From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org,
Shigeru Yoshida <Shigeru.Yoshida@windriver.com>,
Haiqing Bai <Haiqing.Bai@windriver.com>,
Alan Stern <stern@rowland.harvard.edu>
Subject: [PATCH 4.15 37/64] ohci-hcd: Fix race condition caused by ohci_urb_enqueue() and io_watchdog_func()
Date: Mon, 26 Feb 2018 21:22:14 +0100 [thread overview]
Message-ID: <20180226202155.043753482@linuxfoundation.org> (raw)
In-Reply-To: <20180226202153.453363333@linuxfoundation.org>
4.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Shigeru Yoshida <shigeru.yoshida@windriver.com>
commit b2685bdacdaab065c172b97b55ab46c6be77a037 upstream.
Running io_watchdog_func() while ohci_urb_enqueue() is running can
cause a race condition where ohci->prev_frame_no is corrupted and the
watchdog can mis-detect following error:
ohci-platform 664a0800.usb: frame counter not updating; disabled
ohci-platform 664a0800.usb: HC died; cleaning up
Specifically, following scenario causes a race condition:
1. ohci_urb_enqueue() calls spin_lock_irqsave(&ohci->lock, flags)
and enters the critical section
2. ohci_urb_enqueue() calls timer_pending(&ohci->io_watchdog) and it
returns false
3. ohci_urb_enqueue() sets ohci->prev_frame_no to a frame number
read by ohci_frame_no(ohci)
4. ohci_urb_enqueue() schedules io_watchdog_func() with mod_timer()
5. ohci_urb_enqueue() calls spin_unlock_irqrestore(&ohci->lock,
flags) and exits the critical section
6. Later, ohci_urb_enqueue() is called
7. ohci_urb_enqueue() calls spin_lock_irqsave(&ohci->lock, flags)
and enters the critical section
8. The timer scheduled on step 4 expires and io_watchdog_func() runs
9. io_watchdog_func() calls spin_lock_irqsave(&ohci->lock, flags)
and waits on it because ohci_urb_enqueue() is already in the
critical section on step 7
10. ohci_urb_enqueue() calls timer_pending(&ohci->io_watchdog) and it
returns false
11. ohci_urb_enqueue() sets ohci->prev_frame_no to new frame number
read by ohci_frame_no(ohci) because the frame number proceeded
between step 3 and 6
12. ohci_urb_enqueue() schedules io_watchdog_func() with mod_timer()
13. ohci_urb_enqueue() calls spin_unlock_irqrestore(&ohci->lock,
flags) and exits the critical section, then wake up
io_watchdog_func() which is waiting on step 9
14. io_watchdog_func() enters the critical section
15. io_watchdog_func() calls ohci_frame_no(ohci) and set frame_no
variable to the frame number
16. io_watchdog_func() compares frame_no and ohci->prev_frame_no
On step 16, because this calling of io_watchdog_func() is scheduled on
step 4, the frame number set in ohci->prev_frame_no is expected to the
number set on step 3. However, ohci->prev_frame_no is overwritten on
step 11. Because step 16 is executed soon after step 11, the frame
number might not proceed, so ohci->prev_frame_no must equals to
frame_no.
To address above scenario, this patch introduces a special sentinel
value IO_WATCHDOG_OFF and set this value to ohci->prev_frame_no when
the watchdog is not pending or running. When ohci_urb_enqueue()
schedules the watchdog (step 4 and 12 above), it compares
ohci->prev_frame_no to IO_WATCHDOG_OFF so that ohci->prev_frame_no is
not overwritten while io_watchdog_func() is running.
Signed-off-by: Shigeru Yoshida <Shigeru.Yoshida@windriver.com>
Signed-off-by: Haiqing Bai <Haiqing.Bai@windriver.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
drivers/usb/host/ohci-hcd.c | 10 +++++++---
drivers/usb/host/ohci-hub.c | 4 +++-
2 files changed, 10 insertions(+), 4 deletions(-)
--- a/drivers/usb/host/ohci-hcd.c
+++ b/drivers/usb/host/ohci-hcd.c
@@ -74,6 +74,7 @@ static const char hcd_name [] = "ohci_hc
#define STATECHANGE_DELAY msecs_to_jiffies(300)
#define IO_WATCHDOG_DELAY msecs_to_jiffies(275)
+#define IO_WATCHDOG_OFF 0xffffff00
#include "ohci.h"
#include "pci-quirks.h"
@@ -231,7 +232,7 @@ static int ohci_urb_enqueue (
}
/* Start up the I/O watchdog timer, if it's not running */
- if (!timer_pending(&ohci->io_watchdog) &&
+ if (ohci->prev_frame_no == IO_WATCHDOG_OFF &&
list_empty(&ohci->eds_in_use) &&
!(ohci->flags & OHCI_QUIRK_QEMU)) {
ohci->prev_frame_no = ohci_frame_no(ohci);
@@ -501,6 +502,7 @@ static int ohci_init (struct ohci_hcd *o
return 0;
timer_setup(&ohci->io_watchdog, io_watchdog_func, 0);
+ ohci->prev_frame_no = IO_WATCHDOG_OFF;
ohci->hcca = dma_alloc_coherent (hcd->self.controller,
sizeof(*ohci->hcca), &ohci->hcca_dma, GFP_KERNEL);
@@ -730,7 +732,7 @@ static void io_watchdog_func(struct time
u32 head;
struct ed *ed;
struct td *td, *td_start, *td_next;
- unsigned frame_no;
+ unsigned frame_no, prev_frame_no = IO_WATCHDOG_OFF;
unsigned long flags;
spin_lock_irqsave(&ohci->lock, flags);
@@ -835,7 +837,7 @@ static void io_watchdog_func(struct time
}
}
if (!list_empty(&ohci->eds_in_use)) {
- ohci->prev_frame_no = frame_no;
+ prev_frame_no = frame_no;
ohci->prev_wdh_cnt = ohci->wdh_cnt;
ohci->prev_donehead = ohci_readl(ohci,
&ohci->regs->donehead);
@@ -845,6 +847,7 @@ static void io_watchdog_func(struct time
}
done:
+ ohci->prev_frame_no = prev_frame_no;
spin_unlock_irqrestore(&ohci->lock, flags);
}
@@ -973,6 +976,7 @@ static void ohci_stop (struct usb_hcd *h
if (quirk_nec(ohci))
flush_work(&ohci->nec_work);
del_timer_sync(&ohci->io_watchdog);
+ ohci->prev_frame_no = IO_WATCHDOG_OFF;
ohci_writel (ohci, OHCI_INTR_MIE, &ohci->regs->intrdisable);
ohci_usb_reset(ohci);
--- a/drivers/usb/host/ohci-hub.c
+++ b/drivers/usb/host/ohci-hub.c
@@ -311,8 +311,10 @@ static int ohci_bus_suspend (struct usb_
rc = ohci_rh_suspend (ohci, 0);
spin_unlock_irq (&ohci->lock);
- if (rc == 0)
+ if (rc == 0) {
del_timer_sync(&ohci->io_watchdog);
+ ohci->prev_frame_no = IO_WATCHDOG_OFF;
+ }
return rc;
}
next prev parent reply other threads:[~2018-02-26 20:22 UTC|newest]
Thread overview: 74+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-02-26 20:21 [PATCH 4.15 00/64] 4.15.7-stable review Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 01/64] netfilter: drop outermost socket lock in getsockopt() Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 02/64] arm64: mm: dont write garbage into TTBR1_EL1 register Greg Kroah-Hartman
2018-02-26 20:21 ` Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 03/64] kconfig.h: Include compiler types to avoid missed struct attributes Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 04/64] MIPS: boot: Define __ASSEMBLY__ for its.S build Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 05/64] xtensa: fix high memory/reserved memory collision Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 06/64] scsi: ibmvfc: fix misdefined reserved field in ibmvfc_fcp_rsp_info Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 07/64] MIPS: Drop spurious __unused in struct compat_flock Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 08/64] cfg80211: fix cfg80211_beacon_dup Greg Kroah-Hartman
2018-02-26 20:21 ` Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 09/64] i2c: designware: must wait for enable Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 10/64] i2c: bcm2835: Set up the rising/falling edge delays Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 11/64] X.509: fix BUG_ON() when hash algorithm is unsupported Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 12/64] X.509: fix NULL dereference when restricting key with unsupported_sig Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 13/64] PKCS#7: fix certificate chain verification Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 14/64] PKCS#7: fix certificate blacklisting Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 15/64] extcon: int3496: process id-pin first so that we start with the right status Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 16/64] genirq/matrix: Handle CPU offlining proper Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 17/64] RDMA/uverbs: Protect from races between lookup and destroy of uobjects Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 18/64] RDMA/uverbs: Protect from command mask overflow Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 19/64] RDMA/uverbs: Fix bad unlock balance in ib_uverbs_close_xrcd Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 20/64] RDMA/uverbs: Fix circular locking dependency Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 21/64] RDMA/uverbs: Sanitize user entered port numbers prior to access it Greg Kroah-Hartman
2018-02-26 20:21 ` [PATCH 4.15 22/64] iio: adc: stm32: fix stm32h7_adc_enable error handling Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 23/64] iio: srf08: fix link error "devm_iio_triggered_buffer_setup" undefined Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 24/64] iio: buffer: check if a buffer has been set up when poll is called Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 25/64] iio: adis_lib: Initialize trigger before requesting interrupt Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 26/64] Kbuild: always define endianess in kconfig.h Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 27/64] x86/apic/vector: Handle vector release on CPU unplug correctly Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 28/64] x86/oprofile: Fix bogus GCC-8 warning in nmi_setup() Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 29/64] mm, swap, frontswap: fix THP swap if frontswap enabled Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 30/64] mm: dont defer struct page initialization for Xen pv guests Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 31/64] uapi/if_ether.h: move __UAPI_DEF_ETHHDR libc define Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 32/64] irqchip/gic-v3: Use wmb() instead of smb_wmb() in gic_raise_softirq() Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 33/64] irqchip/mips-gic: Avoid spuriously handling masked interrupts Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 34/64] PCI/cxgb4: Extend T3 PCI quirk to T4+ devices Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 35/64] net: thunderbolt: Tear down connection properly on suspend Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 36/64] net: thunderbolt: Run disconnect flow asynchronously when logout is received Greg Kroah-Hartman
2018-02-26 20:22 ` Greg Kroah-Hartman [this message]
2018-02-26 20:22 ` [PATCH 4.15 38/64] usb: ohci: Proper handling of ed_rm_list to handle race condition between usb_kill_urb() and finish_unlinks() Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 39/64] arm64: Remove unimplemented syscall log message Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 40/64] arm64: Disable unhandled signal log messages by default Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 41/64] arm64: cpufeature: Fix CTR_EL0 field definitions Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 42/64] Add delay-init quirk for Corsair K70 RGB keyboards Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 43/64] drm/edid: Add 6 bpc quirk for CPT panel in Asus UX303LA Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 44/64] usb: host: ehci: use correct device pointer for dma ops Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 45/64] usb: dwc3: gadget: Set maxpacket size for ep0 IN Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 46/64] usb: dwc3: ep0: Reset TRB counter " Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 47/64] usb: phy: mxs: Fix NULL pointer dereference on i.MX23/28 Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 48/64] usb: ldusb: add PIDs for new CASSY devices supported by this driver Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 49/64] Revert "usb: musb: host: dont start next rx urb if current one failed" Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 50/64] usb: gadget: f_fs: Process all descriptors during bind Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 51/64] usb: gadget: f_fs: Use config_ep_by_speed() Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 52/64] usb: renesas_usbhs: missed the "running" flag in usb_dmac with rx path Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 53/64] drm/cirrus: Load lut in crtc_commit Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 54/64] drm/atomic: Fix memleak on ERESTARTSYS during non-blocking commits Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 55/64] drm: Handle unexpected holes in color-eviction Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 56/64] drm/amdgpu: disable MMHUB power gating on raven Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 57/64] drm/amdgpu: fix VA hole handling on Vega10 v3 Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 58/64] drm/amdgpu: Add dpm quirk for Jet PRO (v2) Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 59/64] drm/amdgpu: only check mmBIF_IOV_FUNC_IDENTIFIER on tonga/fiji Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 60/64] drm/amdgpu: add atpx quirk handling (v2) Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 61/64] drm/amdgpu: Avoid leaking PM domain on driver unbind (v2) Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 62/64] drm/amdgpu: add new device to use atpx quirk Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 63/64] arm64: __show_regs: Only resolve kernel symbols when running at EL1 Greg Kroah-Hartman
2018-02-26 20:22 ` [PATCH 4.15 64/64] drm/i915/breadcrumbs: Ignore unsubmitted signalers Greg Kroah-Hartman
2018-02-27 0:58 ` [PATCH 4.15 00/64] 4.15.7-stable review Shuah Khan
2018-02-27 13:09 ` Greg Kroah-Hartman
2018-02-27 7:07 ` Naresh Kamboju
2018-02-27 13:09 ` Greg Kroah-Hartman
2018-02-27 7:18 ` kernelci.org bot
2018-02-27 14:58 ` Guenter Roeck
2018-02-27 18:37 ` Greg Kroah-Hartman
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=20180226202155.043753482@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=Haiqing.Bai@windriver.com \
--cc=Shigeru.Yoshida@windriver.com \
--cc=linux-kernel@vger.kernel.org \
--cc=stable@vger.kernel.org \
--cc=stern@rowland.harvard.edu \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.