From patchwork Thu Feb 2 09:08:39 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Chris Wilson X-Patchwork-Id: 9551375 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id CD21160236 for ; Thu, 2 Feb 2017 09:10:04 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id BA3FA28411 for ; Thu, 2 Feb 2017 09:10:04 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id AE2B528420; Thu, 2 Feb 2017 09:10:04 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-4.1 required=2.0 tests=BAYES_00,DKIM_SIGNED, RCVD_IN_DNSWL_MED,T_DKIM_INVALID autolearn=ham version=3.3.1 Received: from gabe.freedesktop.org (gabe.freedesktop.org [131.252.210.177]) (using TLSv1.2 with cipher DHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.wl.linuxfoundation.org (Postfix) with ESMTPS id 1CB1A28411 for ; Thu, 2 Feb 2017 09:10:04 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 498246E984; Thu, 2 Feb 2017 09:09:33 +0000 (UTC) X-Original-To: intel-gfx@lists.freedesktop.org Delivered-To: intel-gfx@lists.freedesktop.org Received: from mail-wm0-x241.google.com (mail-wm0-x241.google.com [IPv6:2a00:1450:400c:c09::241]) by gabe.freedesktop.org (Postfix) with ESMTPS id B9A266E978 for ; Thu, 2 Feb 2017 09:09:30 +0000 (UTC) Received: by mail-wm0-x241.google.com with SMTP id u63so2854620wmu.2 for ; Thu, 02 Feb 2017 01:09:30 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=sender:from:to:cc:subject:date:message-id:in-reply-to:references; bh=/GWRBgi/phOspZMpAEjuGdtFO1rDqTxd+0uat1Yi/Ls=; b=tORZBduBqs/p7V25x2tFk85mVoX1/1gVeLJjRsIW/YSin5Itrwjyx4YW0IqtBTqJRU jH66uRmYjjVb9kUm/y41lJoHzaeLYY4JmcHzRKEGWT0hwmy18DMJjVdq4sA7gRPlCdmF /N+i/sqgPmLn0vhcjzBkWXOarDB06bvJiEQSf6iJuRqIWNd2ZWGPpVAMbSPko3Or8fQi BjGjTWapBUoWZgk4HOypkl3cvdXl2tc4Xrg9W1QzEyzWjErS53ql6aQ7baA/XFqwD9HF uK5ZInbsoDyOv1UVdOrQYttidmIl6DciIAWcSOhGtajYMgrP1D6GDFCL6w7/FiJeYhA2 mDUQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:sender:from:to:cc:subject:date:message-id :in-reply-to:references; bh=/GWRBgi/phOspZMpAEjuGdtFO1rDqTxd+0uat1Yi/Ls=; b=YTLGakwmJm8Rlo/+9VZW+kDxZEDggom+LJFqWAKtivuq8Pvz6ZgRbdye6869TPaQUb XpM0fNQR30vFQIVqhEG30JjL7nNK2nmoFRt45FhULQmuAu/j9QApdTZEBu376S/VCOWe nK703hPkgrELLtRjIagHBycsXy5GtsGXUYXITjvYm1vN619NkoaLwTSJjyQZ7mY+zz/a EpLoPJeMCL0Hrn3S4JJ+ndyzYkFStwieocRskSaFNbx5WkmFTYT78+TPPZjAc2AyaBXN uTMrK7uiuAS8PLZw5mt3ed7YJvKdnPjUVUznHCQXDBZEN2+y8D4Gu3AZ4SBMPFx1CpvX jwzA== X-Gm-Message-State: AIkVDXLUd8VOt1oGIHmBnvPleXyOFjufkubEGnrhsKiya2u7zqGhRkGldW6U5fbD/gV2Nw== X-Received: by 10.223.134.68 with SMTP id 4mr7563152wrw.49.1486026569279; Thu, 02 Feb 2017 01:09:29 -0800 (PST) Received: from haswell.alporthouse.com ([78.156.65.138]) by smtp.gmail.com with ESMTPSA id s17sm38675584wrc.6.2017.02.02.01.09.28 (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Thu, 02 Feb 2017 01:09:28 -0800 (PST) From: Chris Wilson To: intel-gfx@lists.freedesktop.org Date: Thu, 2 Feb 2017 09:08:39 +0000 Message-Id: <20170202090905.29028-21-chris@chris-wilson.co.uk> X-Mailer: git-send-email 2.11.0 In-Reply-To: <20170202090905.29028-1-chris@chris-wilson.co.uk> References: <20170202090905.29028-1-chris@chris-wilson.co.uk> Subject: [Intel-gfx] [PATCH 20/46] drm/i915: Live testing of empty requests X-BeenThere: intel-gfx@lists.freedesktop.org X-Mailman-Version: 2.1.18 Precedence: list List-Id: Intel graphics driver community testing & development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" X-Virus-Scanned: ClamAV using ClamSMTP Primarily to emphasize the difference between just advancing the breadcrumb using a bare request and the overhead of dispatching an execbuffer. Signed-off-by: Chris Wilson Reviewed-by: Joonas Lahtinen --- drivers/gpu/drm/i915/selftests/i915_gem_request.c | 155 ++++++++++++++++++++++ 1 file changed, 155 insertions(+) diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_request.c b/drivers/gpu/drm/i915/selftests/i915_gem_request.c index f9c171d1a05b..92fa55bd68c8 100644 --- a/drivers/gpu/drm/i915/selftests/i915_gem_request.c +++ b/drivers/gpu/drm/i915/selftests/i915_gem_request.c @@ -301,6 +301,160 @@ static int live_nop_request(void *arg) return err; } +static struct i915_vma *empty_batch(struct drm_i915_private *i915) +{ + struct drm_i915_gem_object *obj; + struct i915_vma *vma; + u32 *cmd; + int err; + + obj = i915_gem_object_create_internal(i915, PAGE_SIZE); + if (IS_ERR(obj)) + return ERR_CAST(obj); + + cmd = i915_gem_object_pin_map(obj, I915_MAP_WB); + if (IS_ERR(cmd)) { + err = PTR_ERR(cmd); + goto err; + } + *cmd = MI_BATCH_BUFFER_END; + i915_gem_object_unpin_map(obj); + + err = i915_gem_object_set_to_gtt_domain(obj, false); + if (err) + goto err; + + vma = i915_vma_instance(obj, &i915->ggtt.base, NULL); + if (IS_ERR(vma)) { + err = PTR_ERR(vma); + goto err; + } + + err = i915_vma_pin(vma, 0, 0, PIN_USER | PIN_GLOBAL); + if (err) + goto err; + + return vma; + +err: + i915_gem_object_put(obj); + return ERR_PTR(err); +} + +static struct drm_i915_gem_request * +empty_request(struct intel_engine_cs *engine, + struct i915_vma *batch) +{ + struct drm_i915_gem_request *request; + int err; + + request = i915_gem_request_alloc(engine, + engine->i915->kernel_context); + if (IS_ERR(request)) + return request; + + err = engine->emit_flush(request, EMIT_INVALIDATE); + if (err) + goto out_request; + + err = i915_switch_context(request); + if (err) + goto out_request; + + err = engine->emit_bb_start(request, + batch->node.start, + batch->node.size, + I915_DISPATCH_SECURE); + if (err) + goto out_request; + +out_request: + __i915_add_request(request, err == 0); + return err ? ERR_PTR(err) : request; +} + +static int live_empty_request(void *arg) +{ + struct drm_i915_private *i915 = arg; + struct intel_engine_cs *engine; + struct live_test t; + struct i915_vma *batch; + unsigned int id; + int err = 0; + + /* Submit various sized batches of empty requests, to each engine + * (individually), and wait for the batch to complete. We can check + * the overhead of submitting requests to the hardware. + */ + + mutex_lock(&i915->drm.struct_mutex); + + batch = empty_batch(i915); + if (IS_ERR(batch)) { + err = PTR_ERR(batch); + goto out_unlock; + } + + for_each_engine(engine, i915, id) { + IGT_TIMEOUT(end_time); + struct drm_i915_gem_request *request; + unsigned long n, prime; + ktime_t times[2] = {}; + + err = begin_live_test(&t, i915, __func__, engine->name); + if (err) + goto out_batch; + + /* Warmup / preload */ + request = empty_request(engine, batch); + if (IS_ERR(request)) { + err = PTR_ERR(request); + goto out_batch; + } + i915_wait_request(request, + I915_WAIT_LOCKED, + MAX_SCHEDULE_TIMEOUT); + + for_each_prime_number_from(prime, 1, 8192) { + times[1] = ktime_get_raw(); + + for (n = 0; n < prime; n++) { + request = empty_request(engine, batch); + if (IS_ERR(request)) { + err = PTR_ERR(request); + goto out_batch; + } + } + i915_wait_request(request, + I915_WAIT_LOCKED, + MAX_SCHEDULE_TIMEOUT); + + times[1] = ktime_sub(ktime_get_raw(), times[1]); + if (prime == 1) + times[0] = times[1]; + + if (__igt_timeout(end_time, NULL)) + break; + } + + err = end_live_test(&t); + if (err) + goto out_batch; + + pr_info("Batch latencies on %s: 1 = %lluns, %lu = %lluns\n", + engine->name, + ktime_to_ns(times[0]), + prime, div64_u64(ktime_to_ns(times[1]), prime)); + } + +out_batch: + i915_vma_unpin(batch); + i915_vma_put(batch); +out_unlock: + mutex_unlock(&i915->drm.struct_mutex); + return err; +} + static struct i915_vma *recursive_batch(struct drm_i915_private *i915) { struct i915_gem_context *ctx = i915->kernel_context; @@ -615,6 +769,7 @@ int i915_gem_request_live_selftests(struct drm_i915_private *i915) SUBTEST(live_nop_request), SUBTEST(live_all_engines), SUBTEST(live_sequential_engines), + SUBTEST(live_empty_request), }; return i915_subtests(tests, i915); }