs = i915_gem_object_pin_map(batch->obj, I915_MAP_WC); + if (IS_ERR(cs)) { + i915_vma_put(batch); + return ERR_CAST(cs); + } + + dw = 0; + hw = ce->engine->pinned_default_state; + hw += LRC_STATE_PN * PAGE_SIZE / sizeof(*hw); + do { + u32 lri = hw[dw]; + + if (lri == 0) { + dw++; + continue; + } + + if ((lri & GENMASK(31, 23)) != MI_INSTR(0x22, 0)) { + lri &= 0x7f; + dw += lri + 2; + continue; + } + + lri &= 0x7f; + lri++; + dw++; + + lri /= 2; + *cs++ = MI_LOAD_REGISTER_IMM(lri); + while (lri--) { + *cs++ = hw[dw]; + *cs++ = poison; + dw += 2; + } + } while (dw < PAGE_SIZE / sizeof(u32) && + (hw[dw] & ~BIT(0)) != MI_BATCH_BUFFER_END); + + *cs++ = MI_BATCH_BUFFER_END; + + i915_gem_object_flush_map(batch->obj); + i915_gem_object_unpin_map(batch->obj); + + return batch; +} + +static int poison_registers(struct intel_context *ce, u32 poison, u32 *sema) +{ + struct i915_request *rq; + struct i915_vma *batch; + u32 *cs; + int err; + + batch = load_context(ce, poison); + if (IS_ERR(batch)) + return PTR_ERR(batch); + + rq = intel_context_create_request(ce); + if (IS_ERR(rq)) { + err = PTR_ERR(rq); + goto err_batch; + } + + err = move_to_active(rq, batch, 0); + if (err) + goto err_rq; + + cs = intel_ring_begin(rq, 8); + if (IS_ERR(cs)) { + err = PTR_ERR(cs); + goto err_rq; + } + + *cs++ = MI_ARB_ON_OFF | MI_ARB_DISABLE; + *cs++ = MI_BATCH_BUFFER_START_GEN8 | BIT(8); + *cs++ = lower_32_bits(batch->node.start); + *cs++ = upper_32_bits(batch->node.start); + + *cs++ = MI_STORE_DWORD_IMM_GEN4 | MI_USE_GGTT; + *cs++ = i915_ggtt_offset(ce->engine->status_page.vma) + + offset_in_page(sema); + *cs++ = 0; + *cs++ = 1; + + intel_ring_advance(rq, cs); + + rq->sched.attr.priority = I915_PRIORITY_BARRIER; +err_rq: + i915_request_add(rq); +err_batch: + i915_vma_put(batch); + return err; +} + +static bool is_moving(u32 a, u32 b) +{ + return a != b; +} + +static int compare_isolation(struct intel_engine_cs *engine, + struct i915_vma *ref[2], + struct i915_vma *result[2], + struct intel_context *ce, + u32 poison) +{ + u32 x, dw, *hw, *lrc; + u32 *A[2], *B[2]; + int err = 0; + + A[0] = i915_gem_object_pin_map(ref[0]->obj, I915_MAP_WC); + if (IS_ERR(A[0])) + return PTR_ERR(A[0]); + + A[1] = i915_gem_object_pin_map(ref[1]->obj, I915_MAP_WC); + if (IS_ERR(A[1])) { + err = PTR_ERR(A[1]); + goto err_A0; + } + + B[0] = i915_gem_object_pin_map(result[0]->obj, I915_MAP_WC); + if (IS_ERR(B[0])) { + err = PTR_ERR(B[0]); + goto err_A1; + } + + B[1] = i915_gem_object_pin_map(result[1]->obj, I915_MAP_WC); + if (IS_ERR(B[1])) { + err = PTR_ERR(B[1]); + goto err_B0; + } + + lrc = i915_gem_object_pin_map(ce->state->obj, + i915_coherent_map_type(engine->i915)); + if (IS_ERR(lrc)) { + err = PTR_ERR(lrc); + goto err_B1; + } + lrc += LRC_STATE_PN * PAGE_SIZE / sizeof(*hw); + + x = 0; + dw = 0; + hw = engine->pinned_default_state; + hw += LRC_STATE_PN * PAGE_SIZE / sizeof(*hw); + do { + u32 lri = hw[dw]; + + if (lri == 0) { + dw++; + continue; + } + + if ((lri & GENMASK(31, 23)) != MI_INSTR(0x22, 0)) { + lri &= 0x7f; + dw += lri + 2; + continue; + } + + lri &= 0x7f; + lri++; + dw++; + + while (lri) { + if (!is_moving(A[0][x], A[1][x]) && + (A[0][x] != B[0][x] || A[1][x] != B[1][x])) { + switch (hw[dw] & 4095) { + case 0x30: /* RING_HEAD */ + case 0x34: /* RING_TAIL */ + break; + + default: + pr_err("%s[%d]: Mismatch for register %4x, default %08x, reference %08x, result (%08x, %08x), poison %08x, context %08x\n", + engine->name, x, + hw[dw], hw[dw + 1], + A[0][x], B[0][x], B[1][x], + poison, lrc[dw + 1]); + err = -EINVAL; + break; + } + } + dw += 2; + lri -= 2; + x++; + } + } while (dw < PAGE_SIZE / sizeof(u32) && + (hw[dw] & ~BIT(0)) != MI_BATCH_BUFFER_END); + + i915_gem_object_unpin_map(ce->state->obj); +err_B1: + i915_gem_object_unpin_map(result[1]->obj); +err_B0: + i915_gem_object_unpin_map(result[0]->obj); +err_A1: + i915_gem_object_unpin_map(ref[1]->obj); +err_A0: + i915_gem_object_unpin_map(ref[0]->obj); + return err; +} + +static int __lrc_isolation(struct intel_engine_cs *engine, u32 poison) +{ + u32 *sema = memset32(engine->status_page.addr + 1000, 0, 1); + struct i915_vma *ref[2], *result[2]; + struct intel_context *A, *B; + struct i915_request *rq; + int err; + + A = intel_context_create(engine); + if (IS_ERR(A)) + return PTR_ERR(A); + + B = intel_context_create(engine); + if (IS_ERR(B)) { + err = PTR_ERR(B); + goto err_A; + } + + ref[0] = create_user_vma(A->vm, SZ_64K); + if (IS_ERR(ref[0])) { + err = PTR_ERR(ref[0]); + goto err_B; + } + + ref[1] = create_user_vma(A->vm, SZ_64K); + if (IS_ERR(ref[1])) { + err = PTR_ERR(ref[1]); + goto err_ref0; + } + + rq = record_registers(A, ref[0], ref[1], sema); + if (IS_ERR(rq)) { + err = PTR_ERR(rq); + goto err_ref1; + } + + WRITE_ONCE(*sema, 1); + wmb(); + + if (i915_request_wait(rq, 0, HZ / 2) < 0) { + i915_request_put(rq); + err = -ETIME; + goto err_ref1; + } + i915_request_put(rq); + + result[0] = create_user_vma(A->vm, SZ_64K); + if (IS_ERR(result[0])) { + err = PTR_ERR(result[0]); + goto err_ref1; + } + + result[1] = create_user_vma(A->vm, SZ_64K); + if (IS_ERR(result[1])) { + err = PTR_ERR(result[1]); + goto err_result0; + } + + rq = record_registers(A, result[0], result[1], sema); + if (IS_ERR(rq)) { + err = PTR_ERR(rq); + goto err_result1; + } + + err = poison_registers(B, poison, sema); + if (err) { + WRITE_ONCE(*sema, -1); + i915_request_put(rq); + goto err_result1; + } + + if (i915_request_wait(rq, 0, HZ / 2) < 0) { + i915_request_put(rq); + err = -ETIME; + goto err_result1; + } + i915_request_put(rq); + + err = compare_isolation(engine, ref, result, A, poison); + +err_result1: + i915_vma_put(result[1]); +err_result0: + i915_vma_put(result[0]); +err_ref1: + i915_vma_put(ref[1]); +err_ref0: + i915_vma_put(ref[0]); +err_B: + intel_context_put(B); +err_A: + intel_context_put(A); + return err; +} + +static bool skip_isolation(const struct intel_engine_cs *engine) +{ + if (engine->class == COPY_ENGINE_CLASS && INTEL_GEN(engine->i915) == 9) + return true; + + if (engine->class == RENDER_CLASS && INTEL_GEN(engine->i915) == 11) + return true; + + return false; +} + +static int live_lrc_isolation(void *arg) +{ + struct intel_gt *gt = arg; + struct intel_engine_cs *engine; + enum intel_engine_id id; + const u32 poison[] = { + STACK_MAGIC, + 0x3a3a3a3a, + 0xc5c5c5c5, + 0xffffffff, + }; + + /* + * Our goal is try and verify that per-context state cannot be + * tampered with by another non-privileged client. + * + * We take the list of context registers from the LRI in the default + * context image and attempt to modify that list from a remote context. + */ + + for_each_engine(engine, gt, id) { + int err = 0; + int i; + + /* Just don't even ask */ + if (!IS_ENABLED(CONFIG_DRM_I915_SELFTEST_BROKEN) && + skip_isolation(engine)) + continue; + + intel_engine_pm_get(engine); + if (engine->pinned_default_state) { + for (i = 0; i < ARRAY_SIZE(poison); i++) { + err = __lrc_isolation(engine, poison[i]); + if (err) + break; + } + } + if (igt_flush_test(gt->i915)) + err = -EIO; + intel_engine_pm_put(engine); + if (err) + return err; + } + + return 0; +} + static int __live_pphwsp_runtime(struct intel_engine_cs *engine) { struct intel_context *ce; @@ -4845,6 +5389,7 @@ int intel_lrc_live_selftests(struct drm_i915_private *i915) SUBTEST(live_lrc_fixed), SUBTEST(live_lrc_state), SUBTEST(live_lrc_gpr), + SUBTEST(live_lrc_isolation), SUBTEST(live_lrc_timestamp), SUBTEST(live_pphwsp_runtime), }; From patchwork Wed Feb 26 09:43:04 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405755 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 8FB9F1580 for ; Wed, 26 Feb 2020 09:44:04 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 786722467B for ; Wed, 26 Feb 2020 09:44:04 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 786722467B Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 6D8036E42E; Wed, 26 Feb 2020 09:44:03 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id BE3ED6E408 for ; Wed, 26 Feb 2020 09:43:38 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354703-1500050 for multiple; Wed, 26 Feb 2020 09:43:15 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:04 +0000 Message-Id: <20200226094314.1500667-7-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 07/17] drm/i915/selftests: Check recovery from corrupted LRC X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Matthew Auld Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" Check that we can recover if the LRC is totally corrupted. Based on a very simple theory that anything that can be adjusted via the context (i.e. on behalf of the user), should be under the purview of the per-engine-reset. Signed-off-by: Chris Wilson Reviewed-by: Matthew Auld --- drivers/gpu/drm/i915/gt/selftest_lrc.c | 135 +++++++++++++++++++++++++ 1 file changed, 135 insertions(+) diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c b/drivers/gpu/drm/i915/gt/selftest_lrc.c index 810f7857ad26..d7f98aada626 100644 --- a/drivers/gpu/drm/i915/gt/selftest_lrc.c +++ b/drivers/gpu/drm/i915/gt/selftest_lrc.c @@ -5292,6 +5292,140 @@ static int live_lrc_isolation(void *arg) return 0; } +static void garbage_reset(struct intel_engine_cs *engine, + struct i915_request *rq) +{ + const unsigned int bit = I915_RESET_ENGINE + engine->id; + unsigned long *lock = &engine->gt->reset.flags; + + if (test_and_set_bit(bit, lock)) + return; + + tasklet_disable(&engine->execlists.tasklet); + + if (!rq->fence.error) + intel_engine_reset(engine, NULL); + + tasklet_enable(&engine->execlists.tasklet); + clear_and_wake_up_bit(bit, lock); +} + +static struct i915_request *garbage(struct intel_context *ce, + struct rnd_state *prng) +{ + struct i915_request *rq; + int err; + + err = intel_context_pin(ce); + if (err) + return ERR_PTR(err); + + prandom_bytes_state(prng, + ce->lrc_reg_state, + ce->engine->context_size - + LRC_STATE_PN * PAGE_SIZE); + + rq = intel_context_create_request(ce); + if (IS_ERR(rq)) { + err = PTR_ERR(rq); + goto err_unpin; + } + + i915_request_get(rq); + i915_request_add(rq); + return rq; + +err_unpin: + intel_context_unpin(ce); + return ERR_PTR(err); +} + +static int __lrc_garbage(struct intel_engine_cs *engine, struct rnd_state *prng) +{ + struct intel_context *ce; + struct i915_request *hang; + int err = 0; + + ce = intel_context_create(engine); + if (IS_ERR(ce)) + return PTR_ERR(ce); + + hang = garbage(ce, prng); + if (IS_ERR(hang)) { + err = PTR_ERR(hang); + goto err_ce; + } + + if (wait_for_submit(engine, hang, HZ / 2)) { + i915_request_put(hang); + err = -ETIME; + goto err_ce; + } + + intel_context_set_banned(ce); + garbage_reset(engine, hang); + + intel_engine_flush_submission(engine); + if (!hang->fence.error) { + i915_request_put(hang); + pr_err("%s: corrupted context was not reset\n", + engine->name); + err = -EINVAL; + goto err_ce; + } + + if (i915_request_wait(hang, 0, HZ / 2) < 0) { + pr_err("%s: corrupted context did not recover\n", + engine->name); + i915_request_put(hang); + err = -EIO; + goto err_ce; + } + i915_request_put(hang); + +err_ce: + intel_context_put(ce); + return err; +} + +static int live_lrc_garbage(void *arg) +{ + struct intel_gt *gt = arg; + struct intel_engine_cs *engine; + enum intel_engine_id id; + + /* + * Verify that we can recover if one context state is completely + * corrupted. + */ + + if (!IS_ENABLED(CONFIG_DRM_I915_SELFTEST_BROKEN)) + return 0; + + for_each_engine(engine, gt, id) { + I915_RND_STATE(prng); + int err = 0, i; + + if (!intel_has_reset_engine(engine->gt)) + continue; + + intel_engine_pm_get(engine); + for (i = 0; i < 3; i++) { + err = __lrc_garbage(engine, &prng); + if (err) + break; + } + intel_engine_pm_put(engine); + + if (igt_flush_test(gt->i915)) + err = -EIO; + if (err) + return err; + } + + return 0; +} + static int __live_pphwsp_runtime(struct intel_engine_cs *engine) { struct intel_context *ce; @@ -5391,6 +5525,7 @@ int intel_lrc_live_selftests(struct drm_i915_private *i915) SUBTEST(live_lrc_gpr), SUBTEST(live_lrc_isolation), SUBTEST(live_lrc_timestamp), + SUBTEST(live_lrc_garbage), SUBTEST(live_pphwsp_runtime), }; From patchwork Wed Feb 26 09:43:05 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405753 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id E70AB14D5 for ; Wed, 26 Feb 2020 09:44:03 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id CFB032467B for ; Wed, 26 Feb 2020 09:44:03 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org CFB032467B Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 2A3866E425; Wed, 26 Feb 2020 09:44:03 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id D26C66E40C for ; Wed, 26 Feb 2020 09:43:25 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354704-1500050 for multiple; Wed, 26 Feb 2020 09:43:16 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:05 +0000 Message-Id: <20200226094314.1500667-8-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 08/17] drm/i915/selftests: Wait for the kernel context switch X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" As we require a context switch to ensure that the current context is switched out and saved to memory, perform an explicit switch to the kernel context and wait for it. Signed-off-by: Chris Wilson --- drivers/gpu/drm/i915/gt/selftest_lrc.c | 37 +++++++++++++++++++------- 1 file changed, 28 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c b/drivers/gpu/drm/i915/gt/selftest_lrc.c index d7f98aada626..95da6b880e3f 100644 --- a/drivers/gpu/drm/i915/gt/selftest_lrc.c +++ b/drivers/gpu/drm/i915/gt/selftest_lrc.c @@ -4015,6 +4015,31 @@ static int emit_semaphore_signal(struct intel_context *ce, void *slot) return 0; } +static int context_sync(struct intel_context *ce) +{ + struct i915_request *rq; + struct dma_fence *fence; + int err = 0; + + rq = intel_engine_create_kernel_request(ce->engine); + if (IS_ERR(rq)) + return PTR_ERR(rq); + + fence = i915_active_fence_get(&ce->timeline->last_request); + if (fence) { + i915_request_await_dma_fence(rq, fence); + dma_fence_put(fence); + } + + rq = i915_request_get(rq); + i915_request_add(rq); + if (i915_request_wait(rq, 0, HZ / 2) < 0) + err = -ETIME; + i915_request_put(rq); + + return err; +} + static int live_lrc_layout(void *arg) { struct intel_gt *gt = arg; @@ -4638,16 +4663,10 @@ static int __lrc_timestamp(const struct lrc_timestamp *arg, bool preempt) wmb(); } - if (i915_request_wait(rq, 0, HZ / 2) < 0) { - err = -ETIME; - goto err; - } - - /* and wait for switch to kernel */ - if (igt_flush_test(arg->engine->i915)) { - err = -EIO; + /* and wait for switch to kernel (to save our context to memory) */ + err = context_sync(arg->ce[0]); + if (err) goto err; - } rmb(); From patchwork Wed Feb 26 09:43:06 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405729 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 5537114D5 for ; Wed, 26 Feb 2020 09:43:35 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 3DBA324676 for ; Wed, 26 Feb 2020 09:43:35 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 3DBA324676 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 6DE176E3F2; Wed, 26 Feb 2020 09:43:34 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id B650E6E3F2 for ; Wed, 26 Feb 2020 09:43:26 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354705-1500050 for multiple; Wed, 26 Feb 2020 09:43:16 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:06 +0000 Message-Id: <20200226094314.1500667-9-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 09/17] drm/i915/selftests: Be a little more lenient for reset workers X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" Give the reset worker a kick before losing help when waiting for hang recovery, as the CPU scheduler is a little unreliable. Signed-off-by: Chris Wilson --- drivers/gpu/drm/i915/gt/selftest_lrc.c | 74 ++++++++++++++++++-------- 1 file changed, 52 insertions(+), 22 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/selftest_lrc.c b/drivers/gpu/drm/i915/gt/selftest_lrc.c index 95da6b880e3f..af5b3da6d894 100644 --- a/drivers/gpu/drm/i915/gt/selftest_lrc.c +++ b/drivers/gpu/drm/i915/gt/selftest_lrc.c @@ -90,6 +90,48 @@ static int wait_for_submit(struct intel_engine_cs *engine, return -ETIME; } +static int wait_for_reset(struct intel_engine_cs *engine, + struct i915_request *rq, + unsigned long timeout) +{ + timeout += jiffies; + do { + cond_resched(); + intel_engine_flush_submission(engine); + + if (READ_ONCE(engine->execlists.pending[0])) + continue; + + if (i915_request_completed(rq)) + break; + + if (READ_ONCE(rq->fence.error)) + break; + } while (time_before(jiffies, timeout)); + + flush_scheduled_work(); + + if (rq->fence.error != -EIO) { + pr_err("%s: hanging request %llx:%lld not reset\n", + engine->name, + rq->fence.context, + rq->fence.seqno); + return -EINVAL; + } + + /* Give the request a jiffie to complete after flushing the worker */ + if (i915_request_wait(rq, 0, + max(0l, (long)(timeout - jiffies)) + 1) < 0) { + pr_err("%s: hanging request %llx:%lld did not complete\n", + engine->name, + rq->fence.context, + rq->fence.seqno); + return -ETIME; + } + + return 0; +} + static int live_sanitycheck(void *arg) { struct intel_gt *gt = arg; @@ -1805,14 +1847,9 @@ static int __cancel_active0(struct live_preempt_cancel *arg) if (err) goto out; - if (i915_request_wait(rq, 0, HZ / 5) < 0) { - err = -EIO; - goto out; - } - - if (rq->fence.error != -EIO) { - pr_err("Cancelled inflight0 request did not report -EIO\n"); - err = -EINVAL; + err = wait_for_reset(arg->engine, rq, HZ / 2); + if (err) { + pr_err("Cancelled inflight0 request did not reset\n"); goto out; } @@ -1870,10 +1907,9 @@ static int __cancel_active1(struct live_preempt_cancel *arg) goto out; igt_spinner_end(&arg->a.spin); - if (i915_request_wait(rq[1], 0, HZ / 5) < 0) { - err = -EIO; + err = wait_for_reset(arg->engine, rq[1], HZ / 2); + if (err) goto out; - } if (rq[0]->fence.error != 0) { pr_err("Normal inflight0 request did not complete\n"); @@ -1953,10 +1989,9 @@ static int __cancel_queued(struct live_preempt_cancel *arg) if (err) goto out; - if (i915_request_wait(rq[2], 0, HZ / 5) < 0) { - err = -EIO; + err = wait_for_reset(arg->engine, rq[2], HZ / 2); + if (err) goto out; - } if (rq[0]->fence.error != -EIO) { pr_err("Cancelled inflight0 request did not report -EIO\n"); @@ -2014,14 +2049,9 @@ static int __cancel_hostile(struct live_preempt_cancel *arg) if (err) goto out; - if (i915_request_wait(rq, 0, HZ / 5) < 0) { - err = -EIO; - goto out; - } - - if (rq->fence.error != -EIO) { - pr_err("Cancelled inflight0 request did not report -EIO\n"); - err = -EINVAL; + err = wait_for_reset(arg->engine, rq, HZ / 2); + if (err) { + pr_err("Cancelled inflight0 request did not reset\n"); goto out; } From patchwork Wed Feb 26 09:43:07 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405735 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 49FD914D5 for ; Wed, 26 Feb 2020 09:43:39 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 329E624673 for ; Wed, 26 Feb 2020 09:43:39 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 329E624673 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 12F8A6E417; Wed, 26 Feb 2020 09:43:35 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id CB8826E409 for ; Wed, 26 Feb 2020 09:43:25 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354706-1500050 for multiple; Wed, 26 Feb 2020 09:43:16 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:07 +0000 Message-Id: <20200226094314.1500667-10-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 10/17] drm/i915/selftests: Add request throughput measurement to perf X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" Under ideal circumstances, the driver should be able to keep the GPU fully saturated with work. Measure how close to ideal we get under the harshest of conditions with no user payload. Signed-off-by: Chris Wilson --- .../drm/i915/selftests/i915_perf_selftests.h | 1 + drivers/gpu/drm/i915/selftests/i915_request.c | 280 +++++++++++++++++- 2 files changed, 280 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/selftests/i915_perf_selftests.h b/drivers/gpu/drm/i915/selftests/i915_perf_selftests.h index 3bf7f53e9924..d8da142985eb 100644 --- a/drivers/gpu/drm/i915/selftests/i915_perf_selftests.h +++ b/drivers/gpu/drm/i915/selftests/i915_perf_selftests.h @@ -16,5 +16,6 @@ * Tests are executed in order by igt/i915_selftest */ selftest(engine_cs, intel_engine_cs_perf_selftests) +selftest(request, i915_request_perf_selftests) selftest(blt, i915_gem_object_blt_perf_selftests) selftest(region, intel_memory_region_perf_selftests) diff --git a/drivers/gpu/drm/i915/selftests/i915_request.c b/drivers/gpu/drm/i915/selftests/i915_request.c index f89d9c42f1fa..91f67995f0ac 100644 --- a/drivers/gpu/drm/i915/selftests/i915_request.c +++ b/drivers/gpu/drm/i915/selftests/i915_request.c @@ -23,6 +23,7 @@ */ #include +#include #include "gem/i915_gem_pm.h" #include "gem/selftests/mock_context.h" @@ -1233,7 +1234,7 @@ static int live_parallel_engines(void *arg) struct igt_live_test t; unsigned int idx; - snprintf(name, sizeof(name), "%pS", fn); + snprintf(name, sizeof(name), "%ps", *fn); err = igt_live_test_begin(&t, i915, __func__, name); if (err) break; @@ -1470,3 +1471,280 @@ int i915_request_live_selftests(struct drm_i915_private *i915) return i915_subtests(tests, i915); } + +struct perf_parallel { + struct intel_engine_cs *engine; + unsigned long count; + ktime_t time; + ktime_t busy; + u64 runtime; +}; + +static int switch_to_kernel_sync(struct intel_context *ce, int err) +{ + struct i915_request *rq; + struct dma_fence *fence; + + rq = intel_engine_create_kernel_request(ce->engine); + if (IS_ERR(rq)) + return PTR_ERR(rq); + + fence = i915_active_fence_get(&ce->timeline->last_request); + if (fence) { + i915_request_await_dma_fence(rq, fence); + dma_fence_put(fence); + } + + rq = i915_request_get(rq); + i915_request_add(rq); + if (i915_request_wait(rq, 0, HZ / 2) < 0 && !err) + err = -ETIME; + i915_request_put(rq); + + while (!err && !intel_engine_is_idle(ce->engine)) + intel_engine_flush_submission(ce->engine); + + return err; +} + +static int perf_sync(void *arg) +{ + struct perf_parallel *p = arg; + struct intel_engine_cs *engine = p->engine; + struct intel_context *ce; + IGT_TIMEOUT(end_time); + unsigned long count; + bool busy; + int err = 0; + + ce = intel_context_create(engine); + if (IS_ERR(ce)) + return PTR_ERR(ce); + + err = intel_context_pin(ce); + if (err) { + intel_context_put(ce); + return err; + } + + busy = false; + if (intel_engine_supports_stats(engine) && + !intel_enable_engine_stats(engine)) { + p->busy = intel_engine_get_busy_time(engine); + busy = true; + } + + p->time = ktime_get(); + count = 0; + do { + struct i915_request *rq; + + rq = i915_request_create(ce); + if (IS_ERR(rq)) { + err = PTR_ERR(rq); + break; + } + + i915_request_get(rq); + i915_request_add(rq); + + err = 0; + if (i915_request_wait(rq, 0, HZ / 5) < 0) + err = -ETIME; + i915_request_put(rq); + if (err) + break; + + count++; + } while (!__igt_timeout(end_time, NULL)); + p->time = ktime_sub(ktime_get(), p->time); + + if (busy) { + p->busy = ktime_sub(intel_engine_get_busy_time(engine), + p->busy); + intel_disable_engine_stats(engine); + } + + err = switch_to_kernel_sync(ce, err); + p->runtime = intel_context_get_total_runtime_ns(ce); + p->count = count; + + intel_context_unpin(ce); + intel_context_put(ce); + return err; +} + +static int perf_many(void *arg) +{ + struct perf_parallel *p = arg; + struct intel_engine_cs *engine = p->engine; + struct intel_context *ce; + IGT_TIMEOUT(end_time); + unsigned long count; + int err = 0; + bool busy; + + ce = intel_context_create(engine); + if (IS_ERR(ce)) + return PTR_ERR(ce); + + err = intel_context_pin(ce); + if (err) { + intel_context_put(ce); + return err; + } + + busy = false; + if (intel_engine_supports_stats(engine) && + !intel_enable_engine_stats(engine)) { + p->busy = intel_engine_get_busy_time(engine); + busy = true; + } + + count = 0; + p->time = ktime_get(); + do { + struct i915_request *rq; + + rq = i915_request_create(ce); + if (IS_ERR(rq)) { + err = PTR_ERR(rq); + break; + } + + i915_request_add(rq); + count++; + } while (!__igt_timeout(end_time, NULL)); + p->time = ktime_sub(ktime_get(), p->time); + + if (busy) { + p->busy = ktime_sub(intel_engine_get_busy_time(engine), + p->busy); + intel_disable_engine_stats(engine); + } + + err = switch_to_kernel_sync(ce, err); + p->runtime = intel_context_get_total_runtime_ns(ce); + p->count = count; + + intel_context_unpin(ce); + intel_context_put(ce); + return err; +} + +static int perf_parallel_engines(void *arg) +{ + struct drm_i915_private *i915 = arg; + static int (* const func[])(void *arg) = { + perf_sync, + perf_many, + NULL, + }; + const unsigned int nengines = num_uabi_engines(i915); + struct pm_qos_request qos = {}; + struct intel_engine_cs *engine; + int (* const *fn)(void *arg); + struct { + struct perf_parallel p; + struct task_struct *tsk; + } *engines; + int err = 0; + + engines = kcalloc(nengines, sizeof(*engines), GFP_KERNEL); + if (!engines) + return -ENOMEM; + + pm_qos_add_request(&qos, PM_QOS_CPU_DMA_LATENCY, 0); + + for (fn = func; *fn; fn++) { + char name[KSYM_NAME_LEN]; + struct igt_live_test t; + unsigned int idx; + + snprintf(name, sizeof(name), "%ps", *fn); + err = igt_live_test_begin(&t, i915, __func__, name); + if (err) + break; + + atomic_set(&i915->selftest.counter, nengines); + + idx = 0; + for_each_uabi_engine(engine, i915) { + intel_engine_pm_get(engine); + + memset(&engines[idx].p, 0, sizeof(engines[idx].p)); + engines[idx].p.engine = engine; + + engines[idx].tsk = kthread_run(*fn, &engines[idx].p, + "igt:%s", engine->name); + if (IS_ERR(engines[idx].tsk)) { + err = PTR_ERR(engines[idx].tsk); + intel_engine_pm_put(engine); + break; + } + get_task_struct(engines[idx++].tsk); + } + + yield(); /* start all threads before we kthread_stop() */ + + idx = 0; + for_each_uabi_engine(engine, i915) { + int status; + + if (IS_ERR(engines[idx].tsk)) + break; + + status = kthread_stop(engines[idx].tsk); + if (status && !err) + err = status; + + intel_engine_pm_put(engine); + put_task_struct(engines[idx++].tsk); + } + + if (igt_live_test_end(&t)) + err = -EIO; + if (err) + break; + + idx = 0; + for_each_uabi_engine(engine, i915) { + struct perf_parallel *p = &engines[idx].p; + u64 busy = 100 * ktime_to_ns(p->busy); + u64 dt = ktime_to_ns(p->time); + int integer, decimal; + + if (dt) { + integer = div64_u64(busy, dt); + busy -= integer * dt; + decimal = div64_u64(100 * busy, dt); + } else { + integer = 0; + decimal = 0; + } + + GEM_BUG_ON(engine != p->engine); + pr_info("%s %5s: { count:%lu, busy:%d.%02d%%, runtime:%lldms, walltime:%lldms }\n", + name, engine->name, p->count, integer, decimal, + div_u64(p->runtime, 1000 * 1000), + div_u64(ktime_to_ns(p->time), 1000 * 1000)); + idx++; + } + } + + pm_qos_remove_request(&qos); + kfree(engines); + return err; +} + +int i915_request_perf_selftests(struct drm_i915_private *i915) +{ + static const struct i915_subtest tests[] = { + SUBTEST(perf_parallel_engines), + }; + + if (intel_gt_is_wedged(&i915->gt)) + return 0; + + return i915_subtests(tests, i915); +} From patchwork Wed Feb 26 09:43:08 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405727 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id B3AF214D5 for ; Wed, 26 Feb 2020 09:43:31 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 9C99324670 for ; Wed, 26 Feb 2020 09:43:31 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 9C99324670 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 939626E3FC; Wed, 26 Feb 2020 09:43:27 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id CAE5B6E408 for ; Wed, 26 Feb 2020 09:43:25 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354707-1500050 for multiple; Wed, 26 Feb 2020 09:43:16 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:08 +0000 Message-Id: <20200226094314.1500667-11-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 11/17] drm/i915/gem: Consolidate ctx->engines[] release X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" Use the same engine_idle_release() routine for cleaning all old ctx->engine[] state, closing any potential races with concurrent execbuf submission. Closes: https://gitlab.freedesktop.org/drm/intel/issues/1241 Signed-off-by: Chris Wilson --- Reorder set-closed/engine_idle_release to avoid premature killing Take a reference to prevent racing context free with engine cleanup --- drivers/gpu/drm/i915/gem/i915_gem_context.c | 199 +++++++++++--------- drivers/gpu/drm/i915/gem/i915_gem_context.h | 1 - 2 files changed, 108 insertions(+), 92 deletions(-) diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.c b/drivers/gpu/drm/i915/gem/i915_gem_context.c index b24ee8e104cf..634c770d0548 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_context.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_context.c @@ -244,7 +244,6 @@ static void __free_engines(struct i915_gem_engines *e, unsigned int count) if (!e->engines[count]) continue; - RCU_INIT_POINTER(e->engines[count]->gem_context, NULL); intel_context_put(e->engines[count]); } kfree(e); @@ -257,7 +256,11 @@ static void free_engines(struct i915_gem_engines *e) static void free_engines_rcu(struct rcu_head *rcu) { - free_engines(container_of(rcu, struct i915_gem_engines, rcu)); + struct i915_gem_engines *engines = + container_of(rcu, struct i915_gem_engines, rcu); + + i915_sw_fence_fini(&engines->fence); + free_engines(engines); } static struct i915_gem_engines *default_engines(struct i915_gem_context *ctx) @@ -271,8 +274,6 @@ static struct i915_gem_engines *default_engines(struct i915_gem_context *ctx) if (!e) return ERR_PTR(-ENOMEM); - e->ctx = ctx; - for_each_engine(engine, gt, id) { struct intel_context *ce; @@ -306,7 +307,6 @@ static void i915_gem_context_free(struct i915_gem_context *ctx) list_del(&ctx->link); spin_unlock(&ctx->i915->gem.contexts.lock); - free_engines(rcu_access_pointer(ctx->engines)); mutex_destroy(&ctx->engines_mutex); if (ctx->timeline) @@ -493,30 +493,110 @@ static void kill_engines(struct i915_gem_engines *engines) static void kill_stale_engines(struct i915_gem_context *ctx) { struct i915_gem_engines *pos, *next; - unsigned long flags; - spin_lock_irqsave(&ctx->stale.lock, flags); + spin_lock_irq(&ctx->stale.lock); + GEM_BUG_ON(!i915_gem_context_is_closed(ctx)); list_for_each_entry_safe(pos, next, &ctx->stale.engines, link) { - if (!i915_sw_fence_await(&pos->fence)) + if (!i915_sw_fence_await(&pos->fence)) { + list_del_init(&pos->link); continue; + } - spin_unlock_irqrestore(&ctx->stale.lock, flags); + spin_unlock_irq(&ctx->stale.lock); kill_engines(pos); - spin_lock_irqsave(&ctx->stale.lock, flags); + spin_lock_irq(&ctx->stale.lock); + GEM_BUG_ON(i915_sw_fence_signaled(&pos->fence)); list_safe_reset_next(pos, next, link); list_del_init(&pos->link); /* decouple from FENCE_COMPLETE */ i915_sw_fence_complete(&pos->fence); } - spin_unlock_irqrestore(&ctx->stale.lock, flags); + spin_unlock_irq(&ctx->stale.lock); } static void kill_context(struct i915_gem_context *ctx) { kill_stale_engines(ctx); - kill_engines(__context_engines_static(ctx)); +} + +static int engines_notify(struct i915_sw_fence *fence, + enum i915_sw_fence_notify state) +{ + struct i915_gem_engines *engines = + container_of(fence, typeof(*engines), fence); + + switch (state) { + case FENCE_COMPLETE: + if (!list_empty(&engines->link)) { + struct i915_gem_context *ctx = engines->ctx; + unsigned long flags; + + spin_lock_irqsave(&ctx->stale.lock, flags); + list_del(&engines->link); + spin_unlock_irqrestore(&ctx->stale.lock, flags); + } + break; + + case FENCE_FREE: + i915_gem_context_put(engines->ctx); + init_rcu_head(&engines->rcu); + call_rcu(&engines->rcu, free_engines_rcu); + break; + } + + return NOTIFY_DONE; +} + +static void engines_idle_release(struct i915_gem_context *ctx, + struct i915_gem_engines *engines) +{ + struct i915_gem_engines_iter it; + struct intel_context *ce; + + i915_sw_fence_init(&engines->fence, engines_notify); + INIT_LIST_HEAD(&engines->link); + + engines->ctx = i915_gem_context_get(ctx); + + for_each_gem_engine(ce, engines, it) { + int err = 0; + + RCU_INIT_POINTER(ce->gem_context, NULL); + + if (!ce->timeline) { /* XXX serialisation with execbuf? */ + intel_context_set_banned(ce); + continue; + } + + mutex_lock(&ce->timeline->mutex); + if (!list_empty(&ce->timeline->requests)) { + struct i915_request *rq; + + rq = list_last_entry(&ce->timeline->requests, + typeof(*rq), + link); + + err = i915_sw_fence_await_dma_fence(&engines->fence, + &rq->fence, 0, + GFP_KERNEL); + } + mutex_unlock(&ce->timeline->mutex); + if (err < 0) + goto kill; + } + + spin_lock_irq(&engines->ctx->stale.lock); + if (!i915_gem_context_is_closed(engines->ctx)) + list_add_tail(&engines->link, &engines->ctx->stale.engines); + spin_unlock_irq(&engines->ctx->stale.lock); + +kill: + if (list_empty(&engines->link)) /* raced, already closed */ + kill_engines(engines); + + i915_sw_fence_commit(&engines->fence); } static void set_closed_name(struct i915_gem_context *ctx) @@ -540,11 +620,16 @@ static void context_close(struct i915_gem_context *ctx) { struct i915_address_space *vm; + /* Flush any concurrent set_engines() */ + mutex_lock(&ctx->engines_mutex); + engines_idle_release(ctx, rcu_replace_pointer(ctx->engines, NULL, 1)); i915_gem_context_set_closed(ctx); - set_closed_name(ctx); + mutex_unlock(&ctx->engines_mutex); mutex_lock(&ctx->mutex); + set_closed_name(ctx); + vm = i915_gem_context_vm(ctx); if (vm) i915_vm_close(vm); @@ -1628,77 +1713,6 @@ static const i915_user_extension_fn set_engines__extensions[] = { [I915_CONTEXT_ENGINES_EXT_BOND] = set_engines__bond, }; -static int engines_notify(struct i915_sw_fence *fence, - enum i915_sw_fence_notify state) -{ - struct i915_gem_engines *engines = - container_of(fence, typeof(*engines), fence); - - switch (state) { - case FENCE_COMPLETE: - if (!list_empty(&engines->link)) { - struct i915_gem_context *ctx = engines->ctx; - unsigned long flags; - - spin_lock_irqsave(&ctx->stale.lock, flags); - list_del(&engines->link); - spin_unlock_irqrestore(&ctx->stale.lock, flags); - } - break; - - case FENCE_FREE: - init_rcu_head(&engines->rcu); - call_rcu(&engines->rcu, free_engines_rcu); - break; - } - - return NOTIFY_DONE; -} - -static void engines_idle_release(struct i915_gem_engines *engines) -{ - struct i915_gem_engines_iter it; - struct intel_context *ce; - unsigned long flags; - - GEM_BUG_ON(!engines); - i915_sw_fence_init(&engines->fence, engines_notify); - - INIT_LIST_HEAD(&engines->link); - spin_lock_irqsave(&engines->ctx->stale.lock, flags); - if (!i915_gem_context_is_closed(engines->ctx)) - list_add(&engines->link, &engines->ctx->stale.engines); - spin_unlock_irqrestore(&engines->ctx->stale.lock, flags); - if (list_empty(&engines->link)) /* raced, already closed */ - goto kill; - - for_each_gem_engine(ce, engines, it) { - struct dma_fence *fence; - int err; - - if (!ce->timeline) - continue; - - fence = i915_active_fence_get(&ce->timeline->last_request); - if (!fence) - continue; - - err = i915_sw_fence_await_dma_fence(&engines->fence, - fence, 0, - GFP_KERNEL); - - dma_fence_put(fence); - if (err < 0) - goto kill; - } - goto out; - -kill: - kill_engines(engines); -out: - i915_sw_fence_commit(&engines->fence); -} - static int set_engines(struct i915_gem_context *ctx, const struct drm_i915_gem_context_param *args) @@ -1741,8 +1755,6 @@ set_engines(struct i915_gem_context *ctx, if (!set.engines) return -ENOMEM; - set.engines->ctx = ctx; - for (n = 0; n < num_engines; n++) { struct i915_engine_class_instance ci; struct intel_engine_cs *engine; @@ -1795,6 +1807,11 @@ set_engines(struct i915_gem_context *ctx, replace: mutex_lock(&ctx->engines_mutex); + if (i915_gem_context_is_closed(ctx)) { + mutex_unlock(&ctx->engines_mutex); + free_engines(set.engines); + return -ENOENT; + } if (args->size) i915_gem_context_set_user_engines(ctx); else @@ -1803,7 +1820,7 @@ set_engines(struct i915_gem_context *ctx, mutex_unlock(&ctx->engines_mutex); /* Keep track of old engine sets for kill_context() */ - engines_idle_release(set.engines); + engines_idle_release(ctx, set.engines); return 0; } @@ -2079,8 +2096,6 @@ static int clone_engines(struct i915_gem_context *dst, if (!clone) goto err_unlock; - clone->ctx = dst; - for (n = 0; n < e->num_engines; n++) { struct intel_engine_cs *engine; @@ -2123,8 +2138,7 @@ static int clone_engines(struct i915_gem_context *dst, i915_gem_context_unlock_engines(src); /* Serialised by constructor */ - free_engines(__context_engines_static(dst)); - RCU_INIT_POINTER(dst->engines, clone); + engines_idle_release(dst, rcu_replace_pointer(dst->engines, clone, 1)); if (user_engines) i915_gem_context_set_user_engines(dst); else @@ -2555,6 +2569,9 @@ i915_gem_engines_iter_next(struct i915_gem_engines_iter *it) const struct i915_gem_engines *e = it->engines; struct intel_context *ctx; + if (unlikely(!e)) + return NULL; + do { if (it->idx >= e->num_engines) return NULL; diff --git a/drivers/gpu/drm/i915/gem/i915_gem_context.h b/drivers/gpu/drm/i915/gem/i915_gem_context.h index 3ae61a355d87..57b7ae2893e1 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_context.h +++ b/drivers/gpu/drm/i915/gem/i915_gem_context.h @@ -207,7 +207,6 @@ static inline void i915_gem_engines_iter_init(struct i915_gem_engines_iter *it, struct i915_gem_engines *engines) { - GEM_BUG_ON(!engines); it->engines = engines; it->idx = 0; } From patchwork Wed Feb 26 09:43:09 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405733 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 43EB014D5 for ; Wed, 26 Feb 2020 09:43:38 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 2C90D24670 for ; Wed, 26 Feb 2020 09:43:38 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 2C90D24670 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id D550F6E40C; Wed, 26 Feb 2020 09:43:34 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id B7D6E6E3FC for ; Wed, 26 Feb 2020 09:43:26 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354708-1500050 for multiple; Wed, 26 Feb 2020 09:43:16 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:09 +0000 Message-Id: <20200226094314.1500667-12-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 12/17] drm/i915/gt: Prevent allocation on a banned context X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Matthew Auld Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" If a context is banned even before we submit our first request to it, report the failure before we attempt to allocate any resources for the context. Signed-off-by: Chris Wilson Reviewed-by: Matthew Auld --- drivers/gpu/drm/i915/gt/intel_context.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/gpu/drm/i915/gt/intel_context.c b/drivers/gpu/drm/i915/gt/intel_context.c index 8bb444cda14f..01474d3a558b 100644 --- a/drivers/gpu/drm/i915/gt/intel_context.c +++ b/drivers/gpu/drm/i915/gt/intel_context.c @@ -51,6 +51,11 @@ int intel_context_alloc_state(struct intel_context *ce) return -EINTR; if (!test_bit(CONTEXT_ALLOC_BIT, &ce->flags)) { + if (intel_context_is_banned(ce)) { + err = -EIO; + goto unlock; + } + err = ce->ops->alloc(ce); if (unlikely(err)) goto unlock; From patchwork Wed Feb 26 09:43:10 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405739 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 2A2D614D5 for ; Wed, 26 Feb 2020 09:43:41 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 12D7524673 for ; Wed, 26 Feb 2020 09:43:41 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 12D7524673 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 405526E419; Wed, 26 Feb 2020 09:43:35 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id B69446E3F9 for ; Wed, 26 Feb 2020 09:43:26 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354709-1500050 for multiple; Wed, 26 Feb 2020 09:43:16 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:10 +0000 Message-Id: <20200226094314.1500667-13-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 13/17] drm/i915/gem: Check that the context wasn't closed during setup X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Matthew Auld Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" As setup takes a long time, the user may close the context during the construction of the execbuf. In order to make sure we correctly track all outstanding work with non-persistent contexts, we need to serialise the submission with the context closure and mop up any leaks. Signed-off-by: Chris Wilson Reviewed-by: Matthew Auld --- drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c index fe1e50937fe2..66d45629cc0d 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c @@ -2736,6 +2736,12 @@ i915_gem_do_execbuffer(struct drm_device *dev, goto err_batch_unpin; } + /* Check that the context wasn't destroyed before setup */ + if (!rcu_access_pointer(eb.context->gem_context)) { + err = -ENOENT; + goto err_request; + } + if (in_fence) { err = i915_request_await_dma_fence(eb.request, in_fence); if (err < 0) From patchwork Wed Feb 26 09:43:11 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405751 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 68E801580 for ; Wed, 26 Feb 2020 09:43:53 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 519EF24653 for ; Wed, 26 Feb 2020 09:43:53 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 519EF24653 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 74F4A6E40D; Wed, 26 Feb 2020 09:43:49 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id D9B106E40D for ; Wed, 26 Feb 2020 09:43:25 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354710-1500050 for multiple; Wed, 26 Feb 2020 09:43:17 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:11 +0000 Message-Id: <20200226094314.1500667-14-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 14/17] drm/i915/gt: Declare when we enabled timeslicing X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Kenneth Graunke Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" Let userspace know if they can trust timeslicing by including it as part of the I915_PARAM_HAS_SCHEDULER::I915_SCHEDULER_CAP_TIMESLICING v2: Only declare timeslicing if we can safely preempt userspace. Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing") Signed-off-by: Chris Wilson Cc: Kenneth Graunke --- drivers/gpu/drm/i915/gt/intel_engine.h | 3 ++- drivers/gpu/drm/i915/gt/intel_engine_user.c | 5 +++++ include/uapi/drm/i915_drm.h | 1 + 3 files changed, 8 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/gt/intel_engine.h b/drivers/gpu/drm/i915/gt/intel_engine.h index 29c8c03c5caa..a32dc82a90d4 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine.h +++ b/drivers/gpu/drm/i915/gt/intel_engine.h @@ -326,7 +326,8 @@ intel_engine_has_timeslices(const struct intel_engine_cs *engine) if (!IS_ACTIVE(CONFIG_DRM_I915_TIMESLICE_DURATION)) return false; - return intel_engine_has_semaphores(engine); + return (intel_engine_has_semaphores(engine) && + intel_engine_has_preemption(engine)); } #endif /* _INTEL_RINGBUFFER_H_ */ diff --git a/drivers/gpu/drm/i915/gt/intel_engine_user.c b/drivers/gpu/drm/i915/gt/intel_engine_user.c index 848decee9066..b84fdd722781 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_user.c +++ b/drivers/gpu/drm/i915/gt/intel_engine_user.c @@ -121,6 +121,11 @@ static void set_scheduler_caps(struct drm_i915_private *i915) else disabled |= BIT(map[i].sched); } + + if (intel_engine_has_timeslices(engine)) + enabled |= I915_SCHEDULER_CAP_TIMESLICING; + else + disabled |= I915_SCHEDULER_CAP_TIMESLICING; } i915->caps.scheduler = enabled & ~disabled; diff --git a/include/uapi/drm/i915_drm.h b/include/uapi/drm/i915_drm.h index 2813e579b480..4f903431a3fe 100644 --- a/include/uapi/drm/i915_drm.h +++ b/include/uapi/drm/i915_drm.h @@ -523,6 +523,7 @@ typedef struct drm_i915_irq_wait { #define I915_SCHEDULER_CAP_PREEMPTION (1ul << 2) #define I915_SCHEDULER_CAP_SEMAPHORES (1ul << 3) #define I915_SCHEDULER_CAP_ENGINE_BUSY_STATS (1ul << 4) +#define I915_SCHEDULER_CAP_TIMESLICING (1ul << 5) #define I915_PARAM_HUC_STATUS 42 From patchwork Wed Feb 26 09:43:12 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405745 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 2F39F138D for ; Wed, 26 Feb 2020 09:43:51 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 17DB42467A for ; Wed, 26 Feb 2020 09:43:51 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 17DB42467A Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 45A636E415; Wed, 26 Feb 2020 09:43:49 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id A93616E408 for ; Wed, 26 Feb 2020 09:43:34 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354711-1500050 for multiple; Wed, 26 Feb 2020 09:43:17 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:12 +0000 Message-Id: <20200226094314.1500667-15-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 15/17] drm/i915/gt: Yield the timeslice if caught waiting on a user semaphore X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Kenneth Graunke Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" If we find ourselves waiting on a MI_SEMAPHORE_WAIT, either within the user batch or in our own preamble, the engine raises a GT_WAIT_ON_SEMAPHORE interrupt. We can unmask that interrupt and so respond to a semaphore wait by yielding the timeslice, if we have another context to yield to! The only real complication is that the interrupt is only generated for the start of the semaphore wait, and is asynchronous to our process_csb() -- that is, we may not have registered the timeslice before we see the interrupt. To ensure we don't miss a potential semaphore blocking forward progress (e.g. selftests/live_timeslice_preempt) we mark the interrupt and apply it to the next timeslice regardless of whether it was active at the time. v2: We use semaphores in preempt-to-busy, within the timeslicing implementation itself! Ergo, when we do insert a preemption due to an expired timeslice, the new context may start with the missed semaphore flagged by the retired context and be yielded, ad infinitum. To avoid this, read the context id at the time of the semaphore interrupt and only yield if that context is still active. Fixes: 8ee36e048c98 ("drm/i915/execlists: Minimalistic timeslicing") Signed-off-by: Chris Wilson Cc: Tvrtko Ursulin Cc: Kenneth Graunke --- drivers/gpu/drm/i915/gt/intel_engine_cs.c | 6 +++ drivers/gpu/drm/i915/gt/intel_engine_types.h | 9 +++++ drivers/gpu/drm/i915/gt/intel_gt_irq.c | 13 ++++++- drivers/gpu/drm/i915/gt/intel_lrc.c | 40 +++++++++++++++++--- drivers/gpu/drm/i915/i915_reg.h | 1 + 5 files changed, 61 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_engine_cs.c b/drivers/gpu/drm/i915/gt/intel_engine_cs.c index 119c9cb24fd4..eec6e3245a97 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/gt/intel_engine_cs.c @@ -1288,6 +1288,12 @@ static void intel_engine_print_registers(struct intel_engine_cs *engine, if (engine->id == RENDER_CLASS && IS_GEN_RANGE(dev_priv, 4, 7)) drm_printf(m, "\tCCID: 0x%08x\n", ENGINE_READ(engine, CCID)); + if (HAS_EXECLISTS(dev_priv)) { + drm_printf(m, "\tEL_STAT_HI: 0x%08x\n", + ENGINE_READ(engine, RING_EXECLIST_STATUS_HI)); + drm_printf(m, "\tEL_STAT_LO: 0x%08x\n", + ENGINE_READ(engine, RING_EXECLIST_STATUS_LO)); + } drm_printf(m, "\tRING_START: 0x%08x\n", ENGINE_READ(engine, RING_START)); drm_printf(m, "\tRING_HEAD: 0x%08x\n", diff --git a/drivers/gpu/drm/i915/gt/intel_engine_types.h b/drivers/gpu/drm/i915/gt/intel_engine_types.h index b23366a81048..24cff658e6e5 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_types.h +++ b/drivers/gpu/drm/i915/gt/intel_engine_types.h @@ -156,6 +156,15 @@ struct intel_engine_execlists { */ struct i915_priolist default_priolist; + /** + * @yield: CCID at the time of the last semaphore-wait interrupt. + * + * Instead of leaving a semaphore busy-spinning on an engine, we would + * like to switch to another ready context, i.e. yielding the semaphore + * timeslice. + */ + u32 yield; + /** * @error_interrupt: CS Master EIR * diff --git a/drivers/gpu/drm/i915/gt/intel_gt_irq.c b/drivers/gpu/drm/i915/gt/intel_gt_irq.c index f0e7fd95165a..875bd0392ffc 100644 --- a/drivers/gpu/drm/i915/gt/intel_gt_irq.c +++ b/drivers/gpu/drm/i915/gt/intel_gt_irq.c @@ -39,6 +39,13 @@ cs_irq_handler(struct intel_engine_cs *engine, u32 iir) } } + if (iir & GT_WAIT_SEMAPHORE_INTERRUPT) { + WRITE_ONCE(engine->execlists.yield, + ENGINE_READ_FW(engine, RING_EXECLIST_STATUS_HI)); + if (del_timer(&engine->execlists.timer)) + tasklet = true; + } + if (iir & GT_CONTEXT_SWITCH_INTERRUPT) tasklet = true; @@ -228,7 +235,8 @@ void gen11_gt_irq_postinstall(struct intel_gt *gt) const u32 irqs = GT_CS_MASTER_ERROR_INTERRUPT | GT_RENDER_USER_INTERRUPT | - GT_CONTEXT_SWITCH_INTERRUPT; + GT_CONTEXT_SWITCH_INTERRUPT | + GT_WAIT_SEMAPHORE_INTERRUPT; struct intel_uncore *uncore = gt->uncore; const u32 dmask = irqs << 16 | irqs; const u32 smask = irqs << 16; @@ -366,7 +374,8 @@ void gen8_gt_irq_postinstall(struct intel_gt *gt) const u32 irqs = GT_CS_MASTER_ERROR_INTERRUPT | GT_RENDER_USER_INTERRUPT | - GT_CONTEXT_SWITCH_INTERRUPT; + GT_CONTEXT_SWITCH_INTERRUPT | + GT_WAIT_SEMAPHORE_INTERRUPT; const u32 gt_interrupts[] = { irqs << GEN8_RCS_IRQ_SHIFT | irqs << GEN8_BCS_IRQ_SHIFT, irqs << GEN8_VCS0_IRQ_SHIFT | irqs << GEN8_VCS1_IRQ_SHIFT, diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 35c5cf786726..253e89fecd95 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1728,7 +1728,8 @@ static void defer_active(struct intel_engine_cs *engine) } static bool -need_timeslice(struct intel_engine_cs *engine, const struct i915_request *rq) +need_timeslice(const struct intel_engine_cs *engine, + const struct i915_request *rq) { int hint; @@ -1744,6 +1745,31 @@ need_timeslice(struct intel_engine_cs *engine, const struct i915_request *rq) return hint >= effective_prio(rq); } +static bool +timeslice_yield(const struct intel_engine_execlists *el, + const struct i915_request *rq) +{ + /* + * Once bitten, forever smitten! + * + * If the active context ever busy-waited on a semaphore, + * it will be treated as a hog until the end of its timeslice. + * The HW only sends an interrupt on the first miss, and we + * do know if that semaphore has been signaled, or even if it + * is now stuck on another semaphore. Play safe, yield if it + * might be stuck -- it will be given a fresh timeslice in + * the near future. + */ + return upper_32_bits(rq->context->lrc_desc) == READ_ONCE(el->yield); +} + +static bool +timeslice_expired(const struct intel_engine_execlists *el, + const struct i915_request *rq) +{ + return timer_expired(&el->timer) || timeslice_yield(el, rq); +} + static int switch_prio(struct intel_engine_cs *engine, const struct i915_request *rq) { @@ -1759,8 +1785,7 @@ timeslice(const struct intel_engine_cs *engine) return READ_ONCE(engine->props.timeslice_duration_ms); } -static unsigned long -active_timeslice(const struct intel_engine_cs *engine) +static unsigned long active_timeslice(const struct intel_engine_cs *engine) { const struct i915_request *rq = *engine->execlists.active; @@ -1903,13 +1928,14 @@ static void execlists_dequeue(struct intel_engine_cs *engine) last = NULL; } else if (need_timeslice(engine, last) && - timer_expired(&engine->execlists.timer)) { + timeslice_expired(execlists, last)) { ENGINE_TRACE(engine, - "expired last=%llx:%lld, prio=%d, hint=%d\n", + "expired last=%llx:%lld, prio=%d, hint=%d, yield?=%s\n", last->fence.context, last->fence.seqno, last->sched.attr.priority, - execlists->queue_priority_hint); + execlists->queue_priority_hint, + yesno(timeslice_yield(execlists, last))); ring_set_paused(engine, 1); defer_active(engine); @@ -2169,6 +2195,7 @@ static void execlists_dequeue(struct intel_engine_cs *engine) } clear_ports(port + 1, last_port - port); + WRITE_ONCE(execlists->yield, -1); execlists_submit_ports(engine); set_preempt_timeout(engine); } else { @@ -4417,6 +4444,7 @@ logical_ring_default_irqs(struct intel_engine_cs *engine) engine->irq_enable_mask = GT_RENDER_USER_INTERRUPT << shift; engine->irq_keep_mask = GT_CONTEXT_SWITCH_INTERRUPT << shift; engine->irq_keep_mask |= GT_CS_MASTER_ERROR_INTERRUPT << shift; + engine->irq_keep_mask |= GT_WAIT_SEMAPHORE_INTERRUPT << shift; } static void rcs_submission_override(struct intel_engine_cs *engine) diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index f45b5e86ec63..0b5e669c4867 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -3092,6 +3092,7 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg) #define GT_BSD_CS_ERROR_INTERRUPT (1 << 15) #define GT_BSD_USER_INTERRUPT (1 << 12) #define GT_RENDER_L3_PARITY_ERROR_INTERRUPT_S1 (1 << 11) /* hsw+; rsvd on snb, ivb, vlv */ +#define GT_WAIT_SEMAPHORE_INTERRUPT REG_BIT(11) /* bdw+ */ #define GT_CONTEXT_SWITCH_INTERRUPT (1 << 8) #define GT_RENDER_L3_PARITY_ERROR_INTERRUPT (1 << 5) /* !snb */ #define GT_RENDER_PIPECTL_NOTIFY_INTERRUPT (1 << 4) From patchwork Wed Feb 26 09:43:13 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405743 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 22D9B138D for ; Wed, 26 Feb 2020 09:43:50 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 0BA2824679 for ; Wed, 26 Feb 2020 09:43:50 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 0BA2824679 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 176216E40F; Wed, 26 Feb 2020 09:43:49 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id CAA456E405 for ; Wed, 26 Feb 2020 09:43:25 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354712-1500050 for multiple; Wed, 26 Feb 2020 09:43:17 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:13 +0000 Message-Id: <20200226094314.1500667-16-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 16/17] drm/i915/execlists: Check the sentinel is alone in the ELSP X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" We only use sentinel requests for "preempt-to-idle" passes, so assert that they are the only request in a new submission. Signed-off-by: Chris Wilson --- drivers/gpu/drm/i915/gt/intel_lrc.c | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index 253e89fecd95..d907b5ebc1a4 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1448,6 +1448,7 @@ assert_pending_valid(const struct intel_engine_execlists *execlists, { struct i915_request * const *port, *rq; struct intel_context *ce = NULL; + bool sentinel = false; trace_ports(execlists, msg, execlists->pending); @@ -1481,6 +1482,26 @@ assert_pending_valid(const struct intel_engine_execlists *execlists, } ce = rq->context; + /* + * Sentinels are supposed to be lonely so they flush the + * current exection off the HW. Check that they are the + * only request in the pending submission. + */ + if (sentinel) { + GEM_TRACE_ERR("context:%llx after sentinel in pending[%zd]\n", + ce->timeline->fence_context, + port - execlists->pending); + return false; + } + + sentinel = i915_request_has_sentinel(rq); + if (sentinel && port != execlists->pending) { + GEM_TRACE_ERR("sentinel context:%llx not in prime position[%zd]\n", + ce->timeline->fence_context, + port - execlists->pending); + return false; + } + /* Hold tightly onto the lock to prevent concurrent retires! */ if (!spin_trylock_irqsave(&rq->lock, flags)) continue; From patchwork Wed Feb 26 09:43:14 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 11405741 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 2CB2E14D5 for ; Wed, 26 Feb 2020 09:43:49 +0000 (UTC) Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 1546224679 for ; Wed, 26 Feb 2020 09:43:49 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 1546224679 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=chris-wilson.co.uk Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=intel-gfx-bounces@lists.freedesktop.org Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 73C446E408; Wed, 26 Feb 2020 09:43:48 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from fireflyinternet.com (mail.fireflyinternet.com [109.228.58.192]) by gabe.freedesktop.org (Postfix) with ESMTPS id AC8FE6E3EF for ; Wed, 26 Feb 2020 09:43:26 +0000 (UTC) X-Default-Received-SPF: pass (skip=forwardok (res=PASS)) x-ip-name=78.156.65.138; Received: from haswell.alporthouse.com (unverified [78.156.65.138]) by fireflyinternet.com (Firefly Internet (M1)) with ESMTP id 20354713-1500050 for multiple; Wed, 26 Feb 2020 09:43:17 +0000 From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Wed, 26 Feb 2020 09:43:14 +0000 Message-Id: <20200226094314.1500667-17-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.25.1 In-Reply-To: <20200226094314.1500667-1-chris@chris-wilson.co.uk> References: <20200226094314.1500667-1-chris@chris-wilson.co.uk> MIME-Version: 1.0 Subject: [Intel-gfx] [PATCH 17/17] drm/i915/execlists: Reduce preempt-to-busy roundtrip delay X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" To prevent the context from proceeding past the end of the request as we unwind, we embed a semaphore into the footer of each request. (If the context were to skip past the end of the request as we perform the preemption, next time we reload the context it's RING_HEAD would be past the RING_TAIL and instead of replaying the commands it would read the read of the uninitialised ringbuffer.) However, this requires us to keep the ring paused at the end of the request until we have a change to process the preemption ack and remove the semaphore. Our processing of acks is at the whim of ksoftirqd, and so it is entirely possible that the GPU has to wait for the tasklet before it can proceed with the next request. It was suggested that we could also embed a MI_LOAD_REGISTER_MEM into the footer to read the current RING_TAIL from the context, which would allow us to not only avoid this round trip (and so release the context as soon as we had submitted the preemption request to in ELSP), but also skip using ELSP for lite-restores entirely. That has the nice benefit of dramatically reducing contention and the frequency of interrupts when a client submits two or more execbufs in rapid succession. * This did not work out quite as well as anticipated due to us reloading the new RING_TAIL from the context image moments before the HW acted upon the ELSP. With the calamitous effect that we would submit a preemption request with an identical RING_TAIL as the current RING_HEAD, causing us to fail WaIdleLiteRestore and the HW stop working. However, mmio access to RING_TAIL was defeatured in gen11 so we can only employ this handy trick for gen8/gen9. Signed-off-by: Chris Wilson Cc: Tvrtko Ursulin Cc: Daniele Ceraolo Spurio --- drivers/gpu/drm/i915/gt/intel_engine_types.h | 23 +++-- drivers/gpu/drm/i915/gt/intel_lrc.c | 93 +++++++++++++++++++- 2 files changed, 106 insertions(+), 10 deletions(-) diff --git a/drivers/gpu/drm/i915/gt/intel_engine_types.h b/drivers/gpu/drm/i915/gt/intel_engine_types.h index 24cff658e6e5..ae8724915320 100644 --- a/drivers/gpu/drm/i915/gt/intel_engine_types.h +++ b/drivers/gpu/drm/i915/gt/intel_engine_types.h @@ -488,14 +488,15 @@ struct intel_engine_cs { /* status_notifier: list of callbacks for context-switch changes */ struct atomic_notifier_head context_status_notifier; -#define I915_ENGINE_USING_CMD_PARSER BIT(0) -#define I915_ENGINE_SUPPORTS_STATS BIT(1) -#define I915_ENGINE_HAS_PREEMPTION BIT(2) -#define I915_ENGINE_HAS_SEMAPHORES BIT(3) -#define I915_ENGINE_NEEDS_BREADCRUMB_TASKLET BIT(4) -#define I915_ENGINE_IS_VIRTUAL BIT(5) -#define I915_ENGINE_HAS_RELATIVE_MMIO BIT(6) -#define I915_ENGINE_REQUIRES_CMD_PARSER BIT(7) +#define I915_ENGINE_REQUIRES_CMD_PARSER BIT(0) +#define I915_ENGINE_USING_CMD_PARSER BIT(1) +#define I915_ENGINE_SUPPORTS_STATS BIT(2) +#define I915_ENGINE_HAS_PREEMPTION BIT(3) +#define I915_ENGINE_HAS_SEMAPHORES BIT(4) +#define I915_ENGINE_HAS_TAIL_LRM BIT(5) +#define I915_ENGINE_NEEDS_BREADCRUMB_TASKLET BIT(6) +#define I915_ENGINE_IS_VIRTUAL BIT(7) +#define I915_ENGINE_HAS_RELATIVE_MMIO BIT(8) unsigned int flags; /* @@ -592,6 +593,12 @@ intel_engine_has_semaphores(const struct intel_engine_cs *engine) return engine->flags & I915_ENGINE_HAS_SEMAPHORES; } +static inline bool +intel_engine_has_tail_lrm(const struct intel_engine_cs *engine) +{ + return engine->flags & I915_ENGINE_HAS_TAIL_LRM; +} + static inline bool intel_engine_needs_breadcrumb_tasklet(const struct intel_engine_cs *engine) { diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c index d907b5ebc1a4..4452c3d0ac85 100644 --- a/drivers/gpu/drm/i915/gt/intel_lrc.c +++ b/drivers/gpu/drm/i915/gt/intel_lrc.c @@ -1861,6 +1861,76 @@ static inline void clear_ports(struct i915_request **ports, int count) memset_p((void **)ports, NULL, count); } +static struct i915_request * +skip_lite_restore(struct intel_engine_cs *const engine, + struct i915_request *first, + bool *submit) +{ + struct intel_engine_execlists *const execlists = &engine->execlists; + struct i915_request *last = first; + struct rb_node *rb; + + if (!intel_engine_has_tail_lrm(engine)) + return last; + + GEM_BUG_ON(*submit); + while ((rb = rb_first_cached(&execlists->queue))) { + struct i915_priolist *p = to_priolist(rb); + struct i915_request *rq, *rn; + int i; + + priolist_for_each_request_consume(rq, rn, p, i) { + if (!can_merge_rq(last, rq)) + goto out; + + if (__i915_request_submit(rq)) { + *submit = true; + last = rq; + } + } + + rb_erase_cached(&p->node, &execlists->queue); + i915_priolist_free(p); + } +out: + if (*submit) { + ring_set_paused(engine, 1); + + /* + * If we are quick and the current context hasn't yet completed + * its request, we can just tell it to extend the RING_TAIL + * onto the next without having to submit a new ELSP. + */ + if (!i915_request_completed(first)) { + struct i915_request **port; + + ENGINE_TRACE(engine, + "eliding lite-restore last=%llx:%lld->%lld, current %d\n", + first->fence.context, + first->fence.seqno, + last->fence.seqno, + hwsp_seqno(last)); + GEM_BUG_ON(first->context != last->context); + + for (port = (struct i915_request **)execlists->active; + *port != first; + port++) + ; + + GEM_BUG_ON(first == last); + WRITE_ONCE(*port, i915_request_get(last)); + execlists_update_context(last); + + i915_request_put(first); + *submit = false; + } + + ring_set_paused(engine, 0); + } + + return last; +} + static void execlists_dequeue(struct intel_engine_cs *engine) { struct intel_engine_execlists * const execlists = &engine->execlists; @@ -1998,6 +2068,8 @@ static void execlists_dequeue(struct intel_engine_cs *engine) return; } + + last = skip_lite_restore(engine, last, &submit); } } @@ -4224,15 +4296,28 @@ static u32 *emit_preempt_busywait(struct i915_request *request, u32 *cs) return cs; } +static u32 *emit_lrm_tail(struct i915_request *request, u32 *cs) +{ + *cs++ = MI_LOAD_REGISTER_MEM_GEN8 | MI_USE_GGTT; + *cs++ = i915_mmio_reg_offset(RING_TAIL(request->engine->mmio_base)); + *cs++ = i915_ggtt_offset(request->context->state) + + LRC_STATE_PN * PAGE_SIZE + + CTX_RING_TAIL * sizeof(u32); + *cs++ = 0; + + return cs; +} + static __always_inline u32* -gen8_emit_fini_breadcrumb_footer(struct i915_request *request, - u32 *cs) +gen8_emit_fini_breadcrumb_footer(struct i915_request *request, u32 *cs) { *cs++ = MI_USER_INTERRUPT; *cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE; if (intel_engine_has_semaphores(request->engine)) cs = emit_preempt_busywait(request, cs); + if (intel_engine_has_tail_lrm(request->engine)) + cs = emit_lrm_tail(request, cs); request->tail = intel_ring_offset(request, cs); assert_ring_tail_valid(request->ring, request->tail); @@ -4321,6 +4406,8 @@ static u32 *gen12_emit_preempt_busywait(struct i915_request *request, u32 *cs) static __always_inline u32* gen12_emit_fini_breadcrumb_footer(struct i915_request *request, u32 *cs) { + GEM_BUG_ON(intel_engine_has_tail_lrm(request->engine)); + *cs++ = MI_USER_INTERRUPT; *cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE; @@ -4387,6 +4474,8 @@ void intel_execlists_set_default_submission(struct intel_engine_cs *engine) engine->flags |= I915_ENGINE_HAS_SEMAPHORES; if (HAS_LOGICAL_RING_PREEMPTION(engine->i915)) engine->flags |= I915_ENGINE_HAS_PREEMPTION; + if (INTEL_GEN(engine->i915) < 11) + engine->flags |= I915_ENGINE_HAS_TAIL_LRM; } if (INTEL_GEN(engine->i915) >= 12)