From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org, Michael Neuling <mikey@neuling.org>,
Benjamin Herrenschmidt <benh@kernel.crashing.org>,
Michael Ellerman <mpe@ellerman.id.au>
Subject: [PATCH 4.14 90/91] powerpc/eeh: Fix race with driver un/bind
Date: Mon, 30 Apr 2018 12:25:12 -0700 [thread overview]
Message-ID: <20180430184009.002164454@linuxfoundation.org> (raw)
In-Reply-To: <20180430184004.216234025@linuxfoundation.org>
4.14-stable review patch. If anyone has any objections, please let me know.
------------------
From: Michael Neuling <mikey@neuling.org>
commit f0295e047fcf52ccb42561fb7de6942f5201b676 upstream.
The current EEH callbacks can race with a driver unbind. This can
result in a backtraces like this:
EEH: Frozen PHB#0-PE#1fc detected
EEH: PE location: S000009, PHB location: N/A
CPU: 2 PID: 2312 Comm: kworker/u258:3 Not tainted 4.15.6-openpower1 #2
Workqueue: nvme-wq nvme_reset_work [nvme]
Call Trace:
dump_stack+0x9c/0xd0 (unreliable)
eeh_dev_check_failure+0x420/0x470
eeh_check_failure+0xa0/0xa4
nvme_reset_work+0x138/0x1414 [nvme]
process_one_work+0x1ec/0x328
worker_thread+0x2e4/0x3a8
kthread+0x14c/0x154
ret_from_kernel_thread+0x5c/0xc8
nvme nvme1: Removing after probe failure status: -19
<snip>
cpu 0x23: Vector: 300 (Data Access) at [c000000ff50f3800]
pc: c0080000089a0eb0: nvme_error_detected+0x4c/0x90 [nvme]
lr: c000000000026564: eeh_report_error+0xe0/0x110
sp: c000000ff50f3a80
msr: 9000000000009033
dar: 400
dsisr: 40000000
current = 0xc000000ff507c000
paca = 0xc00000000fdc9d80 softe: 0 irq_happened: 0x01
pid = 782, comm = eehd
Linux version 4.15.6-openpower1 (smc@smc-desktop) (gcc version 6.4.0 (Buildroot 2017.11.2-00008-g4b6188e)) #2 SM P Tue Feb 27 12:33:27 PST 2018
enter ? for help
eeh_report_error+0xe0/0x110
eeh_pe_dev_traverse+0xc0/0xdc
eeh_handle_normal_event+0x184/0x4c4
eeh_handle_event+0x30/0x288
eeh_event_handler+0x124/0x170
kthread+0x14c/0x154
ret_from_kernel_thread+0x5c/0xc8
The first part is an EEH (on boot), the second half is the resulting
crash. nvme probe starts the nvme_reset_work() worker thread. This
worker thread starts touching the device which see a device error
(EEH) and hence queues up an event in the powerpc EEH worker
thread. nvme_reset_work() then continues and runs
nvme_remove_dead_ctrl_work() which results in unbinding the driver
from the device and hence releases all resources. At the same time,
the EEH worker thread starts doing the EEH .error_detected() driver
callback, which no longer works since the resources have been freed.
This fixes the problem in the same way the generic PCIe AER code (in
drivers/pci/pcie/aer/aerdrv_core.c) does. It makes the EEH code hold
the device_lock() while performing the driver EEH callbacks and
associated code. This ensures either the callbacks are no longer
register, or if they are registered the driver will not be removed
from underneath us.
This has been broken forever. The EEH call backs were first introduced
in 2005 (in 77bd7415610) but it's not clear if a lock was needed back
then.
Fixes: 77bd74156101 ("[PATCH] powerpc: PCI Error Recovery: PPC64 core recovery routines")
Cc: stable@vger.kernel.org # v2.6.16+
Signed-off-by: Michael Neuling <mikey@neuling.org>
Reviewed-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
arch/powerpc/kernel/eeh_driver.c | 61 ++++++++++++++++++++++++---------------
1 file changed, 38 insertions(+), 23 deletions(-)
--- a/arch/powerpc/kernel/eeh_driver.c
+++ b/arch/powerpc/kernel/eeh_driver.c
@@ -207,18 +207,18 @@ static void *eeh_report_error(void *data
if (!dev || eeh_dev_removed(edev) || eeh_pe_passed(edev->pe))
return NULL;
+
+ device_lock(&dev->dev);
dev->error_state = pci_channel_io_frozen;
driver = eeh_pcid_get(dev);
- if (!driver) return NULL;
+ if (!driver) goto out_no_dev;
eeh_disable_irq(dev);
if (!driver->err_handler ||
- !driver->err_handler->error_detected) {
- eeh_pcid_put(dev);
- return NULL;
- }
+ !driver->err_handler->error_detected)
+ goto out;
rc = driver->err_handler->error_detected(dev, pci_channel_io_frozen);
@@ -227,7 +227,10 @@ static void *eeh_report_error(void *data
if (*res == PCI_ERS_RESULT_NONE) *res = rc;
edev->in_error = true;
+out:
eeh_pcid_put(dev);
+out_no_dev:
+ device_unlock(&dev->dev);
return NULL;
}
@@ -250,15 +253,14 @@ static void *eeh_report_mmio_enabled(voi
if (!dev || eeh_dev_removed(edev) || eeh_pe_passed(edev->pe))
return NULL;
+ device_lock(&dev->dev);
driver = eeh_pcid_get(dev);
- if (!driver) return NULL;
+ if (!driver) goto out_no_dev;
if (!driver->err_handler ||
!driver->err_handler->mmio_enabled ||
- (edev->mode & EEH_DEV_NO_HANDLER)) {
- eeh_pcid_put(dev);
- return NULL;
- }
+ (edev->mode & EEH_DEV_NO_HANDLER))
+ goto out;
rc = driver->err_handler->mmio_enabled(dev);
@@ -266,7 +268,10 @@ static void *eeh_report_mmio_enabled(voi
if (rc == PCI_ERS_RESULT_NEED_RESET) *res = rc;
if (*res == PCI_ERS_RESULT_NONE) *res = rc;
+out:
eeh_pcid_put(dev);
+out_no_dev:
+ device_unlock(&dev->dev);
return NULL;
}
@@ -289,20 +294,20 @@ static void *eeh_report_reset(void *data
if (!dev || eeh_dev_removed(edev) || eeh_pe_passed(edev->pe))
return NULL;
+
+ device_lock(&dev->dev);
dev->error_state = pci_channel_io_normal;
driver = eeh_pcid_get(dev);
- if (!driver) return NULL;
+ if (!driver) goto out_no_dev;
eeh_enable_irq(dev);
if (!driver->err_handler ||
!driver->err_handler->slot_reset ||
(edev->mode & EEH_DEV_NO_HANDLER) ||
- (!edev->in_error)) {
- eeh_pcid_put(dev);
- return NULL;
- }
+ (!edev->in_error))
+ goto out;
rc = driver->err_handler->slot_reset(dev);
if ((*res == PCI_ERS_RESULT_NONE) ||
@@ -310,7 +315,10 @@ static void *eeh_report_reset(void *data
if (*res == PCI_ERS_RESULT_DISCONNECT &&
rc == PCI_ERS_RESULT_NEED_RESET) *res = rc;
+out:
eeh_pcid_put(dev);
+out_no_dev:
+ device_unlock(&dev->dev);
return NULL;
}
@@ -361,10 +369,12 @@ static void *eeh_report_resume(void *dat
if (!dev || eeh_dev_removed(edev) || eeh_pe_passed(edev->pe))
return NULL;
+
+ device_lock(&dev->dev);
dev->error_state = pci_channel_io_normal;
driver = eeh_pcid_get(dev);
- if (!driver) return NULL;
+ if (!driver) goto out_no_dev;
was_in_error = edev->in_error;
edev->in_error = false;
@@ -374,13 +384,15 @@ static void *eeh_report_resume(void *dat
!driver->err_handler->resume ||
(edev->mode & EEH_DEV_NO_HANDLER) || !was_in_error) {
edev->mode &= ~EEH_DEV_NO_HANDLER;
- eeh_pcid_put(dev);
- return NULL;
+ goto out;
}
driver->err_handler->resume(dev);
+out:
eeh_pcid_put(dev);
+out_no_dev:
+ device_unlock(&dev->dev);
return NULL;
}
@@ -400,22 +412,25 @@ static void *eeh_report_failure(void *da
if (!dev || eeh_dev_removed(edev) || eeh_pe_passed(edev->pe))
return NULL;
+
+ device_lock(&dev->dev);
dev->error_state = pci_channel_io_perm_failure;
driver = eeh_pcid_get(dev);
- if (!driver) return NULL;
+ if (!driver) goto out_no_dev;
eeh_disable_irq(dev);
if (!driver->err_handler ||
- !driver->err_handler->error_detected) {
- eeh_pcid_put(dev);
- return NULL;
- }
+ !driver->err_handler->error_detected)
+ goto out;
driver->err_handler->error_detected(dev, pci_channel_io_perm_failure);
+out:
eeh_pcid_put(dev);
+out_no_dev:
+ device_unlock(&dev->dev);
return NULL;
}
next prev parent reply other threads:[~2018-04-30 19:27 UTC|newest]
Thread overview: 90+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-04-30 19:23 [PATCH 4.14 00/91] 4.14.39-stable review Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 01/91] ext4: prevent right-shifting extents beyond EXT_MAX_BLOCKS Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 02/91] ext4: set h_journal if there is a failure starting a reserved handle Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 04/91] ext4: add validity checks for bitmap block numbers Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 05/91] ext4: fix bitmap position validation Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 06/91] random: set up the NUMA crng instances after the CRNG is fully initialized Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 07/91] random: fix possible sleeping allocation from irq context Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 08/91] random: rate limit unseeded randomness warnings Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 09/91] usbip: usbip_event: fix to not print kernel pointer address Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 10/91] usbip: usbip_host: fix to hold parent lock for device_attach() calls Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 11/91] usbip: vhci_hcd: Fix usb device and sockfd leaks Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 12/91] usbip: vhci_hcd: check rhport before using in vhci_hub_control() Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 13/91] Revert "xhci: plat: Register shutdown for xhci_plat" Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 14/91] xhci: Fix USB ports for Dell Inspiron 5775 Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 15/91] USB: serial: simple: add libtransistor console Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 16/91] USB: serial: ftdi_sio: use jtag quirk for Arrow USB Blaster Greg Kroah-Hartman
2018-04-30 19:23 ` [PATCH 4.14 17/91] USB: serial: cp210x: add ID for NI USB serial console Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 18/91] usb: typec: ucsi: Increase command completion timeout value Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 19/91] usb: core: Add quirk for HP v222w 16GB Mini Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 20/91] USB: Increment wakeup count on remote wakeup Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 21/91] ALSA: usb-audio: Skip broken EU on Dell dock USB-audio Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 22/91] virtio: add ability to iterate over vqs Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 23/91] virtio_console: dont tie bufs to a vq Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 25/91] virtio_console: drop custom control queue cleanup Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 26/91] virtio_console: move removal code Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 27/91] virtio_console: reset on out of memory Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 28/91] drm/virtio: fix vq wait_event condition Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 29/91] tty: Dont call panic() at tty_ldisc_init() Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 30/91] tty: n_gsm: Fix long delays with control frame timeouts in ADM mode Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 31/91] tty: n_gsm: Fix DLCI handling for ADM mode if debug & 2 is not set Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 32/91] tty: Avoid possible error pointer dereference at tty_ldisc_restore() Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 33/91] tty: Use __GFP_NOFAIL for tty_ldisc_get() Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 34/91] ALSA: dice: fix OUI for TC group Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 35/91] ALSA: dice: fix error path to destroy initialized stream data Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 36/91] ALSA: hda - Skip jack and others for non-existing PCM streams Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 37/91] ALSA: opl3: Hardening for potential Spectre v1 Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 38/91] ALSA: asihpi: " Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 39/91] ALSA: hdspm: " Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 40/91] ALSA: rme9652: " Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 41/91] ALSA: control: " Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 42/91] ALSA: pcm: Return negative delays from SNDRV_PCM_IOCTL_DELAY Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 43/91] ALSA: core: Report audio_tstamp in snd_pcm_sync_ptr Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 44/91] ALSA: seq: oss: Fix unbalanced use lock for synth MIDI device Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 45/91] ALSA: seq: oss: Hardening for potential Spectre v1 Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 46/91] ALSA: hda: " Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 47/91] ALSA: hda/realtek - Add some fixes for ALC233 Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 48/91] ALSA: hda/realtek - Update ALC255 depop optimize Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 49/91] ALSA: hda/realtek - change the location for one of two front mics Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 50/91] mtd: spi-nor: cadence-quadspi: Fix page fault kernel panic Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 51/91] mtd: cfi: cmdset_0001: Do not allow read/write to suspend erase block Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 52/91] mtd: cfi: cmdset_0001: Workaround Micron Erase suspend bug Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 53/91] mtd: cfi: cmdset_0002: Do not allow read/write to suspend erase block Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 54/91] mtd: rawnand: tango: Fix struct clk memory leak Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 55/91] kobject: dont use WARN for registration failures Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 56/91] scsi: sd: Defer spinning up drive while SANITIZE is in progress Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 57/91] bfq-iosched: ensure to clear bic/bfqq pointers when preparing request Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 58/91] vfio: ccw: process ssch with interrupts disabled Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 59/91] ANDROID: binder: prevent transactions into own process Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 60/91] PCI: aardvark: Fix logic in advk_pcie_{rd,wr}_conf() Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 61/91] PCI: aardvark: Set PIO_ADDR_LS correctly in advk_pcie_rd_conf() Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 62/91] PCI: aardvark: Use ISR1 instead of ISR0 interrupt in legacy irq mode Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 63/91] PCI: aardvark: Fix PCIe Max Read Request Size setting Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 64/91] ARM: amba: Make driver_override output consistent with other buses Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 65/91] ARM: amba: Fix race condition with driver_override Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 66/91] ARM: amba: Dont read past the end of sysfs "driver_override" buffer Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 67/91] ARM: socfpga_defconfig: Remove QSPI Sector 4K size force Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 68/91] KVM: arm/arm64: Close VMID generation race Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 69/91] powerpc/mm: Flush cache on memory hot(un)plug Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 70/91] powerpc/powernv/npu: Do a PID GPU TLB flush when invalidating a large address range Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 71/91] crypto: drbg - set freed buffers to NULL Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 72/91] ASoC: fsl_esai: Fix divisor calculation failure at lower ratio Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 73/91] libceph: un-backoff on tick when we have a authenticated session Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 74/91] libceph: reschedule a tick in finish_hunting() Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 75/91] libceph: validate con->state at the top of try_write() Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 76/91] fpga-manager: altera-ps-spi: preserve nCONFIG state Greg Kroah-Hartman
2018-04-30 19:24 ` [PATCH 4.14 77/91] earlycon: Use a pointer table to fix __earlycon_table stride Greg Kroah-Hartman
2018-04-30 19:25 ` [PATCH 4.14 78/91] cpufreq: powernv: Fix hardlockup due to synchronous smp_call in timer interrupt Greg Kroah-Hartman
2018-04-30 19:25 ` [PATCH 4.14 79/91] rtc: opal: Fix OPAL RTC driver OPAL_BUSY loops Greg Kroah-Hartman
2018-04-30 19:25 ` [PATCH 4.14 84/91] x86/ipc: Fix x32 version of shmid64_ds and msqid64_ds Greg Kroah-Hartman
2018-04-30 19:25 ` [PATCH 4.14 85/91] x86/smpboot: Dont use mwait_play_dead() on AMD systems Greg Kroah-Hartman
2018-04-30 19:25 ` [PATCH 4.14 86/91] x86/microcode/intel: Save microcode patch unconditionally Greg Kroah-Hartman
2018-04-30 19:25 ` [PATCH 4.14 87/91] x86/microcode: Do not exit early from __reload_late() Greg Kroah-Hartman
2018-04-30 19:25 ` [PATCH 4.14 88/91] tick/sched: Do not mess with an enqueued hrtimer Greg Kroah-Hartman
2018-04-30 19:25 ` [PATCH 4.14 89/91] arm/arm64: KVM: Add PSCI version selection API Greg Kroah-Hartman
2018-04-30 19:25 ` Greg Kroah-Hartman [this message]
2018-04-30 19:25 ` [PATCH 4.14 91/91] net: dont call update_pmtu unconditionally Greg Kroah-Hartman
2018-05-01 0:54 ` [PATCH 4.14 00/91] 4.14.39-stable review Nathan Chancellor
2018-05-01 13:21 ` Guenter Roeck
2018-05-01 14:04 ` Dan Rue
2018-05-01 19:06 ` Shuah Khan
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=20180430184009.002164454@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=benh@kernel.crashing.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mikey@neuling.org \
--cc=mpe@ellerman.id.au \
--cc=stable@vger.kernel.org \
/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).