LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Lu Baolu <baolu.lu@linux.intel.com>
To: Jacob Pan <jacob.jun.pan@linux.intel.com>,
	iommu@lists.linux-foundation.org,
	LKML <linux-kernel@vger.kernel.org>,
	Joerg Roedel <joro@8bytes.org>,
	David Woodhouse <dwmw2@infradead.org>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Alex Williamson <alex.williamson@redhat.com>,
	Jean-Philippe Brucker <jean-philippe.brucker@arm.com>
Cc: Rafael Wysocki <rafael.j.wysocki@intel.com>,
	"Liu, Yi L" <yi.l.liu@intel.com>,
	"Tian, Kevin" <kevin.tian@intel.com>,
	Raj Ashok <ashok.raj@intel.com>,
	Jean Delvare <khali@linux-fr.org>,
	Christoph Hellwig <hch@infradead.org>
Subject: Re: [PATCH v5 15/23] iommu: handle page response timeout
Date: Mon, 14 May 2018 15:43:54 +0800	[thread overview]
Message-ID: <5AF93E3A.2040902@linux.intel.com> (raw)
In-Reply-To: <1526072055-86990-16-git-send-email-jacob.jun.pan@linux.intel.com>

Hi,

On 05/12/2018 04:54 AM, Jacob Pan wrote:
> When IO page faults are reported outside IOMMU subsystem, the page
> request handler may fail for various reasons. E.g. a guest received
> page requests but did not have a chance to run for a long time. The
> irresponsive behavior could hold off limited resources on the pending
> device.
> There can be hardware or credit based software solutions as suggested
> in the PCI ATS Ch-4. To provide a basic safty net this patch
> introduces a per device deferrable timer which monitors the longest
> pending page fault that requires a response. Proper action such as
> sending failure response code could be taken when timer expires but not
> included in this patch. We need to consider the life cycle of page
> groupd ID to prevent confusion with reused group ID by a device.
> For now, a warning message provides clue of such failure.
>
> Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
> Signed-off-by: Ashok Raj <ashok.raj@intel.com>
> ---
>  drivers/iommu/iommu.c | 53 +++++++++++++++++++++++++++++++++++++++++++++++++++
>  include/linux/iommu.h |  4 ++++
>  2 files changed, 57 insertions(+)
>
> diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
> index 02fed3e..1f2f49e 100644
> --- a/drivers/iommu/iommu.c
> +++ b/drivers/iommu/iommu.c
> @@ -827,6 +827,37 @@ int iommu_group_unregister_notifier(struct iommu_group *group,
>  }
>  EXPORT_SYMBOL_GPL(iommu_group_unregister_notifier);
>  
> +static void iommu_dev_fault_timer_fn(struct timer_list *t)
> +{
> +	struct iommu_fault_param *fparam = from_timer(fparam, t, timer);
> +	struct iommu_fault_event *evt;
> +
> +	u64 now;
> +
> +	now = get_jiffies_64();
> +
> +	/* The goal is to ensure driver or guest page fault handler(via vfio)
> +	 * send page response on time. Otherwise, limited queue resources
> +	 * may be occupied by some irresponsive guests or drivers.
> +	 * When per device pending fault list is not empty, we periodically checks
> +	 * if any anticipated page response time has expired.
> +	 *
> +	 * TODO:
> +	 * We could do the following if response time expires:
> +	 * 1. send page response code FAILURE to all pending PRQ
> +	 * 2. inform device driver or vfio
> +	 * 3. drain in-flight page requests and responses for this device
> +	 * 4. clear pending fault list such that driver can unregister fault
> +	 *    handler(otherwise blocked when pending faults are present).
> +	 */
> +	list_for_each_entry(evt, &fparam->faults, list) {
> +		if (time_after64(now, evt->expire))
> +			pr_err("Page response time expired!, pasid %d gid %d exp %llu now %llu\n",
> +				evt->pasid, evt->page_req_group_id, evt->expire, now);
> +	}
> +	mod_timer(t, now + prq_timeout);
> +}
> +

This timer scheme is very rough.

The timer expires every 10 seconds (by default).

0                   10                 20                30                 40            
+---------------+---------------+---------------+---------------+
^  ^   ^  ^                        ^
 |   |     |    |                         |
F0 F1  F2 F3                       (F1,F2,F3 will not be handled until here!)

F0, F1, F2, F3 are four page faults happens during [0, 10s) time
window. F1, F2, F3 timeout won't be handled until the timer expires
again at 20s. That means a fault might be pending there until about
(2 * prq_timeout) seconds later.

Out of curiosity, Why not adding a timer in iommu_fault_event, starting it in
iommu_report_device_fault() and removing it in iommu_page_response()?

Best regards,
Lu Baolu


>  /**
>   * iommu_register_device_fault_handler() - Register a device fault handler
>   * @dev: the device
> @@ -879,6 +910,9 @@ int iommu_register_device_fault_handler(struct device *dev,
>  	param->fault_param->data = data;
>  	INIT_LIST_HEAD(&param->fault_param->faults);
>  
> +	if (prq_timeout)
> +		timer_setup(&param->fault_param->timer, iommu_dev_fault_timer_fn,
> +			TIMER_DEFERRABLE);
>  done_unlock:
>  	mutex_unlock(&param->lock);
>  
> @@ -935,6 +969,8 @@ int iommu_report_device_fault(struct device *dev, struct iommu_fault_event *evt)
>  {
>  	int ret = 0;
>  	struct iommu_fault_event *evt_pending;
> +	struct timer_list *tmr;
> +	u64 exp;
>  	struct iommu_fault_param *fparam;
>  
>  	/* iommu_param is allocated when device is added to group */
> @@ -955,7 +991,17 @@ int iommu_report_device_fault(struct device *dev, struct iommu_fault_event *evt)
>  			ret = -ENOMEM;
>  			goto done_unlock;
>  		}
> +		/* Keep track of response expiration time */
> +		exp = get_jiffies_64() + prq_timeout;
> +		evt_pending->expire = exp;
>  		mutex_lock(&fparam->lock);
> +		if (list_empty(&fparam->faults)) {
> +			/* First pending event, start timer */
> +			tmr = &dev->iommu_param->fault_param->timer;
> +			WARN_ON(timer_pending(tmr));
> +			mod_timer(tmr, exp);
> +		}
> +
>  		list_add_tail(&evt_pending->list, &fparam->faults);
>  		mutex_unlock(&fparam->lock);
>  	}
> @@ -1572,6 +1618,13 @@ int iommu_page_response(struct device *dev,
>  		}
>  	}
>  
> +	/* stop response timer if no more pending request */
> +	if (list_empty(&param->fault_param->faults) &&
> +		timer_pending(&param->fault_param->timer)) {
> +		pr_debug("no pending PRQ, stop timer\n");
> +		del_timer(&param->fault_param->timer);
> +	}
> +
>  done_unlock:
>  	mutex_unlock(&param->fault_param->lock);
>  	return ret;
> diff --git a/include/linux/iommu.h b/include/linux/iommu.h
> index 722b90f..f3665b7 100644
> --- a/include/linux/iommu.h
> +++ b/include/linux/iommu.h
> @@ -360,6 +360,7 @@ enum iommu_fault_reason {
>   * @iommu_private: used by the IOMMU driver for storing fault-specific
>   *                 data. Users should not modify this field before
>   *                 sending the fault response.
> + * @expire: time limit in jiffies will wait for page response
>   */
>  struct iommu_fault_event {
>  	struct list_head list;
> @@ -373,6 +374,7 @@ struct iommu_fault_event {
>  	u32 prot;
>  	u64 device_private;
>  	u64 iommu_private;
> +	u64 expire;
>  };
>  
>  /**
> @@ -380,11 +382,13 @@ struct iommu_fault_event {
>   * @dev_fault_handler: Callback function to handle IOMMU faults at device level
>   * @data: handler private data
>   * @faults: holds the pending faults which needs response, e.g. page response.
> + * @timer: track page request pending time limit
>   * @lock: protect pending PRQ event list
>   */
>  struct iommu_fault_param {
>  	iommu_dev_fault_handler_t handler;
>  	struct list_head faults;
> +	struct timer_list timer;
>  	struct mutex lock;
>  	void *data;
>  };

  reply	other threads:[~2018-05-14  7:43 UTC|newest]

Thread overview: 81+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-05-11 20:53 [PATCH v5 00/23] IOMMU and VT-d driver support for Shared Virtual Address (SVA) Jacob Pan
2018-05-11 20:53 ` [PATCH v5 01/23] iommu: introduce bind_pasid_table API function Jacob Pan
2018-08-23 16:34   ` Auger Eric
2018-08-24 12:47     ` Liu, Yi L
2018-08-24 13:20       ` Auger Eric
2018-08-28 17:04         ` Jacob Pan
2018-08-24 15:00   ` Auger Eric
2018-08-28  5:14     ` Jacob Pan
2018-08-28  8:34       ` Auger Eric
2018-08-28 16:36         ` Jacob Pan
2018-05-11 20:53 ` [PATCH v5 02/23] iommu/vt-d: move device_domain_info to header Jacob Pan
2018-05-11 20:53 ` [PATCH v5 03/23] iommu/vt-d: add a flag for pasid table bound status Jacob Pan
2018-05-13  7:33   ` Lu Baolu
2018-05-14 18:51     ` Jacob Pan
2018-05-13  8:01   ` Lu Baolu
2018-05-14 18:52     ` Jacob Pan
2018-05-11 20:53 ` [PATCH v5 04/23] iommu/vt-d: add bind_pasid_table function Jacob Pan
2018-05-13  9:29   ` Lu Baolu
2018-05-14 20:22     ` Jacob Pan
2018-05-11 20:53 ` [PATCH v5 05/23] iommu: introduce iommu invalidate API function Jacob Pan
2018-05-11 20:53 ` [PATCH v5 06/23] iommu/vt-d: add definitions for PFSID Jacob Pan
2018-05-14  1:36   ` Lu Baolu
2018-05-14 20:30     ` Jacob Pan
2018-05-11 20:53 ` [PATCH v5 07/23] iommu/vt-d: fix dev iotlb pfsid use Jacob Pan
2018-05-14  1:52   ` Lu Baolu
2018-05-14 20:38     ` Jacob Pan
2018-05-11 20:54 ` [PATCH v5 08/23] iommu/vt-d: support flushing more translation cache types Jacob Pan
2018-05-14  2:18   ` Lu Baolu
2018-05-14 20:46     ` Jacob Pan
2018-05-17  8:44   ` kbuild test robot
2018-05-11 20:54 ` [PATCH v5 09/23] iommu/vt-d: add svm/sva invalidate function Jacob Pan
2018-05-14  3:35   ` Lu Baolu
2018-05-14 20:49     ` Jacob Pan
2018-05-11 20:54 ` [PATCH v5 10/23] iommu: introduce device fault data Jacob Pan
2018-09-21 10:07   ` Auger Eric
2018-09-21 17:05     ` Jacob Pan
2018-09-26 10:20       ` Auger Eric
2018-05-11 20:54 ` [PATCH v5 11/23] driver core: add per device iommu param Jacob Pan
2018-05-14  5:27   ` Lu Baolu
2018-05-14 20:52     ` Jacob Pan
2018-05-11 20:54 ` [PATCH v5 12/23] iommu: add a timeout parameter for prq response Jacob Pan
2018-05-11 20:54 ` [PATCH v5 13/23] iommu: introduce device fault report API Jacob Pan
2018-05-14  6:01   ` Lu Baolu
2018-05-14 20:55     ` Jacob Pan
2018-05-15  6:52       ` Lu Baolu
2018-05-17 11:41   ` Liu, Yi L
2018-05-17 15:59     ` Jacob Pan
2018-05-17 23:22       ` Liu, Yi L
2018-05-21 23:03         ` Jacob Pan
2018-09-06  9:25   ` Auger Eric
2018-09-06 12:42     ` Jean-Philippe Brucker
2018-09-06 13:14       ` Auger Eric
2018-09-06 17:06         ` Jean-Philippe Brucker
2018-09-07  7:11           ` Auger Eric
2018-09-07 11:23             ` Jean-Philippe Brucker
2018-09-14 13:24   ` Auger Eric
2018-09-17 16:57     ` Jacob Pan
2018-09-25 14:58   ` Jean-Philippe Brucker
2018-09-25 22:17     ` Jacob Pan
2018-09-26 10:14       ` Jean-Philippe Brucker
2018-05-11 20:54 ` [PATCH v5 14/23] iommu: introduce page response function Jacob Pan
2018-05-14  6:39   ` Lu Baolu
2018-05-29 16:13     ` Jacob Pan
2018-09-10 14:52   ` Auger Eric
2018-09-10 17:50     ` Jacob Pan
2018-09-10 19:06       ` Auger Eric
2018-05-11 20:54 ` [PATCH v5 15/23] iommu: handle page response timeout Jacob Pan
2018-05-14  7:43   ` Lu Baolu [this message]
2018-05-29 16:20     ` Jacob Pan
2018-05-30  7:46       ` Lu Baolu
2018-05-11 20:54 ` [PATCH v5 16/23] iommu/config: add build dependency for dmar Jacob Pan
2018-05-11 20:54 ` [PATCH v5 17/23] iommu/vt-d: report non-recoverable faults to device Jacob Pan
2018-05-14  8:17   ` Lu Baolu
2018-05-29 17:33     ` Jacob Pan
2018-05-11 20:54 ` [PATCH v5 18/23] iommu/intel-svm: report device page request Jacob Pan
2018-05-11 20:54 ` [PATCH v5 19/23] iommu/intel-svm: replace dev ops with fault report API Jacob Pan
2018-05-11 20:54 ` [PATCH v5 20/23] iommu/intel-svm: do not flush iotlb for viommu Jacob Pan
2018-05-11 20:54 ` [PATCH v5 21/23] iommu/vt-d: add intel iommu page response function Jacob Pan
2018-05-11 20:54 ` [PATCH v5 22/23] trace/iommu: add sva trace events Jacob Pan
2018-05-11 20:54 ` [PATCH v5 23/23] iommu: use sva invalidate and device fault trace event Jacob Pan
2018-05-29 15:54 ` [PATCH v5 00/23] IOMMU and VT-d driver support for Shared Virtual Address (SVA) Jacob Pan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5AF93E3A.2040902@linux.intel.com \
    --to=baolu.lu@linux.intel.com \
    --cc=alex.williamson@redhat.com \
    --cc=ashok.raj@intel.com \
    --cc=dwmw2@infradead.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=hch@infradead.org \
    --cc=iommu@lists.linux-foundation.org \
    --cc=jacob.jun.pan@linux.intel.com \
    --cc=jean-philippe.brucker@arm.com \
    --cc=joro@8bytes.org \
    --cc=kevin.tian@intel.com \
    --cc=khali@linux-fr.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=rafael.j.wysocki@intel.com \
    --cc=yi.l.liu@intel.com \
    --subject='Re: [PATCH v5 15/23] iommu: handle page response timeout' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).