Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org 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 smtp.lore.kernel.org (Postfix) with ESMTPS id E22C9EB64D9 for ; Thu, 6 Jul 2023 15:03:21 +0000 (UTC) Received: from gabe.freedesktop.org (localhost [127.0.0.1]) by gabe.freedesktop.org (Postfix) with ESMTP id 9BF6E10E421; Thu, 6 Jul 2023 15:03:21 +0000 (UTC) Received: from mga01.intel.com (mga01.intel.com [192.55.52.88]) by gabe.freedesktop.org (Postfix) with ESMTPS id 96F5210E421 for ; Thu, 6 Jul 2023 15:03:20 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1688655800; x=1720191800; h=date:from:to:cc:subject:message-id:references: content-transfer-encoding:in-reply-to:mime-version; bh=QqvvEPC4K+W5jhsyVzVapPqtHntclNqyWEQP75kxsA8=; b=COYTj/Va2n0wLlekhXWRNM9Je+GqBev4dMAdXTXIIuWk1TDidCvXMe/h QDErfiaKxLNasO7+hLiDRMn9KnaiN6/eV2NRyR+sQylu/JVrd2xqUC5z8 0HkKvpRxJRYmkpagba5FqMi0Bn5VEybgXf8WG8d1zs/mDfo54Zga3UGTj iuMNMQzDZdUQJRTA6eDLwwIgLf85pgcifHHtZaAfiDDTmAjAAzaLn6d1f eIrVpcBp4Uvw6hXW9GBA1IzyThs1PQNK9Htt8Vk8gK1k7WTF38ob4ZV6l X7e0WdFAyjCuTjpHaXRGDerlU6YI/tMdJl049SIPE8hUMJfEf2vi0S357 w==; X-IronPort-AV: E=McAfee;i="6600,9927,10763"; a="394383719" X-IronPort-AV: E=Sophos;i="6.01,185,1684825200"; d="scan'208";a="394383719" Received: from fmsmga002.fm.intel.com ([10.253.24.26]) by fmsmga101.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 06 Jul 2023 08:02:41 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10763"; a="832977329" X-IronPort-AV: E=Sophos;i="6.01,185,1684825200"; d="scan'208";a="832977329" Received: from fmsmsx603.amr.corp.intel.com ([10.18.126.83]) by fmsmga002.fm.intel.com with ESMTP; 06 Jul 2023 08:02:39 -0700 Received: from fmsmsx603.amr.corp.intel.com (10.18.126.83) by fmsmsx603.amr.corp.intel.com (10.18.126.83) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27; Thu, 6 Jul 2023 08:02:39 -0700 Received: from fmsedg602.ED.cps.intel.com (10.1.192.136) by fmsmsx603.amr.corp.intel.com (10.18.126.83) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.2507.27 via Frontend Transport; Thu, 6 Jul 2023 08:02:39 -0700 Received: from NAM10-MW2-obe.outbound.protection.outlook.com (104.47.55.100) by edgegateway.intel.com (192.55.55.71) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.1.2507.27; Thu, 6 Jul 2023 08:02:39 -0700 ARC-Seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=PBzfQY2oYfi67QLH+jpY6V6uq/dlvnrAWc25z+V42nn8fSkTdSXZAhTKlpQzgl/7ZsXUJR0p9tqyDj1OoV4pZYVoiszbiCt4xgy0099/HRfaoG0azkEXcvSxy4SVYytXQ1TdA3KzXpjyRDfmazqVT7ZKEmSdxUltYKOUcZzcUAVv2AjzaQv0LAmA5ItUjEcbVbnljrKLdnr5ymYJNa1LU6fPjivxccF/I0zQ/UHfvoWD0G7RHTS34mP8bDoV7OPXyQZAOd7uRbJnNxO0yQ2SLTMyUPWjyuxOASJ/Q6WUuEPWlLgJx7V4Eh567YRNpa+ZHZAH7qcptaLfNEuzuZcNNg== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-AntiSpam-MessageData-ChunkCount:X-MS-Exchange-AntiSpam-MessageData-0:X-MS-Exchange-AntiSpam-MessageData-1; bh=yjdbEErBIIBGJXrBbPV3JJekdpaltbPuWUEws8S7zqE=; b=fFHk/dwN+2BxifIjPG7zMAFhSUHL1Bxp2G+ynelARlJezpx4UlePWL/5VmwhjPWOzapolyrIDIOKuOXIwp68bwyA9whEEn7UcfAMm0WT8/rm4+GEz+G5knXK9z7LUU11syVmJDHHu5JqE5uI3SQPrHxZ4yPEQ44/+uQa1WQHdXlGx5e1E4Yw8IIXdJ5ZNlOYmOK9r+Oj60qHH0xBJo5sA18le51BKHRIhIwy6jtxfphRyXtqlnwCh+5M3XeP5ZAE9KT4GZ8bJrACRWNn99fNllQI5Kg+PLoZ9FYsTeMmmE7WTutJBMjvwzdVto9uooBxRkZAN58nwh40VEO8RNOa/Q== ARC-Authentication-Results: i=1; mx.microsoft.com 1; spf=pass smtp.mailfrom=intel.com; dmarc=pass action=none header.from=intel.com; dkim=pass header.d=intel.com; arc=none Authentication-Results: dkim=none (message not signed) header.d=none;dmarc=none action=none header.from=intel.com; Received: from PH7PR11MB6522.namprd11.prod.outlook.com (2603:10b6:510:212::12) by SA0PR11MB4766.namprd11.prod.outlook.com (2603:10b6:806:92::20) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.20.6565.17; Thu, 6 Jul 2023 15:02:37 +0000 Received: from PH7PR11MB6522.namprd11.prod.outlook.com ([fe80::c65d:c846:f197:3ca5]) by PH7PR11MB6522.namprd11.prod.outlook.com ([fe80::c65d:c846:f197:3ca5%4]) with mapi id 15.20.6565.019; Thu, 6 Jul 2023 15:02:37 +0000 Date: Thu, 6 Jul 2023 15:01:50 +0000 From: Matthew Brost To: Thomas =?iso-8859-1?Q?Hellstr=F6m?= Message-ID: References: <20230630175804.1096370-1-matthew.brost@intel.com> <20230630175804.1096370-2-matthew.brost@intel.com> <6f79cea3-2c88-868b-3869-3014656cd5ad@linux.intel.com> Content-Type: text/plain; charset="iso-8859-1" Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <6f79cea3-2c88-868b-3869-3014656cd5ad@linux.intel.com> X-ClientProxiedBy: BYAPR11CA0073.namprd11.prod.outlook.com (2603:10b6:a03:f4::14) To PH7PR11MB6522.namprd11.prod.outlook.com (2603:10b6:510:212::12) MIME-Version: 1.0 X-MS-PublicTrafficType: Email X-MS-TrafficTypeDiagnostic: PH7PR11MB6522:EE_|SA0PR11MB4766:EE_ X-MS-Office365-Filtering-Correlation-Id: 10ba17a6-24ba-4c09-c110-08db7e3208f3 X-MS-Exchange-SenderADCheck: 1 X-MS-Exchange-AntiSpam-Relay: 0 X-Microsoft-Antispam: BCL:0; X-Microsoft-Antispam-Message-Info: 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 X-Forefront-Antispam-Report: CIP:255.255.255.255; CTRY:; LANG:en; SCL:1; SRV:; IPV:NLI; SFV:NSPM; H:PH7PR11MB6522.namprd11.prod.outlook.com; PTR:; CAT:NONE; SFS:(13230028)(376002)(366004)(346002)(136003)(396003)(39860400002)(451199021)(6486002)(478600001)(6666004)(186003)(26005)(66946007)(53546011)(6512007)(2906002)(30864003)(66556008)(41300700001)(5660300002)(316002)(66476007)(6916009)(4326008)(44832011)(8936002)(8676002)(38100700002)(83380400001)(82960400001)(6506007)(66574015)(86362001); DIR:OUT; SFP:1102; X-MS-Exchange-AntiSpam-MessageData-ChunkCount: 1 X-MS-Exchange-AntiSpam-MessageData-0: =?iso-8859-1?Q?tOX5shF1ju2cAxMlNx5E3ALXMGlaZkGEiarrN+x8D4CE/H7OVSb+wjhzbp?= =?iso-8859-1?Q?22H8yAluMOOnWWomiN4xfpyggN+I71dnZfxY/TuKh6TXx1ijVL0hABsq2p?= =?iso-8859-1?Q?M6ui/s3wnmWGrhsLiXzS+SEvdbtTN/aFqilxid8/B/jmv/uyw8XWI3F1+p?= =?iso-8859-1?Q?EZhkIZCz0Q6lV9N9BMvWrQrDHw9Cda8ma7hTHwVCsOryLo/h6QGKJkcRJK?= =?iso-8859-1?Q?HFCgLP3NDU5CI97S+AbqNo0GED4uUnPSwECN09GN8nS1e5CqQAKk4t68rj?= =?iso-8859-1?Q?h/MzB+nMBEyntZgqWeATEOwoKaCOl+Y21IjFYM+DxmaYlYq6x1Ag1imid4?= =?iso-8859-1?Q?+78ZW5JLwbGfsD+jc3RRnBQ2N48u0UgxtLoVFa403OkG1OM1hFfvYbmCwa?= =?iso-8859-1?Q?DQWixfWzN0VNPLpR3CR1jc8p8HjUPv6jUmNLUONDVoCrR5E29iOCeC9bW6?= =?iso-8859-1?Q?mSNjTff8Yw9Oj240CUNLnfbvgY93qjWGky355OdzS5YBEEI3MIbMwHXYZG?= =?iso-8859-1?Q?DpALpqnQvsieCge5/8KA4u5xBLvgNMiOA2DBAJAoNde9AopBAS/ES9SQj2?= =?iso-8859-1?Q?J6OyyOTd0ExLdkk0Wy0M1VfUVE7LTL6Qr+Ysmt2cGn43orG9z0Y1nj7jo3?= =?iso-8859-1?Q?tpvr2UoTVBiUwHCw9XeBLJKLfh3H6DaGaeGU67egbLi/zdsaoj6Y1JrKgK?= =?iso-8859-1?Q?7L0Gz2ZwF9ofkRE702GFsowb67zUHBnAw06vnxJxGDmqgHSDaVI5vhQQci?= =?iso-8859-1?Q?5nRc4jJFE5JXz9ttIKvEaZLnb4lQnwSz2iZdfjLRtDCPhSeme+nVBrj3gZ?= =?iso-8859-1?Q?in1AqFQwlp2QFxpL9/YRHSd8nceALCkbRW6DbWMCyvW97Q+g/msS7HIy0T?= =?iso-8859-1?Q?0V9AwZNXcVFuqtSZIajK+Uzgmb3bRtDECnTh/Ve5w6eU4L2ih8NjDmEBPm?= =?iso-8859-1?Q?omQqSDwke1ZfX3yL7irrFAeVQoXgn+moYsWmBd286mmyB/JBr6IHsPjZmv?= =?iso-8859-1?Q?CG7IB4RnDrog28byMpTTD62pr1oL6mvpIGOgoxEZyXnMFNAGno0aUtQpZg?= =?iso-8859-1?Q?550DhUBM91V+4jQwvVTMr4nIzic3O+t3jpQHl6wm6nW59VwGMvLHPx1L9l?= =?iso-8859-1?Q?Fm/QYDd4ssBa6SYiNXN2WaXJ7jdFJtViLZhYA6s6/xQXpwtFpKXkV+2NGL?= =?iso-8859-1?Q?OA3fNtACyDWtbjIPW42CSUVP9mn93TnjFa5ezhcYmkvIm6GWL6ivxh/l5g?= =?iso-8859-1?Q?BEsd069Nj5KCLhbvlNnJ9uMZSxGC5DqhuAbPX2nMbBeSVgS739yf4swnQB?= =?iso-8859-1?Q?Cjb9J6+pHFmBnK+kC8TerAFzm+US+AFx4q4rLawJpgqQV7v/PfqnDAG5Ee?= =?iso-8859-1?Q?8UFAngRh3zpP5qNkOptCtHDORc/8YDDFcTUsbgPhEvxrbbNqJ4Q7TchbAr?= =?iso-8859-1?Q?HLllIsf7dp6btQEwiJsyzsfpHY8DQjtAp/EcHOoJgCYup0qKk3MXDWTqQD?= =?iso-8859-1?Q?O94tI0xjtdm1SnllNo4xvo9OXeoLElWAfgvTm/ItldG6XjS8BHnHCQPAm5?= =?iso-8859-1?Q?bmn2o2LeHqV7k6Nmd3/xkPzhjlFwJ+QXbq7sTwEkf5BflqVJteciNjSElE?= =?iso-8859-1?Q?rhN/SZcNOl4sahkBAyHNxeFXzjXNJxuCVra67oOYVVPagIjXd0mll7vw?= =?iso-8859-1?Q?=3D=3D?= X-MS-Exchange-CrossTenant-Network-Message-Id: 10ba17a6-24ba-4c09-c110-08db7e3208f3 X-MS-Exchange-CrossTenant-AuthSource: PH7PR11MB6522.namprd11.prod.outlook.com X-MS-Exchange-CrossTenant-AuthAs: Internal X-MS-Exchange-CrossTenant-OriginalArrivalTime: 06 Jul 2023 15:02:36.8325 (UTC) X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted X-MS-Exchange-CrossTenant-Id: 46c98d88-e344-4ed4-8496-4ed7712e255d X-MS-Exchange-CrossTenant-MailboxType: HOSTED X-MS-Exchange-CrossTenant-UserPrincipalName: Sv7vKdxXVFIXOy931WhgRJXNxkpdAFlqsZ+e+pL4P4yDGDdAkQGBBezkMYrjaeRapCTuBTe9hjobzOqS2SuSWQ== X-MS-Exchange-Transport-CrossTenantHeadersStamped: SA0PR11MB4766 X-OriginatorOrg: intel.com Subject: Re: [Intel-xe] [PATCH v4 1/9] drm/xe: Ban a VM if rebind worker hits an error X-BeenThere: intel-xe@lists.freedesktop.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Intel Xe graphics driver List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: intel-xe@lists.freedesktop.org Errors-To: intel-xe-bounces@lists.freedesktop.org Sender: "Intel-xe" On Thu, Jul 06, 2023 at 12:31:36PM +0200, Thomas Hellström wrote: > Hi, Matthew. > > On 6/30/23 19:57, Matthew Brost wrote: > > We cannot recover a VM if a rebind worker hits an error, ban the VM if > > happens to ensure we do not attempt to place this VM on the hardware > > again. > > > > A follow up will inform the user if this happens. > > > > v2: Return -ECANCELED in exec VM closed or banned, check for closed or > > banned within VM lock. > > v3: Fix lockdep splat by looking engine outside of vm->lock > > v4: Fix error path when engine lookup fails > > > > Signed-off-by: Matthew Brost > > --- > > drivers/gpu/drm/xe/xe_engine.c | 13 +++++ > > drivers/gpu/drm/xe/xe_exec.c | 6 +- > > drivers/gpu/drm/xe/xe_trace.h | 5 ++ > > drivers/gpu/drm/xe/xe_vm.c | 92 ++++++++++++++++++------------ > > drivers/gpu/drm/xe/xe_vm.h | 11 ++++ > > drivers/gpu/drm/xe/xe_vm_madvise.c | 2 +- > > drivers/gpu/drm/xe/xe_vm_types.h | 5 +- > > 7 files changed, 92 insertions(+), 42 deletions(-) > > > > diff --git a/drivers/gpu/drm/xe/xe_engine.c b/drivers/gpu/drm/xe/xe_engine.c > > index 6e6b2913f766..ada2986c33a2 100644 > > --- a/drivers/gpu/drm/xe/xe_engine.c > > +++ b/drivers/gpu/drm/xe/xe_engine.c > > @@ -597,10 +597,23 @@ int xe_engine_create_ioctl(struct drm_device *dev, void *data, > > if (XE_IOCTL_ERR(xe, !vm)) > > return -ENOENT; > > + err = down_read_interruptible(&vm->lock); > > + if (err) { > > + xe_vm_put(vm); > > + return err; > > + } > > + > > + if (XE_IOCTL_ERR(xe, xe_vm_is_closed_or_banned(vm))) { > > xe_vm_is_closed requires the vm->resv according to comments in the function. > Not sure if that is actually true anylonger but if not, the comment needs an > update, and the function needs an assert. That is wrong, it should be vm->lock but xe_vm_close_and_put needs updating to set closed. Will fix comment and xe_vm_close_and_put. > > Also see below comment on flags usage. > > > + up_read(&vm->lock); > > + xe_vm_put(vm); > > + return -ENOENT; > > + } > > + > > e = xe_engine_create(xe, vm, logical_mask, > > args->width, hwe, > > xe_vm_no_dma_fences(vm) ? 0 : > > ENGINE_FLAG_PERSISTENT); > > + up_read(&vm->lock); > > xe_vm_put(vm); > > if (IS_ERR(e)) > > return PTR_ERR(e); > > diff --git a/drivers/gpu/drm/xe/xe_exec.c b/drivers/gpu/drm/xe/xe_exec.c > > index c52edff9a358..bdf00e59e7a4 100644 > > --- a/drivers/gpu/drm/xe/xe_exec.c > > +++ b/drivers/gpu/drm/xe/xe_exec.c > > @@ -297,9 +297,9 @@ int xe_exec_ioctl(struct drm_device *dev, void *data, struct drm_file *file) > > if (err) > > goto err_unlock_list; > > - if (xe_vm_is_closed(engine->vm)) { > > - drm_warn(&xe->drm, "Trying to schedule after vm is closed\n"); > > - err = -EIO; > > + if (xe_vm_is_closed_or_banned(engine->vm)) { > > + drm_warn(&xe->drm, "Trying to schedule after vm is closed or banned\n"); > > + err = -ECANCELED; > > goto err_engine_end; > > } > > diff --git a/drivers/gpu/drm/xe/xe_trace.h b/drivers/gpu/drm/xe/xe_trace.h > > index 02861c26e145..ca96a0a4bb80 100644 > > --- a/drivers/gpu/drm/xe/xe_trace.h > > +++ b/drivers/gpu/drm/xe/xe_trace.h > > @@ -477,6 +477,11 @@ DECLARE_EVENT_CLASS(xe_vm, > > __entry->asid) > > ); > > +DEFINE_EVENT(xe_vm, xe_vm_kill, > > + TP_PROTO(struct xe_vm *vm), > > + TP_ARGS(vm) > > +); > > + > > DEFINE_EVENT(xe_vm, xe_vm_create, > > TP_PROTO(struct xe_vm *vm), > > TP_ARGS(vm) > > diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c > > index 3bba957e3b89..3c8a48f9a0d3 100644 > > --- a/drivers/gpu/drm/xe/xe_vm.c > > +++ b/drivers/gpu/drm/xe/xe_vm.c > > @@ -514,6 +514,24 @@ void xe_vm_unlock_dma_resv(struct xe_vm *vm, > > #define XE_VM_REBIND_RETRY_TIMEOUT_MS 1000 > > +static void xe_vm_kill(struct xe_vm *vm) > > +{ > > + struct ww_acquire_ctx ww; > > + struct xe_engine *e; > > + > > + lockdep_assert_held(&vm->lock); > > + > > + xe_vm_lock(vm, &ww, 0, false); > > + vm->flags |= XE_VM_FLAG_BANNED; > > Is the vm->flags member always protected by the vm lock? If not, IIRC we > can't access individual bits without using set_bit() and frientds bit-ops. > The vm->flags kerneldoc says all bits are "statically set up a creation > time". All bits aside from XE_VM_FLAG_BANNED can static from creation time, after creation time all changeds should be protected by the lock. Right now that is only XE_VM_FLAG_BANNED. I can update the comment. > > > + trace_xe_vm_kill(vm); > > + > > + list_for_each_entry(e, &vm->preempt.engines, compute.link) > > + e->ops->kill(e); > > + xe_vm_unlock(vm, &ww); > > + > > + /* TODO: Inform user the VM is banned */ > > +} > > + > > static void preempt_rebind_work_func(struct work_struct *w) > > { > > struct xe_vm *vm = container_of(w, struct xe_vm, preempt.rebind_work); > > @@ -533,13 +551,14 @@ static void preempt_rebind_work_func(struct work_struct *w) > > XE_BUG_ON(!xe_vm_in_compute_mode(vm)); > > trace_xe_vm_rebind_worker_enter(vm); > > - if (xe_vm_is_closed(vm)) { > > + down_write(&vm->lock); > > + > > + if (xe_vm_is_closed_or_banned(vm)) { > > + up_write(&vm->lock); > > trace_xe_vm_rebind_worker_exit(vm); > > return; > > } > > - down_write(&vm->lock); > > - > > retry: > > if (vm->async_ops.error) > > goto out_unlock_outer; > > @@ -666,11 +685,12 @@ static void preempt_rebind_work_func(struct work_struct *w) > > goto retry; > > } > > } > > + if (err) > > + xe_vm_kill(vm); > > up_write(&vm->lock); > > free_preempt_fences(&preempt_fences); > > - XE_WARN_ON(err < 0); /* TODO: Kill VM or put in error state */ > > I've often find it useful to have a debug printout here, what the error code > is. Can we use > > drm_debug("VM worker error: %pe\n", ERR_PTR(err)); > > without risking spamming the logs? > Sure. > > trace_xe_vm_rebind_worker_exit(vm); > > } > > @@ -1140,11 +1160,12 @@ xe_vm_find_overlapping_vma(struct xe_vm *vm, const struct xe_vma *vma) > > { > > struct rb_node *node; > > - if (xe_vm_is_closed(vm)) > > + lockdep_assert_held(&vm->lock); > > + > > + if (xe_vm_is_closed_or_banned(vm)) > > return NULL; > > XE_BUG_ON(vma->end >= vm->size); > > - lockdep_assert_held(&vm->lock); > > node = rb_find(vma, &vm->vmas, xe_vma_cmp_vma_cb); > > @@ -3074,30 +3095,35 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) > > if (err) > > return err; > > - vm = xe_vm_lookup(xef, args->vm_id); > > - if (XE_IOCTL_ERR(xe, !vm)) { > > - err = -EINVAL; > > - goto free_objs; > > - } > > - > > - if (XE_IOCTL_ERR(xe, xe_vm_is_closed(vm))) { > > - drm_err(dev, "VM closed while we began looking up?\n"); > > - err = -ENOENT; > > - goto put_vm; > > - } > > - > > if (args->engine_id) { > > e = xe_engine_lookup(xef, args->engine_id); > > if (XE_IOCTL_ERR(xe, !e)) { > > err = -ENOENT; > > - goto put_vm; > > + goto free_objs; > > } > > + > > if (XE_IOCTL_ERR(xe, !(e->flags & ENGINE_FLAG_VM))) { > > err = -EINVAL; > > goto put_engine; > > } > > } > > + vm = xe_vm_lookup(xef, args->vm_id); > > + if (XE_IOCTL_ERR(xe, !vm)) { > > + err = -EINVAL; > > + goto put_engine; > > + } > > + > > + err = down_write_killable(&vm->lock); > > + if (err) > > + goto put_vm; > > + > > + if (XE_IOCTL_ERR(xe, xe_vm_is_closed_or_banned(vm))) { > > + drm_err(dev, "VM closed while we began looking up?\n"); > > Since this path is triggerable from user-space it would be easy for a > malicious client to spam the logs. > Do you suggest I remove this or change the level? I'd lean towards remove as XE_IOCTL_ERR gives us the information we want on a debug build of the kernel. Matt > /Thomas > > > > + err = -ENOENT; > > + goto release_vm_lock; > > + } > > + > > if (VM_BIND_OP(bind_ops[0].op) == XE_VM_BIND_OP_RESTART) { > > if (XE_IOCTL_ERR(xe, !(vm->flags & XE_VM_FLAG_ASYNC_BIND_OPS))) > > err = -EOPNOTSUPP; > > @@ -3107,10 +3133,8 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) > > err = -EPROTO; > > if (!err) { > > - down_write(&vm->lock); > > trace_xe_vm_restart(vm); > > vm_set_async_error(vm, 0); > > - up_write(&vm->lock); > > queue_work(system_unbound_wq, &vm->async_ops.work); > > @@ -3119,13 +3143,13 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) > > xe_vm_queue_rebind_worker(vm); > > } > > - goto put_engine; > > + goto release_vm_lock; > > } > > if (XE_IOCTL_ERR(xe, !vm->async_ops.error && > > async != !!(vm->flags & XE_VM_FLAG_ASYNC_BIND_OPS))) { > > err = -EOPNOTSUPP; > > - goto put_engine; > > + goto release_vm_lock; > > } > > for (i = 0; i < args->num_binds; ++i) { > > @@ -3135,7 +3159,7 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) > > if (XE_IOCTL_ERR(xe, range > vm->size) || > > XE_IOCTL_ERR(xe, addr > vm->size - range)) { > > err = -EINVAL; > > - goto put_engine; > > + goto release_vm_lock; > > } > > if (bind_ops[i].tile_mask) { > > @@ -3144,7 +3168,7 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) > > if (XE_IOCTL_ERR(xe, bind_ops[i].tile_mask & > > ~valid_tiles)) { > > err = -EINVAL; > > - goto put_engine; > > + goto release_vm_lock; > > } > > } > > } > > @@ -3152,13 +3176,13 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) > > bos = kzalloc(sizeof(*bos) * args->num_binds, GFP_KERNEL); > > if (!bos) { > > err = -ENOMEM; > > - goto put_engine; > > + goto release_vm_lock; > > } > > vmas = kzalloc(sizeof(*vmas) * args->num_binds, GFP_KERNEL); > > if (!vmas) { > > err = -ENOMEM; > > - goto put_engine; > > + goto release_vm_lock; > > } > > for (i = 0; i < args->num_binds; ++i) { > > @@ -3213,10 +3237,6 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) > > goto free_syncs; > > } > > - err = down_write_killable(&vm->lock); > > - if (err) > > - goto free_syncs; > > - > > /* Do some error checking first to make the unwind easier */ > > for (i = 0; i < args->num_binds; ++i) { > > u64 range = bind_ops[i].range; > > @@ -3225,7 +3245,7 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) > > err = __vm_bind_ioctl_lookup_vma(vm, bos[i], addr, range, op); > > if (err) > > - goto release_vm_lock; > > + goto free_syncs; > > } > > for (i = 0; i < args->num_binds; ++i) { > > @@ -3345,8 +3365,6 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) > > break; > > } > > } > > -release_vm_lock: > > - up_write(&vm->lock); > > free_syncs: > > while (num_syncs--) { > > if (async && j && > > @@ -3359,11 +3377,13 @@ int xe_vm_bind_ioctl(struct drm_device *dev, void *data, struct drm_file *file) > > put_obj: > > for (i = j; i < args->num_binds; ++i) > > xe_bo_put(bos[i]); > > +release_vm_lock: > > + up_write(&vm->lock); > > +put_vm: > > + xe_vm_put(vm); > > put_engine: > > if (e) > > xe_engine_put(e); > > -put_vm: > > - xe_vm_put(vm); > > free_objs: > > kfree(bos); > > kfree(vmas); > > diff --git a/drivers/gpu/drm/xe/xe_vm.h b/drivers/gpu/drm/xe/xe_vm.h > > index 5edb7771629c..47bc7fbb2f50 100644 > > --- a/drivers/gpu/drm/xe/xe_vm.h > > +++ b/drivers/gpu/drm/xe/xe_vm.h > > @@ -49,6 +49,17 @@ static inline bool xe_vm_is_closed(struct xe_vm *vm) > > return !vm->size; > > } > > +static inline bool xe_vm_is_banned(struct xe_vm *vm) > > +{ > > + return vm->flags & XE_VM_FLAG_BANNED; > > +} > > + > > +static inline bool xe_vm_is_closed_or_banned(struct xe_vm *vm) > > +{ > > + lockdep_assert_held(&vm->lock); > > + return xe_vm_is_closed(vm) || xe_vm_is_banned(vm); > > +} > > + > > struct xe_vma * > > xe_vm_find_overlapping_vma(struct xe_vm *vm, const struct xe_vma *vma); > > diff --git a/drivers/gpu/drm/xe/xe_vm_madvise.c b/drivers/gpu/drm/xe/xe_vm_madvise.c > > index 0f5eef337037..76458f8d57f3 100644 > > --- a/drivers/gpu/drm/xe/xe_vm_madvise.c > > +++ b/drivers/gpu/drm/xe/xe_vm_madvise.c > > @@ -313,7 +313,7 @@ int xe_vm_madvise_ioctl(struct drm_device *dev, void *data, > > if (XE_IOCTL_ERR(xe, !vm)) > > return -EINVAL; > > - if (XE_IOCTL_ERR(xe, xe_vm_is_closed(vm))) { > > + if (XE_IOCTL_ERR(xe, xe_vm_is_closed_or_banned(vm))) { > > err = -ENOENT; > > goto put_vm; > > } > > diff --git a/drivers/gpu/drm/xe/xe_vm_types.h b/drivers/gpu/drm/xe/xe_vm_types.h > > index c148dd49a6ca..286de52160b9 100644 > > --- a/drivers/gpu/drm/xe/xe_vm_types.h > > +++ b/drivers/gpu/drm/xe/xe_vm_types.h > > @@ -183,8 +183,9 @@ struct xe_vm { > > #define XE_VM_FLAG_MIGRATION BIT(3) > > #define XE_VM_FLAG_SCRATCH_PAGE BIT(4) > > #define XE_VM_FLAG_FAULT_MODE BIT(5) > > -#define XE_VM_FLAG_GT_ID(flags) (((flags) >> 6) & 0x3) > > -#define XE_VM_FLAG_SET_TILE_ID(tile) ((tile)->id << 6) > > +#define XE_VM_FLAG_BANNED BIT(6) > > +#define XE_VM_FLAG_GT_ID(flags) (((flags) >> 7) & 0x3) > > +#define XE_VM_FLAG_SET_TILE_ID(tile) ((tile)->id << 7) > > unsigned long flags; > > /** @composite_fence_ctx: context composite fence */