stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	alan@lxorguk.ukuu.org.uk, Huang Ying <ying.huang@intel.com>,
	Bjorn Helgaas <bhelgaas@google.com>,
	"Rafael J. Wysocki" <rafael.j.wysocki@intel.com>,
	Zhang Yanmin <yanmin.zhang@intel.com>
Subject: [ 37/37] PCI/PM: Fix deadlock when unbinding device if parent in D3cold
Date: Fri, 14 Dec 2012 15:01:14 -0800	[thread overview]
Message-ID: <20121214222258.643463632@linuxfoundation.org> (raw)
In-Reply-To: <20121214222252.781413287@linuxfoundation.org>

3.6-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Huang Ying <ying.huang@intel.com>

commit 90b5c1d7c45eeb622302680ff96ed30c1a2b6f0e upstream.

If a PCI device and its parents are put into D3cold, unbinding the
device will trigger deadlock as follow:

- driver_unbind
  - device_release_driver
    - device_lock(dev)				<--- previous lock here
    - __device_release_driver
      - pm_runtime_get_sync
        ...
          - rpm_resume(dev)
            - rpm_resume(dev->parent)
              ...
                - pci_pm_runtime_resume
                  ...
                  - pci_set_power_state
                    - __pci_start_power_transition
                      - pci_wakeup_bus(dev->parent->subordinate)
                        - pci_walk_bus
                          - device_lock(dev)	<--- deadlock here


If we do not do device_lock in pci_walk_bus, we can avoid deadlock.
Device_lock in pci_walk_bus is introduced in commit:
d71374dafbba7ec3f67371d3b7e9f6310a588808, corresponding email thread
is: https://lkml.org/lkml/2006/5/26/38.  The patch author Zhang Yanmin
said device_lock is added to pci_walk_bus because:

  Some error handling functions call pci_walk_bus. For example, PCIe
  aer. Here we lock the device, so the driver wouldn't detach from the
  device, as the cb might call driver's callback function.

So I fixed the deadlock as follows:

- remove device_lock from pci_walk_bus
- add device_lock into callback if callback will call driver's callback

I checked pci_walk_bus users one by one, and found only PCIe aer needs
device lock.

Signed-off-by: Huang Ying <ying.huang@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
CC: stable@vger.kernel.org		# v3.6+
CC: Zhang Yanmin <yanmin.zhang@intel.com>


---
 arch/powerpc/platforms/pseries/eeh_driver.c |   35 ++++++++++++++++++++--------
 drivers/pci/bus.c                           |    3 --
 drivers/pci/pcie/aer/aerdrv_core.c          |   20 ++++++++++++----
 3 files changed, 41 insertions(+), 17 deletions(-)

--- a/arch/powerpc/platforms/pseries/eeh_driver.c
+++ b/arch/powerpc/platforms/pseries/eeh_driver.c
@@ -164,17 +164,18 @@ static int eeh_report_error(struct pci_d
 	enum pci_ers_result rc, *res = userdata;
 	struct pci_driver *driver;
 
+	device_lock(&dev->dev);
 	dev->error_state = pci_channel_io_frozen;
 
 	driver = eeh_pcid_get(dev);
-	if (!driver) return 0;
+	if (!driver) goto out;
 
 	eeh_disable_irq(dev);
 
 	if (!driver->err_handler ||
 	    !driver->err_handler->error_detected) {
 		eeh_pcid_put(dev);
-		return 0;
+		goto out;
 	}
 
 	rc = driver->err_handler->error_detected(dev, pci_channel_io_frozen);
@@ -184,6 +185,8 @@ static int eeh_report_error(struct pci_d
 	if (*res == PCI_ERS_RESULT_NONE) *res = rc;
 
 	eeh_pcid_put(dev);
+out:
+	device_unlock(&dev->dev);
 	return 0;
 }
 
@@ -201,13 +204,14 @@ static int eeh_report_mmio_enabled(struc
 	enum pci_ers_result rc, *res = userdata;
 	struct pci_driver *driver;
 
+	device_lock(&dev->dev);
 	driver = eeh_pcid_get(dev);
-	if (!driver) return 0;
+	if (!driver) goto out;
 
 	if (!driver->err_handler ||
 	    !driver->err_handler->mmio_enabled) {
 		eeh_pcid_put(dev);
-		return 0;
+		goto out;
 	}
 
 	rc = driver->err_handler->mmio_enabled(dev);
@@ -217,6 +221,8 @@ static int eeh_report_mmio_enabled(struc
 	if (*res == PCI_ERS_RESULT_NONE) *res = rc;
 
 	eeh_pcid_put(dev);
+out:
+	device_unlock(&dev->dev);
 	return 0;
 }
 
@@ -235,17 +241,18 @@ static int eeh_report_reset(struct pci_d
 	enum pci_ers_result rc, *res = userdata;
 	struct pci_driver *driver;
 
+	device_lock(&dev->dev);
 	dev->error_state = pci_channel_io_normal;
 
 	driver = eeh_pcid_get(dev);
-	if (!driver) return 0;
+	if (!driver) goto out;
 
 	eeh_enable_irq(dev);
 
 	if (!driver->err_handler ||
 	    !driver->err_handler->slot_reset) {
 		eeh_pcid_put(dev);
-		return 0;
+		goto out;
 	}
 
 	rc = driver->err_handler->slot_reset(dev);
@@ -255,6 +262,8 @@ static int eeh_report_reset(struct pci_d
 	     rc == PCI_ERS_RESULT_NEED_RESET) *res = rc;
 
 	eeh_pcid_put(dev);
+out:
+	device_unlock(&dev->dev);
 	return 0;
 }
 
@@ -271,22 +280,25 @@ static int eeh_report_resume(struct pci_
 {
 	struct pci_driver *driver;
 
+	device_lock(&dev->dev);
 	dev->error_state = pci_channel_io_normal;
 
 	driver = eeh_pcid_get(dev);
-	if (!driver) return 0;
+	if (!driver) goto out;
 
 	eeh_enable_irq(dev);
 
 	if (!driver->err_handler ||
 	    !driver->err_handler->resume) {
 		eeh_pcid_put(dev);
-		return 0;
+		goto out;
 	}
 
 	driver->err_handler->resume(dev);
 
 	eeh_pcid_put(dev);
+out:
+	device_unlock(&dev->dev);
 	return 0;
 }
 
@@ -302,22 +314,25 @@ static int eeh_report_failure(struct pci
 {
 	struct pci_driver *driver;
 
+	device_lock(&dev->dev);
 	dev->error_state = pci_channel_io_perm_failure;
 
 	driver = eeh_pcid_get(dev);
-	if (!driver) return 0;
+	if (!driver) goto out;
 
 	eeh_disable_irq(dev);
 
 	if (!driver->err_handler ||
 	    !driver->err_handler->error_detected) {
 		eeh_pcid_put(dev);
-		return 0;
+		goto out;
 	}
 
 	driver->err_handler->error_detected(dev, pci_channel_io_perm_failure);
 
 	eeh_pcid_put(dev);
+out:
+	device_unlock(&dev->dev);
 	return 0;
 }
 
--- a/drivers/pci/bus.c
+++ b/drivers/pci/bus.c
@@ -316,10 +316,7 @@ void pci_walk_bus(struct pci_bus *top, i
 		} else
 			next = dev->bus_list.next;
 
-		/* Run device routines with the device locked */
-		device_lock(&dev->dev);
 		retval = cb(dev, userdata);
-		device_unlock(&dev->dev);
 		if (retval)
 			break;
 	}
--- a/drivers/pci/pcie/aer/aerdrv_core.c
+++ b/drivers/pci/pcie/aer/aerdrv_core.c
@@ -244,6 +244,7 @@ static int report_error_detected(struct
 	struct aer_broadcast_data *result_data;
 	result_data = (struct aer_broadcast_data *) data;
 
+	device_lock(&dev->dev);
 	dev->error_state = result_data->state;
 
 	if (!dev->driver ||
@@ -262,12 +263,14 @@ static int report_error_detected(struct
 				   dev->driver ?
 				   "no AER-aware driver" : "no driver");
 		}
-		return 0;
+		goto out;
 	}
 
 	err_handler = dev->driver->err_handler;
 	vote = err_handler->error_detected(dev, result_data->state);
 	result_data->result = merge_result(result_data->result, vote);
+out:
+	device_unlock(&dev->dev);
 	return 0;
 }
 
@@ -278,14 +281,17 @@ static int report_mmio_enabled(struct pc
 	struct aer_broadcast_data *result_data;
 	result_data = (struct aer_broadcast_data *) data;
 
+	device_lock(&dev->dev);
 	if (!dev->driver ||
 		!dev->driver->err_handler ||
 		!dev->driver->err_handler->mmio_enabled)
-		return 0;
+		goto out;
 
 	err_handler = dev->driver->err_handler;
 	vote = err_handler->mmio_enabled(dev);
 	result_data->result = merge_result(result_data->result, vote);
+out:
+	device_unlock(&dev->dev);
 	return 0;
 }
 
@@ -296,14 +302,17 @@ static int report_slot_reset(struct pci_
 	struct aer_broadcast_data *result_data;
 	result_data = (struct aer_broadcast_data *) data;
 
+	device_lock(&dev->dev);
 	if (!dev->driver ||
 		!dev->driver->err_handler ||
 		!dev->driver->err_handler->slot_reset)
-		return 0;
+		goto out;
 
 	err_handler = dev->driver->err_handler;
 	vote = err_handler->slot_reset(dev);
 	result_data->result = merge_result(result_data->result, vote);
+out:
+	device_unlock(&dev->dev);
 	return 0;
 }
 
@@ -311,15 +320,18 @@ static int report_resume(struct pci_dev
 {
 	struct pci_error_handlers *err_handler;
 
+	device_lock(&dev->dev);
 	dev->error_state = pci_channel_io_normal;
 
 	if (!dev->driver ||
 		!dev->driver->err_handler ||
 		!dev->driver->err_handler->resume)
-		return 0;
+		goto out;
 
 	err_handler = dev->driver->err_handler;
 	err_handler->resume(dev);
+out:
+	device_unlock(&dev->dev);
 	return 0;
 }
 



  parent reply	other threads:[~2012-12-14 23:01 UTC|newest]

Thread overview: 41+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-12-14 23:00 [ 00/37] 3.6.11-stable review Greg Kroah-Hartman
2012-12-14 23:00 ` [ 01/37] tmpfs: fix shared mempolicy leak Greg Kroah-Hartman
2012-12-14 23:00 ` [ 02/37] Revert misapplied "mmc: sh-mmcif: avoid oops on spurious interrupts" Greg Kroah-Hartman
2012-12-14 23:00 ` [ 03/37] mmc: sh-mmcif: avoid oops on spurious interrupts (second try) Greg Kroah-Hartman
2012-12-14 23:00 ` [ 04/37] ARM: 7566/1: vfp: fix save and restore when running on pre-VFPv3 and CONFIG_VFPv3 set Greg Kroah-Hartman
2012-12-14 23:00 ` [ 05/37] ASoC: dmaengine: Correct Makefile when sound is built as module Greg Kroah-Hartman
2012-12-14 23:00 ` [ 06/37] Staging: ipack/bridges/tpci200: avoid kernel bug when uninstalling a device Greg Kroah-Hartman
2012-12-14 23:00 ` [ 07/37] Input: matrix-keymap - provide proper module license Greg Kroah-Hartman
2012-12-14 23:00 ` [ 08/37] i82975x_edac: Fix dimm label initialization Greg Kroah-Hartman
2012-12-14 23:00 ` [ 09/37] edac: Fix the dimm filling for csrows-based layouts Greg Kroah-Hartman
2012-12-14 23:00 ` [ 10/37] workqueue: convert BUG_ON()s in __queue_delayed_work() to WARN_ON_ONCE()s Greg Kroah-Hartman
2012-12-14 23:00 ` [ 11/37] drm/i915: do not ignore eDP bpc settings from vbt Greg Kroah-Hartman
2012-12-14 23:00 ` [ 12/37] drm/i915: do not default to 18 bpp for eDP if missing from VBT Greg Kroah-Hartman
2012-12-14 23:00 ` [ 13/37] mm: dmapool: use provided gfp flags for all dma_alloc_coherent() calls Greg Kroah-Hartman
2012-12-14 23:00 ` [ 14/37] x86,AMD: Power driver support for AMDs family 16h processors Greg Kroah-Hartman
2012-12-14 23:00 ` [ 15/37] floppy: destroy floppy workqueue before cleaning up the queue Greg Kroah-Hartman
2012-12-14 23:00 ` [ 16/37] x86: hpet: Fix masking of MSI interrupts Greg Kroah-Hartman
2012-12-14 23:00 ` [ 17/37] USB: add new zte 3g-dongles pid to option.c Greg Kroah-Hartman
2012-12-14 23:00 ` [ 18/37] USB: option: blacklist network interface on Huawei E173 Greg Kroah-Hartman
2012-12-14 23:00 ` [ 19/37] USB: ftdi_sio: Add support for Newport AGILIS motor drivers Greg Kroah-Hartman
2012-12-14 23:00 ` [ 20/37] usb: ftdi_sio: fixup BeagleBone A5+ quirk Greg Kroah-Hartman
2012-12-14 23:00 ` [ 21/37] USB: cp210x: add Virtenio Preon32 device id Greg Kroah-Hartman
2012-12-14 23:00 ` [ 22/37] USB: mark uas driver as BROKEN Greg Kroah-Hartman
2012-12-14 23:01 ` [ 23/37] ACPI / battery: Correct battery capacity values on Thinkpads Greg Kroah-Hartman
2012-12-14 23:01 ` [ 24/37] ACPI / PM: Add Sony Vaio VPCEB1S1E to nonvs blacklist Greg Kroah-Hartman
2012-12-14 23:01 ` [ 25/37] ACPI / PNP: Do not crash due to stale pointer use during system resume Greg Kroah-Hartman
2012-12-14 23:01 ` [ 26/37] ACPI / video: ignore BIOS initial backlight value for HP Folio 13-2000 Greg Kroah-Hartman
2012-12-14 23:01 ` [ 27/37] ACPI / video: Add "Asus UL30VT" to ACPI video detect blacklist Greg Kroah-Hartman
2012-12-14 23:01 ` [ 28/37] USB: OHCI: workaround for hardware bug: retired TDs not added to the Done Queue Greg Kroah-Hartman
2012-12-14 23:01 ` [ 29/37] xhci: Extend Fresco Logic MSI quirk Greg Kroah-Hartman
2012-12-14 23:01 ` [ 30/37] ftrace: Clear bits properly in reset_iter_read() Greg Kroah-Hartman
2012-12-14 23:01 ` [ 31/37] ring-buffer: Fix NULL pointer if rb_set_head_page() fails Greg Kroah-Hartman
2012-12-14 23:01 ` [ 32/37] ring-buffer: Fix race between integrity check and readers Greg Kroah-Hartman
2012-12-14 23:01 ` [ 33/37] cdc-acm: implement TIOCSSERIAL to avoid blocking close(2) Greg Kroah-Hartman
2012-12-14 23:01 ` [ 34/37] perf test: fix a build error on builtin-test Greg Kroah-Hartman
2012-12-14 23:01 ` [ 35/37] USB: EHCI: bugfix: urb->hcpriv should not be NULL Greg Kroah-Hartman
2012-12-14 23:01 ` [ 36/37] rcu: Fix batch-limit size problem Greg Kroah-Hartman
2012-12-14 23:01 ` Greg Kroah-Hartman [this message]
2012-12-15 14:24 ` [ 00/37] 3.6.11-stable review Shuah Khan
2012-12-15 20:47   ` Shuah Khan
2012-12-15 14:27 ` Satoru Takeuchi

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=20121214222258.643463632@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=alan@lxorguk.ukuu.org.uk \
    --cc=bhelgaas@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=rafael.j.wysocki@intel.com \
    --cc=stable@vger.kernel.org \
    --cc=yanmin.zhang@intel.com \
    --cc=ying.huang@intel.com \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).