3.15.3-stable review
Subsystems (5)
Greg Kroah-Hartman
maintainercover 3.15.3-stable review
This is the start of the stable review cycle for the 3.15.3 release. There are 139 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.
Responses should be made by Mon Jun 30 17:45:39 UTC 2014. Anything received after that time might be too late.
The whole patch series can be found in one patch at:
kernel.org/pub/linux/kernel/v3.0/stable-review/patch-3.15.3-rc1.gz
and the diffstat can be found below.
thanks,
greg k-h
-------------
Pseudo-Shortlog of commits:
Greg Kroah-Hartman <[email protected]>
Linux 3.15.3-rc1
Andrzej Zaborowski <[email protected]>
efi-pstore: Fix an overflow on 32-bit builds
Fathi Boudra <[email protected]>
builddeb: use $OBJCOPY variable instead of objcopy
Theodore Ts'o <[email protected]>
random: fix nasty entropy accounting bug
Konstantin Khlebnikov <[email protected]>
epoll: fix use-after-free in eventpoll_release_file
Andy Lutomirski <[email protected]>
x86_32, entry: Do syscall exit work on badsys (CVE-2014-4508)
Greg Kroah-Hartman <[email protected]>
lz4: fix another possible overrun
Johan Hedberg <[email protected]>
Bluetooth: Fix properly ignoring LTKs of unknown types
Johan Hedberg <[email protected]>
Bluetooth: Clearly distinguish mgmt LTK type from authenticated property
Eric Sandeen <[email protected]>
btrfs: fix use of uninit "ret" in end_extent_writepage()
Liu Bo <[email protected]>
Btrfs: fix scrub_print_warning to handle skinny metadata extents
Liu Bo <[email protected]>
Btrfs: use right type to get real comparison
Josef Bacik <[email protected]>
Btrfs: don't check nodes for extent items
Rickard Strandqvist <[email protected]>
fs: btrfs: volumes.c: Fix for possible null pointer dereference
Jeff Mahoney <[email protected]>
btrfs: allocate raid type kobjects dynamically
Filipe Manana <[email protected]>
Btrfs: send, use the right limits for xattr names and values
Filipe Manana <[email protected]>
Btrfs: send, don't error in the presence of subvols/snapshots
Wang Shilong <[email protected]>
Btrfs: set right total device count for seeding support
Liu Bo <[email protected]>
Btrfs: mark mapping with error flag to report errors to userspace
Liu Bo <[email protected]>
Btrfs: fix NULL pointer crash of deleting a seed device
Wang Shilong <[email protected]>
Btrfs: make sure there are not any read requests before stopping workers
Filipe Manana <[email protected]>
Btrfs: send, account for orphan directories when building path strings
Miao Xie <[email protected]>
Btrfs: output warning instead of error when loading free space cache failed
Qu Wenruo <[email protected]>
btrfs: Add ctime/mtime update for btrfs device add/remove.
Filipe Manana <[email protected]>
Btrfs: read inode size after acquiring the mutex when punching a hole
Chris Mason <[email protected]>
Btrfs: fix double free in find_lock_delalloc_range
Filipe Manana <[email protected]>
Btrfs: fix leaf corruption caused by ENOSPC while hole punching
Pavel Shilovsky <[email protected]>
CIFS: Fix memory leaks in SMB2_open
Benjamin LaHaise <[email protected]>
aio: fix kernel memory disclosure in io_getevents() introduced in v3.10
Benjamin LaHaise <[email protected]>
aio: fix aio request leak when events are reaped by userspace
Thomas Gleixner <[email protected]>
genirq: Sanitize spurious interrupt detection of threaded irqs
Benjamin Herrenschmidt <[email protected]>
Revert "offb: Add palette hack for little endian"
Alex Deucher <[email protected]>
Revert "drm/radeon: use variable UVD clocks"
Mike Frysinger <[email protected]>
x86, x32: Use compat shims for io_{setup,submit}H. Peter Anvin <[email protected]>
x86-32, espfix: Remove filter for espfix32 due to race
Will Deacon <[email protected]>
arm64: mm: remove broken &= operator from pmd_mknotpresent
Suravee Suthikulpanit <[email protected]>
arm64/dma: Removing ARCH_HAS_DMA_GET_REQUIRED_MASK macro
Will Deacon <[email protected]>
arm64: uid16: fix __kernel_old_{gid,uid}_t definitionsJason Cooper <[email protected]>
ARM: mvebu: DT: fix OpenBlocks AX3-4 RAM size
Jaegeuk Kim <[email protected]>
f2fs: submit bio at the reclaim path
Sagi Grimberg <[email protected]>
TARGET/sbc,loopback: Adjust command data length in case pi exists on the wire
Sagi Grimberg <[email protected]>
libiscsi, iser: Adjust data_length to include protection information
Sagi Grimberg <[email protected]>
scsi_cmnd: Introduce scsi_transfer_length helper
Roland Dreier <[email protected]>
target: Report correct response length for some commands
Sagi Grimberg <[email protected]>
Target/iscsi: Fix sendtargets response pdu for iser transport
Sagi Grimberg <[email protected]>
Target/iser: Fix a wrong dereference in case discovery session is over iser
Nicholas Bellinger <[email protected]>
iscsi-target: Fix ABORT_TASK + connection reset iscsi_queue_req memory leak
Nicholas Bellinger <[email protected]>
target: Use complete_all for se_cmd->t_transport_stop_comp
Nicholas Bellinger <[email protected]>
target: Set CMD_T_ACTIVE bit for Task Management Requests
Sagi Grimberg <[email protected]>
Target/iser: Gracefully reject T10-PI enabled connect request if not supported
Sagi Grimberg <[email protected]>
Target/iser: Wait for proper cleanup before unloading
Sagi Grimberg <[email protected]>
Target/iser: Improve cm events handling
Sagi Grimberg <[email protected]>
Target/iser: Fix hangs in connection teardown
Sagi Grimberg <[email protected]>
Target/iser: Bail from accept_np if np_thread is trying to close
Johan Hedberg <[email protected]>
Bluetooth: Fix missing check for FIPS security level
Johan Hedberg <[email protected]>
Bluetooth: Fix requiring SMP MITM for outgoing connections
Johan Hedberg <[email protected]>
Bluetooth: Fix authentication check for FIPS security level
Jukka Taimisto <[email protected]>
Bluetooth: Fix L2CAP deadlock
Jukka Rissanen <[email protected]>
Bluetooth: 6LoWPAN: Fix MAC address universal/local bit handling
Felipe Balbi <[email protected]>
bluetooth: hci_ldisc: fix deadlock condition
Chander Kashyap <[email protected]>
PM / OPP: fix incorrect OPP count handling in of_init_opp_table
Pekon Gupta <[email protected]>
ARM: OMAP2+: gpmc: enable BCH_HW ecc-scheme for AM43xx platforms
Jianguo Wu <[email protected]>
ARM: 8037/1: mm: support big-endian page tables
Russell King <[email protected]>
ARM: stacktrace: avoid listing stacktrace functions in stacktrace
Hans Verkuil <[email protected]>
media: saa7134: fix regression with tvtime
Pali Rohár <[email protected]>
media: radio-bcm2048: fix wrong overflow check
Olivier Langlois <[email protected]>
media: uvcvideo: Fix clock param realtime setting
Thomas Gleixner <[email protected]>
rtmutex: Plug slow unlock race
Thomas Gleixner <[email protected]>
rtmutex: Handle deadlock detection smarter
Thomas Gleixner <[email protected]>
rtmutex: Detect changes in the pi lock chain
Rafael J. Wysocki <[email protected]>
ACPI / ia64 / sba_iommu: Restore the working initialization ordering
Rafael J. Wysocki <[email protected]>
ACPI / hotplug / PCI: Add hotplug contexts to PCI host bridges
Lv Zheng <[email protected]>
ACPI: Fix conflict between customized DSDT and DSDT local copy
David Binderman <[email protected]>
ACPICA: utstring: Check array index bound before use.
Bjørn Mork <[email protected]>
ACPI: add dynamic_debug support
Sylwester Nawrocki <[email protected]>
media: exynos4-is: Fix compilation for !CONFIG_COMMON_CLK
Sylwester Nawrocki <[email protected]>
media: exynos4-is: Free FIMC-IS CPU memory only when allocated
Ezequiel Garcia <[email protected]>
media: stk1160: Avoid stack-allocated buffer for control URBs
Takashi Iwai <[email protected]>
media: ivtv: Fix Oops when no firmware is loaded
Johan Hovold <[email protected]>
USB: serial: fix potential runtime pm imbalance at device remove
Aleksander Morgado <[email protected]>
usb: qcserial: add additional Sierra Wireless QMI devices
Aleksander Morgado <[email protected]>
usb: qcserial: add Netgear AirCard 341U
Johan Hovold <[email protected]>
USB: sierra: fix remote wakeup
Johan Hovold <[email protected]>
USB: sierra: fix urb and memory leak on disconnect
Johan Hovold <[email protected]>
USB: sierra: fix urb and memory leak in resume error path
Johan Hovold <[email protected]>
USB: sierra: fix use after free at suspend/resume
Johan Hovold <[email protected]>
USB: sierra: fix AA deadlock in open error path
Johan Hovold <[email protected]>
USB: usb_wwan: fix potential blocked I/O after resume
Johan Hovold <[email protected]>
USB: usb_wwan: fix potential NULL-deref at resume
Johan Hovold <[email protected]>
USB: usb_wwan: fix urb leak at shutdown
Johan Hovold <[email protected]>
USB: usb_wwan: fix write and suspend race
xiao jin <[email protected]>
USB: usb_wwan: fix race between write and resume
xiao jin <[email protected]>
USB: usb_wwan: fix urb leak in write error path
Mikulas Patocka <[email protected]>
matroxfb: perform a dummy read of M_STATUS
Maurizio Lombardi <[email protected]>
ext4: fix wrong assert in ext4_mb_normalize_request()
Namjae Jeon <[email protected]>
ext4: fix ZERO_RANGE test failure in data journalling
Jan Kara <[email protected]>
ext4: fix zeroing of page during writeback
Namjae Jeon <[email protected]>
ext4: fix data integrity sync in ordered mode
Krzysztof Kozlowski <[email protected]>
regulator: s2mpa01: Fix accidental enable of buck4 ramp delay
Krzysztof Kozlowski <[email protected]>
regulator: s2mps11: Fix accidental enable of buck6 ramp delay
Krzysztof Kozlowski <[email protected]>
regulator: s2mpa01: Use correct register for buck1 ramp delay
Christian Borntraeger <[email protected]>
s390/lowcore: reserve 96 bytes for IRB in lowcore
Martin Schwidefsky <[email protected]>
s390/time: cast tv_nsec to u64 prior to shift in update_vsyscall
Lai Jiangshan <[email protected]>
idr: fix overflow bug during maximum ID calculation at maximum height
Victor Kamensky <[email protected]>
arm64: ptrace: fix empty registers set in prstatus of aarch32 process core
Will Deacon <[email protected]>
arm64: ptrace: change fs when passing kernel pointer to regset code
Matthew Dempsky <[email protected]>
ptrace: fix fork event messages across pid namespaces
Johannes Weiner <[email protected]>
mm: vmscan: clear kswapd's special reclaim powers before exiting
Kees Cook <[email protected]>
HID: core: fix validation of report id 0
Hugh Dickins <[email protected]>
mm: fix sleeping function warning from __put_anon_vma
Weijie Yang <[email protected]>
zram: correct offset usage in zram_bio_discard
Naoya Horiguchi <[email protected]>
mm/memory-failure.c: support use of a dedicated thread to handle SIGBUS(BUS_MCEERR_AO)
Tony Luck <[email protected]>
mm/memory-failure.c: don't let collect_procs() skip over processes for MF_ACTION_REQUIRED
Tony Luck <[email protected]>
mm/memory-failure.c-failure: send right signal code to correct thread
Mel Gorman <[email protected]>
mm: page_alloc: use word-based accesses for get/set pageblock bitmaps
Michal Hocko <[email protected]>
memcg: do not hang on OOM when killed by userspace OOM access to memory reserves
Mel Gorman <[email protected]>
mm: vmscan: do not throttle based on pfmemalloc reserves if node has no ZONE_NORMAL
Tetsuo Handa <[email protected]>
kthread: fix return value of kthread_create() upon SIGKILL.
Naoya Horiguchi <[email protected]>
hugetlb: restrict hugepage_migration_support() to x86_64
Konstantin Khlebnikov <[email protected]>
tools/vm/page-types.c: catch sigbus if raced with truncate
Johan Hovold <[email protected]>
USB: option: fix runtime PM handling
Alan Stern <[email protected]>
USB: EHCI: avoid BIOS handover on the HASEE E200
Paul Bolle <[email protected]>
ARM: OMAP: replace checks for CONFIG_USB_GADGET_OMAP
Felipe Balbi <[email protected]>
usb: dwc3: gadget: clear stall when disabling endpoint
Paul Bolle <[email protected]>
usb: gadget: rename CONFIG_USB_GADGET_PXA25X
Alan Stern <[email protected]>
USB: usbtest: add a timeout for scatter-gather tests
Huang Rui <[email protected]>
usb: usbtest: fix unlink write error with pattern 1
Dan Carpenter <[email protected]>
applicom: dereferencing NULL on error path
Stephen Boyd <[email protected]>
staging/mt29f_spinand: Terminate of match table
Dan Carpenter <[email protected]>
Staging: rtl8188eu: overflow in update_sta_support_rate()
Paul Bolle <[email protected]>
staging: tidspbridge: check for CONFIG_SND_OMAP_SOC_MCBSP
Antoine Ténart <[email protected]>
phy: exynos-mipi-video: fix check on array index
Stephen Chivers <[email protected]>
printk/of_serial: fix serial console cessation part way through boot.
Alexey Khoroshilov <[email protected]>
w1: do not unlock unheld list_mutex in __w1_remove_master_device()
Krzysztof Kozlowski <[email protected]>
extcon: max14577: Properly handle regmap_irq_get_virq error
Krzysztof Kozlowski <[email protected]>
extcon: max14577: Fix probe failure on successful work queue
Krzysztof Kozlowski <[email protected]>
extcon: max77693: Fix two NULL pointer exceptions on missing pdata
Krzysztof Kozlowski <[email protected]>
extcon: max8997: Fix NULL pointer exception on missing pdata
Ming Lei <[email protected]>
block: virtio_blk: don't hold spin lock during world switch
Russell King <[email protected]>
imx-drm: fix hdmi hotplug detection initial state
-------------
Diffstat:
Documentation/vm/hwpoison.txt | 5 + Makefile | 4 +- arch/arm/boot/dts/armada-xp-openblocks-ax3-4.dts | 2 +- arch/arm/kernel/stacktrace.c | 18 +- arch/arm/mach-omap1/board-h2.c | 2 +- arch/arm/mach-omap1/board-h3.c | 2 +- arch/arm/mach-omap1/board-innovator.c | 2 +- arch/arm/mach-omap1/board-osk.c | 2 +- arch/arm/mach-omap2/gpmc-nand.c | 2 +- arch/arm/mm/hugetlbpage.c | 5 - arch/arm/mm/proc-v7-3level.S | 18 +- arch/arm64/include/asm/Kbuild | 1 - arch/arm64/include/asm/dma-mapping.h | 2 - arch/arm64/include/asm/pgtable.h | 2 +- arch/arm64/include/uapi/asm/posix_types.h | 10 + arch/arm64/kernel/ptrace.c | 32 ++- arch/arm64/mm/hugetlbpage.c | 5 - arch/ia64/hp/common/sba_iommu.c | 64 +++--- arch/ia64/mm/hugetlbpage.c | 5 - arch/metag/mm/hugetlbpage.c | 5 - arch/mips/mm/hugetlbpage.c | 5 - arch/powerpc/mm/hugetlbpage.c | 10 - arch/s390/include/asm/lowcore.h | 11 +- arch/s390/kernel/time.c | 2 +- arch/s390/mm/hugetlbpage.c | 5 - arch/sh/mm/hugetlbpage.c | 5 - arch/sparc/mm/hugetlbpage.c | 5 - arch/tile/mm/hugetlbpage.c | 5 - arch/x86/Kconfig | 4 + arch/x86/kernel/entry_32.S | 15 +- arch/x86/mm/hugetlbpage.c | 10 - arch/x86/syscalls/syscall_64.tbl | 6 +- drivers/acpi/acpica/utstring.c | 2 +- drivers/acpi/bus.c | 7 + drivers/acpi/utils.c | 64 ++++-- drivers/base/power/opp.c | 4 +- drivers/block/virtio_blk.c | 9 +- drivers/block/zram/zram_drv.c | 4 +- drivers/bluetooth/hci_ldisc.c | 24 ++- drivers/bluetooth/hci_uart.h | 1 + drivers/char/applicom.c | 1 - drivers/char/random.c | 17 +- drivers/extcon/extcon-max14577.c | 9 +- drivers/extcon/extcon-max77693.c | 4 +- drivers/extcon/extcon-max8997.c | 2 +- drivers/firmware/efi/efi-pstore.c | 2 +- drivers/gpu/drm/radeon/radeon_pm.c | 3 + drivers/gpu/drm/radeon/radeon_uvd.c | 3 +- drivers/hid/hid-core.c | 12 +- drivers/infiniband/ulp/iser/iser_initiator.c | 34 +--- drivers/infiniband/ulp/isert/ib_isert.c | 70 +++---- drivers/infiniband/ulp/isert/ib_isert.h | 2 +- drivers/media/pci/ivtv/ivtv-alsa-pcm.c | 6 + drivers/media/pci/saa7134/saa7134-video.c | 9 +- drivers/media/platform/exynos4-is/fimc-is.c | 3 + drivers/media/platform/exynos4-is/media-dev.c | 2 +- drivers/media/platform/exynos4-is/media-dev.h | 4 + drivers/media/usb/stk1160/stk1160-core.c | 10 +- drivers/media/usb/stk1160/stk1160.h | 1 - drivers/media/usb/uvc/uvc_video.c | 15 +- drivers/pci/hotplug/acpiphp.h | 10 + drivers/pci/hotplug/acpiphp_glue.c | 60 ++++-- drivers/phy/phy-exynos-mipi-video.c | 2 +- drivers/regulator/s2mpa01.c | 15 +- drivers/regulator/s2mps11.c | 15 +- drivers/scsi/libiscsi.c | 18 +- drivers/staging/imx-drm/imx-hdmi.c | 9 +- drivers/staging/media/bcm2048/radio-bcm2048.c | 2 +- drivers/staging/mt29f_spinand/mt29f_spinand.c | 1 + drivers/staging/rtl8188eu/core/rtw_wlan_util.c | 7 +- drivers/staging/tidspbridge/core/dsp-clock.c | 4 +- drivers/target/iscsi/iscsi_target.c | 17 +- drivers/target/loopback/tcm_loop.c | 15 +- drivers/target/target_core_sbc.c | 19 +- drivers/target/target_core_spc.c | 9 +- drivers/target/target_core_transport.c | 29 ++- drivers/tty/serial/of_serial.c | 1 + drivers/usb/dwc3/gadget.c | 4 + drivers/usb/gadget/inode.c | 2 +- drivers/usb/host/pci-quirks.c | 19 +- drivers/usb/misc/usbtest.c | 26 ++- drivers/usb/phy/phy-isp1301-omap.c | 2 +- drivers/usb/serial/bus.c | 14 +- drivers/usb/serial/option.c | 11 +- drivers/usb/serial/qcserial.c | 21 ++ drivers/usb/serial/sierra.c | 55 +++-- drivers/usb/serial/usb_wwan.c | 125 ++++++------ drivers/video/fbdev/matrox/matroxfb_base.h | 2 +- drivers/video/fbdev/offb.c | 11 +- drivers/w1/w1.c | 2 + drivers/w1/w1_int.c | 4 + fs/aio.c | 6 +- fs/btrfs/backref.c | 37 ++-- fs/btrfs/backref.h | 4 +- fs/btrfs/ctree.h | 8 +- fs/btrfs/disk-io.c | 5 + fs/btrfs/extent-tree.c | 39 ++-- fs/btrfs/extent_io.c | 5 +- fs/btrfs/file.c | 23 ++- fs/btrfs/free-space-cache.c | 4 +- fs/btrfs/scrub.c | 5 +- fs/btrfs/send.c | 67 ++++--- fs/btrfs/sysfs.c | 5 +- fs/btrfs/volumes.c | 44 +++- fs/cifs/smb2pdu.c | 7 +- fs/eventpoll.c | 4 +- fs/ext4/ext4.h | 3 +- fs/ext4/extents.c | 7 + fs/ext4/inode.c | 6 +- fs/ext4/mballoc.c | 2 +- fs/ext4/page-io.c | 32 +-- fs/f2fs/data.c | 2 + include/linux/acpi.h | 22 +- include/linux/hugetlb.h | 13 +- include/linux/irqdesc.h | 4 + include/linux/mempolicy.h | 6 + include/linux/mmzone.h | 6 +- include/linux/page-flags.h | 12 +- include/linux/pageblock-flags.h | 37 +++- include/linux/ptrace.h | 32 +++ include/net/bluetooth/mgmt.h | 3 + include/scsi/scsi_cmnd.h | 17 ++ include/target/iscsi/iscsi_transport.h | 3 +- include/target/target_core_backend.h | 1 + kernel/fork.c | 10 +- kernel/irq/manage.c | 4 +- kernel/irq/spurious.c | 106 +++++++++- kernel/kthread.c | 4 +- kernel/locking/rtmutex-debug.h | 5 + kernel/locking/rtmutex.c | 243 +++++++++++++++++++---- kernel/locking/rtmutex.h | 5 + kernel/printk/printk.c | 1 + lib/idr.c | 8 +- lib/lz4/lz4_decompress.c | 4 +- mm/Kconfig | 3 + mm/memcontrol.c | 3 +- mm/memory-failure.c | 73 +++++-- mm/page-writeback.c | 11 +- mm/page_alloc.c | 52 +++-- mm/rmap.c | 8 +- mm/vmscan.c | 46 ++++- net/bluetooth/6lowpan.c | 65 +++--- net/bluetooth/hci_event.c | 4 +- net/bluetooth/l2cap_sock.c | 5 +- net/bluetooth/mgmt.c | 25 ++- net/bluetooth/smp.c | 7 +- scripts/package/builddeb | 6 +- tools/vm/page-types.c | 35 +++- 148 files changed, 1638 insertions(+), 692 deletions(-)
Greg Kroah-Hartman
maintainerPATCH 1/139 imx-drm: fix hdmi hotplug detection initial state
3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Russell King <[email protected]>
commit 98dbeadaf050335df8655d8d9be7a324b6cd896e upstream.
The initial state at boot is assumed to be disconnected, and we hope to receive an interrupt to update the status. Let's be more explicit about the current state - reading the PHY status register tells us the current level of the hotplug signal, which we can report back in the _detect() method.
--- drivers/staging/imx-drm/imx-hdmi.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-)
drivers/staging/imx-drm/imx-hdmi.c +3 -6
@@ -120,8 +120,6 @@ struct imx_hdmi {struct clk *isfr_clk;struct clk *iahb_clk; -enum drm_connector_status connector_status; -struct hdmi_data_info hdmi_data;int vic;
@@ -1382,7 +1380,9 @@ static enum drm_connector_status imx_hdm { struct imx_hdmi *hdmi = container_of(connector, struct imx_hdmi, connector); -return hdmi->connector_status; + +return hdmi_readb (hdmi,HDMI_PHY_STAT0 ) &HDMI_PHY_HPD ? + connector_status_connected : connector_status_disconnected; }static int imx_hdmi_connector_get_modes (struct drm_connector *connector)
@@ -1524,7 +1524,6 @@ static irqreturn_t imx_hdmi_irq(int irq,hdmi_modb (hdmi,0 ,HDMI_PHY_HPD ,HDMI_PHY_POL0 ); - hdmi->connector_status = connector_status_connected;imx_hdmi_poweron (hdmi); }else {dev_dbg (hdmi->dev,"EVENT=plugout\n" );
@@ -1532,7 +1531,6 @@ static irqreturn_t imx_hdmi_irq(int irq,hdmi_modb (hdmi,HDMI_PHY_HPD ,HDMI_PHY_HPD ,HDMI_PHY_POL0 ); - hdmi->connector_status = connector_status_disconnected;imx_hdmi_poweroff (hdmi); }drm_helper_hpd_irq_event (hdmi->connector.dev);
@@ -1606,7 +1604,6 @@ static int imx_hdmi_bind(struct device *return -ENOMEM ; hdmi->dev = dev; - hdmi->connector_status = connector_status_disconnected; hdmi->sample_rate =48000 ; hdmi->ratio =100 ;
Greg Kroah-Hartman
maintainerPATCH 2/139 block: virtio_blk: dont hold spin lock during world switch
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Ming Lei <[email protected]>
commit e8edca6f7f92234202d6dd163c118ef495244d7c upstream.
Firstly, it isn't necessary to hold lock of vblk->vq_lock when notifying hypervisor about queued I/O.
Secondly, virtqueue_notify() will cause world switch and it may take long time on some hypervisors(such as, qemu-arm), so it isn't good to hold the lock and block other vCPUs.
On arm64 quad core VM(qemu-kvm), the patch can increase I/O performance a lot with VIRTIO_RING_F_EVENT_IDX enabled:
- without the patch: 14K IOPS
- with the patch: 34K IOPS
fio script:
[global] direct=1 bsrange=4k-4k timeout=10 numjobs=4 ioengine=libaio iodepth=64 filename=/dev/vdc group_reporting=1 [f1] rw=randread
--- drivers/block/virtio_blk.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/block/virtio_blk.c +6 -3
@@ -159,6 +159,7 @@ static int virtio_queue_rq(struct blk_mqunsigned int num;const bool last = (req->cmd_flags &REQ_END ) !=0 ;int err; +bool notify = false;BUG_ON (req->nr_phys_segments +2 > vblk->sg_elems);
@@ -211,10 +212,12 @@ static int virtio_queue_rq(struct blk_mqreturn BLK_MQ_RQ_QUEUE_ERROR ; } -if (last) -virtqueue_kick (vblk->vq); - +if (last &&virtqueue_kick_prepare (vblk->vq)) + notify = true;spin_unlock_irqrestore (&vblk->vq_lock, flags); + +if (notify) +virtqueue_notify (vblk->vq);return BLK_MQ_RQ_QUEUE_OK ; }
Greg Kroah-Hartman
maintainerPATCH 3/139 extcon: max8997: Fix NULL pointer exception on missing pdata
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Krzysztof Kozlowski <[email protected]>
commit dfee4111febf3d9ef3a640b2cd6205c75f4e7e3d upstream.
Fix NULL pointer exception when platform data is not supplied. The driver dereferenced pdata pointer where it could be NULL.
--- drivers/extcon/extcon-max8997.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/extcon/extcon-max8997.c +1 -1
@@ -715,7 +715,7 @@ static int max8997_muic_probe(struct pla goto err_irq; } -if (pdata->muic_pdata) { +if (pdata && pdata->muic_pdata) {struct max8997_muic_platform_data *muic_pdata = pdata->muic_pdata;
Greg Kroah-Hartman
maintainerPATCH 4/139 extcon: max77693: Fix two NULL pointer exceptions on missing pdata
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Krzysztof Kozlowski <[email protected]>
commit d5653f2b7304f05eeb45d84f123cf02f840b8537 upstream.
Fix NULL pointer exceptions when platform data is not supplied.
Trace of one exception:
Unable to handle kernel NULL pointer dereference at virtual address 00000008 pgd = c0004000 [00000008] *pgd=00000000 Internal error: Oops: 5 [#1] PREEMPT SMP ARM Modules linked in: CPU: 2 PID: 1 Comm: swapper/0 Not tainted 3.14.0-12045-gead5dd4687a6-dirty #1628 task: eea80000 ti: eea88000 task.ti: eea88000 PC is at max77693_muic_probe+0x27c/0x528 LR is at regmap_write+0x50/0x60 pc : [<c041d1c8>] lr : [<c02eba60>] psr: 20000113
sp : eea89e38 ip : 00000000 fp : c098a834
r10: ee1a5a10 r9 : 00000005 r8 : c098a83c
r7 : 0000000a r6 : c098a774 r5 : 00000005 r4 : eeb006d0 r3 : c0697bd8 r2 : 00000000 r1 : 00000001 r0 : 00000000
Flags: nzCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment kernel
Control: 10c5387d Table: 4000404a DAC: 00000015 Process swapper/0 (pid: 1, stack limit = 0xeea88240)
Stack: (0xeea89e38 to 0xeea8a000)
9e20: c08499fc eeb006d0
9e40: 00000000 00000000 c0915f98 00000001 00000000 ee1a5a10 c098a730 c09a88b8 9e60: 00000000 c098a730 c0915f98 00000000 00000000 c02d6aa0 c02d6a88 ee1a5a10 9e80: c0a712c8 c02d54e4 00001204 c0628b00 ee1a5a10 c098a730 ee1a5a44 00000000 9ea0: eea88000 c02d57b4 00000000 c098a730 c02d5728 c02d3a24 ee813e5c eeb9d534 9ec0: c098a730 ee22f700 c097c720 c02d4b14 c08174ec c098a730 00000006 c098a730 9ee0: 00000006 c092fd30 c09b8500 c02d5df8 00000000 c093cbb8 00000006 c0008928 9f00: 000000c3 ef7fc785 00000000 ef7fc794 00000000 c08af968 00000072 eea89f30 9f20: ef7fc85e c065f198 000000c3 c003e87c 00000003 00000000 c092fd3c 00000000 9f40: c08af618 c0826d58 00000006 00000006 c0956f58 c093cbb8 00000006 c092fd30 9f60: c09b8500 000000c3 c092fd3c c08e8510 00000000 c08e8bb0 00000006 00000006 9f80: c08e8510 c0c0c0c0 00000000 c0628fac 00000000 00000000 00000000 00000000 9fa0: 00000000 c0628fb4 00000000 c000f038 00000000 00000000 00000000 00000000 9fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 9fe0: 00000000 00000000 00000000 00000000 00000013 00000000 c0c0c0c0 c0c0c0c0 [<c041d1c8>] (max77693_muic_probe) from [<c02d6aa0>] (platform_drv_probe+0x18/0x48) [<c02d6aa0>] (platform_drv_probe) from [<c02d54e4>] (driver_probe_device+0x140/0x384) [<c02d54e4>] (driver_probe_device) from [<c02d57b4>] (__driver_attach+0x8c/0x90) [<c02d57b4>] (__driver_attach) from [<c02d3a24>] (bus_for_each_dev+0x54/0x88) [<c02d3a24>] (bus_for_each_dev) from [<c02d4b14>] (bus_add_driver+0xe8/0x204) [<c02d4b14>] (bus_add_driver) from [<c02d5df8>] (driver_register+0x78/0xf4) [<c02d5df8>] (driver_register) from [<c0008928>] (do_one_initcall+0xc4/0x174) [<c0008928>] (do_one_initcall) from [<c08e8bb0>] (kernel_init_freeable+0xfc/0x1c8) [<c08e8bb0>] (kernel_init_freeable) from [<c0628fb4>] (kernel_init+0x8/0xec) [<c0628fb4>] (kernel_init) from [<c000f038>] (ret_from_fork+0x14/0x3c)
Code: caffffe7 e59d200c e3550001 b3a05001 (e5923008) ---[ end trace 85db969ce011bde7 ]---
--- drivers/extcon/extcon-max77693.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/extcon/extcon-max77693.c +2 -2
@@ -1193,7 +1193,7 @@ static int max77693_muic_probe(struct pl/* Initialize MUIC register by using platform data or default data */ -if (pdata->muic_data) { +if (pdata && pdata->muic_data) { init_data = pdata->muic_data->init_data; num_init_data = pdata->muic_data->num_init_data; }else {
@@ -1226,7 +1226,7 @@ static int max77693_muic_probe(struct pl = init_data[i].data; } -if (pdata->muic_data) { +if (pdata && pdata->muic_data) {struct max77693_muic_platform_data *muic_pdata = pdata->muic_data;
Greg Kroah-Hartman
maintainerPATCH 5/139 extcon: max14577: Fix probe failure on successful work queue
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Krzysztof Kozlowski <[email protected]>
commit 12adef5b49e98eb181b4163c36e2998169e1379b upstream.
In probe the driver queued delayed work for cable detection and returned the result of queue_delayed_work() call. However the return value of queue_delayed_work() does not indicate an error and in normal condition it returns true which means successful work queue. This effectively resulted in probe failure:
[ 2.088204] max14577-muic: probe of max77836-muic failed with error 1
--- drivers/extcon/extcon-max14577.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-)
drivers/extcon/extcon-max14577.c +1 -6
@@ -710,13 +710,8 @@ static int max14577_muic_probe(struct pl* driver should notify cable state to upper layer. */ INIT_DELAYED_WORK (&info->wq_detcable, max14577_muic_detect_cable_wq); - ret =queue_delayed_work (system_power_efficient_wq, &info->wq_detcable, +queue_delayed_work (system_power_efficient_wq, &info->wq_detcable, delay_jiffies); -if (ret <0 ) { -dev_err (&pdev->dev, -"failed to schedule delayed work for cable detect\n" ); - goto err_extcon; - }return ret;
Greg Kroah-Hartman
maintainerPATCH 6/139 extcon: max14577: Properly handle regmap_irq_get_virq error
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Krzysztof Kozlowski <[email protected]>
commit 369afd4ba22f5b8de0c9229b6e62b3f9e2207034 upstream.
The regmap_irq_get_virq may return 0 or -EINVAL on error. Fail the probe in both situations.
--- drivers/extcon/extcon-max14577.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/extcon/extcon-max14577.c +1 -1
@@ -650,7 +650,7 @@ static int max14577_muic_probe(struct plunsigned int virq =0 ; virq =regmap_irq_get_virq (max14577->irq_data, muic_irq->irq); -if (!virq) +if (virq <=0 )return -EINVAL ; muic_irq->virq = virq;
Greg Kroah-Hartman
maintainerPATCH 7/139 w1: do not unlock unheld list_mutex in __w1_remove_master_device()
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Alexey Khoroshilov <[email protected]>
commit a0f104644ec27ce5bbb36e950eb426dba9a3ad44 upstream.
w1_process_callbacks() expects to be called with dev->list_mutex held, but it is the fact only in w1_process(). __w1_remove_master_device() calls w1_process_callbacks() after it releases list_mutex.
The patch fixes __w1_remove_master_device() to acquire list_mutex for w1_process_callbacks().
Found by Linux Driver Verification project (linuxtesting.org).
--- drivers/w1/w1.c | 2 ++ drivers/w1/w1_int.c | 4 ++++ 2 files changed, 6 insertions(+)
drivers/w1/w1.c +2 -0
@@ -1078,6 +1078,8 @@ static void w1_search_process(struct w1_* w1_process_callbacks() - execute each dev->async_list callback entry * @dev: w1_master device * +* The w1 master list_mutex must be held. +* * Return: 1 if there were commands to executed 0 otherwise */ int w1_process_callbacks (struct w1_master *dev)
drivers/w1/w1_int.c +4 -0
@@ -219,9 +219,13 @@ void __w1_remove_master_device(struct w1if (msleep_interruptible (1000 ))flush_signals (current); +mutex_lock (&dev->list_mutex);w1_process_callbacks (dev); +mutex_unlock (&dev->list_mutex); } +mutex_lock (&dev->list_mutex);w1_process_callbacks (dev); +mutex_unlock (&dev->list_mutex);memset (&msg,0 ,sizeof (msg)); msg.id.mst.id = dev->id;
Greg Kroah-Hartman
maintainerPATCH 8/139 printk/of_serial: fix serial console cessation part way through boot.
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Stephen Chivers <[email protected]>
commit 7fa21dd8bd191564a195291161d6b43db5d9c350 upstream.
Commit 5f5c9ae56c38942623f69c3e6dc6ec78e4da2076 "serial_core: Unregister console in uart_remove_one_port()" fixed a crash where a serial port was removed but not deregistered as a console.
There is a side effect of that commit for platforms having serial consoles and of_serial configured (CONFIG_SERIAL_OF_PLATFORM). The serial console is disabled midway through the boot process.
This cessation of the serial console affects PowerPC computers such as the MVME5100 and SAM440EP.
The sequence is:
bootconsole [udbg0] enabled
....
serial8250/16550 driver initialises and registers its UARTS, one of these is the serial console.
console [ttyS0] enabled
....
of_serial probes "platform" devices, registering them as it goes. One of these is the serial console.
console [ttyS0] disabled.
The disabling of the serial console is due to:
a. unregister_console in printk not clearing the
CONS_ENABLED bit in the console flags,
even though it has announced that the console is disabled; and
b. of_platform_serial_probe in of_serial not setting the port type
before it registers with serial8250_register_8250_port.
This patch ensures that the serial console is re-enabled when of_serial registers a serial port that corresponds to the designated console.
===
The above failure was identified in Linux-3.15-rc2.
Tested using MVME5100 and SAM440EP PowerPC computers with kernels built from Linux-3.15-rc5 and tty-next.
The continued operation of the serial console is vital for computers such as the MVME5100 as that Single Board Computer does not have any grapical/display hardware.
--- drivers/tty/serial/of_serial.c | 1 + kernel/printk/printk.c | 1 + 2 files changed, 2 insertions(+)
drivers/tty/serial/of_serial.c +1 -0
@@ -173,6 +173,7 @@ static int of_platform_serial_probe(stru {struct uart_8250_port port8250;memset (&port8250,0 ,sizeof (port8250)); + port.type = port_type; port8250.port = port;if (port.fifosize)
kernel/printk/printk.c +1 -0
@@ -2413,6 +2413,7 @@ int unregister_console(struct console *cif (console_drivers !=NULL && console->flags &CON_CONSDEV ) console_drivers->flags |=CON_CONSDEV ; + console->flags &= ~CON_ENABLED ;console_unlock ();console_sysfs_notify ();return res;
Greg Kroah-Hartman
maintainerPATCH 10/139 staging: tidspbridge: check for CONFIG_SND_OMAP_SOC_MCBSP
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Paul Bolle <[email protected]>
commit d3921a03a89acb1b9ca599590c0131c89f8737d8 upstream.
Commit d0f47ff17f29 ("ASoC: OMAP: Build config cleanup for McBSP") removed the Kconfig symbol OMAP_MCBSP. It left two checks for CONFIG_OMAP_MCBSP untouched.
Convert these to checks for CONFIG_SND_OMAP_SOC_MCBSP. That must be correct, since that re-enables calls to functions that are all found in sound/soc/omap/mcbsp.c. And that file is built only if CONFIG_SND_OMAP_SOC_MCBSP is defined.
--- drivers/staging/tidspbridge/core/dsp-clock.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/staging/tidspbridge/core/dsp-clock.c +2 -2
@@ -226,7 +226,7 @@ int dsp_clk_enable(enum dsp_clk_id clk_icase GPT_CLK : status =omap_dm_timer_start (timer[clk_id -1 ]);break ; -#ifdef CONFIG_OMAP_MCBSP +#ifdef CONFIG_SND_OMAP_SOC_MCBSP case MCBSP_CLK :omap_mcbsp_request (MCBSP_ID (clk_id));omap2_mcbsp_set_clks_src (MCBSP_ID (clk_id),MCBSP_CLKS_PAD_SRC );
@@ -302,7 +302,7 @@ int dsp_clk_disable(enum dsp_clk_id clk_case GPT_CLK : status =omap_dm_timer_stop (timer[clk_id -1 ]);break ; -#ifdef CONFIG_OMAP_MCBSP +#ifdef CONFIG_SND_OMAP_SOC_MCBSP case MCBSP_CLK :omap2_mcbsp_set_clks_src (MCBSP_ID (clk_id),MCBSP_CLKS_PRCM_SRC );omap_mcbsp_free (MCBSP_ID (clk_id));
Greg Kroah-Hartman
maintainerPATCH 11/139 Staging: rtl8188eu: overflow in update_sta_support_rate()
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Dan Carpenter <[email protected]>
commit 9dbd79aeb9842144d9a114a979a12c0949ee11eb upstream.
The ->SupportedRates[] array has NDIS_802_11_LENGTH_RATES_EX (16) elements. Since "ie_len" comes from then network and can go up to 255 then it means we should add a range check to prevent memory corruption.
--- drivers/staging/rtl8188eu/core/rtw_wlan_util.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/staging/rtl8188eu/core/rtw_wlan_util.c +6 -1
@@ -1599,13 +1599,18 @@ int update_sta_support_rate(struct adapt pIE = (struct ndis_802_11_var_ie *)rtw_get_ie (pvar_ie, _SUPPORTEDRATES_IE_, &ie_len, var_ie_len);if (pIE ==NULL )return _FAIL; +if (ie_len >NDIS_802_11_LENGTH_RATES_EX ) +return _FAIL;memcpy (pmlmeinfo->FW_sta_info[cam_idx].SupportedRates, pIE->data, ie_len); supportRateNum = ie_len; pIE = (struct ndis_802_11_var_ie *)rtw_get_ie (pvar_ie, _EXT_SUPPORTEDRATES_IE_, &ie_len, var_ie_len); -if (pIE) +if (pIE) { +if (supportRateNum + ie_len >NDIS_802_11_LENGTH_RATES_EX ) +return _FAIL;memcpy ((pmlmeinfo->FW_sta_info[cam_idx].SupportedRates + supportRateNum), pIE->data, ie_len); + }return _SUCCESS; }
Greg Kroah-Hartman
maintainerPATCH 12/139 staging/mt29f_spinand: Terminate of match table
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Stephen Boyd <[email protected]>
commit ffd07de65ef5315053ea16356cd533b7f47c17e9 upstream.
Failure to terminate this match table can lead to boot failures depending on where the compiler places the match table.
--- drivers/staging/mt29f_spinand/mt29f_spinand.c | 1 + 1 file changed, 1 insertion(+)
drivers/staging/mt29f_spinand/mt29f_spinand.c +1 -0
@@ -924,6 +924,7 @@ static int spinand_remove(struct spi_devstatic const struct of_device_id spinand_dt[] = { { .compatible ="spinand,mt29f" , }, + {} };/*
Greg Kroah-Hartman
maintainerPATCH 13/139 applicom: dereferencing NULL on error path
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Dan Carpenter <[email protected]>
commit 8bab797c6e5724a43b7666ad70860712365cdb71 upstream.
This is a static checker fix. The "dev" variable is always NULL after the while statement so we would be dereferencing a NULL pointer here.
--- drivers/char/applicom.c | 1 - 1 file changed, 1 deletion(-)
drivers/char/applicom.c +0 -1
@@ -345,7 +345,6 @@ out:free_irq (apbs[i].irq, &dummy);iounmap (apbs[i].RamIO); } -pci_disable_device (dev);return ret; }
Greg Kroah-Hartman
maintainerPATCH 14/139 usb: usbtest: fix unlink write error with pattern 1
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Huang Rui <[email protected]>
commit e4d58f5dcb7d7be45df8def31881ebfae99c75da upstream.
TEST 12 and TEST 24 unlinks the URB write request for N times. When host and gadget both initialize pattern 1 (mod 63) data series to transfer, the gadget side will complain the wrong data which is not expected. Because in host side, usbtest doesn't fill the data buffer as mod 63 and this patch fixed it.
[20285.488974] dwc3 dwc3.0.auto: ep1out-bulk: Transfer Not Ready [20285.489181] dwc3 dwc3.0.auto: ep1out-bulk: reason Transfer Not Active [20285.489423] dwc3 dwc3.0.auto: ep1out-bulk: req ffff8800aa6cb480 dma aeb50800 length 512 last [20285.489727] dwc3 dwc3.0.auto: ep1out-bulk: cmd 'Start Transfer' params 00000000 a9eaf000 00000000 [20285.490055] dwc3 dwc3.0.auto: Command Complete --> 0 [20285.490281] dwc3 dwc3.0.auto: ep1out-bulk: Transfer Not Ready [20285.490492] dwc3 dwc3.0.auto: ep1out-bulk: reason Transfer Active [20285.490713] dwc3 dwc3.0.auto: ep1out-bulk: endpoint busy [20285.490909] dwc3 dwc3.0.auto: ep1out-bulk: Transfer Complete [20285.491117] dwc3 dwc3.0.auto: request ffff8800aa6cb480 from ep1out-bulk completed 512/512 ===> 0 [20285.491431] zero gadget: bad OUT byte, buf[1] = 0 [20285.491605] dwc3 dwc3.0.auto: ep1out-bulk: cmd 'Set Stall' params 00000000 00000000 00000000 [20285.491915] dwc3 dwc3.0.auto: Command Complete --> 0 [20285.492099] dwc3 dwc3.0.auto: queing request ffff8800aa6cb480 to ep1out-bulk length 512 [20285.492387] dwc3 dwc3.0.auto: ep1out-bulk: Transfer Not Ready [20285.492595] dwc3 dwc3.0.auto: ep1out-bulk: reason Transfer Not Active [20285.492830] dwc3 dwc3.0.auto: ep1out-bulk: req ffff8800aa6cb480 dma aeb51000 length 512 last [20285.493135] dwc3 dwc3.0.auto: ep1out-bulk: cmd 'Start Transfer' params 00000000 a9eaf000 00000000 [20285.493465] dwc3 dwc3.0.auto: Command Complete --> 0
--- drivers/usb/misc/usbtest.c | 10 ++++++++++ 1 file changed, 10 insertions(+)
drivers/usb/misc/usbtest.c +10 -0
@@ -1320,6 +1320,11 @@ static int unlink1(struct usbtest_dev *d urb->context = &completion; urb->complete = unlink1_callback; +if (usb_pipeout (urb->pipe)) { +simple_fill_buf (urb); + urb->transfer_flags |=URB_ZERO_PACKET ; + } +/* keep the endpoint busy. there are lots of hc/hcd-internal * states, and testing should get to all of them over time. *
@@ -1450,6 +1455,11 @@ static int unlink_queued(struct usbtest_ unlink_queued_callback, &ctx); ctx.urbs[i]->transfer_dma = buf_dma; ctx.urbs[i]->transfer_flags =URB_NO_TRANSFER_DMA_MAP ; + +if (usb_pipeout (ctx.urbs[i]->pipe)) { +simple_fill_buf (ctx.urbs[i]); + ctx.urbs[i]->transfer_flags |=URB_ZERO_PACKET ; + } }/* Submit all the URBs and then unlink URBs num - 4 and num - 2. */
Greg Kroah-Hartman
maintainerPATCH 15/139 USB: usbtest: add a timeout for scatter-gather tests
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Alan Stern <[email protected]>
commit 32b36eeae6a859670d2939a7d6136cb5e9ed64f8 upstream.
In usbtest, tests 5 - 8 use the scatter-gather library in usbcore without any sort of timeout. If there's a problem in the gadget or host controller being tested, the test can hang.
This patch adds a 10-second timeout to the tests, so that they will fail gracefully with an ETIMEDOUT error instead of hanging.
--- drivers/usb/misc/usbtest.c | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-)
drivers/usb/misc/usbtest.c +15 -1
@@ -7,7 +7,7 @@#include <linux/moduleparam.h> #include <linux/scatterlist.h> #include <linux/mutex.h> - +#include <linux/timer.h> #include <linux/usb.h> #define SIMPLE_IO_TIMEOUT 10000/* in milliseconds */
@@ -484,6 +484,14 @@ alloc_sglist(int nents, int max, int varreturn sg; } +static void sg_timeout (unsigned long _req) +{ +struct usb_sg_request *req = (struct usb_sg_request *) _req; + + req->status = -ETIMEDOUT ; +usb_sg_cancel (req); +} + static int perform_sglist( struct usbtest_dev *tdev, unsigned iterations,
@@ -495,6 +503,9 @@ static int perform_sglist( {struct usb_device *udev =testdev_to_usbdev (tdev);int retval =0 ; +struct timer_list sg_timer; + +setup_timer_on_stack (&sg_timer, sg_timeout, (unsigned long ) req);while (retval ==0 && iterations-- >0 ) { retval = usb_sg_init(req, udev, pipe,
@@ -505,7 +516,10 @@ static int perform_sglist(if (retval)break ; +mod_timer (&sg_timer, jiffies + +msecs_to_jiffies (SIMPLE_IO_TIMEOUT ));usb_sg_wait (req); +del_timer_sync (&sg_timer); retval = req->status;/* FIXME check resulting data pattern */
Greg Kroah-Hartman
maintainerPATCH 16/139 usb: gadget: rename CONFIG_USB_GADGET_PXA25X
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Paul Bolle <[email protected]>
commit d30f2065d6da377cc76771aca5a9850cfca8723b upstream.
Commit 193ab2a60700 ("usb: gadget: allow multiple gadgets to be built") basically renamed the Kconfig symbol USB_GADGET_PXA25X to USB_PXA25X. It did not rename the related macros in use at that time. Commit c0a39151a405 ("ARM: pxa: fix inconsistent CONFIG_USB_PXA27X") did so for all but one macro. Rename that last macro too now.
--- drivers/usb/gadget/inode.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/usb/gadget/inode.c +1 -1
@@ -1501,7 +1501,7 @@ gadgetfs_setup (struct usb_gadget *gadge }break ; -#ifndef CONFIG_USB_GADGET_PXA25X +#ifndef CONFIG_USB_PXA25X /* PXA automagically handles this request too */ case USB_REQ_GET_CONFIGURATION :if (ctrl->bRequestType !=0x80 )
Greg Kroah-Hartman
maintainerPATCH 17/139 usb: dwc3: gadget: clear stall when disabling endpoint
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Felipe Balbi <[email protected]>
commit 687ef9817df7ed960d14575b9033dde3d04631fe upstream.
so it seems like DWC3 IP doesn't clear stalls automatically when we disable an endpoint, because of that, we _must_ make sure stalls are cleared before clearing the proper bit in DALEPENA register.
--- drivers/usb/dwc3/gadget.c | 4 ++++ 1 file changed, 4 insertions(+)
drivers/usb/dwc3/gadget.c +4 -0
@@ -604,6 +604,10 @@ static int __dwc3_gadget_ep_disable(strudwc3_remove_requests (dwc, dep); +/* make sure HW endpoint isn't stalled */ +if (dep->flags &DWC3_EP_STALL ) +__dwc3_gadget_ep_set_halt (dep,0 ); + reg =dwc3_readl (dwc->regs,DWC3_DALEPENA ); reg &= ~DWC3_DALEPENA_EP (dep->number);dwc3_writel (dwc->regs,DWC3_DALEPENA , reg);
Greg Kroah-Hartman
maintainerPATCH 18/139 ARM: OMAP: replace checks for CONFIG_USB_GADGET_OMAP
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Paul Bolle <[email protected]>
commit 77c2f02edbeda9409a7cf3fd66233015820c213a upstream.
Commit 193ab2a60700 ("usb: gadget: allow multiple gadgets to be built") apparently required that checks for CONFIG_USB_GADGET_OMAP would be replaced with checks for CONFIG_USB_OMAP. Do so now for the remaining checks for CONFIG_USB_GADGET_OMAP, even though these checks have basically been broken since v3.1.
And, since we're touching this code, use the IS_ENABLED() macro, so things will now (hopefully) also work if USB_OMAP is modular.
--- arch/arm/mach-omap1/board-h2.c | 2 +- arch/arm/mach-omap1/board-h3.c | 2 +- arch/arm/mach-omap1/board-innovator.c | 2 +- arch/arm/mach-omap1/board-osk.c | 2 +- drivers/usb/phy/phy-isp1301-omap.c | 2 +- 5 files changed, 5 insertions(+), 5 deletions(-)
arch/arm/mach-omap1/board-h2.c +1 -1
@@ -346,7 +346,7 @@ static struct omap_usb_config h2_usb_con/* usb1 has a Mini-AB port and external isp1301 transceiver */ .otg = 2, -#ifdef CONFIG_USB_GADGET_OMAP +#if IS_ENABLED (CONFIG_USB_OMAP ) .hmc_mode = 19, /* 0:host(off) 1:dev|otg 2:disabled *//* .hmc_mode = 21,*/ /* 0:host(off) 1:dev(loopback) 2:host(loopback) */ #elif defined(CONFIG_USB_OHCI_HCD ) || defined(CONFIG_USB_OHCI_HCD_MODULE )
arch/arm/mach-omap1/board-h3.c +1 -1
@@ -366,7 +366,7 @@ static struct omap_usb_config h3_usb_con/* usb1 has a Mini-AB port and external isp1301 transceiver */ .otg = 2, -#ifdef CONFIG_USB_GADGET_OMAP +#if IS_ENABLED (CONFIG_USB_OMAP ) .hmc_mode = 19, /* 0:host(off) 1:dev|otg 2:disabled */#elif defined(CONFIG_USB_OHCI_HCD ) || defined(CONFIG_USB_OHCI_HCD_MODULE )/* NONSTANDARD CABLE NEEDED (B-to-Mini-B) */
arch/arm/mach-omap1/board-innovator.c +1 -1
@@ -312,7 +312,7 @@ static struct omap_usb_config h2_usb_con/* usb1 has a Mini-AB port and external isp1301 transceiver */ .otg = 2, -#ifdef CONFIG_USB_GADGET_OMAP +#if IS_ENABLED (CONFIG_USB_OMAP ) .hmc_mode = 19, /* 0:host(off) 1:dev|otg 2:disabled *//* .hmc_mode = 21,*/ /* 0:host(off) 1:dev(loopback) 2:host(loopback) */ #elif defined(CONFIG_USB_OHCI_HCD ) || defined(CONFIG_USB_OHCI_HCD_MODULE )
arch/arm/mach-omap1/board-osk.c +1 -1
@@ -283,7 +283,7 @@ static struct omap_usb_config osk_usb_co* be used, with a NONSTANDARD gender-bending cable/dongle, as * a peripheral. */ -#ifdef CONFIG_USB_GADGET_OMAP +#if IS_ENABLED (CONFIG_USB_OMAP ) .register_dev = 1, .hmc_mode = 0,#else
drivers/usb/phy/phy-isp1301-omap.c +1 -1
@@ -1295,7 +1295,7 @@ isp1301_set_host(struct usb_otg *otg, streturn isp1301_otg_enable (isp);return 0 ; -#elif !defined(CONFIG_USB_GADGET_OMAP ) +#elif !IS_ENABLED (CONFIG_USB_OMAP )// FIXME update its refcount otg->host = host;
Greg Kroah-Hartman
maintainerPATCH 19/139 USB: EHCI: avoid BIOS handover on the HASEE E200
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Alan Stern <[email protected]>
commit b0a50e92bda3c4aeb8017d4e6c6e92146ebd5c9b upstream.
Leandro Liptak reports that his HASEE E200 computer hangs when we ask the BIOS to hand over control of the EHCI host controller. This definitely sounds like a bug in the BIOS, but at the moment there is no way to fix it.
This patch works around the problem by avoiding the handoff whenever the motherboard and BIOS version match those of Leandro's computer.
--- drivers/usb/host/pci-quirks.c | 19 ++++++++++++++++--- 1 file changed, 16 insertions(+), 3 deletions(-)
drivers/usb/host/pci-quirks.c +16 -3
@@ -656,6 +656,14 @@ static const struct dmi_system_id ehci_d DMI_MATCH(DMI_BIOS_VERSION, "Lucid-"), }, }, + { + /* HASEE E200 */ + .matches = { + DMI_MATCH(DMI_BOARD_VENDOR, "HASEE"), + DMI_MATCH(DMI_BOARD_NAME, "E210"), + DMI_MATCH(DMI_BIOS_VERSION, "6.00"), + }, + }, { } };
@@ -665,9 +673,14 @@ static void ehci_bios_handoff(struct pci {int try_handoff =1 , tried_handoff =0 ; -/* The Pegatron Lucid tablet sporadically waits for 98 seconds trying -* the handoff on its unused controller. Skip it. */ -if (pdev->vendor ==0x8086 && pdev->device ==0x283a ) { +/* +* The Pegatron Lucid tablet sporadically waits for 98 seconds trying +* the handoff on its unused controller. Skip it. +* +* The HASEE E200 hangs when the semaphore is set (bugzilla #77021). +*/ +if (pdev->vendor ==0x8086 && (pdev->device ==0x283a || + pdev->device ==0x27cc )) {if (dmi_check_system (ehci_dmi_nohandoff_table)) try_handoff =0 ; }
Greg Kroah-Hartman
maintainerPATCH 20/139 USB: option: fix runtime PM handling
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold <[email protected]>
commit acf47d4f9c39b1cba467aa9442fc2efe0b1da741 upstream.
Fix potential I/O while runtime suspended due to missing PM operations in send_setup.
option driver")
--- drivers/usb/serial/option.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-)
drivers/usb/serial/option.c +10 -1
@@ -1925,6 +1925,7 @@ static int option_send_setup(struct usb_struct option_private *priv = intfdata->private;struct usb_wwan_port_private *portdata;int val =0 ; +int res; portdata =usb_get_serial_port_data (port);
@@ -1933,9 +1934,17 @@ static int option_send_setup(struct usb_if (portdata->rts_state) val |=0x02 ; -return usb_control_msg (serial->dev,usb_rcvctrlpipe (serial->dev,0 ), + res =usb_autopm_get_interface (serial->interface); +if (res) +return res; + + res =usb_control_msg (serial->dev,usb_rcvctrlpipe (serial->dev,0 ),0x22 ,0x21 , val, priv->bInterfaceNumber,NULL ,0 ,USB_CTRL_SET_TIMEOUT ); + +usb_autopm_put_interface (serial->interface); + +return res; }MODULE_AUTHOR (DRIVER_AUTHOR );
Greg Kroah-Hartman
maintainerPATCH 21/139 tools/vm/page-types.c: catch sigbus if raced with truncate
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Konstantin Khlebnikov <[email protected]>
commit 1d46598b7903cd5ec83c49adbd741f43bb0ffcdc upstream.
Recently added page-cache dumping is known to be a little bit racy. But after race with truncate it just dies due to unhandled SIGBUS when it tries to poke pages beyond the new end of file. This patch adds handler for SIGBUS which skips the rest of the file.
--- tools/vm/page-types.c | 35 ++++++++++++++++++++++++++++++++--- 1 file changed, 32 insertions(+), 3 deletions(-)
tools/vm/page-types.c +32 -3
@@ -32,6 +32,8 @@#include <assert.h> #include <ftw.h> #include <time.h> +#include <setjmp.h> +#include <signal.h> #include <sys/types.h> #include <sys/errno.h> #include <sys/fcntl.h>
@@ -824,21 +826,38 @@ static void show_file(const char *name, atime, now - st->st_atime); } +static sigjmp_buf sigbus_jmp; + +static void *volatile sigbus_addr; + +static void sigbus_handler (int sig,siginfo_t *info,void *ucontex) +{ + (void )sig; + (void )ucontex; + sigbus_addr = info ? info->si_addr :NULL ; +siglongjmp (sigbus_jmp,1 ); +} + +static struct sigaction sigbus_action = { + .sa_sigaction = sigbus_handler, + .sa_flags =SA_SIGINFO , +}; +static void walk_file (const char *name,const struct stat *st) {uint8_t vec[PAGEMAP_BATCH ];uint64_t buf[PAGEMAP_BATCH ], flags;unsigned long nr_pages, pfn, i; +off_t off, end = st->st_size;int fd; -off_t off;ssize_t len;void *ptr;int first =1 ; fd =checked_open (name,O_RDONLY |O_NOATIME |O_NOFOLLOW ); -for (off =0 ; off < st->st_size; off += len) { - nr_pages = (st->st_size - off + page_size -1 ) / page_size; +for (off =0 ; off < end; off += len) { + nr_pages = (end - off + page_size -1 ) / page_size;if (nr_pages >PAGEMAP_BATCH ) nr_pages =PAGEMAP_BATCH ; len = nr_pages * page_size;
@@ -855,11 +874,19 @@ static void walk_file(const char *name,if (madvise (ptr, len,MADV_RANDOM ))fatal ("madvice failed: %s" , name); +if (sigsetjmp (sigbus_jmp,1 )) { + end = off + sigbus_addr ? sigbus_addr - ptr :0 ; +fprintf (stderr,"got sigbus at offset %lld: %s\n" , + (long long )end, name); + goto got_sigbus; + } +/* populate ptes */ for (i =0 ; i < nr_pages ; i++) {if (vec[i] &1 ) (void )*(volatile int *)(ptr + i * page_size); } +got_sigbus:/* turn off harvesting reference bits */ if (madvise (ptr, len,MADV_SEQUENTIAL ))
@@ -910,6 +937,7 @@ static void walk_page_cache(void) kpageflags_fd =checked_open (PROC_KPAGEFLAGS ,O_RDONLY ); pagemap_fd =checked_open ("/proc/self/pagemap" ,O_RDONLY ); +sigaction (SIGBUS , &sigbus_action,NULL );if (stat (opt_file, &st))fatal ("stat failed: %s\n" , opt_file);
@@ -925,6 +953,7 @@ static void walk_page_cache(void)close (kpageflags_fd);close (pagemap_fd); +signal (SIGBUS ,SIG_DFL ); }static void parse_file (const char *name)
Greg Kroah-Hartman
maintainerPATCH 22/139 hugetlb: restrict hugepage_migration_support() to x86_64
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Naoya Horiguchi <[email protected]>
commit c177c81e09e517bbf75b67762cdab1b83aba6976 upstream.
Currently hugepage migration is available for all archs which support pmd-level hugepage, but testing is done only for x86_64 and there're bugs for other archs. So to avoid breaking such archs, this patch limits the availability strictly to x86_64 until developers of other archs get interested in enabling this feature.
Simply disabling hugepage migration on non-x86_64 archs is not enough to fix the reported problem where sys_move_pages() hits the BUG_ON() in follow_page(FOLL_GET), so let's fix this by checking if hugepage migration is supported in vma_migratable().
--- arch/arm/mm/hugetlbpage.c | 5 ----- arch/arm64/mm/hugetlbpage.c | 5 ----- arch/ia64/mm/hugetlbpage.c | 5 ----- arch/metag/mm/hugetlbpage.c | 5 ----- arch/mips/mm/hugetlbpage.c | 5 ----- arch/powerpc/mm/hugetlbpage.c | 10 ---------- arch/s390/mm/hugetlbpage.c | 5 ----- arch/sh/mm/hugetlbpage.c | 5 ----- arch/sparc/mm/hugetlbpage.c | 5 ----- arch/tile/mm/hugetlbpage.c | 5 ----- arch/x86/Kconfig | 4 ++++ arch/x86/mm/hugetlbpage.c | 10 ---------- include/linux/hugetlb.h | 13 +++++-------- include/linux/mempolicy.h | 6 ++++++ mm/Kconfig | 3 +++ 15 files changed, 18 insertions(+), 73 deletions(-)
arch/arm/mm/hugetlbpage.c +0 -5
@@ -56,8 +56,3 @@ int pmd_huge(pmd_t pmd) {return pmd_val (pmd) && !(pmd_val (pmd) &PMD_TABLE_BIT ); } - -int pmd_huge_support (void ) -{ -return 1 ; -}
arch/arm64/mm/hugetlbpage.c +0 -5
@@ -58,11 +58,6 @@ int pud_huge(pud_t pud)#endif } -int pmd_huge_support (void ) -{ -return 1 ; -} -static __init int setup_hugepagesz (char *opt) {unsigned long ps =memparse (opt, &opt);
arch/ia64/mm/hugetlbpage.c +0 -5
@@ -114,11 +114,6 @@ int pud_huge(pud_t pud)return 0 ; } -int pmd_huge_support (void ) -{ -return 0 ; -} -struct page *follow_huge_pmd (struct mm_struct *mm,unsigned long address,pmd_t *pmd,int write) {
arch/metag/mm/hugetlbpage.c +0 -5
@@ -110,11 +110,6 @@ int pud_huge(pud_t pud)return 0 ; } -int pmd_huge_support (void ) -{ -return 1 ; -} -struct page *follow_huge_pmd (struct mm_struct *mm,unsigned long address,pmd_t *pmd,int write) {
arch/mips/mm/hugetlbpage.c +0 -5
@@ -84,11 +84,6 @@ int pud_huge(pud_t pud)return (pud_val (pud) & _PAGE_HUGE) !=0 ; } -int pmd_huge_support (void ) -{ -return 1 ; -} -struct page *follow_huge_pmd (struct mm_struct *mm,unsigned long address,pmd_t *pmd,int write)
arch/powerpc/mm/hugetlbpage.c +0 -10
@@ -86,11 +86,6 @@ int pgd_huge(pgd_t pgd)*/ return ((pgd_val(pgd) & 0x3) != 0x0); } - -int pmd_huge_support (void ) -{ -return 1 ; -}#else int pmd_huge (pmd_t pmd) {
@@ -106,11 +101,6 @@ int pgd_huge(pgd_t pgd) {return 0 ; } - -int pmd_huge_support (void ) -{ -return 0 ; -}#endif pte_t *huge_pte_offset (struct mm_struct *mm,unsigned long addr)
arch/s390/mm/hugetlbpage.c +0 -5
@@ -220,11 +220,6 @@ int pud_huge(pud_t pud)return 0 ; } -int pmd_huge_support (void ) -{ -return 1 ; -} -struct page *follow_huge_pmd (struct mm_struct *mm,unsigned long address,pmd_t *pmdp,int write) {
arch/sh/mm/hugetlbpage.c +0 -5
@@ -83,11 +83,6 @@ int pud_huge(pud_t pud)return 0 ; } -int pmd_huge_support (void ) -{ -return 0 ; -} -struct page *follow_huge_pmd (struct mm_struct *mm,unsigned long address,pmd_t *pmd,int write) {
arch/sparc/mm/hugetlbpage.c +0 -5
@@ -231,11 +231,6 @@ int pud_huge(pud_t pud)return 0 ; } -int pmd_huge_support (void ) -{ -return 0 ; -} -struct page *follow_huge_pmd (struct mm_struct *mm,unsigned long address,pmd_t *pmd,int write) {
arch/tile/mm/hugetlbpage.c +0 -5
@@ -166,11 +166,6 @@ int pud_huge(pud_t pud)return !!(pud_val (pud) & _PAGE_HUGE_PAGE); } -int pmd_huge_support (void ) -{ -return 1 ; -} -struct page *follow_huge_pmd (struct mm_struct *mm,unsigned long address,pmd_t *pmd,int write) {
arch/x86/Kconfig +4 -0
@@ -1871,6 +1871,10 @@ config ARCH_ENABLE_SPLIT_PMD_PTLOCK def_bool y depends on X86_64 || X86_PAE +config ARCH_ENABLE_HUGEPAGE_MIGRATION + def_bool y + depends on X86_64 && HUGETLB_PAGE && MIGRATION + menu "Power management and ACPI options" config ARCH_HIBERNATION_HEADER
arch/x86/mm/hugetlbpage.c +0 -10
@@ -58,11 +58,6 @@ follow_huge_pmd(struct mm_struct *mm, un {return NULL ; } - -int pmd_huge_support (void ) -{ -return 0 ; -}#else struct page *
@@ -80,11 +75,6 @@ int pud_huge(pud_t pud) {return !!(pud_val (pud) & _PAGE_PSE); } - -int pmd_huge_support (void ) -{ -return 1 ; -}#endif #ifdef CONFIG_HUGETLB_PAGE
include/linux/hugetlb.h +5 -8
@@ -392,15 +392,13 @@ static inline pgoff_t basepage_index(strextern void dissolve_free_huge_pages (unsigned long start_pfn,unsigned long end_pfn); -int pmd_huge_support (void ); -/* -* Currently hugepage migration is enabled only for pmd-based hugepage. -* This function will be updated when hugepage migration is more widely -* supported. -*/ static inline int hugepage_migration_support (struct hstate *h) { -return pmd_huge_support () && (huge_page_shift (h) ==PMD_SHIFT ); +#ifdef CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION +return huge_page_shift (h) ==PMD_SHIFT ; +#else +return 0 ; +#endif } static inline spinlock_t *huge_pte_lockptr(struct hstate *h,
@@ -450,7 +448,6 @@ static inline pgoff_t basepage_index(strreturn page->index; }#define dissolve_free_huge_pages (s, e) do {} while (0) -#define pmd_huge_support () 0#define hugepage_migration_support (h) 0 static inline spinlock_t *huge_pte_lockptr(struct hstate *h,
include/linux/mempolicy.h +6 -0
@@ -175,6 +175,12 @@ static inline int vma_migratable(struct {if (vma->vm_flags & (VM_IO |VM_PFNMAP ))return 0 ; + +#ifndef CONFIG_ARCH_ENABLE_HUGEPAGE_MIGRATION +if (vma->vm_flags &VM_HUGETLB ) +return 0 ; +#endif +/* * Migration allocates pages in the highest zone. If we cannot * do so then migration (at least from node to node) is not
mm/Kconfig +3 -0
@@ -264,6 +264,9 @@ config MIGRATION pages as migration can relocate pages to satisfy a huge page allocation instead of reclaiming. +config ARCH_ENABLE_HUGEPAGE_MIGRATION + boolean + config PHYS_ADDR_T_64BIT def_bool 64BIT || ARCH_PHYS_ADDR_T_64BIT
Greg Kroah-Hartman
maintainerPATCH 23/139 kthread: fix return value of kthread_create() upon SIGKILL.
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Tetsuo Handa <[email protected]>
commit 8fe6929cfd43c44834858a53e129ffdc7c166298 upstream.
Commit 786235eeba0e ("kthread: make kthread_create() killable") meant for allowing kthread_create() to abort as soon as killed by the OOM-killer. But returning -ENOMEM is wrong if killed by SIGKILL from userspace. Change kthread_create() to return -EINTR upon SIGKILL.
--- kernel/kthread.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
kernel/kthread.c +2 -2
@@ -262,7 +262,7 @@ static void create_kthread(struct kthrea* kthread_stop() has been called). The return value should be zero * or a negative error number; it will be passed to kthread_stop(). * -* Returns a task_struct or ERR_PTR(-ENOMEM). +* Returns a task_struct or ERR_PTR(-ENOMEM) or ERR_PTR(-EINTR). */ struct task_struct *kthread_create_on_node(int (*threadfn)(void *data), void *data, int node,
@@ -298,7 +298,7 @@ struct task_struct *kthread_create_on_no* that thread. */ if (xchg (&create->done,NULL )) -return ERR_PTR (-ENOMEM ); +return ERR_PTR (-EINTR );/* * kthreadd (or new kernel thread) will call complete() * shortly.
Greg Kroah-Hartman
maintainerPATCH 24/139 mm: vmscan: do not throttle based on pfmemalloc reserves if node has no ZONE_NORMAL
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Mel Gorman <[email protected]>
commit 675becce15f320337499bc1a9356260409a5ba29 upstream.
throttle_direct_reclaim() is meant to trigger during swap-over-network during which the min watermark is treated as a pfmemalloc reserve. It throttes on the first node in the zonelist but this is flawed.
The user-visible impact is that a process running on CPU whose local memory node has no ZONE_NORMAL will stall for prolonged periods of time, possibly indefintely. This is due to throttle_direct_reclaim thinking the pfmemalloc reserves are depleted when in fact they don't exist on that node.
On a NUMA machine running a 32-bit kernel (I know) allocation requests from CPUs on node 1 would detect no pfmemalloc reserves and the process gets throttled. This patch adjusts throttling of direct reclaim to throttle based on the first node in the zonelist that has a usable ZONE_NORMAL or lower zone.
--- mm/vmscan.c | 43 +++++++++++++++++++++++++++++++++++++------ 1 file changed, 37 insertions(+), 6 deletions(-)
mm/vmscan.c +37 -6
@@ -2525,10 +2525,17 @@ static bool pfmemalloc_watermark_ok(pg_dfor (i =0 ; i <=ZONE_NORMAL ; i++) { zone = &pgdat->node_zones[i]; +if (!populated_zone (zone)) +continue ; + pfmemalloc_reserve +=min_wmark_pages (zone); free_pages +=zone_page_state (zone,NR_FREE_PAGES ); } +/* If there are no reserves (unexpected config) then do not throttle */ +if (!pfmemalloc_reserve) +return true; + wmark_ok = free_pages > pfmemalloc_reserve /2 ;/* kswapd must be awake if processes are being throttled */
@@ -2553,9 +2560,9 @@ static bool pfmemalloc_watermark_ok(pg_dstatic bool throttle_direct_reclaim (gfp_t gfp_mask,struct zonelist *zonelist,nodemask_t *nodemask) { +struct zoneref *z;struct zone *zone; -int high_zoneidx =gfp_zone (gfp_mask); -pg_data_t *pgdat; +pg_data_t *pgdat =NULL ;/* * Kernel threads should not be throttled as they may be indirectly
@@ -2574,10 +2581,34 @@ static bool throttle_direct_reclaim(gfp_if (fatal_signal_pending (current)) goto out; -/* Check if the pfmemalloc reserves are ok */ -first_zones_zonelist (zonelist, high_zoneidx,NULL , &zone); - pgdat = zone->zone_pgdat; -if (pfmemalloc_watermark_ok (pgdat)) +/* +* Check if the pfmemalloc reserves are ok by finding the first node +* with a usable ZONE_NORMAL or lower zone. The expectation is that +* GFP_KERNEL will be required for allocating network buffers when +* swapping over the network so ZONE_HIGHMEM is unusable. +* +* Throttling is based on the first usable node and throttled processes +* wait on a queue until kswapd makes progress and wakes them. There +* is an affinity then between processes waking up and where reclaim +* progress has been made assuming the process wakes on the same node. +* More importantly, processes running on remote nodes will not compete +* for remote pfmemalloc reserves and processes on different nodes +* should make reasonable progress. +*/ +for_each_zone_zonelist_nodemask (zone, z, zonelist, + gfp_mask, nodemask) { +if (zone_idx (zone) >ZONE_NORMAL ) +continue ; + +/* Throttle based on the first usable node */ + pgdat = zone->zone_pgdat; +if (pfmemalloc_watermark_ok (pgdat)) + goto out; +break ; + } + +/* If no zone was usable by the allocation flags then do not throttle */ +if (!pgdat) goto out;/* Account for the throttling */
Greg Kroah-Hartman
maintainerPATCH 25/139 memcg: do not hang on OOM when killed by userspace OOM access to memory reserves
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Michal Hocko <[email protected]>
commit d8dc595ce3909fbc131bdf5ab8c9808fe624b18d upstream.
Eric has reported that he can see task(s) stuck in memcg OOM handler regularly. The only way out is to
echo 0 > $GROUP/memory.oom_control
His usecase is:
- Setup a hierarchy with memory and the freezer (disable kernel oom and have a process watch for oom).
- In that memory cgroup add a process with one thread per cpu.
- In one thread slowly allocate once per second I think it is 16M of ram and mlock and dirty it (just to force the pages into ram and stay there).
- When oom is achieved loop:
* attempt to freeze all of the tasks.
* if frozen send every task SIGKILL, unfreeze, remove the directory in cgroupfs.
Eric has then pinpointed the issue to be memcg specific.
All tasks are sitting on the memcg_oom_waitq when memcg oom is disabled. Those that have received fatal signal will bypass the charge and should continue on their way out. The tricky part is that the exit path might trigger a page fault (e.g. exit_robust_list), thus the memcg charge, while its memcg is still under OOM because nobody has released any charges yet.
Unlike with the in-kernel OOM handler the exiting task doesn't get TIF_MEMDIE set so it doesn't shortcut further charges of the killed task and falls to the memcg OOM again without any way out of it as there are no fatal signals pending anymore.
This patch fixes the issue by checking PF_EXITING early in mem_cgroup_try_charge and bypass the charge same as if it had fatal signal pending or TIF_MEMDIE set.
Normally exiting tasks (aka not killed) will bypass the charge now but this should be OK as the task is leaving and will release memory and increasing the memory pressure just to release it in a moment seems dubious wasting of cycles. Besides that charges after exit_signals should be rare.
I am bringing this patch again (rebased on the current mmotm tree). I hope we can move forward finally. If there is still an opposition then I would really appreciate a concurrent approach so that we can discuss alternatives.
http://comments.gmane.org/gmane.linux.kernel.stable/77650 is a reference to the followup discussion when the patch has been dropped from the mmotm last time.
--- mm/memcontrol.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
mm/memcontrol.c +2 -1
@@ -2684,7 +2684,8 @@ static int mem_cgroup_try_charge(struct* free their memory. */ if (unlikely (test_thread_flag (TIF_MEMDIE ) || -fatal_signal_pending (current))) +fatal_signal_pending (current) || + current->flags &PF_EXITING )) goto bypass;if (unlikely (task_in_memcg_oom (current)))
Greg Kroah-Hartman
maintainerPATCH 26/139 mm: page_alloc: use word-based accesses for get/set pageblock bitmaps
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Mel Gorman <[email protected]>
commit e58469bafd0524e848c3733bc3918d854595e20f upstream.
The test_bit operations in get/set pageblock flags are expensive. This patch reads the bitmap on a word basis and use shifts and masks to isolate the bits of interest. Similarly masks are used to set a local copy of the bitmap and then use cmpxchg to update the bitmap if there have been no other changes made in parallel.
In a test running dd onto tmpfs the overhead of the pageblock-related functions went from 1.27% in profiles to 0.5%.
In addition to the performance benefits, this patch closes races that are possible between:
a) get_ and set_pageblock_migratetype(), where get_pageblock_migratetype() reads part of the bits before and other part of the bits after set_pageblock_migratetype() has updated them.
b) set_pageblock_migratetype() and set_pageblock_skip(), where the non-atomic read-modify-update set bit operation in set_pageblock_skip() will cause lost updates to some bits changed in the set_pageblock_migratetype().
Joonsoo Kim first reported the case a) via code inspection. Vlastimil Babka's testing with a debug patch showed that either a) or b) occurs roughly once per mmtests' stress-highalloc benchmark (although not necessarily in the same pageblock). Furthermore during development of unrelated compaction patches, it was observed that frequent calls to {start,undo}_isolate_page_range() the race occurs several thousands of times and has resulted in NULL pointer dereferences in move_freepages() and free_one_page() in places where free_list[migratetype] is manipulated by e.g. list_move(). Further debugging confirmed that migratetype had invalid value of 6, causing out of bounds access to the free_list array.
That confirmed that the race exist, although it may be extremely rare, and currently only fatal where page isolation is performed due to memory hot remove. Races on pageblocks being updated by set_pageblock_migratetype(), where both old and new migratetype are lower MIGRATE_RESERVE, currently cannot result in an invalid value being observed, although theoretically they may still lead to unexpected creation or destruction of MIGRATE_RESERVE pageblocks. Furthermore, things could get suddenly worse when memory isolation is used more, or when new migratetypes are added.
After this patch, the race has no longer been observed in testing.
--- include/linux/mmzone.h | 6 +++- include/linux/pageblock-flags.h | 37 +++++++++++++++++++++++----- mm/page_alloc.c | 52 ++++++++++++++++++++++++---------------- 3 files changed, 68 insertions(+), 27 deletions(-)
include/linux/mmzone.h +5 -1
@@ -75,9 +75,13 @@ enum {extern int page_group_by_mobility_disabled; +#define NR_MIGRATETYPE_BITS (PB_migrate_end - PB_migrate + 1) +#define MIGRATETYPE_MASK ((1UL << NR_MIGRATETYPE_BITS) - 1) +static inline int get_pageblock_migratetype (struct page *page) { -return get_pageblock_flags_group (page, PB_migrate, PB_migrate_end); +BUILD_BUG_ON (PB_migrate_end - PB_migrate !=2 ); +return get_pageblock_flags_mask (page, PB_migrate_end,MIGRATETYPE_MASK ); }struct free_area {
include/linux/pageblock-flags.h +31 -6
@@ -30,9 +30,12 @@ enum pageblock_bits { PB_migrate, PB_migrate_end = PB_migrate +3 -1 ,/* 3 bits required for migrate types */ -#ifdef CONFIG_COMPACTION PB_migrate_skip,/* If set the block is skipped by compaction */ -#endif /* CONFIG_COMPACTION */ + +/* +* Assume the bits will always align on a word. If this assumption +* changes then get/set pageblock needs updating. +*/ NR_PAGEBLOCK_BITS };
@@ -62,11 +65,33 @@ extern int pageblock_order;/* Forward declaration */ struct page ; +unsigned long get_pageblock_flags_mask (struct page *page, +unsigned long end_bitidx, +unsigned long mask); +void set_pageblock_flags_mask (struct page *page, +unsigned long flags, +unsigned long end_bitidx, +unsigned long mask); +/* Declarations for getting and setting flags. See mm/page_alloc.c */ -unsigned long get_pageblock_flags_group (struct page *page, -int start_bitidx,int end_bitidx); -void set_pageblock_flags_group (struct page *page,unsigned long flags, -int start_bitidx,int end_bitidx); +static inline unsigned long get_pageblock_flags_group (struct page *page, +int start_bitidx,int end_bitidx) +{ +unsigned long nr_flag_bits = end_bitidx - start_bitidx +1 ; +unsigned long mask = (1 << nr_flag_bits) -1 ; + +return get_pageblock_flags_mask (page, end_bitidx, mask); +} + +static inline void set_pageblock_flags_group (struct page *page, +unsigned long flags, +int start_bitidx,int end_bitidx) +{ +unsigned long nr_flag_bits = end_bitidx - start_bitidx +1 ; +unsigned long mask = (1 << nr_flag_bits) -1 ; + +set_pageblock_flags_mask (page, flags, end_bitidx, mask); +}#ifdef CONFIG_COMPACTION #define get_pageblock_skip (page) \
mm/page_alloc.c +32 -20
85 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 27/139 mm/memory-failure.c-failure: send right signal code to correct thread
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Tony Luck <[email protected]>
commit a70ffcac741d31a406c1d2b832ae43d658e7e1cf upstream.
When a thread in a multi-threaded application hits a machine check because of an uncorrectable error in memory - we want to send the SIGBUS with si.si_code = BUS_MCEERR_AR to that thread. Currently we fail to do that if the active thread is not the primary thread in the process. collect_procs() just finds primary threads and this test:
if ((flags & MF_ACTION_REQUIRED) && t == current) {will see that the thread we found isn't the current thread and so send a si.si_code = BUS_MCEERR_AO to the primary (and nothing to the active thread at this time).
We can fix this by checking whether "current" shares the same mm with the process that collect_procs() said owned the page. If so, we send the SIGBUS to current (with code BUS_MCEERR_AR).
--- mm/memory-failure.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
mm/memory-failure.c +2 -2
@@ -204,9 +204,9 @@ static int kill_proc(struct task_struct#endif si.si_addr_lsb =compound_order (compound_head (page)) +PAGE_SHIFT ; -if ((flags &MF_ACTION_REQUIRED ) && t == current) { +if ((flags &MF_ACTION_REQUIRED ) && t->mm == current->mm) { si.si_code =BUS_MCEERR_AR ; - ret =force_sig_info (SIGBUS , &si, t); + ret =force_sig_info (SIGBUS , &si, current); }else {/* * Don't use force here, it's convenient if the signal
Greg Kroah-Hartman
maintainerPATCH 28/139 mm/memory-failure.c: dont let collect_procs() skip over processes for MF_ACTION_REQUIRED
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Tony Luck <[email protected]>
commit 74614de17db6fb472370c426d4f934d8d616edf2 upstream.
When Linux sees an "action optional" machine check (where h/w has reported an error that is not in the current execution path) we generally do not want to signal a process, since most processes do not have a SIGBUS handler - we'd just prematurely terminate the process for a problem that they might never actually see.
task_early_kill() decides whether to consider a process - and it checks whether this specific process has been marked for early signals with "prctl", or if the system administrator has requested early signals for all processes using /proc/sys/vm/memory_failure_early_kill.
But for MF_ACTION_REQUIRED case we must not defer. The error is in the execution path of the current thread so we must send the SIGBUS immediatley.
Fix by passing a flag argument through collect_procs*() to task_early_kill() so it knows whether we can defer or must take action.
--- mm/memory-failure.c | 21 ++++++++++++--------- 1 file changed, 12 insertions(+), 9 deletions(-)
mm/memory-failure.c +12 -9
@@ -380,10 +380,12 @@ static void kill_procs(struct list_head } } -static int task_early_kill (struct task_struct *tsk) +static int task_early_kill (struct task_struct *tsk,int force_early) {if (!tsk->mm)return 0 ; +if (force_early) +return 1 ;if (tsk->flags &PF_MCE_PROCESS )return !!(tsk->flags &PF_MCE_EARLY );return sysctl_memory_failure_early_kill;
@@ -393,7 +395,7 @@ static int task_early_kill(struct task_s* Collect processes when the error hit an anonymous page. */ static void collect_procs_anon (struct page *page,struct list_head *to_kill, -struct to_kill **tkc) +struct to_kill **tkc,int force_early) {struct vm_area_struct *vma;struct task_struct *tsk;
@@ -409,7 +411,7 @@ static void collect_procs_anon(struct pafor_each_process (tsk) {struct anon_vma_chain *vmac; -if (!task_early_kill (tsk)) +if (!task_early_kill (tsk, force_early))continue ;anon_vma_interval_tree_foreach (vmac, &av->rb_root, pgoff, pgoff) {
@@ -428,7 +430,7 @@ static void collect_procs_anon(struct pa* Collect processes when the error hit a file mapped page. */ static void collect_procs_file (struct page *page,struct list_head *to_kill, -struct to_kill **tkc) +struct to_kill **tkc,int force_early) {struct vm_area_struct *vma;struct task_struct *tsk;
@@ -439,7 +441,7 @@ static void collect_procs_file(struct pafor_each_process (tsk) {pgoff_t pgoff = page->index << (PAGE_CACHE_SHIFT -PAGE_SHIFT ); -if (!task_early_kill (tsk)) +if (!task_early_kill (tsk, force_early))continue ; vma_interval_tree_foreach(vma, &mapping->i_mmap, pgoff,
@@ -465,7 +467,8 @@ static void collect_procs_file(struct pa* First preallocate one tokill structure outside the spin locks, * so that we can kill at least one process reasonably reliable. */ -static void collect_procs (struct page *page,struct list_head *tokill) +static void collect_procs (struct page *page,struct list_head *tokill, +int force_early) {struct to_kill *tk;
@@ -476,9 +479,9 @@ static void collect_procs(struct page *pif (!tk)return ;if (PageAnon (page)) -collect_procs_anon (page, tokill, &tk); +collect_procs_anon (page, tokill, &tk, force_early);else -collect_procs_file (page, tokill, &tk); +collect_procs_file (page, tokill, &tk, force_early);kfree (tk); }
@@ -963,7 +966,7 @@ static int hwpoison_user_mappings(struct* there's nothing that can be done. */ if (kill) -collect_procs (ppage, &tokill); +collect_procs (ppage, &tokill, flags &MF_ACTION_REQUIRED ); ret =try_to_unmap (ppage, ttu);if (ret !=SWAP_SUCCESS )
Greg Kroah-Hartman
maintainerPATCH 29/139 mm/memory-failure.c: support use of a dedicated thread to handle SIGBUS(BUS_MCEERR_AO)
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Naoya Horiguchi <[email protected]>
commit 3ba08129e38437561df44c36b7ea9081185d5333 upstream.
Currently memory error handler handles action optional errors in the deferred manner by default. And if a recovery aware application wants to handle it immediately, it can do it by setting PF_MCE_EARLY flag. However, such signal can be sent only to the main thread, so it's problematic if the application wants to have a dedicated thread to handler such signals.
So this patch adds dedicated thread support to memory error handler. We have PF_MCE_EARLY flags for each thread separately, so with this patch AO signal is sent to the thread with PF_MCE_EARLY flag set, not the main thread. If you want to implement a dedicated thread, you call prctl() to set PF_MCE_EARLY on the thread.
Memory error handler collects processes to be killed, so this patch lets it check PF_MCE_EARLY flag on each thread in the collecting routines.
No behavioral change for all non-early kill cases.
Tony said:
: The old behavior was crazy - someone with a multithreaded process might : well expect that if they call prctl(PF_MCE_EARLY) in just one thread, then : that thread would see the SIGBUS with si_code = BUS_MCEERR_A0 - even if : that thread wasn't the main thread for the process.
--- Documentation/vm/hwpoison.txt | 5 +++ mm/memory-failure.c | 56 ++++++++++++++++++++++++++++++++---------- 2 files changed, 48 insertions(+), 13 deletions(-)
Documentation/vm/hwpoison.txt +5 -0
@@ -84,6 +84,11 @@ PR_MCE_KILL PR_MCE_KILL_EARLY: Early kill PR_MCE_KILL_LATE: Late kill PR_MCE_KILL_DEFAULT: Use system global default + Note that if you want to have a dedicated thread which handles + the SIGBUS(BUS_MCEERR_AO) on behalf of the process, you should + call prctl(PR_MCE_KILL_EARLY) on the designated thread. Otherwise, + the SIGBUS is sent to the main thread. + PR_MCE_KILL_GET return current mode
mm/memory-failure.c +43 -13
92 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 30/139 zram: correct offset usage in zram_bio_discard
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Weijie Yang <[email protected]>
commit 38515c73398a4c58059ecf1087e844561b58ee0f upstream.
We want to skip the physical block(PAGE_SIZE) which is partially covered by the discard bio, so we check the remaining size and subtract it if there is a need to goto the next physical block.
The current offset usage in zram_bio_discard is incorrect, it will cause its upper filesystem breakdown. Consider the following scenario:
On some architecture or config, PAGE_SIZE is 64K for example, filesystem is set up on zram disk without PAGE_SIZE aligned, a discard bio leads to a offset = 4K and size=72K, normally, it should not really discard any physical block as it partially cover two physical blocks. However, with the current offset usage, it will discard the second physical block and free its memory, which will cause filesystem breakdown.
This patch corrects the offset usage in zram_bio_discard.
--- drivers/block/zram/zram_drv.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/block/zram/zram_drv.c +2 -2
@@ -572,10 +572,10 @@ static void zram_bio_discard(struct zram* skipping this logical block is appropriate here. */ if (offset) { -if (n < offset) +if (n <= (PAGE_SIZE - offset))return ; - n -= offset; + n -= (PAGE_SIZE - offset); index++; }
Greg Kroah-Hartman
maintainerPATCH 31/139 mm: fix sleeping function warning from __put_anon_vma
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Hugh Dickins <[email protected]>
commit 7f39dda9d86fb4f4f17af0de170decf125726f8c upstream.
Trinity reports BUG:
sleeping function called from invalid context at kernel/locking/rwsem.c:47 in_atomic(): 0, irqs_disabled(): 0, pid: 5787, name: trinity-c27
__might_sleep < down_write < __put_anon_vma < page_get_anon_vma < migrate_pages < compact_zone < compact_zone_order < try_to_compact_pages ..
Right, since conversion to mutex then rwsem, we should not put_anon_vma() from inside an rcu_read_lock()ed section: fix the two places that did so. And add might_sleep() to anon_vma_free(), as suggested by Peter Zijlstra.
--- mm/rmap.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-)
mm/rmap.c +5 -3
@@ -103,6 +103,7 @@ static inline void anon_vma_free(struct* LOCK should suffice since the actual taking of the lock must * happen _before_ what follows. */ +might_sleep ();if (rwsem_is_locked (&anon_vma->root->rwsem)) {anon_vma_lock_write (anon_vma);anon_vma_unlock_write (anon_vma);
@@ -426,8 +427,9 @@ struct anon_vma *page_get_anon_vma(struc* above cannot corrupt). */ if (!page_mapped (page)) { +rcu_read_unlock ();put_anon_vma (anon_vma); - anon_vma =NULL ; +return NULL ; } out:rcu_read_unlock ();
@@ -477,9 +479,9 @@ struct anon_vma *page_lock_anon_vma_read }if (!page_mapped (page)) { +rcu_read_unlock ();put_anon_vma (anon_vma); - anon_vma =NULL ; - goto out; +return NULL ; }/* we pinned the anon_vma, its safe to sleep */
Greg Kroah-Hartman
maintainerPATCH 32/139 HID: core: fix validation of report id 0
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Kees Cook <[email protected]>
commit 1b15d2e5b8077670b1e6a33250a0d9577efff4a5 upstream.
Some drivers use the first HID report in the list instead of using an index. In these cases, validation uses ID 0, which was supposed to mean "first known report". This fixes the problem, which was causing at least the lgff family of devices to stop working since hid_validate_values was being called with ID 0, but the devices used single numbered IDs for their reports:
0x05, 0x01, /* Usage Page (Desktop), */ 0x09, 0x05, /* Usage (Gamepad), */ 0xA1, 0x01, /* Collection (Application), */ 0xA1, 0x02, /* Collection (Logical), */ 0x85, 0x01, /* Report ID (1), */
...
--- drivers/hid/hid-core.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/hid/hid-core.c +11 -1
@@ -842,7 +842,17 @@ struct hid_report *hid_validate_values(s* ->numbered being checked, which may not always be the case when * drivers go to access report values. */ - report = hid->report_enum[type].report_id_hash[id]; +if (id ==0 ) { +/* +* Validating on id 0 means we should examine the first +* report in the list. +*/ + report =list_entry ( + hid->report_enum[type].report_list.next, + struct hid_report, list); + }else { + report = hid->report_enum[type].report_id_hash[id]; + }if (!report) {hid_err (hid,"missing %s %u\n" , hid_report_names[type], id);return NULL ;
Greg Kroah-Hartman
maintainerPATCH 33/139 mm: vmscan: clear kswapds special reclaim powers before exiting
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johannes Weiner <[email protected]>
commit 71abdc15adf8c702a1dd535f8e30df50758848d2 upstream.
When kswapd exits, it can end up taking locks that were previously held by allocating tasks while they waited for reclaim. Lockdep currently warns about this:
Quoted 34 lines · Gu Zheng
inconsistent {RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-R} usage. kswapd2/1151 [HC0[0]:SC0[0]:HE1:SE1] takes:
(&sig->group_rwsem){+++++?}, at:exit_signals+0x24/0x130
{RECLAIM_FS-ON-W} state was registered at:mark_held_locks+0xb9/0x140 lockdep_trace_alloc+0x7a/0xe0 kmem_cache_alloc_trace+0x37/0x240 flex_array_alloc+0x99/0x1a0 cgroup_attach_task+0x63/0x430 attach_task_by_pid+0x210/0x280 cgroup_procs_write+0x16/0x20 cgroup_file_write+0x120/0x2c0 vfs_write+0xc0/0x1f0 SyS_write+0x4c/0xa0 tracesys+0xdd/0xe2 irq event stamp: 49 hardirqs last enabled at (49): _raw_spin_unlock_irqrestore+0x36/0x70 hardirqs last disabled at (48): _raw_spin_lock_irqsave+0x2b/0xa0 softirqs last enabled at (0): copy_process.part.24+0x627/0x15f0 softirqs last disabled at (0): (null)other info that might help us debug this:
Possible unsafe locking scenario:CPU0 ---- lock(&sig->group_rwsem); <Interrupt> lock(&sig->group_rwsem);*** DEADLOCK ***
no locks held by kswapd2/1151.
Quoted 13 lines
stack backtrace: CPU: 30 PID: 1151 Comm: kswapd2 Not tainted 3.10.39+ #4 Call Trace: dump_stack+0x19/0x1b print_usage_bug+0x1f7/0x208 mark_lock+0x21d/0x2a0 __lock_acquire+0x52a/0xb60 lock_acquire+0xa2/0x140 down_read+0x51/0xa0 exit_signals+0x24/0x130 do_exit+0xb5/0xa50 kthread+0xdb/0x100 ret_from_fork+0x7c/0xb0
This is because the kswapd thread is still marked as a reclaimer at the time of exit. But because it is exiting, nobody is actually waiting on it to make reclaim progress anymore, and it's nothing but a regular thread at this point. Be tidy and strip it of all its powers (PF_MEMALLOC, PF_SWAPWRITE, PF_KSWAPD, and the lockdep reclaim state) before returning from the thread function.
--- mm/vmscan.c | 3 +++ 1 file changed, 3 insertions(+)
mm/vmscan.c +3 -0
@@ -3333,7 +3333,10 @@ static int kswapd(void *p) } } + tsk->flags &= ~(PF_MEMALLOC |PF_SWAPWRITE |PF_KSWAPD ); current->reclaim_state =NULL ; +lockdep_clear_current_reclaim_state (); +return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 34/139 ptrace: fix fork event messages across pid namespaces
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Matthew Dempsky <[email protected]>
commit 4e52365f279564cef0ddd41db5237f0471381093 upstream.
When tracing a process in another pid namespace, it's important for fork event messages to contain the child's pid as seen from the tracer's pid namespace, not the parent's. Otherwise, the tracer won't be able to correlate the fork event with later SIGTRAP signals it receives from the child.
We still risk a race condition if a ptracer from a different pid namespace attaches after we compute the pid_t value. However, sending a bogus fork event message in this unlikely scenario is still a vast improvement over the status quo where we always send bogus fork event messages to debuggers in a different pid namespace than the forking process.
--- include/linux/ptrace.h | 32 ++++++++++++++++++++++++++++++++ kernel/fork.c | 10 +++++++--- 2 files changed, 39 insertions(+), 3 deletions(-)
include/linux/ptrace.h +32 -0
@@ -5,6 +5,7 @@#include <linux/sched.h> /* For struct task_struct. */ #include <linux/err.h> /* for IS_ERR_VALUE */ #include <linux/bug.h> /* For BUG_ON. */ +#include <linux/pid_namespace.h> /* For task_active_pid_ns. */ #include <uapi/linux/ptrace.h> /*
@@ -129,6 +130,37 @@ static inline void ptrace_event(int even }/** +* ptrace_event_pid - possibly stop for a ptrace event notification +* @event: %PTRACE_EVENT_* value to report +* @pid: process identifier for %PTRACE_GETEVENTMSG to return +* +* Check whether @event is enabled and, if so, report @event and @pid +* to the ptrace parent. @pid is reported as the pid_t seen from the +* the ptrace parent's pid namespace. +* +* Called without locks. +*/ +static inline void ptrace_event_pid (int event,struct pid *pid) +{ +/* +* FIXME: There's a potential race if a ptracer in a different pid +* namespace than parent attaches between computing message below and +* when we acquire tasklist_lock in ptrace_stop(). If this happens, +* the ptracer will get a bogus pid from PTRACE_GETEVENTMSG. +*/ +unsigned long message =0 ; +struct pid_namespace *ns; + +rcu_read_lock (); + ns =task_active_pid_ns (rcu_dereference (current->parent)); +if (ns) + message =pid_nr_ns (pid, ns); +rcu_read_unlock (); + +ptrace_event (event, message); +} + +/** * ptrace_init_task - initialize ptrace state for a new child * @child: new child task * @ptrace: true if child should be ptrace'd by parent's tracer
kernel/fork.c +7 -3
@@ -1606,10 +1606,12 @@ long do_fork(unsigned long clone_flags,*/ if (!IS_ERR (p)) {struct completion vfork; +struct pid *pid;trace_sched_process_fork (current, p); - nr =task_pid_vnr (p); + pid =get_task_pid (p,PIDTYPE_PID ); + nr =pid_vnr (pid);if (clone_flags &CLONE_PARENT_SETTID )put_user (nr, parent_tidptr);
@@ -1624,12 +1626,14 @@ long do_fork(unsigned long clone_flags,/* forking complete and child started to run, tell ptracer */ if (unlikely (trace)) -ptrace_event (trace, nr); +ptrace_event_pid (trace, pid);if (clone_flags &CLONE_VFORK ) {if (!wait_for_vfork_done (p, &vfork)) -ptrace_event (PTRACE_EVENT_VFORK_DONE , nr); +ptrace_event_pid (PTRACE_EVENT_VFORK_DONE , pid); } + +put_pid (pid); }else { nr =PTR_ERR (p); }
Greg Kroah-Hartman
maintainerPATCH 35/139 arm64: ptrace: change fs when passing kernel pointer to regset code
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Will Deacon <[email protected]>
commit c168870704bcde6bb63d05f7882b620dd3985a46 upstream.
Our compat PTRACE_POKEUSR implementation simply passes the user data to regset_copy_from_user after some simple range checking. Unfortunately, the data in question has already been copied to the kernel stack by this point, so the subsequent access_ok check fails and the ptrace request returns -EFAULT. This causes problems tracing fork() with older versions of strace.
This patch briefly changes the fs to KERNEL_DS, so that the access_ok check passes even with a kernel address.
--- arch/arm64/kernel/ptrace.c | 4 ++++ 1 file changed, 4 insertions(+)
arch/arm64/kernel/ptrace.c +4 -0
@@ -821,6 +821,7 @@ static int compat_ptrace_write_user(stru compat_ulong_t val) {int ret; +mm_segment_t old_fs =get_fs ();if (off &3 || off >=COMPAT_USER_SZ )return -EIO ;
@@ -828,10 +829,13 @@ static int compat_ptrace_write_user(struif (off >=sizeof (compat_elf_gregset_t))return 0 ; +set_fs (KERNEL_DS ); ret =copy_regset_from_user (tsk, &user_aarch32_view,REGSET_COMPAT_GPR , off,sizeof (compat_ulong_t), &val); +set_fs (old_fs); +return ret; }
Greg Kroah-Hartman
maintainerPATCH 36/139 arm64: ptrace: fix empty registers set in prstatus of aarch32 process core
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Victor Kamensky <[email protected]>
commit 2227901a0230d8fde81ba9c602d649839390f56b upstream.
Currently core file of aarch32 process prstatus note has empty registers set. As result aarch32 core files create by V8 kernel are not very useful.
It happens because compat_gpr_get and compat_gpr_set functions can copy registers values to/from either kbuf or ubuf. ELF core file collection function fill_thread_core_info calls compat_gpr_get with kbuf set and ubuf set to 0. But current compat_gpr_get and compat_gpr_set function handle copy to/from only ubuf case.
Fix is to handle kbuf and ubuf as two separate cases in similar way as other functions like user_regset_copyout, user_regset_copyin do.
--- arch/arm64/kernel/ptrace.c | 26 ++++++++++++++++++-------- 1 file changed, 18 insertions(+), 8 deletions(-)
arch/arm64/kernel/ptrace.c +18 -8
@@ -650,11 +650,16 @@ static int compat_gpr_get(struct task_st reg =task_pt_regs (target)->regs[idx]; } - ret =copy_to_user (ubuf, ®,sizeof (reg)); -if (ret) -break ; +if (kbuf) { +memcpy (kbuf, ®,sizeof (reg)); + kbuf +=sizeof (reg); + }else { + ret =copy_to_user (ubuf, ®,sizeof (reg)); +if (ret) +break ; - ubuf +=sizeof (reg); + ubuf +=sizeof (reg); + } }return ret;
@@ -684,11 +689,16 @@ static int compat_gpr_set(struct task_stunsigned int idx = start + i;compat_ulong_t reg; - ret =copy_from_user (®, ubuf,sizeof (reg)); -if (ret) -return ret; +if (kbuf) { +memcpy (®, kbuf,sizeof (reg)); + kbuf +=sizeof (reg); + }else { + ret =copy_from_user (®, ubuf,sizeof (reg)); +if (ret) +return ret; - ubuf +=sizeof (reg); + ubuf +=sizeof (reg); + }switch (idx) {case 15 :
Greg Kroah-Hartman
maintainerPATCH 37/139 idr: fix overflow bug during maximum ID calculation at maximum height
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Lai Jiangshan <[email protected]>
commit 3afb69cb5572b3c8c898c00880803cf1a49852c4 upstream.
idr_replace() open-codes the logic to calculate the maximum valid ID given the height of the idr tree; unfortunately, the open-coded logic doesn't account for the fact that the top layer may have unused slots and over-shifts the limit to zero when the tree is at its maximum height.
The following test code shows it fails to replace the value for id=((1<<27)+42):
static void test5(void)
{
int id;
DEFINE_IDR(test_idr);
#define TEST5_START ((1<<27)+42) /* use the highest layer */ printk(KERN_INFO "Start test5\n");
id = idr_alloc(&test_idr, (void *)1, TEST5_START, 0, GFP_KERNEL);
BUG_ON(id != TEST5_START);
TEST_BUG_ON(idr_replace(&test_idr, (void *)2, TEST5_START) != (void *)1);
idr_destroy(&test_idr);
printk(KERN_INFO "End of test5\n");}
Fix the bug by using idr_max() which correctly takes into account the maximum allowed shift.
sub_alloc() shares the same problem and may incorrectly fail with -EAGAIN; however, this bug doesn't affect correct operation because idr_get_empty_slot(), which already uses idr_max(), retries with the increased @id in such cases.
--- lib/idr.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-)
lib/idr.c +3 -5
@@ -249,7 +249,7 @@ static int sub_alloc(struct idr *idp, in id = (id | ((1 << (IDR_BITS * l)) -1 )) +1 ;/* if already at the top layer, we need to grow */ -if (id >=1 << (idp->layers *IDR_BITS )) { +if (id >idr_max (idp->layers)) { *starting_id = id;return -EAGAIN ; }
@@ -811,12 +811,10 @@ void *idr_replace(struct idr *idp, voidif (!p)return ERR_PTR (-EINVAL ); - n = (p->layer+1 ) *IDR_BITS ; - -if (id >= (1 << n)) +if (id >idr_max (p->layer +1 ))return ERR_PTR (-EINVAL ); - n -=IDR_BITS ; + n = p->layer *IDR_BITS ;while ((n >0 ) && p) { p = p->ary[(id >> n) &IDR_MASK ]; n -=IDR_BITS ;
Greg Kroah-Hartman
maintainerPATCH 38/139 s390/time: cast tv_nsec to u64 prior to shift in update_vsyscall
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Martin Schwidefsky <[email protected]>
commit b6f4296279ab3ada554d993d12844272fd86b36a upstream.
Analog to git commit 28b92e09e25bdc0ae864b22eacf195a74f861389 first cast tk->wall_to_monotonic.tv_nsec to u64 before doing the shift with tk->shift to avoid loosing relevant bits on a 32-bit kernel.
--- arch/s390/kernel/time.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/s390/kernel/time.c +1 -1
@@ -226,7 +226,7 @@ void update_vsyscall(struct timekeeper * vdso_data->wtom_clock_sec = tk->xtime_sec + tk->wall_to_monotonic.tv_sec; vdso_data->wtom_clock_nsec = tk->xtime_nsec + - + (tk->wall_to_monotonic.tv_nsec << tk->shift); + + ((u64 ) tk->wall_to_monotonic.tv_nsec << tk->shift); nsecps = (u64 )NSEC_PER_SEC << tk->shift;while (vdso_data->wtom_clock_nsec >= nsecps) { vdso_data->wtom_clock_nsec -= nsecps;
Greg Kroah-Hartman
maintainerPATCH 39/139 s390/lowcore: reserve 96 bytes for IRB in lowcore
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Christian Borntraeger <[email protected]>
commit 993072ee67aa179c48c85eb19869804e68887d86 upstream.
The IRB might be 96 bytes if the extended-I/O-measurement facility is used. This feature is currently not used by Linux, but struct irb already has the emw defined. So let's make the irb in lowcore match the size of the internal data structure to be future proof.
We also have to add a pad, to correctly align the paste.
The bigger irb field also circumvents a bug in some QEMU versions that always write the emw field on test subchannel and therefore destroy the paste definitions of this CPU. Running under these QEMU version broke some timing functions in the VDSO and all users of these functions, e.g. some JREs.
--- arch/s390/include/asm/lowcore.h | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-)
arch/s390/include/asm/lowcore.h +6 -5
@@ -142,9 +142,9 @@ struct _lowcore {__u8 pad_0x02fc[0x0300 -0x02fc ];/* 0x02fc */ /* Interrupt response block */ -__u8 irb[64 ];/* 0x0300 */ +__u8 irb[96 ];/* 0x0300 */ -__u8 pad_0x0340[0x0e00 -0x0340 ];/* 0x0340 */ +__u8 pad_0x0360[0x0e00 -0x0360 ];/* 0x0360 */ /* * 0xe00 contains the address of the IPL Parameter Information
@@ -288,12 +288,13 @@ struct _lowcore {__u8 pad_0x03a0[0x0400 -0x03a0 ];/* 0x03a0 */ /* Interrupt response block. */ -__u8 irb[64 ];/* 0x0400 */ +__u8 irb[96 ];/* 0x0400 */ +__u8 pad_0x0460[0x0480 -0x0460 ];/* 0x0460 */ /* Per cpu primary space access list */ -__u32 paste[16 ];/* 0x0440 */ +__u32 paste[16 ];/* 0x0480 */ -__u8 pad_0x0480[0x0e00 -0x0480 ];/* 0x0480 */ +__u8 pad_0x04c0[0x0e00 -0x04c0 ];/* 0x04c0 */ /* * 0xe00 contains the address of the IPL Parameter Information
Greg Kroah-Hartman
maintainerPATCH 40/139 regulator: s2mpa01: Use correct register for buck1 ramp delay
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Krzysztof Kozlowski <[email protected]>
commit 112da5cb43427b843e49b8710f53ecdbb3471d9f upstream.
Fix the register for ramp delay of buck1 regulator. Buck1 and buck6 share the field (offset 4) in ramp delay register S2MPA01_REG_RAMP2.
The driver used the same register and field for ramp delay of buck3 and buck1. This lead to updating of ramp delay of buck3 when setting buck1 and actually the ramp delay of buck1 was never set.
--- drivers/regulator/s2mpa01.c | 1 - 1 file changed, 1 deletion(-)
drivers/regulator/s2mpa01.c +0 -1
@@ -116,7 +116,6 @@ static int s2mpa01_set_ramp_delay(struct ramp_delay = s2mpa01->ramp_delay16; ramp_shift =S2MPA01_BUCK16_RAMP_SHIFT ; - ramp_reg =S2MPA01_REG_RAMP1 ;break ;case S2MPA01_BUCK2 : enable_shift =S2MPA01_BUCK2_RAMP_EN_SHIFT ;
Greg Kroah-Hartman
maintainerPATCH 41/139 regulator: s2mps11: Fix accidental enable of buck6 ramp delay
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Krzysztof Kozlowski <[email protected]>
commit b203e0dfe1a2b0ae5e2681e9285056e4ae8560af upstream.
S2MPS11 supports enabling/disabling ramp delay only for buck[2346]. Other bucks have ramp delay enabled always.
However the bit shift for enabling buck6 ramp delay in register is equal to 0. When ramp delay was set for the bucks unsupporting enable/disable (buck[15789] and buck10), the ramp delay for buck6 was also enabled.
--- drivers/regulator/s2mps11.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-)
drivers/regulator/s2mps11.c +10 -5
@@ -202,11 +202,16 @@ static int s2mps11_set_ramp_delay(structif (!ramp_enable) goto ramp_disable; - ret =regmap_update_bits (rdev->regmap,S2MPS11_REG_RAMP , -1 << enable_shift,1 << enable_shift); -if (ret) { -dev_err (&rdev->dev,"failed to enable ramp rate\n" ); -return ret; +/* Ramp delay can be enabled/disabled only for buck[2346] */ +if ((rdev_get_id (rdev) >=S2MPS11_BUCK2 && +rdev_get_id (rdev) <=S2MPS11_BUCK4 ) || +rdev_get_id (rdev) ==S2MPS11_BUCK6 ) { + ret =regmap_update_bits (rdev->regmap,S2MPS11_REG_RAMP , +1 << enable_shift,1 << enable_shift); +if (ret) { +dev_err (&rdev->dev,"failed to enable ramp rate\n" ); +return ret; + } } ramp_val =get_ramp_delay (ramp_delay);
Greg Kroah-Hartman
maintainerPATCH 42/139 regulator: s2mpa01: Fix accidental enable of buck4 ramp delay
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Krzysztof Kozlowski <[email protected]>
commit 51e2fc0a251ba64c68207e4c6f6ac33c891b2465 upstream.
S2MPA01 supports enabling/disabling ramp delay only for buck[1234]. Other bucks have ramp delay enabled always.
However the bit shift for enabling buck4 ramp delay in register is equal to 0. When ramp delay was set for the bucks unsupporting enable/disable (buck[56789] and buck10), the ramp delay for buck4 was also enabled.
--- drivers/regulator/s2mpa01.c | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-)
drivers/regulator/s2mpa01.c +9 -5
@@ -191,11 +191,15 @@ static int s2mpa01_set_ramp_delay(structif (!ramp_enable) goto ramp_disable; - ret =regmap_update_bits (rdev->regmap,S2MPA01_REG_RAMP1 , -1 << enable_shift,1 << enable_shift); -if (ret) { -dev_err (&rdev->dev,"failed to enable ramp rate\n" ); -return ret; +/* Ramp delay can be enabled/disabled only for buck[1234] */ +if (rdev_get_id (rdev) >=S2MPA01_BUCK1 && +rdev_get_id (rdev) <=S2MPA01_BUCK4 ) { + ret =regmap_update_bits (rdev->regmap,S2MPA01_REG_RAMP1 , +1 << enable_shift,1 << enable_shift); +if (ret) { +dev_err (&rdev->dev,"failed to enable ramp rate\n" ); +return ret; + } } ramp_val =get_ramp_delay (ramp_delay);
Greg Kroah-Hartman
maintainerPATCH 43/139 ext4: fix data integrity sync in ordered mode
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Namjae Jeon <[email protected]>
commit 1c8349a17137b93f0a83f276c764a6df1b9a116e upstream.
When we perform a data integrity sync we tag all the dirty pages with PAGECACHE_TAG_TOWRITE at start of ext4_da_writepages. Later we check for this tag in write_cache_pages_da and creates a struct mpage_da_data containing contiguously indexed pages tagged with this tag and sync these pages with a call to mpage_da_map_and_submit. This process is done in while loop until all the PAGECACHE_TAG_TOWRITE pages are synced. We also do journal start and stop in each iteration. journal_stop could initiate journal commit which would call ext4_writepage which in turn will call ext4_bio_write_page even for delayed OR unwritten buffers. When ext4_bio_write_page is called for such buffers, even though it does not sync them but it clears the PAGECACHE_TAG_TOWRITE of the corresponding page and hence these pages are also not synced by the currently running data integrity sync. We will end up with dirty pages although sync is completed.
This could cause a potential data loss when the sync call is followed by a truncate_pagecache call, which is exactly the case in collapse_range. (It will cause generic/127 failure in xfstests)
To avoid this issue, we can use set_page_writeback_keepwrite instead of set_page_writeback, which doesn't clear TOWRITE tag.
--- fs/ext4/ext4.h | 3 ++- fs/ext4/inode.c | 6 ++++-- fs/ext4/page-io.c | 8 ++++++-- include/linux/page-flags.h | 12 +++++++++++- mm/page-writeback.c | 11 ++++++----- 5 files changed, 29 insertions(+), 11 deletions(-)
fs/ext4/ext4.h +2 -1
@@ -2771,7 +2771,8 @@ extern void ext4_io_submit(struct ext4_iextern int ext4_bio_write_page (struct ext4_io_submit *io,struct page *page,int len, -struct writeback_control *wbc); +struct writeback_control *wbc, +bool keep_towrite);/* mmp.c */ extern int ext4_multi_mount_protect (struct super_block *,ext4_fsblk_t );
fs/ext4/inode.c +4 -2
@@ -1846,6 +1846,7 @@ static int ext4_writepage(struct page *pstruct buffer_head *page_bufs =NULL ;struct inode *inode = page->mapping->host;struct ext4_io_submit io_submit; +bool keep_towrite = false;trace_ext4_writepage (page); size =i_size_read (inode);
@@ -1876,6 +1877,7 @@ static int ext4_writepage(struct page *punlock_page (page);return 0 ; } + keep_towrite = true; }if (PageChecked (page) &&ext4_should_journal_data (inode))
@@ -1892,7 +1894,7 @@ static int ext4_writepage(struct page *punlock_page (page);return -ENOMEM ; } - ret =ext4_bio_write_page (&io_submit, page, len, wbc); + ret =ext4_bio_write_page (&io_submit, page, len, wbc, keep_towrite);ext4_io_submit (&io_submit);/* Drop io_end reference we got from init */ ext4_put_io_end_defer (io_submit.io_end);
@@ -1911,7 +1913,7 @@ static int mpage_submit_page(struct mpagelse len =PAGE_CACHE_SIZE ;clear_page_dirty_for_io (page); - err =ext4_bio_write_page (&mpd->io_submit, page, len, mpd->wbc); + err =ext4_bio_write_page (&mpd->io_submit, page, len, mpd->wbc, false);if (!err) mpd->wbc->nr_to_write--; mpd->first_page++;
fs/ext4/page-io.c +6 -2
@@ -401,7 +401,8 @@ submit_and_retry:int ext4_bio_write_page (struct ext4_io_submit *io,struct page *page,int len, -struct writeback_control *wbc) +struct writeback_control *wbc, +bool keep_towrite) {struct inode *inode = page->mapping->host;unsigned block_start, blocksize;
@@ -414,7 +415,10 @@ int ext4_bio_write_page(struct ext4_io_sBUG_ON (!PageLocked (page));BUG_ON (PageWriteback (page)); -set_page_writeback (page); +if (keep_towrite) +set_page_writeback_keepwrite (page); +else +set_page_writeback (page);ClearPageError (page);/*
include/linux/page-flags.h +11 -1
@@ -317,13 +317,23 @@ CLEARPAGEFLAG(Uptodate, uptodate)extern void cancel_dirty_page (struct page *page,unsigned int account_size);int test_clear_page_writeback (struct page *page); -int test_set_page_writeback (struct page *page); +int __test_set_page_writeback (struct page *page,bool keep_write); + +#define test_set_page_writeback (page) \ + __test_set_page_writeback(page, false) +#define test_set_page_writeback_keepwrite (page) \ + __test_set_page_writeback(page, true)static inline void set_page_writeback (struct page *page) {test_set_page_writeback (page); } +static inline void set_page_writeback_keepwrite (struct page *page) +{ +test_set_page_writeback_keepwrite (page); +} +#ifdef CONFIG_PAGEFLAGS_EXTENDED /* * System with lots of page flags available. This allows separate
mm/page-writeback.c +6 -5
@@ -2398,7 +2398,7 @@ int test_clear_page_writeback(struct pagreturn ret; } -int test_set_page_writeback (struct page *page) +int __test_set_page_writeback (struct page *page,bool keep_write) {struct address_space *mapping =page_mapping (page);int ret;
@@ -2423,9 +2423,10 @@ int test_set_page_writeback(struct pageradix_tree_tag_clear (&mapping->page_tree,page_index (page),PAGECACHE_TAG_DIRTY ); -radix_tree_tag_clear (&mapping->page_tree, -page_index (page), -PAGECACHE_TAG_TOWRITE ); +if (!keep_write) +radix_tree_tag_clear (&mapping->page_tree, +page_index (page), +PAGECACHE_TAG_TOWRITE );spin_unlock_irqrestore (&mapping->tree_lock, flags); }else { ret =TestSetPageWriteback (page);
@@ -2436,7 +2437,7 @@ int test_set_page_writeback(struct pagereturn ret; } -EXPORT_SYMBOL (test_set_page_writeback); +EXPORT_SYMBOL (__test_set_page_writeback);/* * Return true if any of the pages in the mapping are marked with the
Greg Kroah-Hartman
maintainerPATCH 44/139 ext4: fix zeroing of page during writeback
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jan Kara <[email protected]>
commit eeece469dedadf3918bad50ad80f4616a0064e90 upstream.
Tail of a page straddling inode size must be zeroed when being written out due to POSIX requirement that modifications of mmaped page beyond inode size must not be written to the file. ext4_bio_write_page() did this only for blocks fully beyond inode size but didn't properly zero blocks partially beyond inode size. Fix this.
The problem has been uncovered by mmap_11-4 test in openposix test suite (part of LTP).
--- fs/ext4/page-io.c | 24 +++++++++++------------- 1 file changed, 11 insertions(+), 13 deletions(-)
fs/ext4/page-io.c +11 -13
@@ -422,6 +422,17 @@ int ext4_bio_write_page(struct ext4_io_sClearPageError (page);/* +* Comments copied from block_write_full_page_endio: +* +* The page straddles i_size. It must be zeroed out on each and every +* writepage invocation because it may be mmapped. "A file is mapped +* in multiples of the page size. For a file that is not a multiple of +* the page size, the remaining memory is zeroed when mapped, and +* writes to that region are not written out to the file." +*/ +if (len <PAGE_CACHE_SIZE ) +zero_user_segment (page, len,PAGE_CACHE_SIZE ); +/* * In the first loop we prepare and mark buffers to submit. We have to * mark all buffers in the page before submitting so that * end_page_writeback() cannot be called from ext4_bio_end_io() when IO
@@ -432,19 +443,6 @@ int ext4_bio_write_page(struct ext4_io_s do { block_start = bh_offset(bh); if (block_start >= len) { - /* - * Comments copied from block_write_full_page_endio: - * - * The page straddles i_size. It must be zeroed out on - * each and every writepage invocation because it may - * be mmapped. "A file is mapped in multiples of the - * page size. For a file that is not a multiple of - * the page size, the remaining memory is zeroed when - * mapped, and writes to that region are not written - * out to the file." - */ - zero_user_segment(page, block_start, - block_start + blocksize); clear_buffer_dirty(bh); set_buffer_uptodate(bh); continue;
Greg Kroah-Hartman
maintainerPATCH 45/139 ext4: fix ZERO_RANGE test failure in data journalling
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Namjae Jeon <[email protected]>
commit e1ee60fd89670da61b0a4bda59f8ffb2b8abea63 upstream.
xfstests generic/091 is failing when mounting ext4 with data=journal. I think that this regression is same problem that occurred prior to collapse range issue. So ZERO RANGE also need to call ext4_force_commit as collapse range.
--- fs/ext4/extents.c | 7 +++++++ 1 file changed, 7 insertions(+)
fs/ext4/extents.c +7 -0
@@ -4744,6 +4744,13 @@ static long ext4_zero_range(struct fileif (!S_ISREG (inode->i_mode))return -EINVAL ; +/* Call ext4_force_commit to flush all data in case of data=journal. */ +if (ext4_should_journal_data (inode)) { + ret =ext4_force_commit (inode->i_sb); +if (ret) +return ret; + } +/* * Write out all dirty pages to avoid race conditions * Then release them.
Greg Kroah-Hartman
maintainerPATCH 46/139 ext4: fix wrong assert in ext4_mb_normalize_request()
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Maurizio Lombardi <[email protected]>
commit b5b60778558cafad17bbcbf63e0310bd3c68eb17 upstream.
The variable "size" is expressed as number of blocks and not as number of clusters, this could trigger a kernel panic when using ext4 with the size of a cluster different from the size of a block.
--- fs/ext4/mballoc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
fs/ext4/mballoc.c +1 -1
@@ -3145,7 +3145,7 @@ ext4_mb_normalize_request(struct ext4_al }BUG_ON (start + size <= ac->ac_o_ex.fe_logical && start > ac->ac_o_ex.fe_logical); -BUG_ON (size <=0 || size >EXT4_CLUSTERS_PER_GROUP (ac->ac_sb)); +BUG_ON (size <=0 || size >EXT4_BLOCKS_PER_GROUP (ac->ac_sb));/* now prepare goal request */
Greg Kroah-Hartman
maintainerPATCH 47/139 matroxfb: perform a dummy read of M_STATUS
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Mikulas Patocka <[email protected]>
commit 972754cfaee94d6e25acf94a497bc0a864d91b7e upstream.
I had occasional screen corruption with the matrox framebuffer driver and I found out that the reason for the corruption is that the hardware blitter accesses the videoram while it is being written to.
The matrox driver has a macro WaitTillIdle() that should wait until the blitter is idle, but it sometimes doesn't work. I added a dummy read mga_inl(M_STATUS) to WaitTillIdle() to fix the problem. The dummy read will flush the write buffer in the PCI chipset, and the next read of M_STATUS will return the hardware status.
Since applying this patch, I had no screen corruption at all.
--- drivers/video/fbdev/matrox/matroxfb_base.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/video/fbdev/matrox/matroxfb_base.h +1 -1
@@ -698,7 +698,7 @@ void matroxfb_unregister_driver(struct m#define mga_fifo (n) do {} while ((mga_inl(M_FIFOSTATUS) & 0xFF) < (n)) -#define WaitTillIdle () do {} while (mga_inl(M_STATUS) & 0x10000) +#define WaitTillIdle () do { mga_inl(M_STATUS); do {} while (mga_inl(M_STATUS) & 0x10000); } while (0)/* code speedup */ #ifdef CONFIG_FB_MATROX_MILLENIUM
Greg Kroah-Hartman
maintainerPATCH 48/139 USB: usb_wwan: fix urb leak in write error path
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: xiao jin <[email protected]>
commit db0904737947d509844e171c9863ecc5b4534005 upstream.
When enable usb serial for modem data, sometimes the tty is blocked in tty_wait_until_sent because portdata->out_busy always is set and have no chance to be cleared.
We find a bug in write error path. usb_wwan_write set portdata->out_busy firstly, then try autopm async with error. No out urb submit and no usb_wwan_outdat_callback to this write, portdata->out_busy can't be cleared.
This patch clear portdata->out_busy if usb_wwan_write try autopm async with error.
option driver")
--- drivers/usb/serial/usb_wwan.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/usb/serial/usb_wwan.c +3 -1
@@ -228,8 +228,10 @@ int usb_wwan_write(struct tty_struct *tt usb_pipeendpoint(this_urb->pipe), i); err =usb_autopm_get_interface_async (port->serial->interface); -if (err <0 ) +if (err <0 ) { +clear_bit (i, &portdata->out_busy);break ; + }/* send the data */ memcpy (this_urb->transfer_buffer, buf, todo);
Greg Kroah-Hartman
maintainerPATCH 49/139 USB: usb_wwan: fix race between write and resume
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: xiao jin <[email protected]>
commit d9e93c08d8d985e5ef89436ebc9f4aad7e31559f upstream.
We find a race between write and resume. usb_wwan_resume run play_delayed() and spin_unlock, but intfdata->suspended still is not set to zero. At this time usb_wwan_write is called and anchor the urb to delay list. Then resume keep running but the delayed urb have no chance to be commit until next resume. If the time of next resume is far away, tty will be blocked in tty_wait_until_sent during time. The race also can lead to writes being reordered.
This patch put play_Delayed and intfdata->suspended together in the spinlock, it's to avoid the write race during resume.
option driver")
--- drivers/usb/serial/usb_wwan.c | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-)
drivers/usb/serial/usb_wwan.c +2 -6
@@ -660,17 +660,15 @@ int usb_wwan_resume(struct usb_serial *s } } +spin_lock_irq (&intfdata->susp_lock);for (i =0 ; i < serial->num_ports; i++) {/* walk all ports */ port = serial->port[i]; portdata =usb_get_serial_port_data (port);/* skip closed ports */ -spin_lock_irq (&intfdata->susp_lock); -if (!portdata || !portdata->opened) { -spin_unlock_irq (&intfdata->susp_lock); +if (!portdata || !portdata->opened)continue ; - }for (j =0 ; j <N_IN_URB ; j++) { urb = portdata->in_urbs[j];
@@ -683,9 +681,7 @@ int usb_wwan_resume(struct usb_serial *s } }play_delayed (port); -spin_unlock_irq (&intfdata->susp_lock); } -spin_lock_irq (&intfdata->susp_lock); intfdata->suspended =0 ;spin_unlock_irq (&intfdata->susp_lock); err_out:
Greg Kroah-Hartman
maintainerPATCH 50/139 USB: usb_wwan: fix write and suspend race
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold <[email protected]>
commit 170fad9e22df0063eba0701adb966786d7a4ec5a upstream.
Fix race between write() and suspend() which could lead to writes being dropped (or I/O while suspended) if the device is runtime suspended while a write request is being processed.
Specifically, suspend() releases the susp_lock after determining the device is idle but before setting the suspended flag, thus leaving a window where a concurrent write() can submit an urb.
option driver")
--- drivers/usb/serial/usb_wwan.c | 13 +++++-------- 1 file changed, 5 insertions(+), 8 deletions(-)
drivers/usb/serial/usb_wwan.c +5 -8
@@ -579,20 +579,17 @@ static void stop_read_write_urbs(structint usb_wwan_suspend (struct usb_serial *serial,pm_message_t message) {struct usb_wwan_intf_private *intfdata = serial->private; -int b; +spin_lock_irq (&intfdata->susp_lock);if (PMSG_IS_AUTO (message)) { -spin_lock_irq (&intfdata->susp_lock); - b = intfdata->in_flight; -spin_unlock_irq (&intfdata->susp_lock); - -if (b) +if (intfdata->in_flight) { +spin_unlock_irq (&intfdata->susp_lock);return -EBUSY ; + } } - -spin_lock_irq (&intfdata->susp_lock); intfdata->suspended =1 ;spin_unlock_irq (&intfdata->susp_lock); +stop_read_write_urbs (serial);return 0 ;
Greg Kroah-Hartman
maintainerPATCH 51/139 USB: usb_wwan: fix urb leak at shutdown
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold <[email protected]>
commit 79eed03e77d481b55d85d1cfe5a1636a0d3897fd upstream.
The delayed-write queue was never emptied at shutdown (close), something which could lead to leaked urbs if the port is closed before being runtime resumed due to a write.
When this happens the output buffer would not drain on close (closing_wait timeout), and after consecutive opens, writes could be corrupted with previously buffered data, transfered with reduced throughput or completely blocked.
Note that unbusy_queued_urb() was simply moved out of CONFIG_PM.
option driver")
--- drivers/usb/serial/usb_wwan.c | 34 ++++++++++++++++++++++------------ 1 file changed, 22 insertions(+), 12 deletions(-)
drivers/usb/serial/usb_wwan.c +22 -12
@@ -414,12 +414,26 @@ int usb_wwan_open(struct tty_struct *tty }EXPORT_SYMBOL (usb_wwan_open); +static void unbusy_queued_urb (struct urb *urb, +struct usb_wwan_port_private *portdata) +{ +int i; + +for (i =0 ; i <N_OUT_URB ; i++) { +if (urb == portdata->out_urbs[i]) { +clear_bit (i, &portdata->out_busy); +break ; + } + } +} +void usb_wwan_close (struct usb_serial_port *port) {int i;struct usb_serial *serial = port->serial;struct usb_wwan_port_private *portdata;struct usb_wwan_intf_private *intfdata = port->serial->private; +struct urb *urb; portdata =usb_get_serial_port_data (port);
@@ -428,6 +442,14 @@ void usb_wwan_close(struct usb_serial_po portdata->opened =0 ;spin_unlock_irq (&intfdata->susp_lock); +for (;;) { + urb =usb_get_from_anchor (&portdata->delayed); +if (!urb) +break ; +unbusy_queued_urb (urb, portdata); +usb_autopm_put_interface_async (serial->interface); + } +for (i =0 ; i <N_IN_URB ; i++)usb_kill_urb (portdata->in_urbs[i]);for (i =0 ; i <N_OUT_URB ; i++)
@@ -596,18 +618,6 @@ int usb_wwan_suspend(struct usb_serial * }EXPORT_SYMBOL (usb_wwan_suspend); -static void unbusy_queued_urb (struct urb *urb,struct usb_wwan_port_private *portdata) -{ -int i; - -for (i =0 ; i <N_OUT_URB ; i++) { -if (urb == portdata->out_urbs[i]) { -clear_bit (i, &portdata->out_busy); -break ; - } - } -} -static void play_delayed (struct usb_serial_port *port) {struct usb_wwan_intf_private *data;
Greg Kroah-Hartman
maintainerPATCH 52/139 USB: usb_wwan: fix potential NULL-deref at resume
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold <[email protected]>
commit 9096f1fbba916c2e052651e9de82fcfb98d4bea7 upstream.
The interrupt urb was submitted unconditionally at resume, something which could lead to a NULL-pointer dereference in the urb completion handler as resume may be called after the port and port data is gone.
Fix this by making sure the interrupt urb is only submitted and active when the port is open.
option driver")
--- drivers/usb/serial/usb_wwan.c | 43 ++++++++++++++++++------------------------ 1 file changed, 19 insertions(+), 24 deletions(-)
drivers/usb/serial/usb_wwan.c +19 -24
@@ -388,6 +388,14 @@ int usb_wwan_open(struct tty_struct *tty portdata =usb_get_serial_port_data (port); intfdata = serial->private; +if (port->interrupt_in_urb) { + err =usb_submit_urb (port->interrupt_in_urb,GFP_KERNEL ); +if (err) { +dev_dbg (&port->dev,"%s: submit int urb failed: %d\n" , + __func__, err); + } + } +/* Start reading from the IN endpoint */ for (i =0 ; i <N_IN_URB ; i++) { urb = portdata->in_urbs[i];
@@ -454,6 +462,7 @@ void usb_wwan_close(struct usb_serial_pousb_kill_urb (portdata->in_urbs[i]);for (i =0 ; i <N_OUT_URB ; i++)usb_kill_urb (portdata->out_urbs[i]); +usb_kill_urb (port->interrupt_in_urb);/* balancing - important as an error cannot be handled*/ usb_autopm_get_interface_no_resume (serial->interface);
@@ -487,7 +496,6 @@ int usb_wwan_port_probe(struct usb_seriastruct usb_wwan_port_private *portdata;struct urb *urb;u8 *buffer; -int err;int i;if (!port->bulk_in_size || !port->bulk_out_size)
@@ -527,13 +535,6 @@ int usb_wwan_port_probe(struct usb_seriausb_set_serial_port_data (port, portdata); -if (port->interrupt_in_urb) { - err =usb_submit_urb (port->interrupt_in_urb,GFP_KERNEL ); -if (err) -dev_dbg (&port->dev,"%s: submit irq_in urb failed %d\n" , - __func__, err); - } -return 0 ; bail_out_error2:
@@ -651,22 +652,6 @@ int usb_wwan_resume(struct usb_serial *sstruct urb *urb;int err =0 ; -/* get the interrupt URBs resubmitted unconditionally */ -for (i =0 ; i < serial->num_ports; i++) { - port = serial->port[i]; -if (!port->interrupt_in_urb) { -dev_dbg (&port->dev,"%s: No interrupt URB for port\n" , __func__); -continue ; - } - err =usb_submit_urb (port->interrupt_in_urb,GFP_NOIO ); -dev_dbg (&port->dev,"Submitted interrupt URB for port (result %d)\n" , err); -if (err <0 ) { -dev_err (&port->dev,"%s: Error %d for interrupt URB\n" , - __func__, err); - goto err_out; - } - } -spin_lock_irq (&intfdata->susp_lock);for (i =0 ; i < serial->num_ports; i++) {/* walk all ports */
@@ -677,6 +662,16 @@ int usb_wwan_resume(struct usb_serial *sif (!portdata || !portdata->opened)continue ; +if (port->interrupt_in_urb) { + err =usb_submit_urb (port->interrupt_in_urb, +GFP_ATOMIC ); +if (err) { +dev_err (&port->dev, +"%s: submit int urb failed: %d\n" , + __func__, err); + } + } +for (j =0 ; j <N_IN_URB ; j++) { urb = portdata->in_urbs[j]; err =usb_submit_urb (urb,GFP_ATOMIC );
Greg Kroah-Hartman
maintainerPATCH 53/139 USB: usb_wwan: fix potential blocked I/O after resume
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold <[email protected]>
commit fb7ad4f93d9f0f7d49beda32f5e7becb94b29a4d upstream.
Keep trying to submit urbs rather than bail out on first read-urb submission error, which would also prevent I/O for any further ports from being resumed.
Instead keep an error count, for all types of failed submissions, and let USB core know that something went wrong.
Also make sure to always clear the suspended flag. Currently a failed read-urb submission would prevent cached writes as well as any subsequent writes from being submitted until next suspend-resume cycle, something which may not even necessarily happen.
Note that USB core currently only logs an error if an interface resume failed.
option driver")
--- drivers/usb/serial/usb_wwan.c | 25 +++++++++++++++++-------- 1 file changed, 17 insertions(+), 8 deletions(-)
drivers/usb/serial/usb_wwan.c +17 -8
@@ -619,12 +619,12 @@ int usb_wwan_suspend(struct usb_serial * }EXPORT_SYMBOL (usb_wwan_suspend); -static void play_delayed (struct usb_serial_port *port) +static int play_delayed (struct usb_serial_port *port) {struct usb_wwan_intf_private *data;struct usb_wwan_port_private *portdata;struct urb *urb; -int err; +int err =0 ; portdata =usb_get_serial_port_data (port); data = port->serial->private;
@@ -641,6 +641,8 @@ static void play_delayed(struct usb_seribreak ; } } + +return err; }int usb_wwan_resume (struct usb_serial *serial)
@@ -650,7 +652,8 @@ int usb_wwan_resume(struct usb_serial *sstruct usb_wwan_intf_private *intfdata = serial->private;struct usb_wwan_port_private *portdata;struct urb *urb; -int err =0 ; +int err; +int err_count =0 ;spin_lock_irq (&intfdata->susp_lock);for (i =0 ; i < serial->num_ports; i++) {
@@ -669,25 +672,31 @@ int usb_wwan_resume(struct usb_serial *sdev_err (&port->dev,"%s: submit int urb failed: %d\n" , __func__, err); + err_count++; } } + err =play_delayed (port); +if (err) + err_count++; +for (j =0 ; j <N_IN_URB ; j++) { urb = portdata->in_urbs[j]; err =usb_submit_urb (urb,GFP_ATOMIC );if (err <0 ) {dev_err (&port->dev,"%s: Error %d for bulk URB %d\n" , __func__, err, i); -spin_unlock_irq (&intfdata->susp_lock); - goto err_out; + err_count++; } } -play_delayed (port); } intfdata->suspended =0 ;spin_unlock_irq (&intfdata->susp_lock); -err_out: -return err; + +if (err_count) +return -EIO ; + +return 0 ; }EXPORT_SYMBOL (usb_wwan_resume);#endif
Greg Kroah-Hartman
maintainerPATCH 54/139 USB: sierra: fix AA deadlock in open error path
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold <[email protected]>
commit 353fe198602e8b4d1c7bdcceb8e60955087201b1 upstream.
Fix AA deadlock in open error path that would call close() and try to grab the already held disc_mutex.
--- drivers/usb/serial/sierra.c | 21 +++++++++++++-------- 1 file changed, 13 insertions(+), 8 deletions(-)
drivers/usb/serial/sierra.c +13 -8
@@ -816,14 +816,9 @@ static int sierra_open(struct tty_struct usb_sndbulkpipe(serial->dev, endpoint) | USB_DIR_IN); err =sierra_submit_rx_urbs (port,GFP_KERNEL ); -if (err) { -/* get rid of everything as in close */ -sierra_close (port); -/* restore balance for autopm */ -if (!serial->disconnected) -usb_autopm_put_interface (serial->interface); -return err; - } +if (err) + goto err_submit; +sierra_send_setup (port); serial->interface->needs_remote_wakeup =1 ;
@@ -833,6 +828,16 @@ static int sierra_open(struct tty_structusb_autopm_put_interface (serial->interface);return 0 ; + +err_submit: +sierra_stop_rx_urbs (port); + +for (i =0 ; i < portdata->num_in_urbs; i++) { +sierra_release_urb (portdata->in_urbs[i]); + portdata->in_urbs[i] =NULL ; + } + +return err; }
Greg Kroah-Hartman
maintainerPATCH 55/139 USB: sierra: fix use after free at suspend/resume
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold <[email protected]>
commit 8452727de70f6ad850cd6d0aaa18b5d9050aa63b upstream.
Fix use after free or NULL-pointer dereference during suspend and resume.
The port data may never have been allocated (port probe failed) or may already have been released by port_remove (e.g. driver is unloaded) when suspend and resume are called.
online")
--- drivers/usb/serial/sierra.c | 6 ++++++ 1 file changed, 6 insertions(+)
drivers/usb/serial/sierra.c +6 -0
@@ -933,6 +933,7 @@ static int sierra_port_remove(struct usbstruct sierra_port_private *portdata; portdata =usb_get_serial_port_data (port); +usb_set_serial_port_data (port,NULL );kfree (portdata);return 0 ;
@@ -949,6 +950,8 @@ static void stop_read_write_urbs(structfor (i =0 ; i < serial->num_ports; ++i) { port = serial->port[i]; portdata =usb_get_serial_port_data (port); +if (!portdata) +continue ;sierra_stop_rx_urbs (port);usb_kill_anchored_urbs (&portdata->active); }
@@ -991,6 +994,9 @@ static int sierra_resume(struct usb_seri port = serial->port[i]; portdata =usb_get_serial_port_data (port); +if (!portdata) +continue ; +while ((urb =usb_get_from_anchor (&portdata->delayed))) {usb_anchor_urb (urb, &portdata->active); intfdata->in_flight++;
Greg Kroah-Hartman
maintainerPATCH 56/139 USB: sierra: fix urb and memory leak in resume error path
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold <[email protected]>
commit 7fdd26a01eb7b6cb6855ff8f69ef4a720720dfcb upstream.
Neither the transfer buffer or the urb itself were released in the resume error path for delayed writes. Also on errors, the remainder of the queue was not even processed, which leads to further urb and buffer leaks.
The same error path also failed to balance the outstanding-urb counter, something which results in degraded throughput or completely blocked writes.
Fix this by releasing urb and buffer and balancing counters on errors, and by always processing the whole queue even when submission of one urb fails.
online")
--- drivers/usb/serial/sierra.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
drivers/usb/serial/sierra.c +6 -2
@@ -1004,8 +1004,12 @@ static int sierra_resume(struct usb_seriif (err <0 ) { intfdata->in_flight--;usb_unanchor_urb (urb); -usb_scuttle_anchored_urbs (&portdata->delayed); -break ; +kfree (urb->transfer_buffer); +usb_free_urb (urb); +spin_lock (&portdata->lock); + portdata->outstanding_urbs--; +spin_unlock (&portdata->lock); +continue ; } }
Greg Kroah-Hartman
maintainerPATCH 57/139 USB: sierra: fix urb and memory leak on disconnect
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold <[email protected]>
commit 014333f77c0b71123d6ef7d31a9724e0699c9548 upstream.
The delayed-write queue was never emptied on disconnect, something which would lead to leaked urbs and transfer buffers if the device is disconnected before being runtime resumed due to a write.
online")
--- drivers/usb/serial/sierra.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)
drivers/usb/serial/sierra.c +13 -0
@@ -759,6 +759,7 @@ static void sierra_close(struct usb_seristruct usb_serial *serial = port->serial;struct sierra_port_private *portdata;struct sierra_intf_private *intfdata = port->serial->private; +struct urb *urb; portdata =usb_get_serial_port_data (port);
@@ -780,6 +781,18 @@ static void sierra_close(struct usb_seri portdata->opened =0 ;spin_unlock_irq (&intfdata->susp_lock); +for (;;) { + urb =usb_get_from_anchor (&portdata->delayed); +if (!urb) +break ; +kfree (urb->transfer_buffer); +usb_free_urb (urb); +usb_autopm_put_interface_async (serial->interface); +spin_lock (&portdata->lock); + portdata->outstanding_urbs--; +spin_unlock (&portdata->lock); + } +sierra_stop_rx_urbs (port);for (i =0 ; i < portdata->num_in_urbs; i++) {sierra_release_urb (portdata->in_urbs[i]);
Greg Kroah-Hartman
maintainerPATCH 58/139 USB: sierra: fix remote wakeup
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold <[email protected]>
commit 80cc0fcbdaeaf10d04ba27779a2d7ceb73d2717a upstream.
Make sure that needs_remote_wake up is always set when there are open ports.
Currently close() would unconditionally set needs_remote_wakeup to 0 even though there might still be open ports. This could lead to blocked input and possibly dropped data on devices that do not support remote wakeup (and which must therefore not be runtime suspended while open).
Add an open_ports counter (protected by the susp_lock) and only clear needs_remote_wakeup when the last port is closed.
online")
--- drivers/usb/serial/sierra.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)
drivers/usb/serial/sierra.c +5 -2
@@ -58,6 +58,7 @@ struct sierra_intf_private {spinlock_t susp_lock; unsigned int suspended:1;int in_flight; +unsigned int open_ports; };static int sierra_set_power_state (struct usb_device *udev,__u16 swiState)
@@ -768,7 +769,6 @@ static void sierra_close(struct usb_serimutex_lock (&serial->disc_mutex);if (!serial->disconnected) { - serial->interface->needs_remote_wakeup =0 ;/* odd error handling due to pm counters */ if (!usb_autopm_get_interface (serial->interface))sierra_send_setup (port);
@@ -779,6 +779,8 @@ static void sierra_close(struct usb_serimutex_unlock (&serial->disc_mutex);spin_lock_irq (&intfdata->susp_lock); portdata->opened =0 ; +if (--intfdata->open_ports ==0 ) + serial->interface->needs_remote_wakeup =0 ;spin_unlock_irq (&intfdata->susp_lock);for (;;) {
@@ -834,9 +836,10 @@ static int sierra_open(struct tty_structsierra_send_setup (port); - serial->interface->needs_remote_wakeup =1 ;spin_lock_irq (&intfdata->susp_lock); portdata->opened =1 ; +if (++intfdata->open_ports ==1 ) + serial->interface->needs_remote_wakeup =1 ;spin_unlock_irq (&intfdata->susp_lock);usb_autopm_put_interface (serial->interface);
Greg Kroah-Hartman
maintainerPATCH 59/139 usb: qcserial: add Netgear AirCard 341U
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Aleksander Morgado <[email protected]>
commit ff1fcd50bc2459744e6f948310bc18eb7d6e8c72 upstream.
--- drivers/usb/serial/qcserial.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/usb/serial/qcserial.c +3 -0
@@ -151,6 +151,9 @@ static const struct usb_device_id id_tab {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9051, 0)}, /* Netgear AirCard 340U Device Management */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9051, 2)}, /* Netgear AirCard 340U NMEA */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9051, 3)}, /* Netgear AirCard 340U Modem */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9055, 0)}, /* Netgear AirCard 341U Device Management */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9055, 2)}, /* Netgear AirCard 341U NMEA */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9055, 3)}, /* Netgear AirCard 341U Modem */ {USB_DEVICE_INTERFACE_NUMBER(0x413c, 0x81a2, 0)}, /* Dell Wireless 5806 Gobi(TM) 4G LTE Mobile Broadband Card Device Management */ {USB_DEVICE_INTERFACE_NUMBER(0x413c, 0x81a2, 2)}, /* Dell Wireless 5806 Gobi(TM) 4G LTE Mobile Broadband Card NMEA */ {USB_DEVICE_INTERFACE_NUMBER(0x413c, 0x81a2, 3)}, /* Dell Wireless 5806 Gobi(TM) 4G LTE Mobile Broadband Card Modem */
Greg Kroah-Hartman
maintainerPATCH 60/139 usb: qcserial: add additional Sierra Wireless QMI devices
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Aleksander Morgado <[email protected]>
commit 0ce5fb58564fd85aa8fd2d24209900e2e845317b upstream.
A set of new VID/PIDs retrieved from the out-of-tree GobiNet/GobiSerial Sierra Wireless drivers.
--- drivers/usb/serial/qcserial.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+)
drivers/usb/serial/qcserial.c +18 -0
@@ -145,15 +145,33 @@ static const struct usb_device_id id_tab {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x901f, 0)}, /* Sierra Wireless EM7355 Device Management */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x901f, 2)}, /* Sierra Wireless EM7355 NMEA */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x901f, 3)}, /* Sierra Wireless EM7355 Modem */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9040, 0)}, /* Sierra Wireless Modem Device Management */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9040, 2)}, /* Sierra Wireless Modem NMEA */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9040, 3)}, /* Sierra Wireless Modem Modem */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9041, 0)}, /* Sierra Wireless MC7305/MC7355 Device Management */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9041, 2)}, /* Sierra Wireless MC7305/MC7355 NMEA */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9041, 3)}, /* Sierra Wireless MC7305/MC7355 Modem */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9051, 0)}, /* Netgear AirCard 340U Device Management */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9051, 2)}, /* Netgear AirCard 340U NMEA */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9051, 3)}, /* Netgear AirCard 340U Modem */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9053, 0)}, /* Sierra Wireless Modem Device Management */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9053, 2)}, /* Sierra Wireless Modem NMEA */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9053, 3)}, /* Sierra Wireless Modem Modem */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9054, 0)}, /* Sierra Wireless Modem Device Management */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9054, 2)}, /* Sierra Wireless Modem NMEA */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9054, 3)}, /* Sierra Wireless Modem Modem */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9055, 0)}, /* Netgear AirCard 341U Device Management */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9055, 2)}, /* Netgear AirCard 341U NMEA */ {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9055, 3)}, /* Netgear AirCard 341U Modem */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9056, 0)}, /* Sierra Wireless Modem Device Management */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9056, 2)}, /* Sierra Wireless Modem NMEA */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9056, 3)}, /* Sierra Wireless Modem Modem */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9060, 0)}, /* Sierra Wireless Modem Device Management */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9060, 2)}, /* Sierra Wireless Modem NMEA */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9060, 3)}, /* Sierra Wireless Modem Modem */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9061, 0)}, /* Sierra Wireless Modem Device Management */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9061, 2)}, /* Sierra Wireless Modem NMEA */ + {USB_DEVICE_INTERFACE_NUMBER(0x1199, 0x9061, 3)}, /* Sierra Wireless Modem Modem */ {USB_DEVICE_INTERFACE_NUMBER(0x413c, 0x81a2, 0)}, /* Dell Wireless 5806 Gobi(TM) 4G LTE Mobile Broadband Card Device Management */ {USB_DEVICE_INTERFACE_NUMBER(0x413c, 0x81a2, 2)}, /* Dell Wireless 5806 Gobi(TM) 4G LTE Mobile Broadband Card NMEA */ {USB_DEVICE_INTERFACE_NUMBER(0x413c, 0x81a2, 3)}, /* Dell Wireless 5806 Gobi(TM) 4G LTE Mobile Broadband Card Modem */
Greg Kroah-Hartman
maintainerPATCH 61/139 USB: serial: fix potential runtime pm imbalance at device remove
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold <[email protected]>
commit c14829fad88dbeda57253590695b85ba51270621 upstream.
Only call usb_autopm_put_interface() if the corresponding usb_autopm_get_interface() was successful.
This prevents a potential runtime PM counter imbalance should usb_autopm_get_interface() fail. Note that the USB PM usage counter is reset when the interface is unbound, but that the runtime PM counter may be left unbalanced.
Also add comment on why we don't need to worry about racing resume/suspend on autopm_get failures.
against probe/remove")
--- drivers/usb/serial/bus.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-)
drivers/usb/serial/bus.c +11 -3
@@ -97,13 +97,19 @@ static int usb_serial_device_remove(strustruct usb_serial_port *port;int retval =0 ;int minor; +int autopm_err; port =to_usb_serial_port (dev);if (!port)return -ENODEV ; -/* make sure suspend/resume doesn't race against port_remove */ -usb_autopm_get_interface (port->serial->interface); +/* +* Make sure suspend/resume doesn't race against port_remove. +* +* Note that no further runtime PM callbacks will be made if +* autopm_get fails. +*/ + autopm_err =usb_autopm_get_interface (port->serial->interface); minor = port->minor;tty_unregister_device (usb_serial_tty_driver, minor);
@@ -117,7 +123,9 @@ static int usb_serial_device_remove(strudev_info (dev,"%s converter now disconnected from ttyUSB%d\n" , driver->description, minor); -usb_autopm_put_interface (port->serial->interface); +if (!autopm_err) +usb_autopm_put_interface (port->serial->interface); +return retval; }
Greg Kroah-Hartman
maintainerPATCH 62/139 media: ivtv: Fix Oops when no firmware is loaded
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Takashi Iwai <[email protected]>
commit deb29e90221a6d4417aa67be971613c353180331 upstream.
When ivtv PCM device is accessed at the state where no firmware is loaded, it oopses like:
BUG: unable to handle kernel NULL pointer dereference at 0000000000000050
IP: [<ffffffffa049a881>] try_mailbox.isra.0+0x11/0x50 [ivtv]
Call Trace:
[<ffffffffa049aa20>] ivtv_api_call+0x160/0x6b0 [ivtv]
[<ffffffffa049af86>] ivtv_api+0x16/0x40 [ivtv]
[<ffffffffa049b10c>] ivtv_vapi+0xac/0xc0 [ivtv]
[<ffffffffa049d40d>] ivtv_start_v4l2_encode_stream+0x19d/0x630 [ivtv]
[<ffffffffa0530653>] snd_ivtv_pcm_capture_open+0x173/0x1c0 [ivtv_alsa]
[<ffffffffa04526f1>] snd_pcm_open_substream+0x51/0x100 [snd_pcm]
[<ffffffffa0452853>] snd_pcm_open+0xb3/0x260 [snd_pcm]
[<ffffffffa0452a37>] snd_pcm_capture_open+0x37/0x50 [snd_pcm]
[<ffffffffa033f557>] snd_open+0xa7/0x1e0 [snd]
[<ffffffff8118a628>] chrdev_open+0x88/0x1d0
[<ffffffff811840be>] do_dentry_open+0x1de/0x270
[<ffffffff81193a73>] do_last+0x1c3/0xec0
[<ffffffff81194826>] path_openat+0xb6/0x670
[<ffffffff81195b65>] do_filp_open+0x35/0x80
[<ffffffff81185449>] do_sys_open+0x129/0x210
[<ffffffff815b782d>] system_call_fastpath+0x1a/0x1fThis patch adds the check of firmware at PCM open callback like other open callbacks of this driver.
Bugzilla: https://apibugzilla.novell.com/show_bug.cgi?id=875440
--- drivers/media/pci/ivtv/ivtv-alsa-pcm.c | 6 ++++++ 1 file changed, 6 insertions(+)
drivers/media/pci/ivtv/ivtv-alsa-pcm.c +6 -0
@@ -159,6 +159,12 @@ static int snd_ivtv_pcm_capture_open(str/* Instruct the CX2341[56] to start sending packets */ snd_ivtv_lock (itvsc); + +if (ivtv_init_on_first_open (itv)) { +snd_ivtv_unlock (itvsc); +return -ENXIO ; + } + s = &itv->streams[IVTV_ENC_STREAM_TYPE_PCM ];v4l2_fh_init (&item.fh, s->vdev);
Greg Kroah-Hartman
maintainerPATCH 63/139 media: stk1160: Avoid stack-allocated buffer for control URBs
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Ezequiel Garcia <[email protected]>
commit 85ac1a1772bb41da895bad83a81f6a62c8f293f6 upstream.
Currently stk1160_read_reg() uses a stack-allocated char to get the read control value. This is wrong because usb_control_msg() requires a kmalloc-ed buffer.
This commit fixes such issue by kmalloc'ating a 1-byte buffer to receive the read value.
While here, let's remove the urb_buf array which was meant for a similar purpose, but never really used.
--- drivers/media/usb/stk1160/stk1160-core.c | 10 +++++++++- drivers/media/usb/stk1160/stk1160.h | 1 - 2 files changed, 9 insertions(+), 2 deletions(-)
drivers/media/usb/stk1160/stk1160-core.c +9 -1
@@ -67,17 +67,25 @@ int stk1160_read_reg(struct stk1160 *dev {int ret;int pipe =usb_rcvctrlpipe (dev->udev,0 ); +u8 *buf; *value =0 ; + + buf =kmalloc (sizeof (u8),GFP_KERNEL ); +if (!buf) +return -ENOMEM ; ret =usb_control_msg (dev->udev, pipe,0x00 ,USB_DIR_IN |USB_TYPE_VENDOR |USB_RECIP_DEVICE , -0x00 , reg, value,sizeof (u8),HZ ); +0x00 , reg, buf,sizeof (u8),HZ );if (ret <0 ) {stk1160_err ("read failed on reg 0x%x (%d)\n" , reg, ret); +kfree (buf);return ret; } + *value = *buf; +kfree (buf);return 0 ; }
drivers/media/usb/stk1160/stk1160.h +0 -1
@@ -143,7 +143,6 @@ struct stk1160 {int num_alt;struct stk1160_isoc_ctl isoc_ctl; -char urb_buf[255 ];/* urb control msg buffer */ /* frame properties */ int width;/* current frame width */
Greg Kroah-Hartman
maintainerPATCH 64/139 media: exynos4-is: Free FIMC-IS CPU memory only when allocated
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sylwester Nawrocki <[email protected]>
commit 404a90abc60f60df2757cb272660e003d326881f upstream.
Ensure dma_free_coherent() is not called with incorrect arguments and only when the memory was actually allocated. This will prevent possible crashes on error paths of the top level media device driver, when fimc-is device gets unregistered and its driver detached.
--- drivers/media/platform/exynos4-is/fimc-is.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/media/platform/exynos4-is/fimc-is.c +3 -0
@@ -367,6 +367,9 @@ static void fimc_is_free_cpu_memory(stru {struct device *dev = &is->pdev->dev; +if (is->memory.vaddr ==NULL ) +return ; +dma_free_coherent (dev, is->memory.size, is->memory.vaddr, is->memory.paddr); }
Greg Kroah-Hartman
maintainerPATCH 65/139 media: exynos4-is: Fix compilation for !CONFIG_COMMON_CLK
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sylwester Nawrocki <[email protected]>
commit f486e7c3cb9849b6a661931fa8c51a43d477046b upstream.
CONFIG_COMMON_CLK is not enabled on S5PV210 platform, so include some clk API data structures conditionally to avoid compilation errors. These #ifdefs will be removed for next kernel release, when the S5PV210 platform moves to DT and the common clk API.
--- drivers/media/platform/exynos4-is/media-dev.c | 2 +- drivers/media/platform/exynos4-is/media-dev.h | 4 ++++ 2 files changed, 5 insertions(+), 1 deletion(-)
drivers/media/platform/exynos4-is/media-dev.c +1 -1
@@ -1520,7 +1520,7 @@ err: }#else #define fimc_md_register_clk_provider (fmd) (0) -#define fimc_md_unregister_clk_provider (fmd) (0) +#define fimc_md_unregister_clk_provider (fmd)#endif static int subdev_notifier_bound(struct v4l2_async_notifier *notifier,
drivers/media/platform/exynos4-is/media-dev.h +4 -0
@@ -94,7 +94,9 @@ struct fimc_sensor_info { };struct cam_clk { +#ifdef CONFIG_COMMON_CLK struct clk_hw hw; +#endif struct fimc_md *fmd; };#define to_cam_clk (_hw) container_of(_hw, struct cam_clk, hw)
@@ -142,7 +144,9 @@ struct fimc_md {struct cam_clk_provider {struct clk *clks[FIMC_MAX_CAMCLKS ]; +#ifdef CONFIG_COMMON_CLK struct clk_onecell_data clk_data; +#endif struct device_node *of_node;struct cam_clk camclk[FIMC_MAX_CAMCLKS ];int num_clocks;
Greg Kroah-Hartman
maintainerPATCH 67/139 ACPICA: utstring: Check array index bound before use.
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: David Binderman <[email protected]>
commit 5d42b0fa25df7ef2f575107597c1aaebe2407d10 upstream.
ACPICA BZ 1077. David Binderman.
References: https://bugs.acpica.org/show_bug.cgi?id=1077
--- drivers/acpi/acpica/utstring.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/acpi/acpica/utstring.c +1 -1
@@ -353,7 +353,7 @@ void acpi_ut_print_string(char *string, }acpi_os_printf ("\"" ); -for (i =0 ; string[i] && (i < max_length); i++) { +for (i =0 ; (i < max_length) && string[i]; i++) {/* Escape sequences */
Greg Kroah-Hartman
maintainerPATCH 68/139 ACPI: Fix conflict between customized DSDT and DSDT local copy
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Lv Zheng <[email protected]>
commit 73577d1df8e1f31f6b1a5eebcdbc334eb0330e47 upstream.
This patch fixes the following issue:
If DSDT is customized, no local DSDT copy is needed.
References: https://bugzilla.kernel.org/show_bug.cgi?id=69711
--- drivers/acpi/bus.c | 7 +++++++ 1 file changed, 7 insertions(+)
drivers/acpi/bus.c +7 -0
@@ -52,6 +52,12 @@ struct proc_dir_entry *acpi_root_dir;EXPORT_SYMBOL (acpi_root_dir); #ifdef CONFIG_X86 +#ifdef CONFIG_ACPI_CUSTOM_DSDT +static inline int set_copy_dsdt(const struct dmi_system_id *id) +{ + return 0; +} +#else static int set_copy_dsdt(const struct dmi_system_id *id) { printk(KERN_NOTICE "%s detected - "
@@ -59,6 +65,7 @@ static int set_copy_dsdt(const struct dm acpi_gbl_copy_dsdt_locally =1 ;return 0 ; } +#endif static struct dmi_system_id dsdt_dmi_table[]__initdata = {/*
Greg Kroah-Hartman
maintainerPATCH 69/139 ACPI / hotplug / PCI: Add hotplug contexts to PCI host bridges
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: "Rafael J. Wysocki" <[email protected]>
commit 882d18a702c66404fcb62b84748f719f9b47441c upstream.
After relatively recent changes in the ACPI-based PCI hotplug (ACPIPHP) code, the acpiphp_check_host_bridge() executed for PCI host bridges via acpi_pci_root_scan_dependent() doesn't do anything useful, because those bridges do not have hotplug contexts. That happens by mistake, so fix it by making acpiphp_enumerate_slots() add hotplug contexts to PCI host bridges too and modify acpiphp_remove_slots() to drop those contexts for host bridges as appropriate.
--- drivers/pci/hotplug/acpiphp.h | 10 ++++++ drivers/pci/hotplug/acpiphp_glue.c | 60 +++++++++++++++++++++++++------------ 2 files changed, 52 insertions(+), 18 deletions(-)
drivers/pci/hotplug/acpiphp.h +10 -0
@@ -142,6 +142,16 @@ static inline acpi_handle func_to_handlereturn func_to_acpi_device (func)->handle; } +struct acpiphp_root_context { +struct acpi_hotplug_context hp; +struct acpiphp_bridge *root_bridge; +}; + +static inline struct acpiphp_root_context *to_acpiphp_root_context (struct acpi_hotplug_context *hp) +{ + return container_of(hp, struct acpiphp_root_context, hp); +} +/* * struct acpiphp_attention_info - device specific attention registration *
drivers/pci/hotplug/acpiphp_glue.c +42 -18
101 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 71/139 rtmutex: Detect changes in the pi lock chain
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Thomas Gleixner <[email protected]>
commit 82084984383babe728e6e3c9a8e5c46278091315 upstream.
When we walk the lock chain, we drop all locks after each step. So the lock chain can change under us before we reacquire the locks. That's harmless in principle as we just follow the wrong lock path. But it can lead to a false positive in the dead lock detection logic:
T0 holds L0
T0 blocks on L1 held by T1
T1 blocks on L2 held by T2
T2 blocks on L3 held by T3
T4 blocks on L4 held by T4
Now we walk the chain
lock T1 -> lock L2 -> adjust L2 -> unlock T1 ->
lock T2 -> adjust T2 -> drop locks
T2 times out and blocks on L0
Now we continue:
lock T2 -> lock L0 -> deadlock detected, but it's not a deadlock at all.
Brad tried to work around that in the deadlock detection logic itself, but the more I looked at it the less I liked it, because it's crystal ball magic after the fact.
We actually can detect a chain change very simple:
lock T1 -> lock L2 -> adjust L2 -> unlock T1 -> lock T2 -> adjust T2 ->
next_lock = T2->pi_blocked_on->lock;
drop locks
T2 times out and blocks on L0
Now we continue:
lock T2 ->
if (next_lock != T2->pi_blocked_on->lock)
return;So if we detect that T2 is now blocked on a different lock we stop the chain walk. That's also correct in the following scenario:
lock T1 -> lock L2 -> adjust L2 -> unlock T1 -> lock T2 -> adjust T2 ->
next_lock = T2->pi_blocked_on->lock;
drop locks
T3 times out and drops L3
T2 acquires L3 and blocks on L4 now
Now we continue:
lock T2 ->
if (next_lock != T2->pi_blocked_on->lock)
return;We don't have to follow up the chain at that point, because T2 propagated our priority up to T4 already.
[ Folded a cleanup patch from peterz ]
--- kernel/locking/rtmutex.c | 95 +++++++++++++++++++++++++++++++++++------------ 1 file changed, 71 insertions(+), 24 deletions(-)
kernel/locking/rtmutex.c +71 -24
@@ -260,27 +260,36 @@ static void rt_mutex_adjust_prio(struct*/ int max_lock_depth =1024 ; +static inline struct rt_mutex *task_blocked_on_lock (struct task_struct *p) +{ +return p->pi_blocked_on ? p->pi_blocked_on->lock :NULL ; +} +/* * Adjust the priority chain. Also used for deadlock detection. * Decreases task's usage by one - may thus free the task. * -* @task: the task owning the mutex (owner) for which a chain walk is probably -* needed +* @task: the task owning the mutex (owner) for which a chain walk is +* probably needed * @deadlock_detect: do we have to carry out deadlock detection? -* @orig_lock: the mutex (can be NULL if we are walking the chain to recheck -* things for a task that has just got its priority adjusted, and -* is waiting on a mutex) +* @orig_lock: the mutex (can be NULL if we are walking the chain to recheck +* things for a task that has just got its priority adjusted, and +* is waiting on a mutex) +* @next_lock: the mutex on which the owner of @orig_lock was blocked before +* we dropped its pi_lock. Is never dereferenced, only used for +* comparison to detect lock chain changes. * @orig_waiter: rt_mutex_waiter struct for the task that has just donated -* its priority to the mutex owner (can be NULL in the case -* depicted above or if the top waiter is gone away and we are -* actually deboosting the owner) -* @top_task: the current top waiter +* its priority to the mutex owner (can be NULL in the case +* depicted above or if the top waiter is gone away and we are +* actually deboosting the owner) +* @top_task: the current top waiter * * Returns 0 or -EDEADLK. */ static int rt_mutex_adjust_prio_chain (struct task_struct *task,int deadlock_detect,struct rt_mutex *orig_lock, +struct rt_mutex *next_lock,struct rt_mutex_waiter *orig_waiter,struct task_struct *top_task) {
@@ -339,6 +348,18 @@ static int rt_mutex_adjust_prio_chain(st goto out_unlock_pi;/* +* We dropped all locks after taking a refcount on @task, so +* the task might have moved on in the lock chain or even left +* the chain completely and blocks now on an unrelated lock or +* on @orig_lock. +* +* We stored the lock on which @task was blocked in @next_lock, +* so we can detect the chain change. +*/ +if (next_lock != waiter->lock) + goto out_unlock_pi; + +/* * Drop out, when the task has no waiters. Note, * top_waiter can be NULL, when we are in the deboosting * mode!
@@ -422,11 +443,26 @@ static int rt_mutex_adjust_prio_chain(st__rt_mutex_adjust_prio (task); } +/* +* Check whether the task which owns the current lock is pi +* blocked itself. If yes we store a pointer to the lock for +* the lock chain change detection above. After we dropped +* task->pi_lock next_lock cannot be dereferenced anymore. +*/ + next_lock =task_blocked_on_lock (task); +raw_spin_unlock_irqrestore (&task->pi_lock, flags); top_waiter =rt_mutex_top_waiter (lock);raw_spin_unlock (&lock->wait_lock); +/* +* We reached the end of the lock chain. Stop right here. No +* point to go back just to figure that out. +*/ +if (!next_lock) + goto out_put_task; +if (!detect_deadlock && waiter != top_waiter) goto out_put_task;
@@ -536,8 +572,9 @@ static int task_blocks_on_rt_mutex(struc {struct task_struct *owner =rt_mutex_owner (lock);struct rt_mutex_waiter *top_waiter = waiter; -unsigned long flags; +struct rt_mutex *next_lock;int chain_walk =0 , res; +unsigned long flags;/* * Early deadlock detection. We really don't want the task to
@@ -569,20 +606,28 @@ static int task_blocks_on_rt_mutex(strucif (!owner)return 0 ; +raw_spin_lock_irqsave (&owner->pi_lock, flags);if (waiter ==rt_mutex_top_waiter (lock)) { -raw_spin_lock_irqsave (&owner->pi_lock, flags);rt_mutex_dequeue_pi (owner, top_waiter);rt_mutex_enqueue_pi (owner, waiter);__rt_mutex_adjust_prio (owner);if (owner->pi_blocked_on) chain_walk =1 ; -raw_spin_unlock_irqrestore (&owner->pi_lock, flags); - } -else if (debug_rt_mutex_detect_deadlock (waiter, detect_deadlock)) + }else if (debug_rt_mutex_detect_deadlock (waiter, detect_deadlock)) { chain_walk =1 ; + } -if (!chain_walk) +/* Store the lock on which owner is blocked or NULL */ + next_lock =task_blocked_on_lock (owner); + +raw_spin_unlock_irqrestore (&owner->pi_lock, flags); +/* +* Even if full deadlock detection is on, if the owner is not +* blocked itself, we can avoid finding this out in the chain +* walk. +*/ +if (!chain_walk || !next_lock)return 0 ;/*
@@ -594,8 +639,8 @@ static int task_blocks_on_rt_mutex(strucraw_spin_unlock (&lock->wait_lock); - res =rt_mutex_adjust_prio_chain (owner, detect_deadlock, lock, waiter, - task); + res =rt_mutex_adjust_prio_chain (owner, detect_deadlock, lock, + next_lock, waiter, task);raw_spin_lock (&lock->wait_lock);
@@ -644,8 +689,8 @@ static void remove_waiter(struct rt_mute {int first = (waiter ==rt_mutex_top_waiter (lock));struct task_struct *owner =rt_mutex_owner (lock); +struct rt_mutex *next_lock =NULL ;unsigned long flags; -int chain_walk =0 ;raw_spin_lock_irqsave (¤t->pi_lock, flags);rt_mutex_dequeue (lock, waiter);
@@ -669,13 +714,13 @@ static void remove_waiter(struct rt_mute }__rt_mutex_adjust_prio (owner); -if (owner->pi_blocked_on) - chain_walk =1 ; +/* Store the lock on which owner is blocked or NULL */ + next_lock =task_blocked_on_lock (owner);raw_spin_unlock_irqrestore (&owner->pi_lock, flags); } -if (!chain_walk) +if (!next_lock)return ;/* gets dropped in rt_mutex_adjust_prio_chain()! */
@@ -683,7 +728,7 @@ static void remove_waiter(struct rt_muteraw_spin_unlock (&lock->wait_lock); -rt_mutex_adjust_prio_chain (owner,0 , lock,NULL , current); +rt_mutex_adjust_prio_chain (owner,0 , lock, next_lock,NULL , current);raw_spin_lock (&lock->wait_lock); }
@@ -696,6 +741,7 @@ static void remove_waiter(struct rt_mutevoid rt_mutex_adjust_pi (struct task_struct *task) {struct rt_mutex_waiter *waiter; +struct rt_mutex *next_lock;unsigned long flags;raw_spin_lock_irqsave (&task->pi_lock, flags);
@@ -706,12 +752,13 @@ void rt_mutex_adjust_pi(struct task_struraw_spin_unlock_irqrestore (&task->pi_lock, flags);return ; } - + next_lock = waiter->lock;raw_spin_unlock_irqrestore (&task->pi_lock, flags);/* gets dropped in rt_mutex_adjust_prio_chain()! */ get_task_struct (task); -rt_mutex_adjust_prio_chain (task,0 ,NULL ,NULL , task); + +rt_mutex_adjust_prio_chain (task,0 ,NULL , next_lock,NULL , task); }/**
Greg Kroah-Hartman
maintainerPATCH 72/139 rtmutex: Handle deadlock detection smarter
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Thomas Gleixner <[email protected]>
commit 3d5c9340d1949733eb37616abd15db36aef9a57c upstream.
Even in the case when deadlock detection is not requested by the caller, we can detect deadlocks. Right now the code stops the lock chain walk and keeps the waiter enqueued, even on itself. Silly not to yell when such a scenario is detected and to keep the waiter enqueued.
Return -EDEADLK unconditionally and handle it at the call sites.
The futex calls return -EDEADLK. The non futex ones dequeue the waiter, throw a warning and put the task into a schedule loop.
Tagged for stable as it makes the code more robust.
--- kernel/locking/rtmutex-debug.h | 5 +++++ kernel/locking/rtmutex.c | 33 ++++++++++++++++++++++++++++----- kernel/locking/rtmutex.h | 5 +++++ 3 files changed, 38 insertions(+), 5 deletions(-)
kernel/locking/rtmutex-debug.h +5 -0
@@ -31,3 +31,8 @@ static inline int debug_rt_mutex_detect_ {return (waiter !=NULL ); } + +static inline void rt_mutex_print_deadlock (struct rt_mutex_waiter *w) +{ +debug_rt_mutex_print_deadlock (w); +}
kernel/locking/rtmutex.c +28 -5
@@ -323,7 +323,7 @@ static int rt_mutex_adjust_prio_chain(st }put_task_struct (task); -return deadlock_detect ? -EDEADLK :0 ; +return -EDEADLK ; } retry:/*
@@ -398,7 +398,7 @@ static int rt_mutex_adjust_prio_chain(stif (lock == orig_lock ||rt_mutex_owner (lock) == top_task) {debug_rt_mutex_deadlock (deadlock_detect, orig_waiter, lock);raw_spin_unlock (&lock->wait_lock); - ret = deadlock_detect ? -EDEADLK :0 ; + ret = -EDEADLK ; goto out_unlock_pi; }
@@ -585,7 +585,7 @@ static int task_blocks_on_rt_mutex(struc* which is wrong, as the other waiter is not in a deadlock * situation. */ -if (detect_deadlock && owner == task) +if (owner == task)return -EDEADLK ;raw_spin_lock_irqsave (&task->pi_lock, flags);
@@ -810,6 +810,26 @@ __rt_mutex_slowlock(struct rt_mutex *locreturn ret; } +static void rt_mutex_handle_deadlock (int res,int detect_deadlock, +struct rt_mutex_waiter *w) +{ +/* +* If the result is not -EDEADLOCK or the caller requested +* deadlock detection, nothing to do here. +*/ +if (res != -EDEADLOCK || detect_deadlock) +return ; + +/* +* Yell lowdly and stop the task right here. +*/ +rt_mutex_print_deadlock (w); +while (1 ) { +set_current_state (TASK_INTERRUPTIBLE ); +schedule (); + } +} +/* * Slow path lock function: */
@@ -849,8 +869,10 @@ rt_mutex_slowlock(struct rt_mutex *lock,set_current_state (TASK_RUNNING ); -if (unlikely (ret)) +if (unlikely (ret)) {remove_waiter (lock, &waiter); +rt_mutex_handle_deadlock (ret, detect_deadlock, &waiter); + }/* * try_to_take_rt_mutex() sets the waiter bit
@@ -1159,7 +1181,8 @@ int rt_mutex_start_proxy_lock(struct rt_return 1 ; } - ret =task_blocks_on_rt_mutex (lock, waiter, task, detect_deadlock); +/* We enforce deadlock detection for futexes */ + ret =task_blocks_on_rt_mutex (lock, waiter, task,1 );if (ret && !rt_mutex_owner (lock)) {/*
kernel/locking/rtmutex.h +5 -0
@@ -24,3 +24,8 @@#define debug_rt_mutex_print_deadlock (w) do { } while (0)#define debug_rt_mutex_detect_deadlock (w,d) (d)#define debug_rt_mutex_reset_waiter (w) do { } while (0) + +static inline void rt_mutex_print_deadlock (struct rt_mutex_waiter *w) +{ +WARN (1 ,"rtmutex deadlock detected\n" ); +}
Greg Kroah-Hartman
maintainerPATCH 73/139 rtmutex: Plug slow unlock race
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Thomas Gleixner <[email protected]>
commit 27e35715df54cbc4f2d044f681802ae30479e7fb upstream.
When the rtmutex fast path is enabled the slow unlock function can create the following situation:
spin_lock(foo->m->wait_lock);
foo->m->owner = NULL; rt_mutex_lock(foo->m); <-- fast path free = atomic_dec_and_test(foo->refcnt); rt_mutex_unlock(foo->m); <-- fast path if (free) kfree(foo);
spin_unlock(foo->m->wait_lock); <--- Use after free.
Plug the race by changing the slow unlock to the following scheme:
while (!rt_mutex_has_waiters(m)) {
/* Clear the waiters bit in m->owner */
clear_rt_mutex_waiters(m);
owner = rt_mutex_owner(m);
spin_unlock(m->wait_lock);
if (cmpxchg(m->owner, owner, 0) == owner)
return;
spin_lock(m->wait_lock);
}So in case of a new waiter incoming while the owner tries the slow path unlock we have two situations:
unlock(wait_lock); lock(wait_lock); cmpxchg(p, owner, 0) == owner mark_rt_mutex_waiters(lock); acquire(lock);
Or:
unlock(wait_lock); lock(wait_lock); mark_rt_mutex_waiters(lock); cmpxchg(p, owner, 0) != owner enqueue_waiter(); unlock(wait_lock); lock(wait_lock); wakeup_next waiter(); unlock(wait_lock); lock(wait_lock); acquire(lock);
If the fast path is disabled, then the simple
m->owner = NULL; unlock(m->wait_lock);
is sufficient as all access to m->owner is serialized via m->wait_lock;
Also document and clarify the wakeup_next_waiter function as suggested by Oleg Nesterov.
--- kernel/locking/rtmutex.c | 115 ++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 109 insertions(+), 6 deletions(-)
kernel/locking/rtmutex.c +109 -6
@@ -83,6 +83,47 @@ static inline void mark_rt_mutex_waiters owner = *p; }while (cmpxchg (p, owner, owner |RT_MUTEX_HAS_WAITERS ) != owner); } + +/* +* Safe fastpath aware unlock: +* 1) Clear the waiters bit +* 2) Drop lock->wait_lock +* 3) Try to unlock the lock with cmpxchg +*/ +static inline bool unlock_rt_mutex_safe (struct rt_mutex *lock) + __releases(lock->wait_lock) +{ +struct task_struct *owner =rt_mutex_owner (lock); + +clear_rt_mutex_waiters (lock); +raw_spin_unlock (&lock->wait_lock); +/* +* If a new waiter comes in between the unlock and the cmpxchg +* we have two situations: +* +* unlock(wait_lock); +* lock(wait_lock); +* cmpxchg(p, owner, 0) == owner +* mark_rt_mutex_waiters(lock); +* acquire(lock); +* or: +* +* unlock(wait_lock); +* lock(wait_lock); +* mark_rt_mutex_waiters(lock); +* +* cmpxchg(p, owner, 0) != owner +* enqueue_waiter(); +* unlock(wait_lock); +* lock(wait_lock); +* wake waiter(); +* unlock(wait_lock); +* lock(wait_lock); +* acquire(lock); +*/ +return rt_mutex_cmpxchg (lock, owner,NULL ); +} +#else # define rt_mutex_cmpxchg (l,c,n) (0)static inline void mark_rt_mutex_waiters (struct rt_mutex *lock)
@@ -90,6 +131,17 @@ static inline void mark_rt_mutex_waiters lock->owner = (struct task_struct *) ((unsigned long)lock->owner | RT_MUTEX_HAS_WAITERS); } + +/* +* Simple slow path only version: lock->owner is protected by lock->wait_lock. +*/ +static inline bool unlock_rt_mutex_safe (struct rt_mutex *lock) + __releases(lock->wait_lock) +{ + lock->owner =NULL ; +raw_spin_unlock (&lock->wait_lock); +return true; +}#endif static inline int
@@ -650,7 +702,8 @@ static int task_blocks_on_rt_mutex(struc/* * Wake up the next waiter on the lock. * -* Remove the top waiter from the current tasks waiter list and wake it up. +* Remove the top waiter from the current tasks pi waiter list and +* wake it up. * * Called with lock->wait_lock held. */
@@ -671,10 +724,23 @@ static void wakeup_next_waiter(struct rt*/ rt_mutex_dequeue_pi (current, waiter); -rt_mutex_set_owner (lock,NULL ); +/* +* As we are waking up the top waiter, and the waiter stays +* queued on the lock until it gets the lock, this lock +* obviously has waiters. Just set the bit here and this has +* the added benefit of forcing all new tasks into the +* slow path making sure no task of lower priority than +* the top waiter can steal this lock. +*/ + lock->owner = (void *)RT_MUTEX_HAS_WAITERS ;raw_spin_unlock_irqrestore (¤t->pi_lock, flags); +/* +* It's safe to dereference waiter as it cannot go away as +* long as we hold lock->wait_lock. The waiter task needs to +* acquire it in order to dequeue the waiter. +*/ wake_up_process (waiter->task); }
@@ -928,12 +994,49 @@ rt_mutex_slowunlock(struct rt_mutex *locrt_mutex_deadlock_account_unlock (current); -if (!rt_mutex_has_waiters (lock)) { - lock->owner =NULL ; -raw_spin_unlock (&lock->wait_lock); -return ; +/* +* We must be careful here if the fast path is enabled. If we +* have no waiters queued we cannot set owner to NULL here +* because of: +* +* foo->lock->owner = NULL; +* rtmutex_lock(foo->lock); <- fast path +* free = atomic_dec_and_test(foo->refcnt); +* rtmutex_unlock(foo->lock); <- fast path +* if (free) +* kfree(foo); +* raw_spin_unlock(foo->lock->wait_lock); +* +* So for the fastpath enabled kernel: +* +* Nothing can set the waiters bit as long as we hold +* lock->wait_lock. So we do the following sequence: +* +* owner = rt_mutex_owner(lock); +* clear_rt_mutex_waiters(lock); +* raw_spin_unlock(&lock->wait_lock); +* if (cmpxchg(&lock->owner, owner, 0) == owner) +* return; +* goto retry; +* +* The fastpath disabled variant is simple as all access to +* lock->owner is serialized by lock->wait_lock: +* +* lock->owner = NULL; +* raw_spin_unlock(&lock->wait_lock); +*/ +while (!rt_mutex_has_waiters (lock)) { +/* Drops lock->wait_lock ! */ +if (unlock_rt_mutex_safe (lock) == true) +return ; +/* Relock the rtmutex and try again */ +raw_spin_lock (&lock->wait_lock); } +/* +* The wakeup next waiter path does not suffer from the above +* race. See the comments there. +*/ wakeup_next_waiter (lock);raw_spin_unlock (&lock->wait_lock);
Greg Kroah-Hartman
maintainerPATCH 74/139 media: uvcvideo: Fix clock param realtime setting
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Olivier Langlois <[email protected]>
commit 3b35fc81e7ec552147a4fd843d0da0bbbe4ef253 upstream.
timestamps in v4l2 buffers returned to userspace are updated in uvc_video_clock_update() which uses timestamps fetched from uvc_video_clock_decode() by calling unconditionally ktime_get_ts().
Hence setting the module clock param to realtime has no effect before this patch.
This has been tested with ffmpeg:
ffmpeg -y -f v4l2 -input_format yuyv422 -video_size 640x480 -framerate 30 -i /dev/video0 \
-f alsa -acodec pcm_s16le -ar 16000 -ac 1 -i default \
-c:v libx264 -preset ultrafast \
-c:a libfdk_aac \
out.mkv
and inspecting the v4l2 input starting timestamp.
--- drivers/media/usb/uvc/uvc_video.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-)
drivers/media/usb/uvc/uvc_video.c +10 -5
@@ -361,6 +361,14 @@ static int uvc_commit_video(struct uvc_s* Clocks and timestamps */ +static inline void uvc_video_get_ts (struct timespec *ts) +{ +if (uvc_clock_param ==CLOCK_MONOTONIC ) +ktime_get_ts (ts); +else +ktime_get_real_ts (ts); +} +static void uvc_video_clock_decode (struct uvc_streaming *stream,struct uvc_buffer *buf,const __u8 *data,int len)
@@ -420,7 +428,7 @@ uvc_video_clock_decode(struct uvc_stream stream->clock.last_sof = dev_sof; host_sof =usb_get_current_frame_number (stream->dev->udev); -ktime_get_ts (&ts); +uvc_video_get_ts (&ts);/* The UVC specification allows device implementations that can't obtain * the USB frame number to keep their own frame counters as long as they
@@ -1011,10 +1019,7 @@ static int uvc_video_decode_start(structreturn -ENODATA ; } -if (uvc_clock_param ==CLOCK_MONOTONIC ) -ktime_get_ts (&ts); -else -ktime_get_real_ts (&ts); +uvc_video_get_ts (&ts); buf->buf.v4l2_buf.sequence = stream->sequence; buf->buf.v4l2_buf.timestamp.tv_sec = ts.tv_sec;
Greg Kroah-Hartman
maintainerPATCH 76/139 media: saa7134: fix regression with tvtime
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Hans Verkuil <[email protected]>
commit 17e7f1b515803e1a79b246688aacbddd2e34165d upstream.
This solves this bug:
https://bugzilla.kernel.org/show_bug.cgi?id=73361
The problem is that when you quit tvtime it calls STREAMOFF, but then it queues a bunch of buffers for no good reason before closing the file descriptor.
In the past closing the fd would free the vb queue since that was part of the file handle struct. Since that was moved to the global struct that no longer happened.
This wouldn't be a problem, but the extra QBUF calls that tvtime does meant that the buffer list in videobuf (q->stream) contained buffers, so REQBUFS would fail with -EBUSY.
The solution is to init the list head explicitly when releasing the file descriptor and to not free the video resource when calling streamoff.
The real fix will hopefully go into kernel 3.16 when the vb2 conversion is merged. Basically the saa7134 driver with the old videobuf is so full of holes it ain't funny anymore, so consider this a band-aid for kernels 3.14 and 15.
--- drivers/media/pci/saa7134/saa7134-video.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-)
drivers/media/pci/saa7134/saa7134-video.c +3 -6
@@ -1243,6 +1243,7 @@ static int video_release(struct file *fivideobuf_streamoff (&dev->cap);res_free (dev, fh,RESOURCE_VIDEO );videobuf_mmap_free (&dev->cap); +INIT_LIST_HEAD (&dev->cap.stream); }if (dev->cap.read_buf) {buffer_release (&dev->cap, dev->cap.read_buf);
@@ -1254,6 +1255,7 @@ static int video_release(struct file *fivideobuf_stop (&dev->vbi);res_free (dev, fh,RESOURCE_VBI );videobuf_mmap_free (&dev->vbi); +INIT_LIST_HEAD (&dev->vbi.stream); }/* ts-capture will not work in planar mode, so turn it off Hac: 04.05*/
@@ -1987,17 +1989,12 @@ int saa7134_streamoff(struct file *file, enum v4l2_buf_type type) {struct saa7134_dev *dev =video_drvdata (file); -int err;int res =saa7134_resource (file);if (res !=RESOURCE_EMPRESS )pm_qos_remove_request (&dev->qos_request); - err =videobuf_streamoff (saa7134_queue (file)); -if (err <0 ) -return err; -res_free (dev, priv, res); -return 0 ; +return videobuf_streamoff (saa7134_queue (file)); }EXPORT_SYMBOL_GPL (saa7134_streamoff);
Greg Kroah-Hartman
maintainerPATCH 77/139 ARM: stacktrace: avoid listing stacktrace functions in stacktrace
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Russell King <[email protected]>
commit 3683f44c42e991d313dc301504ee0fca1aeb8580 upstream.
While debugging the FEC ethernet driver using stacktrace, it was noticed that the stacktraces always begin as follows:
[<c00117b4>] save_stack_trace_tsk+0x0/0x98 [<c0011870>] save_stack_trace+0x24/0x28 ...
This is because the stack trace code includes the stack frames for itself. This is incorrect behaviour, and also leads to "skip" doing the wrong thing (which is the number of stack frames to avoid recording.)
Perversely, it does the right thing when passed a non-current thread. Fix this by ensuring that we have a known constant number of frames above the main stack trace function, and always skip these.
--- arch/arm/kernel/stacktrace.c | 18 +++++++++++++----- 1 file changed, 13 insertions(+), 5 deletions(-)
arch/arm/kernel/stacktrace.c +13 -5
@@ -83,13 +83,16 @@ static int save_trace(struct stackframereturn trace->nr_entries >= trace->max_entries; } -void save_stack_trace_tsk (struct task_struct *tsk,struct stack_trace *trace) +/* This must be noinline to so that our skip calculation works correctly */ +static noinline void __save_stack_trace(struct task_struct *tsk, +struct stack_trace *trace,unsigned int nosched) {struct stack_trace_data data;struct stackframe frame; data.trace = trace; data.skip = trace->skip; + data.no_sched_functions = nosched;if (tsk != current) {#ifdef CONFIG_SMP
@@ -102,7 +105,6 @@ void save_stack_trace_tsk(struct task_st trace->entries[trace->nr_entries++] =ULONG_MAX ;return ;#else - data.no_sched_functions =1 ; frame.fp =thread_saved_fp (tsk); frame.sp =thread_saved_sp (tsk); frame.lr =0 ;/* recovered from the stack */
@@ -111,11 +113,12 @@ void save_stack_trace_tsk(struct task_st }else { registerunsigned long current_sp asm ("sp" ); - data.no_sched_functions =0 ; +/* We don't want this function nor the caller */ + data.skip +=2 ; frame.fp = (unsigned long )__builtin_frame_address (0 ); frame.sp = current_sp; frame.lr = (unsigned long )__builtin_return_address (0 ); - frame.pc = (unsigned long )save_stack_trace_tsk; + frame.pc = (unsigned long )__save_stack_trace; }walk_stackframe (&frame, save_trace, &data);
@@ -123,9 +126,14 @@ void save_stack_trace_tsk(struct task_st trace->entries[trace->nr_entries++] = ULONG_MAX; } +void save_stack_trace_tsk (struct task_struct *tsk,struct stack_trace *trace) +{ +__save_stack_trace (tsk, trace,1 ); +} +void save_stack_trace (struct stack_trace *trace) { -save_stack_trace_tsk (current, trace); +__save_stack_trace (current, trace,0 ); }EXPORT_SYMBOL_GPL (save_stack_trace);#endif
Greg Kroah-Hartman
maintainerPATCH 78/139 ARM: 8037/1: mm: support big-endian page tables
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jianguo Wu <[email protected]>
commit 86f40622af7329375e38f282f6c0aab95f3e5f72 upstream.
When enable LPAE and big-endian in a hisilicon board, while specify mem=384M mem=512M@7680M, will get bad page state:
Freeing unused kernel memory: 180K (c0466000 - c0493000)
BUG: Bad page state in process init pfn:fa442 page:c7749840 count:0 mapcount:-1 mapping: (null) index:0x0 page flags: 0x40000400(reserved) Modules linked in: CPU: 0 PID: 1 Comm: init Not tainted 3.10.27+ #66 [<c000f5f0>] (unwind_backtrace+0x0/0x11c) from [<c000cbc4>] (show_stack+0x10/0x14) [<c000cbc4>] (show_stack+0x10/0x14) from [<c009e448>] (bad_page+0xd4/0x104) [<c009e448>] (bad_page+0xd4/0x104) from [<c009e520>] (free_pages_prepare+0xa8/0x14c) [<c009e520>] (free_pages_prepare+0xa8/0x14c) from [<c009f8ec>] (free_hot_cold_page+0x18/0xf0) [<c009f8ec>] (free_hot_cold_page+0x18/0xf0) from [<c00b5444>] (handle_pte_fault+0xcf4/0xdc8) [<c00b5444>] (handle_pte_fault+0xcf4/0xdc8) from [<c00b6458>] (handle_mm_fault+0xf4/0x120) [<c00b6458>] (handle_mm_fault+0xf4/0x120) from [<c0013754>] (do_page_fault+0xfc/0x354) [<c0013754>] (do_page_fault+0xfc/0x354) from [<c0008400>] (do_DataAbort+0x2c/0x90) [<c0008400>] (do_DataAbort+0x2c/0x90) from [<c0008fb4>] (__dabt_usr+0x34/0x40)
The bad pfn:fa442 is not system memory(mem=384M mem=512M@7680M), after debugging, I find in page fault handler, will get wrong pfn from pte just after set pte, as follow:do_anonymous_page()
{
...
set_pte_at(mm, address, page_table, entry);
//debug code
pfn = pte_pfn(entry);
pr_info("pfn:0x%lx, pte:0x%llxn", pfn, pte_val(entry));
//read out the pte just set
new_pte = pte_offset_map(pmd, address);
new_pfn = pte_pfn(*new_pte);
pr_info("new pfn:0x%lx, new pte:0x%llxn", pfn, pte_val(entry));
...
}pfn: 0x1fa4f5, pte:0xc00001fa4f575f
new_pfn:0xfa4f5, new_pte:0xc00000fa4f5f5f //new pfn/pte is wrong.
The bug is happened in cpu_v7_set_pte_ext(ptep, pte):
An LPAE PTE is a 64bit quantity, passed to cpu_v7_set_pte_ext in the r2 and r3 registers. On an LE kernel, r2 contains the LSB of the PTE, and r3 the MSB.
On a BE kernel, the assignment is reversed.
Unfortunately, the current code always assumes the LE case, leading to corruption of the PTE when clearing/setting bits.
This patch fixes this issue much like it has been done already in the cpu_v7_switch_mm case.
--- arch/arm/mm/proc-v7-3level.S | 18 +++++++++++++----- 1 file changed, 13 insertions(+), 5 deletions(-)
arch/arm/mm/proc-v7-3level.S +13 -5
@@ -64,6 +64,14 @@ ENTRY(cpu_v7_switch_mm) mov pc, lr ENDPROC(cpu_v7_switch_mm) +#ifdef __ARMEB__ +#define rl r3 +#define rh r2 +#else +#define rl r2 +#define rh r3 +#endif + /* * cpu_v7_set_pte_ext(ptep, pte) *
@@ -73,13 +81,13 @@ ENDPROC(cpu_v7_switch_mm) */ ENTRY(cpu_v7_set_pte_ext) #ifdef CONFIG_MMU - tst r2, #L_PTE_VALID + tst rl, #L_PTE_VALID beq 1f - tst r3, #1 << (57 - 32) @ L_PTE_NONE - bicne r2, #L_PTE_VALID + tst rh, #1 << (57 - 32) @ L_PTE_NONE + bicne rl, #L_PTE_VALID bne 1f - tst r3, #1 << (55 - 32) @ L_PTE_DIRTY - orreq r2, #L_PTE_RDONLY + tst rh, #1 << (55 - 32) @ L_PTE_DIRTY + orreq rl, #L_PTE_RDONLY 1: strd r2, r3, [r0] ALT_SMP(W(nop)) ALT_UP (mcr p15, 0, r0, c7, c10, 1) @ flush_pte
Greg Kroah-Hartman
maintainerPATCH 79/139 ARM: OMAP2+: gpmc: enable BCH_HW ecc-scheme for AM43xx platforms
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Pekon Gupta <[email protected]>
commit 2e091d13e65d26f21159323b95b426e5bc42670c upstream.
ARM: OMAP2+: gpmc: update gpmc_hwecc_bch_capable() for new platforms and ECC schemes
Though the commit log of above commit mentions AM43xx platforms, but code change missed AM43xx. This patch adds AM43xx to list of those SoC which have built-in ELM hardware engine, so that BCH ecc-schemes with hardware error-correction can be enabled on AM43xx devices.
--- arch/arm/mach-omap2/gpmc-nand.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/mach-omap2/gpmc-nand.c +1 -1
@@ -46,7 +46,7 @@ static struct platform_device gpmc_nand_static bool gpmc_hwecc_bch_capable (enum omap_ecc ecc_opt) {/* platforms which support all ECC schemes */ -if (soc_is_am33xx () ||cpu_is_omap44xx () || +if (soc_is_am33xx () ||soc_is_am43xx () ||cpu_is_omap44xx () ||soc_is_omap54xx () ||soc_is_dra7xx ())return 1 ;
Greg Kroah-Hartman
maintainerPATCH 80/139 PM / OPP: fix incorrect OPP count handling in of_init_opp_table
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Chander Kashyap <[email protected]>
commit 086abb58590a4df73e8a6ed71fd418826937cd46 upstream.
In of_init_opp_table function, if a failure to add an OPP is detected, the count of OPPs, yet to be added is not updated. Fix this by decrementing this count on failure as well.
--- drivers/base/power/opp.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-)
drivers/base/power/opp.c +1 -3
@@ -734,11 +734,9 @@ int of_init_opp_table(struct device *devunsigned long freq =be32_to_cpup (val++) *1000 ;unsigned long volt =be32_to_cpup (val++); -if (dev_pm_opp_add (dev, freq, volt)) { +if (dev_pm_opp_add (dev, freq, volt))dev_warn (dev,"%s: Failed to add OPP %ld\n" , __func__, freq); -continue ; - } nr -=2 ; }
Greg Kroah-Hartman
maintainerPATCH 82/139 Bluetooth: 6LoWPAN: Fix MAC address universal/local bit handling
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jukka Rissanen <[email protected]>
commit 62bbd5b35994eaf30519f126765d7f6af9cd3526 upstream.
The universal/local bit handling was incorrectly done in the code.
So when setting EUI address from BD address we do this:
- If BD address type is PUBLIC, then we clear the universal bit in EUI address. If the address type is RANDOM, then the universal bit is set (BT 6lowpan draft chapter 3.2.2)
- After this we invert the universal/local bit according to RFC 2464
When figuring out BD address we do the reverse:
- Take EUI address from stateless IPv6 address, invert the universal/local bit according to RFC 2464
- If universal bit is 1 in this modified EUI address, then address type is set to RANDOM, otherwise it is PUBLIC
Note that 6lowpan_iphc.[ch] does the final toggling of U/L bit before sending or receiving the network packet.
--- net/bluetooth/6lowpan.c | 65 +++++++++++++++++++++++++----------------------- 1 file changed, 35 insertions(+), 30 deletions(-)
net/bluetooth/6lowpan.c +35 -30
@@ -420,12 +420,18 @@ static int conn_send(struct l2cap_conn *return 0 ; } -static void get_dest_bdaddr (struct in6_addr *ip6_daddr, -bdaddr_t *addr,u8 *addr_type) +static u8 get_addr_type_from_eui64 (u8 byte) { -u8 *eui64; +/* Is universal(0) or local(1) bit, */ +if (byte &0x02 ) +return ADDR_LE_DEV_RANDOM ; - eui64 = ip6_daddr->s6_addr +8 ; +return ADDR_LE_DEV_PUBLIC ; +} + +static void copy_to_bdaddr (struct in6_addr *ip6_daddr,bdaddr_t *addr) +{ +u8 *eui64 = ip6_daddr->s6_addr +8 ; addr->b[0 ] = eui64[7 ]; addr->b[1 ] = eui64[6 ];
@@ -433,16 +439,19 @@ static void get_dest_bdaddr(struct in6_a addr->b[3 ] = eui64[2 ]; addr->b[4 ] = eui64[1 ]; addr->b[5 ] = eui64[0 ]; +} + +static void convert_dest_bdaddr (struct in6_addr *ip6_daddr, +bdaddr_t *addr,u8 *addr_type) +{ +copy_to_bdaddr (ip6_daddr, addr); - addr->b[5 ] ^=2 ; +/* We need to toggle the U/L bit that we got from IPv6 address +* so that we get the proper address and type of the BD address. +*/ + addr->b[5 ] ^=0x02 ; -/* Set universal/local bit to 0 */ -if (addr->b[5 ] &1 ) { - addr->b[5 ] &= ~1 ; - *addr_type =ADDR_LE_DEV_PUBLIC ; - }else { - *addr_type =ADDR_LE_DEV_RANDOM ; - } + *addr_type =get_addr_type_from_eui64 (addr->b[5 ]); } static int header_create(struct sk_buff *skb, struct net_device *netdev,
@@ -473,9 +482,11 @@ static int header_create(struct sk_buff/* Get destination BT device from skb. * If there is no such peer then discard the packet. */ -get_dest_bdaddr (&hdr->daddr, &addr, &addr_type); +convert_dest_bdaddr (&hdr->daddr, &addr, &addr_type); -BT_DBG ("dest addr %pMR type %d" , &addr, addr_type); +BT_DBG ("dest addr %pMR type %s IP %pI6c" , &addr, + addr_type ==ADDR_LE_DEV_PUBLIC ?"PUBLIC" :"RANDOM" , + &hdr->daddr);read_lock_irqsave (&devices_lock, flags); peer =peer_lookup_ba (dev, &addr, addr_type);
@@ -556,7 +567,7 @@ static netdev_tx_t bt_xmit(struct sk_buf }else {unsigned long flags; -get_dest_bdaddr (&lowpan_cb (skb)->addr, &addr, &addr_type); +convert_dest_bdaddr (&lowpan_cb (skb)->addr, &addr, &addr_type); eui64_addr =lowpan_cb (skb)->addr.s6_addr +8 ; dev =lowpan_dev (netdev);
@@ -564,8 +575,10 @@ static netdev_tx_t bt_xmit(struct sk_buf peer =peer_lookup_ba (dev, &addr, addr_type);read_unlock_irqrestore (&devices_lock, flags); -BT_DBG ("xmit from %s to %pMR (%pI6c) peer %p" , netdev->name, - &addr, &lowpan_cb (skb)->addr, peer); +BT_DBG ("xmit %s to %pMR type %s IP %pI6c peer %p" , + netdev->name, &addr, + addr_type ==ADDR_LE_DEV_PUBLIC ?"PUBLIC" :"RANDOM" , + &lowpan_cb (skb)->addr, peer);if (peer && peer->conn) err = send_pkt(peer->conn, netdev->dev_addr,
@@ -620,13 +633,13 @@ static void set_addr(u8 *eui, u8 *addr, eui[6 ] = addr[1 ]; eui[7 ] = addr[0 ]; - eui[0 ] ^=2 ; - -/* Universal/local bit set, RFC 4291 */ +/* Universal/local bit set, BT 6lowpan draft ch. 3.2.1 */ if (addr_type ==ADDR_LE_DEV_PUBLIC ) - eui[0 ] |=1 ; + eui[0 ] &= ~0x02 ;else - eui[0 ] &= ~1 ; + eui[0 ] |=0x02 ; + +BT_DBG ("type %d addr %*phC" , addr_type,8 , eui); } static void set_dev_addr(struct net_device *netdev, bdaddr_t *addr,
@@ -634,7 +647,6 @@ static void set_dev_addr(struct net_devi { netdev->addr_assign_type =NET_ADDR_PERM ;set_addr (netdev->dev_addr, addr->b, addr_type); - netdev->dev_addr[0 ] ^=2 ; }static void ifup (struct net_device *netdev)
@@ -684,13 +696,6 @@ static int add_peer_conn(struct l2cap_comemcpy (&peer->eui64_addr, (u8 *)&peer->peer_addr.s6_addr +8 ,EUI64_ADDR_LEN ); - peer->eui64_addr[0 ] ^=2 ;/* second bit-flip (Universe/Local) -* is done according RFC2464 -*/ - -raw_dump_inline (__func__,"peer IPv6 address" , - (unsigned char *)&peer->peer_addr,16 ); -raw_dump_inline (__func__,"peer EUI64 address" , peer->eui64_addr,8 );write_lock_irqsave (&devices_lock, flags);INIT_LIST_HEAD (&peer->list);
Greg Kroah-Hartman
maintainerPATCH 84/139 Bluetooth: Fix authentication check for FIPS security level
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hedberg <[email protected]>
commit 7e3691e13ab51f3491e996e2edaf99b173621288 upstream.
When checking whether we need to request authentication or not we should include HCI_SECURITY_FIPS to the levels that always need authentication. This patch fixes check for it in the hci_outgoing_auth_needed() function.
--- net/bluetooth/hci_event.c | 1 + 1 file changed, 1 insertion(+)
net/bluetooth/hci_event.c +1 -0
@@ -1342,6 +1342,7 @@ static int hci_outgoing_auth_needed(stru* is requested. */ if (!hci_conn_ssp_enabled (conn) && !(conn->auth_type &0x01 ) && + conn->pending_sec_level !=BT_SECURITY_FIPS && conn->pending_sec_level !=BT_SECURITY_HIGH && conn->pending_sec_level !=BT_SECURITY_MEDIUM )return 0 ;
Greg Kroah-Hartman
maintainerPATCH 85/139 Bluetooth: Fix requiring SMP MITM for outgoing connections
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hedberg <[email protected]>
commit 79897d2097a629179e142014ecd3cdce6eac7f0e upstream.
Due to recent changes to the way that the MITM requirement is set for outgoing pairing attempts we can no longer rely on the hcon->auth_type variable (which is actually good since it was formed from BR/EDR concepts that don't really exist for SMP).
To match the logic that BR/EDR now uses simply rely on the local IO capability and/or needed security level to set the MITM requirement for outgoing pairing requests.
--- net/bluetooth/smp.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-)
net/bluetooth/smp.c +4 -3
@@ -908,10 +908,11 @@ int smp_conn_security(struct hci_conn *h authreq =seclevel_to_authreq (sec_level); -/* hcon->auth_type is set by pair_device in mgmt.c. If the MITM -* flag is set we should also set it for the SMP request. +/* Require MITM if IO Capability allows or the security level +* requires it. */ -if ((hcon->auth_type &0x01 )) +if (hcon->io_capability !=HCI_IO_NO_INPUT_OUTPUT || + sec_level >BT_SECURITY_MEDIUM ) authreq |=SMP_AUTH_MITM ;if (hcon->link_mode &HCI_LM_MASTER ) {
Greg Kroah-Hartman
maintainerPATCH 86/139 Bluetooth: Fix missing check for FIPS security level
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hedberg <[email protected]>
commit f3fb0b58c85666f73139963a7a04d7878f3d22c9 upstream.
When checking whether a legacy link key provides at least HIGH security level we also need to check for FIPS level which is one step above HIGH. This patch fixes a missing check in the hci_link_key_request_evt() function.
--- net/bluetooth/hci_event.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
net/bluetooth/hci_event.c +2 -1
@@ -2958,7 +2958,8 @@ static void hci_link_key_request_evt(str }if (key->type ==HCI_LK_COMBINATION && key->pin_len <16 && - conn->pending_sec_level ==BT_SECURITY_HIGH ) { + (conn->pending_sec_level ==BT_SECURITY_HIGH || + conn->pending_sec_level ==BT_SECURITY_FIPS )) {BT_DBG ("%s ignoring key unauthenticated for high security" , hdev->name); goto not_found;
Greg Kroah-Hartman
maintainerPATCH 87/139 Target/iser: Bail from accept_np if np_thread is trying to close
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit e346ab343f4f58c12a96725c7b13df9cc2ad56f6 upstream.
In case np_thread state is in RESET/SHUTDOWN/EXIT states, no point for isert to stall there as we may get a hang in case no one will wake it up later.
--- drivers/infiniband/ulp/isert/ib_isert.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
drivers/infiniband/ulp/isert/ib_isert.c +7 -2
@@ -3156,9 +3156,14 @@ accept_wait:return -ENODEV ;spin_lock_bh (&np->np_thread_lock); -if (np->np_thread_state ==ISCSI_NP_THREAD_RESET ) { +if (np->np_thread_state >=ISCSI_NP_THREAD_RESET ) {spin_unlock_bh (&np->np_thread_lock); -pr_debug ("ISCSI_NP_THREAD_RESET for isert_accept_np\n" ); +pr_debug ("np_thread_state %d for isert_accept_np\n" , + np->np_thread_state); +/** +* No point in stalling here when np_thread +* is in state RESET/SHUTDOWN/EXIT - bail +**/ return -ENODEV ; }spin_unlock_bh (&np->np_thread_lock);
Greg Kroah-Hartman
maintainerPATCH 88/139 Target/iser: Fix hangs in connection teardown
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit 9d49f5e284e700576f3b65f1e28dea8539da6661 upstream.
In ungraceful teardowns isert close flows seem racy such that isert_wait_conn hangs as RDMA_CM_EVENT_DISCONNECTED never gets invoked (no one called rdma_disconnect).
Both graceful and ungraceful teardowns will have rx flush errors (isert posts a batch once connection is established). Once all flush errors are consumed we invoke isert_wait_conn and it will be responsible for calling rdma_disconnect. This way it can be sure that rdma_disconnect was called and it won't wait forever.
This patch also removes the logout_posted indicator. either the logout completion was consumed and no problem decrementing the post_send_buf_count, or it was consumed as a flush error. no point of keeping it for isert_wait_conn as there is no danger that isert_conn will be accidentally removed while it is running.
(Drop unnecessary sleep_on_conn_wait_comp check in
isert_cq_rx_comp_err - nab)
--- drivers/infiniband/ulp/isert/ib_isert.c | 31 ++++++++++--------------------- drivers/infiniband/ulp/isert/ib_isert.h | 1 - 2 files changed, 10 insertions(+), 22 deletions(-)
drivers/infiniband/ulp/isert/ib_isert.c +10 -21
@@ -787,14 +787,10 @@ isert_disconnect_work(struct work_structisert_put_conn (isert_conn);return ; } -if (!isert_conn->logout_posted) { -pr_debug ("Calling rdma_disconnect for !logout_posted from" -" isert_disconnect_work\n" ); -rdma_disconnect (isert_conn->conn_cm_id); -mutex_unlock (&isert_conn->conn_mutex); -iscsit_cause_connection_reinstatement (isert_conn->conn,0 ); - goto wake_up; - } + +/* Send DREQ/DREP towards our initiator */ +rdma_disconnect (isert_conn->conn_cm_id); +mutex_unlock (&isert_conn->conn_mutex); wake_up:
@@ -1824,11 +1820,8 @@ isert_do_control_comp(struct work_structbreak ;case ISTATE_SEND_LOGOUTRSP :pr_debug ("Calling iscsit_logout_post_handler >>>>>>>>>>>>>>\n" ); -/* -* Call atomic_dec(&isert_conn->post_send_buf_count) -* from isert_wait_conn() -*/ - isert_conn->logout_posted = true; + +atomic_dec (&isert_conn->post_send_buf_count);iscsit_logout_post_handler (cmd, cmd->conn);break ;case ISTATE_SEND_TEXTRSP :
@@ -2034,6 +2027,8 @@ isert_cq_rx_comp_err(struct isert_conn * isert_conn->state =ISER_CONN_DOWN ;mutex_unlock (&isert_conn->conn_mutex); +iscsit_cause_connection_reinstatement (isert_conn->conn,0 ); +complete (&isert_conn->conn_wait_comp_err); }
@@ -3213,15 +3208,9 @@ static void isert_wait_conn(struct iscsistruct isert_conn *isert_conn = conn->context;pr_debug ("isert_wait_conn: Starting \n" ); -/* -* Decrement post_send_buf_count for special case when called -* from isert_do_control_comp() -> iscsit_logout_post_handler() -*/ -mutex_lock (&isert_conn->conn_mutex); -if (isert_conn->logout_posted) -atomic_dec (&isert_conn->post_send_buf_count); -if (isert_conn->conn_cm_id && isert_conn->state !=ISER_CONN_DOWN ) { +mutex_lock (&isert_conn->conn_mutex); +if (isert_conn->conn_cm_id) {pr_debug ("Calling rdma_disconnect from isert_wait_conn\n" );rdma_disconnect (isert_conn->conn_cm_id); }
drivers/infiniband/ulp/isert/ib_isert.h +0 -1
@@ -116,7 +116,6 @@ struct isert_device;struct isert_conn {enum iser_conn_state state; -bool logout_posted;int post_recv_buf_count;atomic_t post_send_buf_count;u32 responder_resources;
Greg Kroah-Hartman
maintainerPATCH 89/139 Target/iser: Improve cm events handling
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit 88c4015fda6d014392f76d3b1688347950d7a12d upstream.
There are 4 RDMA_CM events that all basically mean that the user should teardown the IB connection:
- DISCONNECTED
- ADDR_CHANGE
- DEVICE_REMOVAL
- TIMEWAIT_EXIT
Only in DISCONNECTED/ADDR_CHANGE it makes sense to call rdma_disconnect (send DREQ/DREP to our initiator). So we keep the same teardown handler for all of them but only indicate calling rdma_disconnect for the relevant events.
This patch also removes redundant debug prints for each single event.
v2 changes:
- Call isert_disconnected_handler() for DEVICE_REMOVAL (Or + Sag)
--- drivers/infiniband/ulp/isert/ib_isert.c | 26 ++++++++++++++------------ drivers/infiniband/ulp/isert/ib_isert.h | 1 + 2 files changed, 15 insertions(+), 12 deletions(-)
drivers/infiniband/ulp/isert/ib_isert.c +14 -12
@@ -788,8 +788,10 @@ isert_disconnect_work(struct work_structreturn ; } -/* Send DREQ/DREP towards our initiator */ -rdma_disconnect (isert_conn->conn_cm_id); +if (isert_conn->disconnect) { +/* Send DREQ/DREP towards our initiator */ +rdma_disconnect (isert_conn->conn_cm_id); + }mutex_unlock (&isert_conn->conn_mutex);
@@ -799,10 +801,11 @@ wake_up: }static void -isert_disconnected_handler (struct rdma_cm_id *cma_id) +isert_disconnected_handler (struct rdma_cm_id *cma_id,bool disconnect) {struct isert_conn *isert_conn = (struct isert_conn *)cma_id->context; + isert_conn->disconnect = disconnect;INIT_WORK (&isert_conn->conn_logout_work, isert_disconnect_work);schedule_work (&isert_conn->conn_logout_work); }
@@ -811,29 +814,28 @@ static int isert_cma_handler(struct rdma_cm_id *cma_id, struct rdma_cm_event *event) { int ret = 0; +bool disconnect = false;pr_debug ("isert_cma_handler: event %d status %d conn %p id %p\n" , event->event, event->status, cma_id->context, cma_id);switch (event->event) {case RDMA_CM_EVENT_CONNECT_REQUEST : -pr_debug ("RDMA_CM_EVENT_CONNECT_REQUEST: >>>>>>>>>>>>>>>\n" ); ret =isert_connect_request (cma_id, event);break ;case RDMA_CM_EVENT_ESTABLISHED : -pr_debug ("RDMA_CM_EVENT_ESTABLISHED >>>>>>>>>>>>>>\n" );isert_connected_handler (cma_id);break ; -case RDMA_CM_EVENT_DISCONNECTED : -pr_debug ("RDMA_CM_EVENT_DISCONNECTED: >>>>>>>>>>>>>>\n" ); -isert_disconnected_handler (cma_id); -break ; -case RDMA_CM_EVENT_DEVICE_REMOVAL : -case RDMA_CM_EVENT_ADDR_CHANGE : +case RDMA_CM_EVENT_ADDR_CHANGE :/* FALLTHRU */ +case RDMA_CM_EVENT_DISCONNECTED :/* FALLTHRU */ +case RDMA_CM_EVENT_DEVICE_REMOVAL :/* FALLTHRU */ + disconnect = true; +case RDMA_CM_EVENT_TIMEWAIT_EXIT :/* FALLTHRU */ +isert_disconnected_handler (cma_id, disconnect);break ;case RDMA_CM_EVENT_CONNECT_ERROR :default : -pr_err ("Unknown RDMA CMA event: %d\n" , event->event); +pr_err ("Unhandled RDMA CMA event: %d\n" , event->event);break ; }
drivers/infiniband/ulp/isert/ib_isert.h +1 -0
@@ -150,6 +150,7 @@ struct isert_conn {#define ISERT_COMP_BATCH_COUNT 8int conn_comp_batch;struct llist_head conn_comp_llist; +bool disconnect; };#define ISERT_MAX_CQ 64
Greg Kroah-Hartman
maintainerPATCH 90/139 Target/iser: Wait for proper cleanup before unloading
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit f5ebec9629cf78eeeea4b8258882a9f439ab2404 upstream.
disconnected_handler works are scheduled on system_wq. When attempting to unload, first make sure all works have cleaned up.
--- drivers/infiniband/ulp/isert/ib_isert.c | 1 + 1 file changed, 1 insertion(+)
drivers/infiniband/ulp/isert/ib_isert.c +1 -0
@@ -3289,6 +3289,7 @@ destroy_rx_wq: static void __exit isert_exit(void) { +flush_scheduled_work ();destroy_workqueue (isert_comp_wq);destroy_workqueue (isert_rx_wq);iscsit_unregister_transport (&iser_target_transport);
Greg Kroah-Hartman
maintainerPATCH 91/139 Target/iser: Gracefully reject T10-PI enabled connect request if not supported
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit 5f80ff8eccba50832dcc640ac89add4c7fced963 upstream.
In case user chose to set T10-PI enable on the target while the IB device does not support it, gracefully reject the request.
--- drivers/infiniband/ulp/isert/ib_isert.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/infiniband/ulp/isert/ib_isert.c +3 -2
@@ -663,8 +663,9 @@ isert_connect_request(struct rdma_cm_id pi_support = np->tpg_np->tpg->tpg_attrib.t10_pi;if (pi_support && !device->pi_capable) { -pr_err ("Protection information requested but not supported\n" ); - ret = -EINVAL ; +pr_err ("Protection information requested but not supported, " +"rejecting connect request\n" ); + ret =rdma_reject (cma_id,NULL ,0 ); goto out_mr; }
Greg Kroah-Hartman
maintainerPATCH 92/139 target: Set CMD_T_ACTIVE bit for Task Management Requests
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Nicholas Bellinger <[email protected]>
commit f15e9cd910c4d9da7de43f2181f362082fc45f0f upstream.
This patch fixes a bug where se_cmd descriptors associated with a Task Management Request (TMR) where not setting CMD_T_ACTIVE before being dispatched into target_tmr_work() process context.
This is required in order for transport_generic_free_cmd() -> transport_wait_for_tasks() to wait on se_cmd->t_transport_stop_comp if a session reset event occurs while an ABORT_TASK is outstanding waiting for another I/O to complete.
--- drivers/target/target_core_transport.c | 6 ++++++ 1 file changed, 6 insertions(+)
drivers/target/target_core_transport.c +6 -0
@@ -2938,6 +2938,12 @@ static void target_tmr_work(struct work_int transport_generic_handle_tmr (struct se_cmd *cmd) { +unsigned long flags; + +spin_lock_irqsave (&cmd->t_state_lock, flags); + cmd->transport_state |=CMD_T_ACTIVE ; +spin_unlock_irqrestore (&cmd->t_state_lock, flags); +INIT_WORK (&cmd->work, target_tmr_work);queue_work (cmd->se_dev->tmr_wq, &cmd->work);return 0 ;
Greg Kroah-Hartman
maintainerPATCH 93/139 target: Use complete_all for se_cmd->t_transport_stop_comp
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Nicholas Bellinger <[email protected]>
commit a95d6511303b848da45ee27b35018bb58087bdc6 upstream.
This patch fixes a bug where multiple waiters on ->t_transport_stop_comp occurs due to a concurrent ABORT_TASK and session reset both invoking transport_wait_for_tasks(), while waiting for the associated se_cmd descriptor backend processing to complete.
For this case, complete_all() should be invoked in order to wake up both waiters in core_tmr_abort_task() + transport_generic_free_cmd() process contexts.
--- drivers/target/target_core_transport.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/target/target_core_transport.c +3 -3
@@ -562,7 +562,7 @@ static int transport_cmd_check_stop(struspin_unlock_irqrestore (&cmd->t_state_lock, flags); -complete (&cmd->t_transport_stop_comp); +complete_all (&cmd->t_transport_stop_comp);return 1 ; }
@@ -687,7 +687,7 @@ void target_complete_cmd(struct se_cmd *if (cmd->transport_state &CMD_T_ABORTED && cmd->transport_state &CMD_T_STOP ) {spin_unlock_irqrestore (&cmd->t_state_lock, flags); -complete (&cmd->t_transport_stop_comp); +complete_all (&cmd->t_transport_stop_comp);return ; }else if (!success) {INIT_WORK (&cmd->work, target_complete_failure_work);
@@ -1761,7 +1761,7 @@ void target_execute_cmd(struct se_cmd *c cmd->se_tfo->get_task_tag(cmd));spin_unlock_irq (&cmd->t_state_lock); -complete (&cmd->t_transport_stop_comp); +complete_all (&cmd->t_transport_stop_comp);return ; }
Greg Kroah-Hartman
maintainerPATCH 94/139 iscsi-target: Fix ABORT_TASK + connection reset iscsi_queue_req memory leak
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Nicholas Bellinger <[email protected]>
commit bbc050488525e1ab1194c27355f63c66814385b8 upstream.
This patch fixes a iscsi_queue_req memory leak when ABORT_TASK response has been queued by TFO->queue_tm_rsp() -> lio_queue_tm_rsp() after a long standing I/O completes, but the connection has already reset and waiting for cleanup to complete in iscsit_release_commands_from_conn() -> transport_generic_free_cmd() -> transport_wait_for_tasks() code.
It moves iscsit_free_queue_reqs_for_conn() after the per-connection command list has been released, so that the associated se_cmd tag can be completed + released by target-core before freeing any remaining iscsi_queue_req memory for the connection generated by lio_queue_tm_rsp().
--- drivers/target/iscsi/iscsi_target.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/target/iscsi/iscsi_target.c +1 -2
@@ -4234,8 +4234,6 @@ int iscsit_close_connection(if (conn->conn_transport->iscsit_wait_conn) conn->conn_transport->iscsit_wait_conn (conn); -iscsit_free_queue_reqs_for_conn (conn); -/* * During Connection recovery drop unacknowledged out of order * commands for this connection, and prepare the other commands
@@ -4252,6 +4250,7 @@ int iscsit_close_connection(iscsit_clear_ooo_cmdsns_for_conn (conn);iscsit_release_commands_from_conn (conn); } +iscsit_free_queue_reqs_for_conn (conn);/* * Handle decrementing session or connection usage count if
Greg Kroah-Hartman
maintainerPATCH 95/139 Target/iser: Fix a wrong dereference in case discovery session is over iser
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit e0546fc1ba66c90cb38a5764357366267d3e58e4 upstream.
In case the discovery session is carried over iser, we can't access the assumed network portal since the default portal is used. In this case we don't really need to allocate the fastreg pool, just prepare to the text pdu that will follow.
--- drivers/infiniband/ulp/isert/ib_isert.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/infiniband/ulp/isert/ib_isert.c +3 -1
@@ -1053,7 +1053,9 @@ isert_put_login_tx(struct iscsi_conn *co }if (!login->login_failed) {if (login->login_complete) { -if (isert_conn->conn_device->use_fastreg) { +if (!conn->sess->sess_ops->SessionType && + isert_conn->conn_device->use_fastreg) { +/* Normal Session and fastreg is used */ u8 pi_support = login->np->tpg_np->tpg->tpg_attrib.t10_pi; ret = isert_conn_create_fastreg_pool(isert_conn,
Greg Kroah-Hartman
maintainerPATCH 96/139 Target/iscsi: Fix sendtargets response pdu for iser transport
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit 22c7aaa57e80853b4904a46c18f97db0036a3b97 upstream.
In case the transport is iser we should not include the iscsi target info in the sendtargets text response pdu. This causes sendtargets response to include the target info twice.
Modify iscsit_build_sendtargets_response to filter transport types that don't match.
--- drivers/infiniband/ulp/isert/ib_isert.c | 2 +- drivers/target/iscsi/iscsi_target.c | 14 ++++++++++---- include/target/iscsi/iscsi_transport.h | 3 ++- 3 files changed, 13 insertions(+), 6 deletions(-)
drivers/infiniband/ulp/isert/ib_isert.c +1 -1
@@ -2320,7 +2320,7 @@ isert_put_text_rsp(struct iscsi_cmd *cmdint rc;isert_create_send_desc (isert_conn, isert_cmd, &isert_cmd->tx_desc); - rc =iscsit_build_text_rsp (cmd, conn, hdr); + rc =iscsit_build_text_rsp (cmd, conn, hdr,ISCSI_INFINIBAND );if (rc <0 )return rc;
drivers/target/iscsi/iscsi_target.c +10 -4
@@ -3390,7 +3390,9 @@ static bool iscsit_check_inaddr_any(stru#define SENDTARGETS_BUF_LIMIT 32768U -static int iscsit_build_sendtargets_response (struct iscsi_cmd *cmd) +static int +iscsit_build_sendtargets_response (struct iscsi_cmd *cmd, +enum iscsit_transport_type network_transport) {char *payload =NULL ;struct iscsi_conn *conn = cmd->conn;
@@ -3467,6 +3469,9 @@ static int iscsit_build_sendtargets_respstruct iscsi_np *np = tpg_np->tpg_np;bool inaddr_any =iscsit_check_inaddr_any (np); +if (np->np_network_transport != network_transport) +continue ; +if (!target_name_printed) { len =sprintf (buf,"TargetName=%s" , tiqn->tiqn);
@@ -3520,11 +3525,12 @@ eob:int iscsit_build_text_rsp (struct iscsi_cmd *cmd,struct iscsi_conn *conn, -struct iscsi_text_rsp *hdr) +struct iscsi_text_rsp *hdr, +enum iscsit_transport_type network_transport) {int text_length, padding; - text_length =iscsit_build_sendtargets_response (cmd); + text_length =iscsit_build_sendtargets_response (cmd, network_transport);if (text_length <0 )return text_length;
@@ -3562,7 +3568,7 @@ static int iscsit_send_text_rsp(u32 tx_size =0 ;int text_length, iov_count =0 , rc; - rc =iscsit_build_text_rsp (cmd, conn, hdr); + rc =iscsit_build_text_rsp (cmd, conn, hdr,ISCSI_TCP );if (rc <0 )return rc;
include/target/iscsi/iscsi_transport.h +2 -1
@@ -70,7 +70,8 @@ extern void iscsit_build_nopin_rsp(strucextern void iscsit_build_task_mgt_rsp (struct iscsi_cmd *,struct iscsi_conn *,struct iscsi_tm_rsp *);extern int iscsit_build_text_rsp (struct iscsi_cmd *,struct iscsi_conn *, -struct iscsi_text_rsp *); +struct iscsi_text_rsp *, +enum iscsit_transport_type );extern void iscsit_build_reject (struct iscsi_cmd *,struct iscsi_conn *,struct iscsi_reject *); extern int iscsit_build_logout_rsp(struct iscsi_cmd *, struct iscsi_conn *,
Greg Kroah-Hartman
maintainerPATCH 97/139 target: Report correct response length for some commands
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Roland Dreier <[email protected]>
commit 2426bd456a61407388b6e61fc5f98dbcbebc50e2 upstream.
When an initiator sends an allocation length bigger than what its command consumes, the target should only return the actual response data and set the residual length to the unused part of the allocation length.
Add a helper function that command handlers (INQUIRY, READ CAPACITY, etc) can use to do this correctly, and use this code to get the correct residual for commands that don't use the full initiator allocation in the handlers for READ CAPACITY, READ CAPACITY(16), INQUIRY, MODE SENSE and REPORT LUNS.
This addresses a handful of failures as reported by Christophe with the Windows Certification Kit:
http://permalink.gmane.org/gmane.linux.scsi.target.devel/6515
--- drivers/target/target_core_sbc.c | 4 ++-- drivers/target/target_core_spc.c | 9 ++++++--- drivers/target/target_core_transport.c | 17 +++++++++++++++++ include/target/target_core_backend.h | 1 + 4 files changed, 26 insertions(+), 5 deletions(-)
drivers/target/target_core_sbc.c +2 -2
@@ -81,7 +81,7 @@ sbc_emulate_readcapacity(struct se_cmd *transport_kunmap_data_sg (cmd); } -target_complete_cmd (cmd,GOOD ); +target_complete_cmd_with_length (cmd,GOOD ,8 );return 0 ; }
@@ -137,7 +137,7 @@ sbc_emulate_readcapacity_16(struct se_cmtransport_kunmap_data_sg (cmd); } -target_complete_cmd (cmd,GOOD ); +target_complete_cmd_with_length (cmd,GOOD ,32 );return 0 ; }
drivers/target/target_core_spc.c +6 -3
@@ -721,6 +721,7 @@ spc_emulate_inquiry(struct se_cmd *cmd)unsigned char *buf;sense_reason_t ret;int p; +int len =0 ; buf =kzalloc (SE_INQUIRY_BUF ,GFP_KERNEL );if (!buf) {
@@ -742,6 +743,7 @@ spc_emulate_inquiry(struct se_cmd *cmd) } ret =spc_emulate_inquiry_std (cmd, buf); + len = buf[4 ] +5 ; goto out; }
@@ -749,6 +751,7 @@ spc_emulate_inquiry(struct se_cmd *cmd)if (cdb[2 ] == evpd_handlers[p].page) { buf[1 ] = cdb[2 ]; ret = evpd_handlers[p].emulate (cmd, buf); + len =get_unaligned_be16 (&buf[2 ]) +4 ; goto out; } }
@@ -765,7 +768,7 @@ out:kfree (buf);if (!ret) -target_complete_cmd (cmd,GOOD ); +target_complete_cmd_with_length (cmd,GOOD , len);return ret; }
@@ -1103,7 +1106,7 @@ set_length:transport_kunmap_data_sg (cmd); } -target_complete_cmd (cmd,GOOD ); +target_complete_cmd_with_length (cmd,GOOD , length);return 0 ; }
@@ -1279,7 +1282,7 @@ done: buf[3 ] = (lun_count &0xff );transport_kunmap_data_sg (cmd); -target_complete_cmd (cmd,GOOD ); +target_complete_cmd_with_length (cmd,GOOD ,8 + lun_count *8 );return 0 ; }EXPORT_SYMBOL (spc_emulate_report_luns);
drivers/target/target_core_transport.c +17 -0
@@ -703,6 +703,23 @@ void target_complete_cmd(struct se_cmd * }EXPORT_SYMBOL (target_complete_cmd); +void target_complete_cmd_with_length (struct se_cmd *cmd,u8 scsi_status,int length) +{ +if (scsi_status ==SAM_STAT_GOOD && length < cmd->data_length) { +if (cmd->se_cmd_flags &SCF_UNDERFLOW_BIT ) { + cmd->residual_count += cmd->data_length - length; + }else { + cmd->se_cmd_flags |=SCF_UNDERFLOW_BIT ; + cmd->residual_count = cmd->data_length - length; + } + + cmd->data_length = length; + } + +target_complete_cmd (cmd, scsi_status); +} +EXPORT_SYMBOL (target_complete_cmd_with_length); +static void target_add_to_state_list (struct se_cmd *cmd) {struct se_device *dev = cmd->se_dev;
include/target/target_core_backend.h +1 -0
@@ -59,6 +59,7 @@ int transport_subsystem_register(structvoid transport_subsystem_release (struct se_subsystem_api *);void target_complete_cmd (struct se_cmd *,u8 ); +void target_complete_cmd_with_length (struct se_cmd *,u8 ,int );sense_reason_t spc_parse_cdb (struct se_cmd *cmd,unsigned int *size);sense_reason_t spc_emulate_report_luns (struct se_cmd *cmd);
Greg Kroah-Hartman
maintainerPATCH 98/139 scsi_cmnd: Introduce scsi_transfer_length helper
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit 8846bab180fa2bcfe02d4ba5288fbaba12c8f4f3 upstream.
In case protection information exists on the wire scsi transports should include it in the transfer byte count (even if protection information does not exist in the host memory space). This helper will compute the total transfer length from the scsi command data length and protection attributes.
--- include/scsi/scsi_cmnd.h | 17 +++++++++++++++++ 1 file changed, 17 insertions(+)
include/scsi/scsi_cmnd.h +17 -0
@@ -7,6 +7,7 @@#include <linux/types.h> #include <linux/timer.h> #include <linux/scatterlist.h> +#include <scsi/scsi_device.h> struct Scsi_Host ;struct scsi_device ;
@@ -306,4 +307,20 @@ static inline void set_driver_byte(struc cmd->result = (cmd->result & 0x00ffffff) | (status << 24); } +static inline unsigned scsi_transfer_length (struct scsi_cmnd *scmd) +{ +unsigned int xfer_len =blk_rq_bytes (scmd->request); +unsigned int prot_op =scsi_get_prot_op (scmd); +unsigned int sector_size = scmd->device->sector_size; + +switch (prot_op) { +case SCSI_PROT_NORMAL : +case SCSI_PROT_WRITE_STRIP : +case SCSI_PROT_READ_INSERT : +return xfer_len; + } + +return xfer_len + (xfer_len >>ilog2 (sector_size)) *8 ; +} +#endif /* _SCSI_SCSI_CMND_H */
Greg Kroah-Hartman
maintainerPATCH 99/139 libiscsi, iser: Adjust data_length to include protection information
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit d77e65350f2d82dfa0557707d505711f5a43c8fd upstream.
In case protection information exists over the wire iscsi header data length is required to include it. Use protection information aware scsi helpers to set the correct transfer length.
In order to avoid breakage, remove iser transfer length checks for each task as they are not always true and somewhat redundant anyway.
--- drivers/infiniband/ulp/iser/iser_initiator.c | 34 +++++++-------------------- drivers/scsi/libiscsi.c | 18 +++++++------- 2 files changed, 19 insertions(+), 33 deletions(-)
drivers/infiniband/ulp/iser/iser_initiator.c +10 -24
@@ -41,11 +41,11 @@#include "iscsi_iser.h" /* Register user buffer memory and initialize passive rdma -* dto descriptor. Total data size is stored in -* iser_task->data[ISER_DIR_IN].data_len +* dto descriptor. Data size is stored in +* task->data[ISER_DIR_IN].data_len, Protection size +* os stored in task->prot[ISER_DIR_IN].data_len */ -static int iser_prepare_read_cmd (struct iscsi_task *task, -unsigned int edtl) +static int iser_prepare_read_cmd (struct iscsi_task *task) {struct iscsi_iser_task *iser_task = task->dd_data;
@@ -73,14 +73,6 @@ static int iser_prepare_read_cmd(structreturn err; } -if (edtl > iser_task->data[ISER_DIR_IN ].data_len) { -iser_err ("Total data length: %ld, less than EDTL: " -"%d, in READ cmd BHS itt: %d, conn: 0x%p\n" , - iser_task->data[ISER_DIR_IN ].data_len, edtl, - task->itt, iser_task->ib_conn); -return -EINVAL ; - } - err = device->iser_reg_rdma_mem (iser_task,ISER_DIR_IN );if (err) {iser_err ("Failed to set up Data-IN RDMA\n" );
@@ -100,8 +92,9 @@ static int iser_prepare_read_cmd(struct }/* Register user buffer memory and initialize passive rdma -* dto descriptor. Total data size is stored in -* task->data[ISER_DIR_OUT].data_len +* dto descriptor. Data size is stored in +* task->data[ISER_DIR_OUT].data_len, Protection size +* is stored at task->prot[ISER_DIR_OUT].data_len */ static int iser_prepare_write_cmd(struct iscsi_task *task,
@@ -135,14 +128,6 @@ iser_prepare_write_cmd(struct iscsi_taskreturn err; } -if (edtl > iser_task->data[ISER_DIR_OUT ].data_len) { -iser_err ("Total data length: %ld, less than EDTL: %d, " -"in WRITE cmd BHS itt: %d, conn: 0x%p\n" , - iser_task->data[ISER_DIR_OUT ].data_len, - edtl, task->itt, task->conn); -return -EINVAL ; - } - err = device->iser_reg_rdma_mem (iser_task,ISER_DIR_OUT );if (err !=0 ) {iser_err ("Failed to register write cmd RDMA mem\n" );
@@ -417,11 +402,12 @@ int iser_send_command(struct iscsi_connif (scsi_prot_sg_count (sc)) { prot_buf->buf =scsi_prot_sglist (sc); prot_buf->size =scsi_prot_sg_count (sc); - prot_buf->data_len = sc->prot_sdb->length; + prot_buf->data_len = data_buf->data_len >> +ilog2 (sc->device->sector_size) *8 ; }if (hdr->flags &ISCSI_FLAG_CMD_READ ) { - err =iser_prepare_read_cmd (task, edtl); + err =iser_prepare_read_cmd (task);if (err) goto send_command_error; }
drivers/scsi/libiscsi.c +9 -9
@@ -338,7 +338,7 @@ static int iscsi_prep_scsi_cmd_pdu(strucstruct iscsi_session *session = conn->session;struct scsi_cmnd *sc = task->sc;struct iscsi_scsi_req *hdr; -unsigned hdrlength, cmd_len; +unsigned hdrlength, cmd_len, transfer_length;itt_t itt;int rc;
@@ -391,11 +391,11 @@ static int iscsi_prep_scsi_cmd_pdu(strucif (scsi_get_prot_op (sc) !=SCSI_PROT_NORMAL ) task->protected = true; + transfer_length =scsi_transfer_length (sc); + hdr->data_length =cpu_to_be32 (transfer_length);if (sc->sc_data_direction ==DMA_TO_DEVICE ) { -unsigned out_len =scsi_out (sc)->length;struct iscsi_r2t_info *r2t = &task->unsol_r2t; - hdr->data_length =cpu_to_be32 (out_len); hdr->flags |=ISCSI_FLAG_CMD_WRITE ;/* * Write counters:
@@ -414,18 +414,19 @@ static int iscsi_prep_scsi_cmd_pdu(strucmemset (r2t,0 ,sizeof (*r2t));if (session->imm_data_en) { -if (out_len >= session->first_burst) +if (transfer_length >= session->first_burst) task->imm_count =min (session->first_burst, conn->max_xmit_dlength);else - task->imm_count =min (out_len, - conn->max_xmit_dlength); + task->imm_count =min (transfer_length, + conn->max_xmit_dlength);hton24 (hdr->dlength, task->imm_count); }else zero_data (hdr->dlength);if (!session->initial_r2t_en) { - r2t->data_length =min (session->first_burst, out_len) - + r2t->data_length =min (session->first_burst, + transfer_length) - task->imm_count; r2t->data_offset = task->imm_count; r2t->ttt =cpu_to_be32 (ISCSI_RESERVED_TAG );
@@ -438,7 +439,6 @@ static int iscsi_prep_scsi_cmd_pdu(struc }else { hdr->flags |=ISCSI_FLAG_CMD_FINAL ;zero_data (hdr->dlength); - hdr->data_length =cpu_to_be32 (scsi_in (sc)->length);if (sc->sc_data_direction ==DMA_FROM_DEVICE ) hdr->flags |=ISCSI_FLAG_CMD_READ ;
@@ -466,7 +466,7 @@ static int iscsi_prep_scsi_cmd_pdu(strucscsi_bidi_cmnd (sc) ?"bidirectional" : sc->sc_data_direction ==DMA_TO_DEVICE ?"write" :"read" , conn->id, sc, sc->cmnd[0 ], - task->itt,scsi_bufflen (sc), + task->itt, transfer_length,scsi_bidi_cmnd (sc) ?scsi_in (sc)->length :0 , session->cmdsn, session->max_cmdsn - session->exp_cmdsn + 1);
Re: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
Quoted 23 lines · Greg Kroah-Hartman
3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit
d77e65350f2d82dfa0557707d505711f5a43c8fdupstream.In case protection information exists over the wire iscsi header data length is required to include it. Use protection information aware scsi helpers to set the correct transfer length.
In order to avoid breakage, remove iser transfer length checks for each task as they are not always true and somewhat redundant anyway.
Signed-off-by Sagi Grimberg <[email protected]>Reviewed-by Mike Christie <[email protected]>Acked-by Mike Christie <[email protected]>Signed-off-by Nicholas Bellinger <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>
Quoted 4 lines
--- drivers/infiniband/ulp/iser/iser_initiator.c | 34 +++++++-------------------- drivers/scsi/libiscsi.c | 18 +++++++------- 2 files changed, 19 insertions(+), 33 deletions(-)
I'm a bit late to object, but this is causing a regression for me on 3.15.3 and 3.16-rc. Using an iscsi_tcp rootfs, I'm seeing the NULL pointer dereference below. Reverting this patch resolves it. Thanks,
Alex
[ 19.672549] BUG: unable to handle kernel NULL pointer dereference at 000000000000000c [ 19.680405] IP: [<ffffffffa0019d38>] iscsi_tcp_segment_done+0x2a8/0x370 [libiscsi_tcp] [ 19.688338] PGD 0 [ 19.690365] Oops: 0000 [#1] SMP [ 19.693624] Modules linked in: cfg80211 snd_hda_codec_analog snd_hda_codec_generic nfsd auth_rpcgss nfs_acl lockd sunrpc snd_hda_intel snd_hda_controller coretemp snd_hda_codec kvm_intel kvm snd_hwdep snd_seq i7core_edac ioatdma snd_seq_device lpc_ich edac_core snd_pcm serio_raw snd_timer shpchp tpm_tis pcspkr i2c_i801 snd mfd_core microcode tpm soundcore acpi_cpufreq nouveau video mxm_wmi drm_kms_helper igb tg3 ttm dca drm i2c_algo_bit ptp crc32c_intel pps_core i2c_core wmi iscsi_tcp libiscsi_tcp libiscsi scsi_transport_iscsi [ 19.741125] CPU: 3 PID: 98 Comm: kworker/u16:5 Tainted: G I 3.15.3 #23 [ 19.748602] Hardware name: LENOVO 4157CTO/LENOVO, BIOS 60KT41AUS 01/04/2011 [ 19.755558] Workqueue: iscsi_q_6 iscsi_xmitworker [libiscsi] [ 19.761231] task: ffff8801e5cb0890 ti: ffff8801e5cc8000 task.ti: ffff8801e5cc8000 [ 19.768702] RIP: 0010:[<ffffffffa0019d38>] [<ffffffffa0019d38>] iscsi_tcp_segment_done+0x2a8/0x370 [libiscsi_tcp] [ 19.779060] RSP: 0018:ffff8801e5ccbc48 EFLAGS: 00010206 [ 19.784365] RAX: 0000000000000000 RBX: ffff8801e5d9dec0 RCX: 0000000000000200 [ 19.791490] RDX: 0000000000000e00 RSI: 0000000000000000 RDI: ffff8801e5d9dec0 [ 19.798615] RBP: ffff8801e5ccbca8 R08: 0000000000000200 R09: 0000000000016640 [ 19.805747] R10: ffff8801e3d48120 R11: 00000000000005a8 R12: 0000000000000200 [ 19.812872] R13: ffff8801e5d9dd30 R14: 0000000000000000 R15: ffff8801e5d9dec0 [ 19.819997] FS: 0000000000000000(0000) GS:ffff88037fcc0000(0000) knlGS:0000000000000000 [ 19.828075] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 19.833812] CR2: 000000000000000c CR3: 0000000001c11000 CR4: 00000000000007e0 [ 19.840938] Stack: [ 19.842947] ffff8801e3d48000 ffff88036dde5400 ffff8801e5d9deb0 ffff8801e5ccbce8 [ 19.850402] ffff8801e5d9dec0 ffff8801e5ccbca8 00000000ef5a5f2a 0000000000000200 [ 19.857856] ffff8801e5d9dd30 ffff8801e5d9deb0 ffff8801e5ccbce8 ffff8801e5d9dec0 [ 19.865310] Call Trace: [ 19.867760] [<ffffffffa003456e>] iscsi_sw_tcp_pdu_xmit+0xde/0x2d0 [iscsi_tcp] [ 19.874979] [<ffffffffa001952a>] iscsi_tcp_task_xmit+0xaa/0x2a0 [libiscsi_tcp] [ 19.882285] [<ffffffff810ac2be>] ? dequeue_task_fair+0x42e/0x640 [ 19.888379] [<ffffffffa0025763>] iscsi_xmit_task+0x53/0xc0 [libiscsi] [ 19.894906] [<ffffffffa00281e0>] iscsi_xmitworker+0x260/0x2f0 [libiscsi] [ 19.901691] [<ffffffff8108d465>] process_one_work+0x175/0x430 [ 19.907522] [<ffffffff8108e0db>] worker_thread+0x11b/0x3a0 [ 19.913096] [<ffffffff8108dfc0>] ? rescuer_thread+0x390/0x390 [ 19.918931] [<ffffffff81094af1>] kthread+0xe1/0x100 [ 19.923896] [<ffffffff81094a10>] ? insert_kthread_work+0x40/0x40 [ 19.929991] [<ffffffff816a6cbc>] ret_from_fork+0x7c/0xb0 [ 19.935389] [<ffffffff81094a10>] ? insert_kthread_work+0x40/0x40 [ 19.941481] Code: b8 00 00 e9 25 fe ff ff 0f 1f 00 48 8b 7b 30 e8 7f 9e 31 e1 8b 53 10 2b 53 14 44 89 f6 48 89 43 30 c7 43 40 00 00 00 00 48 89 df <8b> 40 0c 48 c7 03 00 00 00 00 39 c2 0f 47 d0 89 53 08 e8 41 fc [ 19.961432] RIP [<ffffffffa0019d38>] iscsi_tcp_segment_done+0x2a8/0x370 [libiscsi_tcp] [ 19.969451] RSP <ffff8801e5ccbc48> [ 19.972935] CR2: 000000000000000c [ 19.976247] ---[ end trace f418fccfb3747914 ]---
Re: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
Quoted 24 lines · Alex Williamson
Quoted 23 lines · Greg Kroah-Hartman
Quoted 23 lines
3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit
d77e65350f2d82dfa0557707d505711f5a43c8fdupstream.In case protection information exists over the wire iscsi header data length is required to include it. Use protection information aware scsi helpers to set the correct transfer length.
In order to avoid breakage, remove iser transfer length checks for each task as they are not always true and somewhat redundant anyway.
Signed-off-by Sagi Grimberg <[email protected]>Reviewed-by Mike Christie <[email protected]>Acked-by Mike Christie <[email protected]>Signed-off-by Nicholas Bellinger <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>
Quoted 6 lines
Quoted 4 lines
Quoted 4 lines
--- drivers/infiniband/ulp/iser/iser_initiator.c | 34 +++++++-------------------- drivers/scsi/libiscsi.c | 18 +++++++------- 2 files changed, 19 insertions(+), 33 deletions(-)I'm a bit late to object, but this is causing a regression for me on 3.15.3 and 3.16-rc. Using an
iscsi_tcprootfs, I'm seeing the NULL
A fix for this is in Christoph's tree. Here is the status of the patch.
Re: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
Quoted 29 lines · Mike Christie
Quoted 28 lines · Alex Williamson
Quoted 27 lines · Greg Kroah-Hartman
Quoted 27 lines
3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit
d77e65350f2d82dfa0557707d505711f5a43c8fdupstream.In case protection information exists over the wire iscsi header data length is required to include it. Use protection information aware scsi helpers to set the correct transfer length.
In order to avoid breakage, remove iser transfer length checks for each task as they are not always true and somewhat redundant anyway.
Signed-off-by Sagi Grimberg <[email protected]>Reviewed-by Mike Christie <[email protected]>Acked-by Mike Christie <[email protected]>Signed-off-by Nicholas Bellinger <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>--- drivers/infiniband/ulp/iser/iser_initiator.c | 34 +++++++-------------------- drivers/scsi/libiscsi.c | 18 +++++++------- 2 files changed, 19 insertions(+), 33 deletions(-)
Quoted 7 lines
Quoted 2 lines
I'm a bit late to object, but this is causing a regression for me on 3.15.3 and 3.16-rc. Using an
iscsi_tcprootfs, I'm seeing the NULLA fix for this is in Christoph's tree. Here is the status of the patch.
Thanks for the pointer. It's curious why there was no objection to this going into stable if it's been known to be broken for a week.
Re: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
Quoted 2 lines · Alex Williamson
Thanks for the pointer. It's curious why there was no objection to this going into stable if it's been known to be broken for a week.
I've never seen the stable patch on linux-scsi. IMHO it should never have been proposed for -stable with all the issues it caused.
James Bottomley
maintainerRe: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
In reply to Mike ChristieQuoted 29 lines · Mike Christie
Quoted 28 lines · Alex Williamson
Quoted 27 lines · Greg Kroah-Hartman
Quoted 27 lines
3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit
d77e65350f2d82dfa0557707d505711f5a43c8fdupstream.In case protection information exists over the wire iscsi header data length is required to include it. Use protection information aware scsi helpers to set the correct transfer length.
In order to avoid breakage, remove iser transfer length checks for each task as they are not always true and somewhat redundant anyway.
Signed-off-by Sagi Grimberg <[email protected]>Reviewed-by Mike Christie <[email protected]>Acked-by Mike Christie <[email protected]>Signed-off-by Nicholas Bellinger <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>--- drivers/infiniband/ulp/iser/iser_initiator.c | 34 +++++++-------------------- drivers/scsi/libiscsi.c | 18 +++++++------- 2 files changed, 19 insertions(+), 33 deletions(-)
Quoted 7 lines
Quoted 2 lines
I'm a bit late to object, but this is causing a regression for me on 3.15.3 and 3.16-rc. Using an
iscsi_tcprootfs, I'm seeing the NULLA fix for this is in Christoph's tree. Here is the status of the patch.
I assume it's this fix:
commit 7bafc240e126c37f60b040fa429ae4a710b88ab5
Author: Martin K. Petersen <[email protected]> Date: Tue Jun 24 16:59:35 2014 +0200
scsi: use the scsi data buffer length to extract transfer size
In which case it's not marked for stable backport ... I assume it should be?
James
Re: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
Quoted 40 lines · James Bottomley
Quoted 29 lines · Mike Christie
Quoted 28 lines · Alex Williamson
Quoted 27 lines · Greg Kroah-Hartman
Quoted 27 lines
3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit
d77e65350f2d82dfa0557707d505711f5a43c8fdupstream.In case protection information exists over the wire iscsi header data length is required to include it. Use protection information aware scsi helpers to set the correct transfer length.
In order to avoid breakage, remove iser transfer length checks for each task as they are not always true and somewhat redundant anyway.
Signed-off-by Sagi Grimberg <[email protected]>Reviewed-by Mike Christie <[email protected]>Acked-by Mike Christie <[email protected]>Signed-off-by Nicholas Bellinger <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>--- drivers/infiniband/ulp/iser/iser_initiator.c | 34 +++++++-------------------- drivers/scsi/libiscsi.c | 18 +++++++------- 2 files changed, 19 insertions(+), 33 deletions(-)Quoted 7 lines
Quoted 2 lines
I'm a bit late to object, but this is causing a regression for me on 3.15.3 and 3.16-rc. Using an
iscsi_tcprootfs, I'm seeing the NULLA fix for this is in Christoph's tree. Here is the status of the patch.
I assume it's this fix:
Quoted 8 lines
commit
7bafc240e126c37f60b040fa429ae4a710b88ab5Author: Martin K. Petersen <[email protected]> Date: Tue Jun 24 16:59:35 2014 +0200scsi: use the scsi data buffer length to extract transfer size
In which case it's not marked for stable backport ... I assume it should be?
Yes, that's the one. It definitely needs a 3.15 stable backport. Thanks,
Alex
Re: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
Quoted 45 lines · Alex Williamson
Quoted 44 lines · James Bottomley
Quoted 29 lines · Mike Christie
Quoted 28 lines · Alex Williamson
Quoted 27 lines · Greg Kroah-Hartman
Quoted 27 lines
3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit
d77e65350f2d82dfa0557707d505711f5a43c8fdupstream.In case protection information exists over the wire iscsi header data length is required to include it. Use protection information aware scsi helpers to set the correct transfer length.
In order to avoid breakage, remove iser transfer length checks for each task as they are not always true and somewhat redundant anyway.
Signed-off-by Sagi Grimberg <[email protected]>Reviewed-by Mike Christie <[email protected]>Acked-by Mike Christie <[email protected]>Signed-off-by Nicholas Bellinger <[email protected]>Signed-off-by Greg Kroah-Hartman <[email protected]>--- drivers/infiniband/ulp/iser/iser_initiator.c | 34 +++++++-------------------- drivers/scsi/libiscsi.c | 18 +++++++------- 2 files changed, 19 insertions(+), 33 deletions(-)Quoted 7 lines
Quoted 2 lines
I'm a bit late to object, but this is causing a regression for me on 3.15.3 and 3.16-rc. Using an
iscsi_tcprootfs, I'm seeing the NULLA fix for this is in Christoph's tree. Here is the status of the patch.
I assume it's this fix:
commit
7bafc240e126c37f60b040fa429ae4a710b88ab5Author: Martin K. Petersen <[email protected]> Date: Tue Jun 24 16:59:35 2014 +0200
Quoted 8 lines
Quoted 4 lines
scsi: use the scsi data buffer length to extract transfer size
In which case it's not marked for stable backport ... I assume it should be?
Yes, that's the one. It definitely needs a 3.15 stable backport. Thanks,
Yes. James, please mark this for v3.15.y and push to Linus ASAP.
Thank you,
--nab
Re: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
In reply to James BottomleyQuoted 2 lines · James Bottomley
In which case it's not marked for stable backport ... I assume it should be?
Only if the patch that regression went to stable, which it shouldn't,
James Bottomley
maintainerRe: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
Quoted 5 lines · Christoph Hellwig
Quoted 2 lines · James Bottomley
In which case it's not marked for stable backport ... I assume it should be?
Only if the patch that regression went to stable, which it shouldn't,
Well it did:
The stable process is a bit broken from my point of view since it doesn't copy linux-scsi on SCSI commits ... but this time even I didn't get a cc, so really no-one who knew there was a problem had a chance to comment.
The choice is either request revert from stable or add fix.
James
Greg Kroah-Hartman
maintainerRe: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
Quoted 5 lines · James Bottomley
Quoted 4 lines · Christoph Hellwig
Quoted 2 lines · James Bottomley
In which case it's not marked for stable backport ... I assume it should be?
Quoted 8 lines
Quoted 1 line
Only if the patch that regression went to stable, which it shouldn't,
Well it did:
The stable process is a bit broken from my point of view since it doesn't copy linux-scsi on SCSI commits
I copy everyone on the patch. You were not on that patch as it doesn't look like it went through your tree at all.
Quoted 2 lines
... but this time even I didn't get a cc, so really no-one who knew there was a problem had a chance to comment.
The target people (Nicholas) asked for the patch to be applied, they knew what was going on here.
Quoted 1 line
The choice is either request revert from stable or add fix.
Agreed, just let me know.
greg k-h
Re: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
Quoted 2 lines · Greg Kroah-Hartman
I copy everyone on the patch. You were not on that patch as it doesn't look like it went through your tree at all.
It went through the target tree despite exclusive touching scsi initiator side code. I'm still not sure how this happened, but we should take care to avoid this in the future. In addition to this regression due to a complete lack of testing it also caused various merge issues.
Re: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
Quoted 9 lines · Christoph Hellwig
Quoted 2 lines · Greg Kroah-Hartman
I copy everyone on the patch. You were not on that patch as it doesn't look like it went through your tree at all.
It went through the target tree despite exclusive touching scsi initiator side code. I'm still not sure how this happened, but we should take care to avoid this in the future. In addition to this regression due to a complete lack of testing it also caused various merge issues.
OK, Nic cc'd
Please explain how this commit:
commit d77e65350f2d82dfa0557707d505711f5a43c8fd
Author: Sagi Grimberg <[email protected]> Date: Wed Jun 11 12:09:58 2014 +0300
libiscsi, iser: Adjust data_length to include protection information
In case protection information exists over the wire iscsi header data length is required to include it. Use protection information aware scsi helpers to set the correct transfer length.
In order to avoid breakage, remove iser transfer length checks for each task as they are not always true and somewhat redundant anyway.
drivers/infiniband/ulp/iser/iser_initiator.c | 34 ++++++++-------------------- drivers/scsi/libiscsi.c | 18 +++++++--------
Came to go through the target tree even though it's initiator only and how come it was tagged for stable?
James
Re: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
Quoted 34 lines · James Bottomley
Quoted 8 lines · Christoph Hellwig
Quoted 2 lines · Greg Kroah-Hartman
I copy everyone on the patch. You were not on that patch as it doesn't look like it went through your tree at all.
It went through the target tree despite exclusive touching scsi initiator side code. I'm still not sure how this happened, but we should take care to avoid this in the future. In addition to this regression due to a complete lack of testing it also caused various merge issues.
OK, Nic cc'd
Please explain how this commit:
commit
d77e65350f2d82dfa0557707d505711f5a43c8fdAuthor: Sagi Grimberg <[email protected]> Date: Wed Jun 11 12:09:58 2014 +0300libiscsi, iser: Adjust
data_lengthto include protection informationIn case protection information exists over the wire iscsi header data length is required to include it. Use protection information aware scsi helpers to set the correct transfer length.
In order to avoid breakage, remove iser transfer length checks for each task as they are not always true and somewhat redundant anyway.
Signed-off-by Sagi Grimberg <[email protected]>Reviewed-by Mike Christie <[email protected]>Acked-by Mike Christie <[email protected]>Cc [email protected] # 3.15+Signed-off-by Nicholas Bellinger <[email protected]>
Quoted 5 lines
drivers/infiniband/ulp/iser/iser_initiator.c | 34 ++++++++-------------------- drivers/scsi/libiscsi.c | 18 +++++++--------Came to go through the target tree even though it's initiator only and how come it was tagged for stable?
Hey James,
I think it's because of the cover-letter comment I gave in my initial patchset:
"Although these patches involve 3 subsystems with different maintainers (scsi, iser, target) I would prefer seeing these patches included together." The set involved initiator-target wire protocol dependency.
I removed this comment in v1, v2 due to Roland's comment:
"Why? Because they break wire compatibility? I hate to say it but even if they're merged at the same time, you can't guarantee that targets and initiators will be updated together."
So I guess Nic just followed up on my request in order to avoid wire-protocol breakage. I assume the stable tag was there for the same reason.
Anyway, Sorry for all the fuss... I'll try to avoid such mistakes in the future.
Sagi.
Re: [PATCH 3.15 099/139] libiscsi, iser: Adjust data_length to include protection information
In reply to James BottomleyQuoted 10 lines · James Bottomley
Well it did:
The stable process is a bit broken from my point of view since it doesn't copy linux-scsi on SCSI commits ... but this time even I didn't get a cc, so really no-one who knew there was a problem had a chance to comment.
The choice is either request revert from stable or add fix.
Unfortunately the fix hasn't even made it to mainline yet, despite lots of people tripping over it. If you're too busy I'm happy to send it to Linus, it's been sitting in my tree for over a week now.
So if we can get it in ASAP we can also send it to stable, otherwise a revert is the only option.
Greg Kroah-Hartman
maintainerPATCH 100/139 TARGET/sbc,loopback: Adjust command data length in case pi exists on the wire
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Sagi Grimberg <[email protected]>
commit e2a4f55c6498b59a17a85a1bb6db122a993ffe02 upstream.
In various areas of the code, it is assumed that se_cmd->data_length describes pure data. In case that protection information exists over the wire (protect bits is are on) the target core re-calculates the data length from the CDB and the backed device block size (instead of each transport peeking in the cdb).
Modify loopback device to include protection information in the transferred data length (like other scsi transports).
--- drivers/target/loopback/tcm_loop.c | 15 ++++++++++++--- drivers/target/target_core_sbc.c | 15 +++++++++++++-- 2 files changed, 25 insertions(+), 5 deletions(-)
drivers/target/loopback/tcm_loop.c +12 -3
@@ -179,7 +179,7 @@ static void tcm_loop_submission_work(strstruct tcm_loop_hba *tl_hba;struct tcm_loop_tpg *tl_tpg;struct scatterlist *sgl_bidi =NULL ; -u32 sgl_bidi_count =0 ; +u32 sgl_bidi_count =0 , transfer_length;int rc; tl_hba = *(struct tcm_loop_hba **)shost_priv (sc->device->host);
@@ -213,12 +213,21 @@ static void tcm_loop_submission_work(str } -if (!scsi_prot_sg_count (sc) &&scsi_get_prot_op (sc) !=SCSI_PROT_NORMAL ) + transfer_length =scsi_transfer_length (sc); +if (!scsi_prot_sg_count (sc) && +scsi_get_prot_op (sc) !=SCSI_PROT_NORMAL ) { se_cmd->prot_pto = true; +/* +* loopback transport doesn't support +* WRITE_GENERATE, READ_STRIP protection +* information operations, go ahead unprotected. +*/ + transfer_length =scsi_bufflen (sc); + } rc = target_submit_cmd_map_sgls(se_cmd, tl_nexus->se_sess, sc->cmnd, &tl_cmd->tl_sense_buf[0], tl_cmd->sc->device->lun, - scsi_bufflen(sc), tcm_loop_sam_attr(sc), + transfer_length, tcm_loop_sam_attr(sc), sc->sc_data_direction, 0, scsi_sglist(sc), scsi_sg_count(sc), sgl_bidi, sgl_bidi_count,
drivers/target/target_core_sbc.c +13 -2
@@ -665,8 +665,19 @@ sbc_check_prot(struct se_device *dev, st cmd->prot_type = dev->dev_attrib.pi_prot_type; cmd->prot_length = dev->prot_length * sectors; -pr_debug ("%s: prot_type=%d, prot_length=%d prot_op=%d prot_checks=%d\n" , - __func__, cmd->prot_type, cmd->prot_length, + +/** +* In case protection information exists over the wire +* we modify command data length to describe pure data. +* The actual transfer length is data length + protection +* length +**/ +if (protect) + cmd->data_length = sectors * dev->dev_attrib.block_size; + +pr_debug ("%s: prot_type=%d, data_length=%d, prot_length=%d " +"prot_op=%d prot_checks=%d\n" , + __func__, cmd->prot_type, cmd->data_length, cmd->prot_length, cmd->prot_op, cmd->prot_checks);return true;
Greg Kroah-Hartman
maintainerPATCH 101/139 f2fs: submit bio at the reclaim path
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jaegeuk Kim <[email protected]>
commit 2aea39eca6b68d6ae7eb545332df0695f56a3d3f upstream.
If f2fs_write_data_page is called through the reclaim path, we should submit the bio right away.
This patch resolves the following issue that Marc Dietrich reported. "It took me a while to bisect a problem which causes my ARM (tegra2) netbook to frequently stall for 5-10 seconds when I enable EXA acceleration (opentegra experimental ddx)."
And this patch fixes that.
--- fs/f2fs/data.c | 2 ++ 1 file changed, 2 insertions(+)
fs/f2fs/data.c +2 -0
@@ -835,6 +835,8 @@ out:unlock_page (page);if (need_balance_fs)f2fs_balance_fs (sbi); +if (wbc->for_reclaim) +f2fs_submit_merged_bio (sbi,DATA ,WRITE );return 0 ; redirty_out:
Greg Kroah-Hartman
maintainerPATCH 102/139 ARM: mvebu: DT: fix OpenBlocks AX3-4 RAM size
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jason Cooper <[email protected]>
commit e47043aea3853a74a9aa5726a1faa916d7462ab7 upstream.
The OpenBlocks AX3-4 has a non-DT bootloader. It also comes with 1GB of soldered on RAM, and a DIMM slot for expansion.
Unfortunately, atags_to_fdt() doesn't work in big-endian mode, so we see the following failure when attempting to boot a big-endian kernel:
686 slab pages
17 pages shared
0 pages swap cached
[ pid ] uid tgid total_vm rss nr_ptes swapents oom_score_adj name
Kernel panic - not syncing: Out of memory and no killable processes... CPU: 1 PID: 351 Comm: kworker/u4:0 Not tainted 3.15.0-rc8-next-20140603 #1 [<c0215a54>] (unwind_backtrace) from [<c021160c>] (show_stack+0x10/0x14) [<c021160c>] (show_stack) from [<c0802500>] (dump_stack+0x78/0x94) [<c0802500>] (dump_stack) from [<c0800068>] (panic+0x90/0x21c) [<c0800068>] (panic) from [<c02b5704>] (out_of_memory+0x320/0x340) [<c02b5704>] (out_of_memory) from [<c02b93a0>] (__alloc_pages_nodemask+0x874/0x930) [<c02b93a0>] (__alloc_pages_nodemask) from [<c02d446c>] (handle_mm_fault+0x744/0x96c) [<c02d446c>] (handle_mm_fault) from [<c02cf250>] (__get_user_pages+0xd0/0x4c0) [<c02cf250>] (__get_user_pages) from [<c02f3598>] (get_arg_page+0x54/0xbc) [<c02f3598>] (get_arg_page) from [<c02f3878>] (copy_strings+0x278/0x29c) [<c02f3878>] (copy_strings) from [<c02f38bc>] (copy_strings_kernel+0x20/0x28) [<c02f38bc>] (copy_strings_kernel) from [<c02f4f1c>] (do_execve+0x3a8/0x4c8) [<c02f4f1c>] (do_execve) from [<c025ac10>] (____call_usermodehelper+0x15c/0x194) [<c025ac10>] (____call_usermodehelper) from [<c020e9b8>] (ret_from_fork+0x14/0x3c) CPU0: stopping CPU: 0 PID: 0 Comm: swapper/0 Not tainted 3.15.0-rc8-next-20140603 #1 [<c0215a54>] (unwind_backtrace) from [<c021160c>] (show_stack+0x10/0x14) [<c021160c>] (show_stack) from [<c0802500>] (dump_stack+0x78/0x94) [<c0802500>] (dump_stack) from [<c021429c>] (handle_IPI+0x138/0x174) [<c021429c>] (handle_IPI) from [<c02087f0>] (armada_370_xp_handle_irq+0xb0/0xcc) [<c02087f0>] (armada_370_xp_handle_irq) from [<c0212100>] (__irq_svc+0x40/0x50) Exception stack(0xc0b6bf68 to 0xc0b6bfb0) bf60: e9fad598 00000000 00f509a3 00000000 c0b6a000 c0b724c4 bf80: c0b72458 c0b6a000 00000000 00000000 c0b66da0 c0b6a000 00000000 c0b6bfb0 bfa0: c027bb94 c027bb24 60000313 ffffffff [<c0212100>] (__irq_svc) from [<c027bb24>] (cpu_startup_entry+0x54/0x214) [<c027bb24>] (cpu_startup_entry) from [<c0ac5b30>] (start_kernel+0x318/0x37c) [<c0ac5b30>] (start_kernel) from [<00208078>] (0x208078) ---[ end Kernel panic - not syncing: Out of memory and no killable processes...
A similar failure will also occur if ARM_ATAG_DTB_COMPAT isn't selected.
Fix this by setting a sane default (1 GB) in the dts file.
--- arch/arm/boot/dts/armada-xp-openblocks-ax3-4.dts | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/boot/dts/armada-xp-openblocks-ax3-4.dts +1 -1
@@ -25,7 +25,7 @@ memory { device_type = "memory"; - reg = <0 0x00000000 0 0xC0000000>; /* 3 GB */ + reg = <0 0x00000000 0 0x40000000>; /* 1 GB soldered on */ }; soc {
Greg Kroah-Hartman
maintainerPATCH 103/139 arm64: uid16: fix __kernel_old_{gid,uid}_t definitions
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Will Deacon <[email protected]>
commit 34c65c43f1518bf85f93526ad373adc6a683b4c5 upstream.
Whilst native arm64 applications don't have the 16-bit UID/GID syscalls wired up, compat tasks can still access them. The 16-bit wrappers for these syscalls use __kernel_old_uid_t and __kernel_old_gid_t, which must be 16-bit data types to maintain compatibility with the 16-bit UIDs used by compat applications.
This patch defines 16-bit __kernel_old_{gid,uid}_t types for arm64 instead of using the 32-bit types provided by asm-generic.
--- arch/arm64/include/asm/Kbuild | 1 - arch/arm64/include/uapi/asm/posix_types.h | 10 ++++++++++ 2 files changed, 10 insertions(+), 1 deletion(-)
arch/arm64/include/asm/Kbuild +0 -1
@@ -30,7 +30,6 @@ generic-y += msgbuf.h generic-y += mutex.h generic-y += pci.h generic-y += poll.h -generic-y += posix_types.h generic-y += preempt.h generic-y += resource.h generic-y += rwsem.h
arch/arm64/include/uapi/asm/posix_types.h new file +10 -0
@@ -0,0 +1,10 @@ +#ifndef __ASM_POSIX_TYPES_H +#define __ASM_POSIX_TYPES_H + +typedef unsigned short __kernel_old_uid_t ; +typedef unsigned short __kernel_old_gid_t ; +#define __kernel_old_uid_t __kernel_old_uid_t + +#include <asm-generic/posix_types.h> + +#endif /* __ASM_POSIX_TYPES_H */
Greg Kroah-Hartman
maintainerPATCH 104/139 arm64/dma: Removing ARCH_HAS_DMA_GET_REQUIRED_MASK macro
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Suravee Suthikulpanit <[email protected]>
commit f3a183cb422574014538017b5b291a416396f97e upstream.
Arm64 does not define dma_get_required_mask() function. Therefore, it should not define the ARCH_HAS_DMA_GET_REQUIRED_MASK. This causes build errors in some device drivers (e.g. mpt2sas)
--- arch/arm64/include/asm/dma-mapping.h | 2 -- 1 file changed, 2 deletions(-)
arch/arm64/include/asm/dma-mapping.h +0 -2
@@ -26,8 +26,6 @@#include <xen/xen.h> #include <asm/xen/hypervisor.h> -#define ARCH_HAS_DMA_GET_REQUIRED_MASK -#define DMA_ERROR_CODE (~(dma_addr_t)0)extern struct dma_map_ops *dma_ops;extern struct dma_map_ops coherent_swiotlb_dma_ops;
Greg Kroah-Hartman
maintainerPATCH 105/139 arm64: mm: remove broken &= operator from pmd_mknotpresent
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Will Deacon <[email protected]>
commit e3a920afc3482e954834a4ed95908c4bc5e4c000 upstream.
This should be a plain old '&' and could easily lead to undefined behaviour if the target of a pmd_mknotpresent invocation was the same as the parameter.
--- arch/arm64/include/asm/pgtable.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm64/include/asm/pgtable.h +1 -1
@@ -253,7 +253,7 @@ static inline pmd_t pte_pmd(pte_t pte)#define pmd_mkwrite (pmd) pte_pmd(pte_mkwrite(pmd_pte(pmd)))#define pmd_mkdirty (pmd) pte_pmd(pte_mkdirty(pmd_pte(pmd)))#define pmd_mkyoung (pmd) pte_pmd(pte_mkyoung(pmd_pte(pmd))) -#define pmd_mknotpresent (pmd) (__pmd(pmd_val(pmd) &= ~PMD_TYPE_MASK)) +#define pmd_mknotpresent (pmd) (__pmd(pmd_val(pmd) & ~PMD_TYPE_MASK))#define __HAVE_ARCH_PMD_WRITE#define pmd_write (pmd) pte_write(pmd_pte(pmd))
Greg Kroah-Hartman
maintainerPATCH 106/139 x86-32, espfix: Remove filter for espfix32 due to race
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: "H. Peter Anvin" <[email protected]>
commit 246f2d2ee1d715e1077fc47d61c394569c8ee692 upstream.
It is not safe to use LAR to filter when to go down the espfix path, because the LDT is per-process (rather than per-thread) and another thread might change the descriptors behind our back. Fortunately it is always *safe* (if a bit slow) to go down the espfix path, and a 32-bit LDT stack segment is extremely rare.
--- arch/x86/kernel/entry_32.S | 5 ----- 1 file changed, 5 deletions(-)
arch/x86/kernel/entry_32.S +0 -5
@@ -551,11 +551,6 @@ ENTRY(iret_exc) CFI_RESTORE_STATE ldt_ss: - larl PT_OLDSS(%esp), %eax - jnz restore_nocheck - testl $0x00400000, %eax # returning to 32bit stack? - jnz restore_nocheck # allright, normal return - #ifdef CONFIG_PARAVIRT /* * The kernel can't run on a non-flat stack if paravirt mode
Greg Kroah-Hartman
maintainerPATCH 107/139 x86, x32: Use compat shims for io_{setup,submit}
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Mike Frysinger <[email protected]>
commit 7fd44dacdd803c0bbf38bf478d51d280902bb0f1 upstream.
The io_setup takes a pointer to a context id of type aio_context_t. This in turn is typed to a __kernel_ulong_t. We could tweak the exported headers to define this as a 64bit quantity for specific ABIs, but since we already have a 32bit compat shim for the x86 ABI, let's just re-use that logic. The libaio package is also written to expect this as a pointer type, so a compat shim would simplify that.
The io_submit func operates on an array of pointers to iocb structs. Padding out the array to be 64bit aligned is a huge pain, so convert it over to the existing compat shim too.
We don't convert io_getevents to the compat func as its only purpose is to handle the timespec struct, and the x32 ABI uses 64bit times.
With this change, the libaio package can now pass its testsuite when built for the x32 ABI.
--- arch/x86/syscalls/syscall_64.tbl | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
arch/x86/syscalls/syscall_64.tbl +4 -2
@@ -212,10 +212,10 @@ 203 common sched_setaffinity sys_sched_setaffinity 204 common sched_getaffinity sys_sched_getaffinity 205 64 set_thread_area -206 common io_setup sys_io_setup +206 64 io_setup sys_io_setup 207 common io_destroy sys_io_destroy 208 common io_getevents sys_io_getevents -209 common io_submit sys_io_submit +209 64 io_submit sys_io_submit 210 common io_cancel sys_io_cancel 211 64 get_thread_area 212 common lookup_dcookie sys_lookup_dcookie
@@ -359,3 +359,5 @@ 540 x32 process_vm_writev compat_sys_process_vm_writev 541 x32 setsockopt compat_sys_setsockopt 542 x32 getsockopt compat_sys_getsockopt +543 x32 io_setup compat_sys_io_setup +544 x32 io_submit compat_sys_io_submit
Greg Kroah-Hartman
maintainerPATCH 108/139 Revert "drm/radeon: use variable UVD clocks"
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Alex Deucher <[email protected]>
commit 0690a229c69f40a6c9c459ab455c85df49822525 upstream.
This caused reduced performance for some users with advanced post processing enabled. We need a better method to pick the UVD state based on the amount of post processing required or tune the advanced post processing to fit within the lower power state envelope.
This reverts commit 14a9579ddbf15dd1992a9481a4ec80b0b91656d5.
--- drivers/gpu/drm/radeon/radeon_pm.c | 3 +++ drivers/gpu/drm/radeon/radeon_uvd.c | 3 ++- 2 files changed, 5 insertions(+), 1 deletion(-)
drivers/gpu/drm/radeon/radeon_pm.c +3 -0
@@ -984,6 +984,8 @@ void radeon_dpm_enable_uvd(struct radeonif (enable) {mutex_lock (&rdev->pm.mutex); rdev->pm.dpm.uvd_active = true; +/* disable this for now */ +#if 0 if ((rdev->pm.dpm.sd ==1 ) && (rdev->pm.dpm.hd ==0 )) dpm_state =POWER_STATE_TYPE_INTERNAL_UVD_SD ;else if ((rdev->pm.dpm.sd ==2 ) && (rdev->pm.dpm.hd ==0 ))
@@ -993,6 +995,7 @@ void radeon_dpm_enable_uvd(struct radeon else if ((rdev->pm.dpm.sd == 0) && (rdev->pm.dpm.hd == 2)) dpm_state = POWER_STATE_TYPE_INTERNAL_UVD_HD2; else +#endif dpm_state =POWER_STATE_TYPE_INTERNAL_UVD ; rdev->pm.dpm.state = dpm_state;mutex_unlock (&rdev->pm.mutex);
drivers/gpu/drm/radeon/radeon_uvd.c +2 -1
@@ -812,7 +812,8 @@ void radeon_uvd_note_usage(struct radeon (rdev->pm.dpm.hd != hd)) { rdev->pm.dpm.sd = sd; rdev->pm.dpm.hd = hd; - streams_changed = true; +/* disable this for now */ +/*streams_changed = true;*/ } }
Greg Kroah-Hartman
maintainerPATCH 109/139 Revert "offb: Add palette hack for little endian"
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Benjamin Herrenschmidt <[email protected]>
commit 68986c9f0f4552c34c248501eb0c690553866d6e upstream.
This reverts commit e1edf18b20076da83dd231dbd2146cbbc31c0b14.
This patch was a misguided attempt at fixing offb for LE ppc64 kernels on BE qemu but is just wrong ... it breaks real LE/LE setups, LE with real HW, and existing mixed endian systems that did the fight thing with the appropriate device-tree property. Bad reviewing on my part, sorry.
The right fix is to either make qemu change its endian when the guest changes endian (working on that) or to use the existing foreign endian support.
--- drivers/video/fbdev/offb.c | 11 +---------- 1 file changed, 1 insertion(+), 10 deletions(-)
drivers/video/fbdev/offb.c +1 -10
@@ -91,15 +91,6 @@ extern boot_infos_t *boot_infos;#define AVIVO_DC_LUTB_WHITE_OFFSET_GREEN 0x6cd4#define AVIVO_DC_LUTB_WHITE_OFFSET_RED 0x6cd8 -#define FB_RIGHT_POS (p, bpp) (fb_be_math(p) ? 0 : (32 - (bpp))) - -static inline u32 offb_cmap_byteswap (struct fb_info *info,u32 value) -{ -u32 bpp = info->var.bits_per_pixel; - -return cpu_to_be32 (value) >>FB_RIGHT_POS (info, bpp); -} -/* * Set a single color register. The values supplied are already * rounded down to the hardware's capabilities (according to the
@@ -129,7 +120,7 @@ static int offb_setcolreg(u_int regno, u mask <<= info->var.transp.offset; value |= mask; } - pal[regno] =offb_cmap_byteswap (info, value); + pal[regno] = value;return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 110/139 genirq: Sanitize spurious interrupt detection of threaded irqs
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Thomas Gleixner <[email protected]>
commit 1e77d0a1ed7417d2a5a52a7b8d32aea1833faa6c upstream.
Till reported that the spurious interrupt detection of threaded interrupts is broken in two ways:
- note_interrupt() is called for each action thread of a shared interrupt line. That's wrong as we are only interested whether none of the device drivers felt responsible for the interrupt, but by calling multiple times for a single interrupt line we account IRQ_NONE even if one of the drivers felt responsible.
- note_interrupt() when called from the thread handler is not serialized. That leaves the members of irq_desc which are used for the spurious detection unprotected.
To solve this we need to defer the spurious detection of a threaded interrupt to the next hardware interrupt context where we have implicit serialization.
If note_interrupt is called with action_ret == IRQ_WAKE_THREAD, we check whether the previous interrupt requested a deferred check. If not, we request a deferred check for the next hardware interrupt and return.
If set, we check whether one of the interrupt threads signaled success. Depending on this information we feed the result into the spurious detector.
If one primary handler of a shared interrupt returns IRQ_HANDLED we disable the deferred check of irq threads on the same line, as we have found at least one device driver who cared.
--- include/linux/irqdesc.h | 4 + kernel/irq/manage.c | 4 - kernel/irq/spurious.c | 106 ++++++++++++++++++++++++++++++++++++++++++++++-- 3 files changed, 108 insertions(+), 6 deletions(-)
include/linux/irqdesc.h +4 -0
@@ -27,6 +27,8 @@ struct irq_desc;* @irq_count: stats field to detect stalled irqs * @last_unhandled: aging timer for unhandled count * @irqs_unhandled: stats field for spurious unhandled interrupts +* @threads_handled: stats field for deferred spurious detection of threaded handlers +* @threads_handled_last: comparator field for deferred spurious detection of theraded handlers * @lock: locking for SMP * @affinity_hint: hint to user space for preferred irq affinity * @affinity_notify: context for notification of affinity changes
@@ -52,6 +54,8 @@ struct irq_desc {unsigned int irq_count;/* For detecting broken IRQs */ unsigned long last_unhandled;/* Aging timer for unhandled count */ unsigned int irqs_unhandled; +atomic_t threads_handled; +int threads_handled_last;raw_spinlock_t lock;struct cpumask *percpu_enabled;#ifdef CONFIG_SMP
kernel/irq/manage.c +2 -2
@@ -886,8 +886,8 @@ static int irq_thread(void *data)irq_thread_check_affinity (desc, action); action_ret =handler_fn (desc, action); -if (!noirqdebug) -note_interrupt (action->irq, desc, action_ret); +if (action_ret ==IRQ_HANDLED ) +atomic_inc (&desc->threads_handled);wake_threads_waitq (desc); }
kernel/irq/spurious.c +102 -4
123 lines not on this page · raw message
Greg Kroah-Hartman
maintainerPATCH 111/139 aio: fix aio request leak when events are reaped by userspace
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Benjamin LaHaise <[email protected]>
commit f8567a3845ac05bb28f3c1b478ef752762bd39ef upstream.
The aio cleanups and optimizations by kmo that were merged into the 3.10 tree added a regression for userspace event reaping. Specifically, the reference counts are not decremented if the event is reaped in userspace, leading to the application being unable to submit further aio requests. This patch applies to 3.12+. A separate backport is required for 3.10/3.11. This issue was uncovered as part of CVE-2014-0206.
--- fs/aio.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
fs/aio.c +1 -2
@@ -1021,6 +1021,7 @@ void aio_complete(struct kiocb *iocb, lo/* everything turned out well, dispose of the aiocb. */ kiocb_free (iocb); +put_reqs_available (ctx,1 );/* * We have to order our ring_info tail store above and test
@@ -1100,8 +1101,6 @@ static long aio_read_events_ring(structflush_dcache_page (ctx->ring_pages[0 ]);pr_debug ("%li h%u t%u\n" , ret, head, tail); - -put_reqs_available (ctx, ret); out:mutex_unlock (&ctx->ring_lock);
Greg Kroah-Hartman
maintainerPATCH 112/139 aio: fix kernel memory disclosure in io_getevents() introduced in v3.10
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Benjamin LaHaise <[email protected]>
commit edfbbf388f293d70bf4b7c0bc38774d05e6f711a upstream.
A kernel memory disclosure was introduced in aio_read_events_ring() in v3.10 by commit a31ad380bed817aa25f8830ad23e1a0480fef797. The changes made to aio_read_events_ring() failed to correctly limit the index into ctx->ring_pages[], allowing an attacked to cause the subsequent kmap() of an arbitrary page with a copy_to_user() to copy the contents into userspace. This vulnerability has been assigned CVE-2014-0206. Thanks to Mateusz and Petr for disclosing this issue.
This patch applies to v3.12+. A separate backport is needed for 3.10/3.11.
--- fs/aio.c | 3 +++ 1 file changed, 3 insertions(+)
fs/aio.c +3 -0
@@ -1063,6 +1063,9 @@ static long aio_read_events_ring(structif (head == tail) goto out; + head %= ctx->nr_events; + tail %= ctx->nr_events; +while (ret < nr) {long avail;struct io_event *ev;
Greg Kroah-Hartman
maintainerPATCH 113/139 CIFS: Fix memory leaks in SMB2_open
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Pavel Shilovsky <[email protected]>
commit 663a962151593c69374776e8651238d0da072459 upstream.
--- fs/cifs/smb2pdu.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
fs/cifs/smb2pdu.c +6 -1
@@ -1089,6 +1089,7 @@ SMB2_open(const unsigned int xid, structint rc =0 ;unsigned int num_iovecs =2 ;__u32 file_attributes =0 ; +char *dhc_buf =NULL , *lc_buf =NULL ;cifs_dbg (FYI ,"create/open\n" );
@@ -1155,6 +1156,7 @@ SMB2_open(const unsigned int xid, structkfree (copy_path);return rc; } + lc_buf = iov[num_iovecs-1 ].iov_base; }if (*oplock ==SMB2_OPLOCK_LEVEL_BATCH ) {
@@ -1169,9 +1171,10 @@ SMB2_open(const unsigned int xid, structif (rc) {cifs_small_buf_release (req);kfree (copy_path); -kfree (iov[num_iovecs-1 ].iov_base); +kfree (lc_buf);return rc; } + dhc_buf = iov[num_iovecs-1 ].iov_base; } rc =SendReceive2 (xid, ses, iov, num_iovecs, &resp_buftype,0 );
@@ -1203,6 +1206,8 @@ SMB2_open(const unsigned int xid, struct *oplock = rsp->OplockLevel; creat_exit:kfree (copy_path); +kfree (lc_buf); +kfree (dhc_buf);free_rsp_buf (resp_buftype, rsp);return rc; }
Greg Kroah-Hartman
maintainerPATCH 114/139 Btrfs: fix leaf corruption caused by ENOSPC while hole punching
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Filipe Manana <[email protected]>
commit fc19c5e73645f95d3eca12b4e91e7b56faf1e4a4 upstream.
While running a stress test with multiple threads writing to the same btrfs file system, I ended up with a situation where a leaf was corrupted in that it had 2 file extent item keys that had the same exact key. I was able to detect this quickly thanks to the following patch which triggers an assertion as soon as a leaf is marked dirty if there are duplicated keys or out of order keys:
Btrfs: check if items are ordered when a leaf is marked dirty
(https://patchwork.kernel.org/patch/3955431/)
Basically while running the test, I got the following in dmesg:
[28877.415877] WARNING: CPU: 2 PID: 10706 at fs/btrfs/file.c:553 btrfs_drop_extent_cache+0x435/0x440 [btrfs]()
(...)
[28877.415917] Call Trace:
[28877.415922] [<ffffffff816f1189>] dump_stack+0x4e/0x68
[28877.415926] [<ffffffff8104a32c>] warn_slowpath_common+0x8c/0xc0
[28877.415929] [<ffffffff8104a37a>] warn_slowpath_null+0x1a/0x20
[28877.415944] [<ffffffffa03775a5>] btrfs_drop_extent_cache+0x435/0x440 [btrfs]
[28877.415949] [<ffffffff8118e7be>] ? kmem_cache_alloc+0xfe/0x1c0
[28877.415962] [<ffffffffa03777d9>] fill_holes+0x229/0x3e0 [btrfs]
[28877.415972] [<ffffffffa0345865>] ? block_rsv_add_bytes+0x55/0x80 [btrfs]
[28877.415984] [<ffffffffa03792cb>] btrfs_fallocate+0xb6b/0xc20 [btrfs]
(...)
[29854.132560] BTRFS critical (device sdc): corrupt leaf, bad key order: block=955232256,root=1, slot=24
[29854.132565] BTRFS info (device sdc): leaf 955232256 total ptrs 40 free space 778
(...)
[29854.132637] item 23 key (3486 108 667648) itemoff 2694 itemsize 53
[29854.132638] extent data disk bytenr 14574411776 nr 286720
[29854.132639] extent data offset 0 nr 286720 ram 286720
[29854.132640] item 24 key (3486 108 954368) itemoff 2641 itemsize 53
[29854.132641] extent data disk bytenr 0 nr 0
[29854.132643] extent data offset 0 nr 0 ram 0
[29854.132644] item 25 key (3486 108 954368) itemoff 2588 itemsize 53
[29854.132645] extent data disk bytenr 8699670528 nr 77824
[29854.132646] extent data offset 0 nr 77824 ram 77824
[29854.132647] item 26 key (3486 108 1146880) itemoff 2535 itemsize 53
[29854.132648] extent data disk bytenr 8699670528 nr 77824
[29854.132649] extent data offset 0 nr 77824 ram 77824
(...)
[29854.132707] kernel BUG at fs/btrfs/ctree.h:3901!
(...)
[29854.132771] Call Trace:
[29854.132779] [<ffffffffa0342b5c>] setup_items_for_insert+0x2dc/0x400 [btrfs]
[29854.132791] [<ffffffffa0378537>] __btrfs_drop_extents+0xba7/0xdd0 [btrfs]
[29854.132794] [<ffffffff8109c0d6>] ? trace_hardirqs_on_caller+0x16/0x1d0
[29854.132797] [<ffffffff8109c29d>] ? trace_hardirqs_on+0xd/0x10
[29854.132800] [<ffffffff8118e7be>] ? kmem_cache_alloc+0xfe/0x1c0
[29854.132810] [<ffffffffa036783b>] insert_reserved_file_extent.constprop.66+0xab/0x310 [btrfs]
[29854.132820] [<ffffffffa036a6c6>] __btrfs_prealloc_file_range+0x116/0x340 [btrfs]
[29854.132830] [<ffffffffa0374d53>] btrfs_prealloc_file_range+0x23/0x30 [btrfs]
(...)So this is caused by getting an -ENOSPC error while punching a file hole, more specifically, we get -ENOSPC error from __btrfs_drop_extents in the while loop of file.c:btrfs_punch_hole() when it's unable to modify the btree to delete one or more file extent items due to lack of enough free space. When this happens, in btrfs_punch_hole(), we attempt to reclaim free space by switching our transaction block reservation object to root->fs_info->trans_block_rsv, end our transaction and start a new transaction basically - and, we keep increasing our current offset (cur_offset) as long as it's smaller than the end of the target range (lockend) - this makes use leave the loop with cur_offset == drop_end which in turn makes us call fill_holes() for inserting a file extent item that represents a 0 bytes range hole (and this insertion succeeds, as in the meanwhile more space became available).
This 0 bytes file hole extent item is a problem because any subsequent caller of __btrfs_drop_extents (regular file writes, or fallocate calls for e.g.), with a start file offset that is equal to the offset of the hole, will not remove this extent item due to the following conditional in the while loop of __btrfs_drop_extents:
if (extent_end <= search_start) {
path->slots[0]++;
goto next_slot;
}This later makes the call to setup_items_for_insert() (at the very end of __btrfs_drop_extents), insert a new file extent item with the same offset as the 0 bytes file hole extent item that follows it. Needless is to say that this causes chaos, either when reading the leaf from disk (btree_readpage_end_io_hook), where we perform leaf sanity checks or in subsequent operations that manipulate file extent items, as in the fallocate call as shown by the dmesg trace above.
Without my other patch to perform the leaf sanity checks once a leaf is marked as dirty (if the integrity checker is enabled), it would have been much harder to debug this issue.
This change might fix a few similar issues reported by users in the mailing list regarding assertion failures in btrfs_set_item_key_safe calls performed by __btrfs_drop_extents, such as the following report:
http://comments.gmane.org/gmane.comp.file-systems.btrfs/32938
Asking fill_holes() to create a 0 bytes wide file hole item also produced the first warning in the trace above, as we passed a range to btrfs_drop_extent_cache that has an end smaller (by -1) than its start.
On 3.14 kernels this issue manifests itself through leaf corruption, as we get duplicated file extent item keys in a leaf when calling setup_items_for_insert(), but on older kernels, setup_items_for_insert() isn't called by __btrfs_drop_extents(), instead we have callers of __btrfs_drop_extents(), namely the functions inode.c:insert_inline_extent() and inode.c:insert_reserved_file_extent(), calling btrfs_insert_empty_item() to insert the new file extent item, which would fail with error -EEXIST, instead of inserting a duplicated key - which is still a serious issue as it would make all similar file extent item replace operations keep failing if they target the same file range.
--- fs/btrfs/file.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-)
fs/btrfs/file.c +19 -1
@@ -780,6 +780,18 @@ next_slot: extent_end = search_start; } +/* +* Don't skip extent items representing 0 byte lengths. They +* used to be created (bug) if while punching holes we hit +* -ENOSPC condition. So if we find one here, just ensure we +* delete it, otherwise we would insert a new file extent item +* with the same key (offset) as that 0 bytes length file +* extent item in the call to setup_items_for_insert() later +* in this function. +*/ +if (extent_end == key.offset && extent_end >= search_start) + goto delete_extent_item; +if (extent_end <= search_start) { path->slots[0 ]++; goto next_slot;
@@ -893,6 +905,7 @@ next_slot:* | ------ extent ------ | */ if (start <= key.offset && end >= extent_end) { +delete_extent_item:if (del_nr ==0 ) { del_slot = path->slots[0 ]; del_nr =1 ;
@@ -2347,7 +2360,12 @@ static int btrfs_punch_hole(struct inode } trans->block_rsv = &root->fs_info->trans_block_rsv; -if (cur_offset < ino_size) { +/* +* Don't insert file hole extent item if it's for a range beyond eof +* (because it's useless) or if it represents a 0 bytes range (when +* cur_offset == drop_end). +*/ +if (cur_offset < ino_size && cur_offset < drop_end) { ret =fill_holes (trans, inode, path, cur_offset, drop_end);if (ret) { err = ret;
Greg Kroah-Hartman
maintainerPATCH 115/139 Btrfs: fix double free in find_lock_delalloc_range
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Chris Mason <[email protected]>
commit 7d78874273463a784759916fc3e0b4e2eb141c70 upstream.
We need to NULL the cached_state after freeing it, otherwise we might free it again if find_delalloc_range doesn't find anything.
--- fs/btrfs/extent_io.c | 1 + 1 file changed, 1 insertion(+)
fs/btrfs/extent_io.c +1 -0
@@ -1693,6 +1693,7 @@ again:* shortening the size of the delalloc range we're searching */ free_extent_state (cached_state); + cached_state =NULL ;if (!loops) { max_bytes =PAGE_CACHE_SIZE ; loops =1 ;
Greg Kroah-Hartman
maintainerPATCH 116/139 Btrfs: read inode size after acquiring the mutex when punching a hole
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Filipe Manana <[email protected]>
commit a1a50f60a6bf4f861eb94793420274bc1ccd409a upstream.
In a previous change, commit 12870f1c9b2de7d475d22e73fd7db1b418599725, I accidentally moved the roundup of inode->i_size to outside of the critical section delimited by the inode mutex, which is not atomic and not correct since the size can be changed by other task before we acquire the mutex. Therefore fix it.
--- fs/btrfs/file.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
fs/btrfs/file.c +2 -1
@@ -2200,13 +2200,14 @@ static int btrfs_punch_hole(struct inodebool same_page = ((offset >>PAGE_CACHE_SHIFT ) == ((offset + len -1 ) >>PAGE_CACHE_SHIFT ));bool no_holes =btrfs_fs_incompat (root->fs_info,NO_HOLES ); -u64 ino_size =round_up (inode->i_size,PAGE_CACHE_SIZE ); +u64 ino_size; ret =btrfs_wait_ordered_range (inode, offset, len);if (ret)return ret;mutex_lock (&inode->i_mutex); + ino_size =round_up (inode->i_size,PAGE_CACHE_SIZE );/* * We needn't truncate any page which is beyond the end of the file * because we are sure there is no data there.
Greg Kroah-Hartman
maintainerPATCH 117/139 btrfs: Add ctime/mtime update for btrfs device add/remove.
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Qu Wenruo <[email protected]>
commit 5a1972bd9fd4b2fb1bac8b7a0b636d633d8717e3 upstream.
Btrfs will send uevent to udev inform the device change,
but ctime/mtime for the block device inode is not udpated, which cause libblkid used by btrfs-progs unable to detect device change and use old cache, causing 'btrfs dev scan; btrfs dev rmove; btrfs dev scan' give an error message.
--- fs/btrfs/volumes.c | 26 ++++++++++++++++++++++++-- 1 file changed, 24 insertions(+), 2 deletions(-)
fs/btrfs/volumes.c +24 -2
@@ -1452,6 +1452,22 @@ out:return ret; } +/* +* Function to update ctime/mtime for a given device path. +* Mainly used for ctime/mtime based probe like libblkid. +*/ +static void update_dev_time (char *path_name) +{ +struct file *filp; + + filp =filp_open (path_name,O_RDWR ,0 ); +if (!filp) +return ; +file_update_time (filp); +filp_close (filp,NULL ); +return ; +} +static int btrfs_rm_dev_item (struct btrfs_root *root,struct btrfs_device *device) {
@@ -1704,10 +1720,14 @@ int btrfs_rm_device(struct btrfs_root *r ret =0 ; -/* Notify udev that device has changed */ -if (bdev) +if (bdev) { +/* Notify udev that device has changed */ btrfs_kobject_uevent (bdev,KOBJ_CHANGE ); +/* Update ctime/mtime for device path for libblkid */ +update_dev_time (device_path); + } + error_brelse:brelse (bh);if (bdev)
@@ -2146,6 +2166,8 @@ int btrfs_init_new_device(struct btrfs_r ret =btrfs_commit_transaction (trans, root); } +/* Update ctime/mtime for libblkid */ +update_dev_time (device_path);return ret; error_trans:
Greg Kroah-Hartman
maintainerPATCH 118/139 Btrfs: output warning instead of error when loading free space cache failed
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Miao Xie <[email protected]>
commit 32d6b47fe6fc1714d5f1bba1b9f38e0ab0ad58a8 upstream.
If we fail to load a free space cache, we can rebuild it from the extent tree, so it is not a serious error, we should not output a error message that would make the users uncomfortable. This patch uses warning message instead of it.
--- fs/btrfs/free-space-cache.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
fs/btrfs/free-space-cache.c +2 -2
@@ -831,7 +831,7 @@ int load_free_space_cache(struct btrfs_fif (!matched) {__btrfs_remove_free_space_cache (ctl); -btrfs_err (fs_info,"block group %llu has wrong amount of free space" , +btrfs_warn (fs_info,"block group %llu has wrong amount of free space" , block_group->key.objectid); ret =-1 ; }
@@ -843,7 +843,7 @@ out:spin_unlock (&block_group->lock); ret =0 ; -btrfs_err (fs_info,"failed to load free space cache for block group %llu" , +btrfs_warn (fs_info,"failed to load free space cache for block group %llu, rebuild it now" , block_group->key.objectid); }
Greg Kroah-Hartman
maintainerPATCH 119/139 Btrfs: send, account for orphan directories when building path strings
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Filipe Manana <[email protected]>
commit c992ec94f24c3e7135d6c23860615f269f0b1d87 upstream.
If we have directories with a pending move/rename operation, we must take into account any orphan directories that got created before executing the pending move/rename. Those orphan directories are directories with an inode number higher then the current send progress and that don't exist in the parent snapshot, they are created before current progress reaches their inode number, with a generated name of the form oN-M-I and at the root of the filesystem tree, and later when progress matches their inode number, moved/renamed to their final location.
Reproducer:
$ mkfs.btrfs -f /dev/sdd
$ mount /dev/sdd /mnt $ mkdir -p /mnt/a/b/c/d
$ mkdir /mnt/a/b/e
$ mv /mnt/a/b/c /mnt/a/b/e/CC
$ mkdir /mnt/a/b/e/CC/d/f
$ mkdir /mnt/a/g $ btrfs subvolume snapshot -r /mnt /mnt/snap1
$ btrfs send /mnt/snap1 -f /tmp/base.send $ mkdir /mnt/a/g/h
$ mv /mnt/a/b/e /mnt/a/g/h/EE
$ mv /mnt/a/g/h/EE/CC/d /mnt/a/g/h/EE/DD $ btrfs subvolume snapshot -r /mnt /mnt/snap2
$ btrfs send -p /mnt/snap1 /mnt/snap2 -f /tmp/incremental.sendThe second receive command failed with the following error:
ERROR: rename a/b/e/CC/d -> o264-7-0/EE/DD failed. No such file or directory
A test case for xfstests follows soon.
--- fs/btrfs/send.c | 33 +++++++++------------------------ 1 file changed, 9 insertions(+), 24 deletions(-)
fs/btrfs/send.c +9 -24
@@ -3054,33 +3054,18 @@ static int apply_dir_move(struct send_ctif (ret <0 ) goto out; -if (parent_ino == sctx->cur_ino) { -/* child only renamed, not moved */ -ASSERT (parent_gen == sctx->cur_inode_gen); - ret =get_cur_path (sctx, sctx->cur_ino, sctx->cur_inode_gen, - from_path); -if (ret <0 ) - goto out; - ret =fs_path_add_path (from_path, name); -if (ret <0 ) - goto out; - }else { -/* child moved and maybe renamed too */ - sctx->send_progress = pm->ino; - ret =get_cur_path (sctx, pm->ino, pm->gen, from_path); -if (ret <0 ) - goto out; - } + ret =get_cur_path (sctx, parent_ino, parent_gen, + from_path); +if (ret <0 ) + goto out; + ret =fs_path_add_path (from_path, name); +if (ret <0 ) + goto out; -fs_path_free (name); +fs_path_reset (name); + to_path = name; name =NULL ; - to_path =fs_path_alloc (); -if (!to_path) { - ret = -ENOMEM ; - goto out; - } - sctx->send_progress = sctx->cur_ino +1 ; ret =get_cur_path (sctx, pm->ino, pm->gen, to_path);if (ret <0 )
Greg Kroah-Hartman
maintainerPATCH 120/139 Btrfs: make sure there are not any read requests before stopping workers
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Wang Shilong <[email protected]>
commit de348ee022175401e77d7662b7ca6e231a94e3fd upstream.
In close_ctree(), after we have stopped all workers,there maybe still some read requests(for example readahead) to submit and this *maybe* trigger an oops that user reported before:
kernel BUG at fs/btrfs/async-thread.c:619!
By hacking codes, i can reproduce this problem with one cpu available. We fix this potential problem by invalidating all btree inode pages before stopping all workers.
Thanks to Miao for pointing out this problem.
--- fs/btrfs/disk-io.c | 5 +++++ 1 file changed, 5 insertions(+)
fs/btrfs/disk-io.c +5 -0
@@ -3633,6 +3633,11 @@ int close_ctree(struct btrfs_root *root)btrfs_free_block_groups (fs_info); +/* +* we must make sure there is not any read request to +* submit after we stopping all workers. +*/ +invalidate_inode_pages2 (fs_info->btree_inode->i_mapping);btrfs_stop_all_workers (fs_info);free_root_pointers (fs_info,1 );
Greg Kroah-Hartman
maintainerPATCH 121/139 Btrfs: fix NULL pointer crash of deleting a seed device
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Liu Bo <[email protected]>
commit 29cc83f69c8338ff8fd1383c9be263d4bdf52d73 upstream.
Same as normal devices, seed devices should be initialized with fs_info->dev_root as well, otherwise we'll get a NULL pointer crash.
--- fs/btrfs/volumes.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-)
fs/btrfs/volumes.c +8 -4
@@ -6080,10 +6080,14 @@ void btrfs_init_devices_late(struct btrfstruct btrfs_fs_devices *fs_devices = fs_info->fs_devices;struct btrfs_device *device; -mutex_lock (&fs_devices->device_list_mutex); -list_for_each_entry (device, &fs_devices->devices, dev_list) - device->dev_root = fs_info->dev_root; -mutex_unlock (&fs_devices->device_list_mutex); +while (fs_devices) { +mutex_lock (&fs_devices->device_list_mutex); +list_for_each_entry (device, &fs_devices->devices, dev_list) + device->dev_root = fs_info->dev_root; +mutex_unlock (&fs_devices->device_list_mutex); + + fs_devices = fs_devices->seed; + } }static void __btrfs_reset_dev_stats (struct btrfs_device *dev)
Greg Kroah-Hartman
maintainerPATCH 122/139 Btrfs: mark mapping with error flag to report errors to userspace
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Liu Bo <[email protected]>
commit 5dca6eea91653e9949ce6eb9e9acab6277e2f2c4 upstream.
According to commit 865ffef3797da2cac85b3354b5b6050dc9660978 (fs: fix fsync() error reporting),
it's not stable to just check error pages because pages can be truncated or invalidated, we should also mark mapping with error flag so that a later fsync can catch the error.
--- fs/btrfs/extent_io.c | 2 ++ 1 file changed, 2 insertions(+)
fs/btrfs/extent_io.c +2 -0
@@ -2368,6 +2368,8 @@ int end_extent_writepage(struct page *paif (!uptodate) {ClearPageUptodate (page);SetPageError (page); + ret = ret <0 ? ret : -EIO ; +mapping_set_error (page->mapping, ret); }return 0 ; }
Greg Kroah-Hartman
maintainerPATCH 123/139 Btrfs: set right total device count for seeding support
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Wang Shilong <[email protected]>
commit 298658414a2f0bea1f05a81876a45c1cd96aa2e0 upstream.
Seeding device support allows us to create a new filesystem based on existed filesystem.
However newly created filesystem's @total_devices should include seed devices. This patch fix the following problem:
# mkfs.btrfs -f /dev/sdb # btrfstune -S 1 /dev/sdb # mount /dev/sdb /mnt # btrfs device add -f /dev/sdc /mnt --->fs_devices->total_devices = 1 # umount /mnt # mount /dev/sdc /mnt --->fs_devices->total_devices = 2
This is because we record right @total_devices in superblock, but @fs_devices->total_devices is reset to be 0 in btrfs_prepare_sprout().
Fix this problem by not resetting @fs_devices->total_devices.
--- fs/btrfs/volumes.c | 1 - 1 file changed, 1 deletion(-)
fs/btrfs/volumes.c +0 -1
@@ -1903,7 +1903,6 @@ static int btrfs_prepare_sprout(struct b fs_devices->seeding =0 ; fs_devices->num_devices =0 ; fs_devices->open_devices =0 ; - fs_devices->total_devices =0 ; fs_devices->seed = seed_devices;generate_random_uuid (fs_devices->fsid);
Greg Kroah-Hartman
maintainerPATCH 124/139 Btrfs: send, dont error in the presence of subvols/snapshots
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Filipe Manana <[email protected]>
commit 1af56070e3ef9477dbc7eba3b9ad7446979c7974 upstream.
If we are doing an incremental send and the base snapshot has a directory with name X that doesn't exist anymore in the second snapshot and a new subvolume/snapshot exists in the second snapshot that has the same name as the directory (name X), the incremental send would fail with -ENOENT error. This is because it attempts to lookup for an inode with a number matching the objectid of a root, which doesn't exist.
Steps to reproduce:
mkfs.btrfs -f /dev/sdd
mount /dev/sdd /mnt
mkdir /mnt/testdir
btrfs subvolume snapshot -r /mnt /mnt/mysnap1
rmdir /mnt/testdir
btrfs subvolume create /mnt/testdir
btrfs subvolume snapshot -r /mnt /mnt/mysnap2
btrfs send -p /mnt/mysnap1 /mnt/mysnap2 -f /tmp/send.data
A test case for xfstests follows.
--- fs/btrfs/send.c | 4 ++++ 1 file changed, 4 insertions(+)
fs/btrfs/send.c +4 -0
@@ -1628,6 +1628,10 @@ static int lookup_dir_item_inode(struct goto out; }btrfs_dir_item_key_to_cpu (path->nodes[0 ], di, &key); +if (key.type ==BTRFS_ROOT_ITEM_KEY ) { + ret = -ENOENT ; + goto out; + } *found_inode = key.objectid; *found_type =btrfs_dir_type (path->nodes[0 ], di);
Greg Kroah-Hartman
maintainerPATCH 125/139 Btrfs: send, use the right limits for xattr names and values
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Filipe Manana <[email protected]>
commit 7e3ae33efad1490d01040f552ef50e58ed6376ca upstream.
We were limiting the sum of the xattr name and value lengths to PATH_MAX, which is not correct, specially on filesystems created with btrfs-progs v3.12 or higher, where the default leaf size is max(16384, PAGE_SIZE), or systems with page sizes larger than 4096 bytes.
Xattrs have their own specific maximum name and value lengths, which depend on the leaf size, therefore use these limits to be able to send xattrs with sizes larger than PATH_MAX.
A test case for xfstests follows.
--- fs/btrfs/send.c | 30 +++++++++++++++++++++++------- 1 file changed, 23 insertions(+), 7 deletions(-)
fs/btrfs/send.c +23 -7
@@ -975,7 +975,7 @@ static int iterate_dir_item(struct btrfsstruct btrfs_dir_item *di;struct btrfs_key di_key;char *buf =NULL ; -const int buf_len =PATH_MAX ; +int buf_len;u32 name_len;u32 data_len;u32 cur;
@@ -985,6 +985,11 @@ static int iterate_dir_item(struct btrfsint num;u8 type; +if (found_key->type ==BTRFS_XATTR_ITEM_KEY ) + buf_len =BTRFS_MAX_XATTR_SIZE (root); +else + buf_len =PATH_MAX ; + buf =kmalloc (buf_len,GFP_NOFS );if (!buf) { ret = -ENOMEM ;
@@ -1006,12 +1011,23 @@ static int iterate_dir_item(struct btrfs type =btrfs_dir_type (eb, di);btrfs_dir_item_key_to_cpu (eb, di, &di_key); -/* -* Path too long -*/ -if (name_len + data_len > buf_len) { - ret = -ENAMETOOLONG ; - goto out; +if (type ==BTRFS_FT_XATTR ) { +if (name_len >XATTR_NAME_MAX ) { + ret = -ENAMETOOLONG ; + goto out; + } +if (name_len + data_len > buf_len) { + ret = -E2BIG ; + goto out; + } + }else { +/* +* Path too long +*/ +if (name_len + data_len > buf_len) { + ret = -ENAMETOOLONG ; + goto out; + } } read_extent_buffer(eb, buf, (unsigned long)(di + 1),
Greg Kroah-Hartman
maintainerPATCH 126/139 btrfs: allocate raid type kobjects dynamically
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Jeff Mahoney <[email protected]>
commit c1895442be01c58449e3bf9272f22062a670e08f upstream.
We are currently allocating space_info objects in an array when we allocate space_info. When a user does something like:
# btrfs balance start -mconvert=raid1 -dconvert=raid1 /mnt # btrfs balance start -mconvert=single -dconvert=single /mnt -f # btrfs balance start -mconvert=raid1 -dconvert=raid1 /
We can end up with memory corruption since the kobject hasn't been reinitialized properly and the name pointer was left set.
The rationale behind allocating them statically was to avoid creating a separate kobject container that just contained the raid type. It used the index in the array to determine the index.
Ultimately, though, this wastes more memory than it saves in all but the most complex scenarios and introduces kobject lifetime questions.
This patch allocates the kobjects dynamically instead. Note that we also remove the kobject_get/put of the parent kobject since kobject_add and kobject_del do that internally.
--- fs/btrfs/ctree.h | 8 +++++++- fs/btrfs/extent-tree.c | 39 ++++++++++++++++++++++++++------------- fs/btrfs/sysfs.c | 5 +++-- 3 files changed, 36 insertions(+), 16 deletions(-)
fs/btrfs/ctree.h +7 -1
@@ -1113,6 +1113,12 @@ struct btrfs_qgroup_limit_item {__le64 rsv_excl; } __attribute__ ((__packed__)); +/* For raid type sysfs entries */ +struct raid_kobject { +int raid_type; +struct kobject kobj; +}; +struct btrfs_space_info {spinlock_t lock;
@@ -1163,7 +1169,7 @@ struct btrfs_space_info {wait_queue_head_t wait;struct kobject kobj; -struct kobject block_group_kobjs[BTRFS_NR_RAID_TYPES ]; +struct kobject *block_group_kobjs[BTRFS_NR_RAID_TYPES ]; };#define BTRFS_BLOCK_RSV_GLOBAL 1
fs/btrfs/extent-tree.c +26 -13
@@ -3401,10 +3401,8 @@ static int update_space_info(struct btrfreturn ret; } -for (i =0 ; i <BTRFS_NR_RAID_TYPES ; i++) { +for (i =0 ; i <BTRFS_NR_RAID_TYPES ; i++)INIT_LIST_HEAD (&found->block_groups[i]); -kobject_init (&found->block_group_kobjs[i], &btrfs_raid_ktype); - }init_rwsem (&found->groups_sem);spin_lock_init (&found->lock); found->flags = flags &BTRFS_BLOCK_GROUP_TYPE_MASK ;
@@ -8327,8 +8325,9 @@ int btrfs_free_block_groups(struct btrfslist_del (&space_info->list);for (i =0 ; i <BTRFS_NR_RAID_TYPES ; i++) {struct kobject *kobj; - kobj = &space_info->block_group_kobjs[i]; -if (kobj->parent) { + kobj = space_info->block_group_kobjs[i]; + space_info->block_group_kobjs[i] =NULL ; +if (kobj) {kobject_del (kobj);kobject_put (kobj); }
@@ -8352,17 +8351,26 @@ static void __link_block_group(struct btup_write (&space_info->groups_sem);if (first) { -struct kobject *kobj = &space_info->block_group_kobjs[index]; +struct raid_kobject *rkobj;int ret; -kobject_get (&space_info->kobj);/* put in release */ - ret =kobject_add (kobj, &space_info->kobj,"%s" , -get_raid_name (index)); + rkobj =kzalloc (sizeof (*rkobj),GFP_NOFS ); +if (!rkobj) + goto out_err; + rkobj->raid_type = index; +kobject_init (&rkobj->kobj, &btrfs_raid_ktype); + ret =kobject_add (&rkobj->kobj, &space_info->kobj, +"%s" ,get_raid_name (index));if (ret) { -pr_warn ("BTRFS: failed to add kobject for block cache. ignoring.\n" ); -kobject_put (&space_info->kobj); +kobject_put (&rkobj->kobj); + goto out_err; } + space_info->block_group_kobjs[index] = &rkobj->kobj; } + +return ; +out_err: +pr_warn ("BTRFS: failed to add kobject for block cache. ignoring.\n" ); } static struct btrfs_block_group_cache *
@@ -8697,6 +8705,7 @@ int btrfs_remove_block_group(struct btrfstruct btrfs_root *tree_root = root->fs_info->tree_root;struct btrfs_key key;struct inode *inode; +struct kobject *kobj =NULL ;int ret;int index;int factor;
@@ -8796,11 +8805,15 @@ int btrfs_remove_block_group(struct btrf*/ list_del_init (&block_group->list);if (list_empty (&block_group->space_info->block_groups[index])) { -kobject_del (&block_group->space_info->block_group_kobjs[index]); -kobject_put (&block_group->space_info->block_group_kobjs[index]); + kobj = block_group->space_info->block_group_kobjs[index]; + block_group->space_info->block_group_kobjs[index] =NULL ;clear_avail_alloc_bits (root->fs_info, block_group->flags); }up_write (&block_group->space_info->groups_sem); +if (kobj) { +kobject_del (kobj); +kobject_put (kobj); + }if (block_group->cached ==BTRFS_CACHE_STARTED )wait_block_group_cache_done (block_group);
fs/btrfs/sysfs.c +3 -2
@@ -254,6 +254,7 @@ static ssize_t global_rsv_reserved_show(BTRFS_ATTR (global_rsv_reserved,0444 , global_rsv_reserved_show);#define to_space_info (_kobj) container_of(_kobj, struct btrfs_space_info, kobj) +#define to_raid_kobj (_kobj) container_of(_kobj, struct raid_kobject, kobj)static ssize_t raid_bytes_show (struct kobject *kobj,struct kobj_attribute *attr,char *buf);
@@ -266,7 +267,7 @@ static ssize_t raid_bytes_show(struct ko {struct btrfs_space_info *sinfo =to_space_info (kobj->parent);struct btrfs_block_group_cache *block_group; -int index = kobj - sinfo->block_group_kobjs; +int index =to_raid_kobj (kobj)->raid_type;u64 val =0 ;down_read (&sinfo->groups_sem);
@@ -288,7 +289,7 @@ static struct attribute *raid_attributesstatic void release_raid_kobj (struct kobject *kobj) { -kobject_put (kobj->parent); +kfree (to_raid_kobj (kobj)); }struct kobj_type btrfs_raid_ktype = {
Greg Kroah-Hartman
maintainerPATCH 127/139 fs: btrfs: volumes.c: Fix for possible null pointer dereference
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Rickard Strandqvist <[email protected]>
commit 8321cf2596d283821acc466377c2b85bcd3422b7 upstream.
There is otherwise a risk of a possible null pointer dereference.
Was largely found by using a static code analysis program called cppcheck.
--- fs/btrfs/volumes.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
fs/btrfs/volumes.c +3 -2
@@ -1690,11 +1690,12 @@ int btrfs_rm_device(struct btrfs_root *rstruct btrfs_fs_devices *fs_devices; fs_devices = root->fs_info->fs_devices;while (fs_devices) { -if (fs_devices->seed == cur_devices) +if (fs_devices->seed == cur_devices) { + fs_devices->seed = cur_devices->seed;break ; + } fs_devices = fs_devices->seed; } - fs_devices->seed = cur_devices->seed; cur_devices->seed =NULL ;lock_chunks (root);__btrfs_close_devices (cur_devices);
Greg Kroah-Hartman
maintainerPATCH 128/139 Btrfs: dont check nodes for extent items
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Josef Bacik <[email protected]>
commit 8a56457f5f8fa7c2698ffae8545214c5b96a2cb5 upstream.
The backref code was looking at nodes as well as leaves when we tried to populate extent item entries. This is not good, and although we go away with it for the most part because we'd skip where disk_bytenr != random_memory, sometimes random_memory would match and suddenly boom. This fixes that problem. Thanks,
--- fs/btrfs/backref.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
fs/btrfs/backref.c +3 -2
@@ -984,11 +984,12 @@ again: goto out; }if (ref->count && ref->parent) { -if (extent_item_pos && !ref->inode_list) { +if (extent_item_pos && !ref->inode_list && + ref->level ==0 ) {u32 bsz;struct extent_buffer *eb; bsz =btrfs_level_size (fs_info->extent_root, - info_level); + ref->level); eb =read_tree_block (fs_info->extent_root, ref->parent, bsz,0 );if (!eb || !extent_buffer_uptodate (eb)) {
Greg Kroah-Hartman
maintainerPATCH 129/139 Btrfs: use right type to get real comparison
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Liu Bo <[email protected]>
commit cd857dd6bc2ae9ecea14e75a34e8a8fdc158e307 upstream.
We want to make sure the point is still within the extent item, not to verify the memory it's pointing to.
--- fs/btrfs/backref.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
fs/btrfs/backref.c +1 -1
@@ -1424,7 +1424,7 @@ static int __get_extent_inline_ref(unsig *out_eiref = (struct btrfs_extent_inline_ref *)(ei +1 ); } *ptr = (unsigned long )*out_eiref; -if ((void *)*ptr >= (void *)ei + item_size) +if ((unsigned long )(*ptr) >= (unsigned long )ei + item_size)return -ENOENT ; }
Greg Kroah-Hartman
maintainerPATCH 130/139 Btrfs: fix scrub_print_warning to handle skinny metadata extents
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Liu Bo <[email protected]>
commit 6eda71d0c030af0fc2f68aaa676e6d445600855b upstream.
The skinny extents are intepreted incorrectly in scrub_print_warning(), and end up hitting the BUG() in btrfs_extent_inline_ref_size.
--- fs/btrfs/backref.c | 30 +++++++++++++++++++----------- fs/btrfs/backref.h | 4 ++-- fs/btrfs/scrub.c | 5 +++-- 3 files changed, 24 insertions(+), 15 deletions(-)
fs/btrfs/backref.c +19 -11
@@ -1405,9 +1405,10 @@ int extent_from_logical(struct btrfs_fs_* returns <0 on error */ static int __get_extent_inline_ref (unsigned long *ptr,struct extent_buffer *eb, -struct btrfs_extent_item *ei,u32 item_size, -struct btrfs_extent_inline_ref **out_eiref, -int *out_type) +struct btrfs_key *key, +struct btrfs_extent_item *ei,u32 item_size, +struct btrfs_extent_inline_ref **out_eiref, +int *out_type) {unsigned long end;u64 flags;
@@ -1417,9 +1418,16 @@ static int __get_extent_inline_ref(unsig/* first call */ flags =btrfs_extent_flags (eb, ei);if (flags &BTRFS_EXTENT_FLAG_TREE_BLOCK ) { - info = (struct btrfs_tree_block_info *)(ei +1 ); - *out_eiref = - (struct btrfs_extent_inline_ref *)(info +1 ); +if (key->type ==BTRFS_METADATA_ITEM_KEY ) { +/* a skinny metadata extent */ + *out_eiref = + (struct btrfs_extent_inline_ref *)(ei +1 ); + }else { +WARN_ON (key->type !=BTRFS_EXTENT_ITEM_KEY ); + info = (struct btrfs_tree_block_info *)(ei +1 ); + *out_eiref = + (struct btrfs_extent_inline_ref *)(info +1 ); + } }else { *out_eiref = (struct btrfs_extent_inline_ref *)(ei +1 ); }
@@ -1429,7 +1437,7 @@ static int __get_extent_inline_ref(unsig } end = (unsigned long )ei + item_size; - *out_eiref = (struct btrfs_extent_inline_ref *)*ptr; + *out_eiref = (struct btrfs_extent_inline_ref *)(*ptr); *out_type =btrfs_extent_inline_ref_type (eb, *out_eiref); *ptr +=btrfs_extent_inline_ref_size (*out_type);
@@ -1448,8 +1456,8 @@ static int __get_extent_inline_ref(unsig* <0 on error. */ int tree_backref_for_extent (unsigned long *ptr,struct extent_buffer *eb, -struct btrfs_extent_item *ei,u32 item_size, -u64 *out_root,u8 *out_level) +struct btrfs_key *key,struct btrfs_extent_item *ei, +u32 item_size,u64 *out_root,u8 *out_level) {int ret;int type;
@@ -1460,8 +1468,8 @@ int tree_backref_for_extent(unsigned lonreturn 1 ;while (1 ) { - ret =__get_extent_inline_ref (ptr, eb, ei, item_size, - &eiref, &type); + ret =__get_extent_inline_ref (ptr, eb, key, ei, item_size, + &eiref, &type);if (ret <0 )return ret;
fs/btrfs/backref.h +2 -2
@@ -40,8 +40,8 @@ int extent_from_logical(struct btrfs_fs_ u64 *flags);int tree_backref_for_extent (unsigned long *ptr,struct extent_buffer *eb, -struct btrfs_extent_item *ei,u32 item_size, -u64 *out_root,u8 *out_level); +struct btrfs_key *key,struct btrfs_extent_item *ei, +u32 item_size,u64 *out_root,u8 *out_level); int iterate_extent_inodes(struct btrfs_fs_info *fs_info, u64 extent_item_objectid,
fs/btrfs/scrub.c +3 -2
@@ -588,8 +588,9 @@ static void scrub_print_warning(const chif (flags &BTRFS_EXTENT_FLAG_TREE_BLOCK ) { do { - ret = tree_backref_for_extent(&ptr, eb, ei, item_size, - &ref_root, &ref_level); + ret = tree_backref_for_extent(&ptr, eb, &found_key, ei, + item_size, &ref_root, + &ref_level); printk_in_rcu(KERN_WARNING "BTRFS: %s at logical %llu on dev %s, " "sector %llu: metadata %s (level %d) in tree "
Greg Kroah-Hartman
maintainerPATCH 131/139 btrfs: fix use of uninit "ret" in end_extent_writepage()
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Eric Sandeen <[email protected]>
commit 3e2426bd0eb980648449e7a2f5a23e3cd3c7725c upstream.
If this condition in end_extent_writepage() is false:
if (tree->ops && tree->ops->writepage_end_io_hook)
we will then test an uninitialized "ret" at:
ret = ret < 0 ? ret : -EIO;
The test for ret is for the case where ->writepage_end_io_hook failed, and we'd choose that ret as the error; but if there is no ->writepage_end_io_hook, nothing sets ret.
Initializing ret to 0 should be sufficient; if writepage_end_io_hook wasn't set, (!uptodate) means non-zero err was passed in, so we choose -EIO in that case.
--- fs/btrfs/extent_io.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
fs/btrfs/extent_io.c +1 -1
@@ -2354,7 +2354,7 @@ int end_extent_writepage(struct page *pa {int uptodate = (err ==0 );struct extent_io_tree *tree; -int ret; +int ret =0 ; tree = &BTRFS_I (page->mapping->host)->io_tree;
Greg Kroah-Hartman
maintainerPATCH 132/139 Bluetooth: Clearly distinguish mgmt LTK type from authenticated property
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hedberg <[email protected]>
commit d7b2545023ecfde94d3ea9c03c5480ac18da96c9 upstream.
On the mgmt level we have a key type parameter which currently accepts two possible values: 0x00 for unauthenticated and 0x01 for authenticated. However, in the internal struct smp_ltk representation we have an explicit "authenticated" boolean value.
To make this distinction clear, add defines for the possible mgmt values and do conversion to and from the internal authenticated value.
--- include/net/bluetooth/mgmt.h | 3 +++ net/bluetooth/mgmt.c | 19 ++++++++++++++++--- 2 files changed, 19 insertions(+), 3 deletions(-)
include/net/bluetooth/mgmt.h +3 -0
@@ -181,6 +181,9 @@ struct mgmt_cp_load_link_keys { } __packed;#define MGMT_LOAD_LINK_KEYS_SIZE 3 +#define MGMT_LTK_UNAUTHENTICATED 0x00 +#define MGMT_LTK_AUTHENTICATED 0x01 +struct mgmt_ltk_info {struct mgmt_addr_info addr;__u8 type;
net/bluetooth/mgmt.c +16 -3
@@ -4530,7 +4530,7 @@ static int load_long_term_keys(struct sofor (i =0 ; i < key_count; i++) {struct mgmt_ltk_info *key = &cp->keys[i]; -u8 type, addr_type; +u8 type, addr_type, authenticated;if (key->addr.type ==BDADDR_LE_PUBLIC ) addr_type =ADDR_LE_DEV_PUBLIC ;
@@ -4542,8 +4542,13 @@ static int load_long_term_keys(struct soelse type =HCI_SMP_LTK_SLAVE ; +if (key->type ==MGMT_LTK_UNAUTHENTICATED ) + authenticated =0x00 ; +else + authenticated =0x01 ; +hci_add_ltk (hdev, &key->addr.bdaddr, addr_type, type, - key->type, key->val, key->enc_size, key->ediv, + authenticated, key->val, key->enc_size, key->ediv, key->rand); }
@@ -5005,6 +5010,14 @@ void mgmt_new_link_key(struct hci_dev *hmgmt_event (MGMT_EV_NEW_LINK_KEY , hdev, &ev,sizeof (ev),NULL ); } +static u8 mgmt_ltk_type (struct smp_ltk *ltk) +{ +if (ltk->authenticated) +return MGMT_LTK_AUTHENTICATED ; + +return MGMT_LTK_UNAUTHENTICATED ; +} +void mgmt_new_ltk (struct hci_dev *hdev,struct smp_ltk *key,bool persistent) {struct mgmt_ev_new_long_term_key ev;
@@ -5030,7 +5043,7 @@ void mgmt_new_ltk(struct hci_dev *hdev,bacpy (&ev.key.addr.bdaddr, &key->bdaddr); ev.key.addr.type =link_to_bdaddr (LE_LINK , key->bdaddr_type); - ev.key.type = key->authenticated; + ev.key.type =mgmt_ltk_type (key); ev.key.enc_size = key->enc_size; ev.key.ediv = key->ediv; ev.key.rand = key->rand;
Greg Kroah-Hartman
maintainerPATCH 133/139 Bluetooth: Fix properly ignoring LTKs of unknown types
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hedberg <[email protected]>
commit 61b433579b6ffecb1d3534fd482dcd48535277c8 upstream.
In case there are new LTK types in the future we shouldn't just blindly assume that != MGMT_LTK_UNAUTHENTICATED means that the key is authenticated. This patch adds explicit checks for each allowed key type in the form of a switch statement and skips any key which has an unknown value.
--- net/bluetooth/mgmt.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)
net/bluetooth/mgmt.c +8 -2
@@ -4542,10 +4542,16 @@ static int load_long_term_keys(struct soelse type =HCI_SMP_LTK_SLAVE ; -if (key->type ==MGMT_LTK_UNAUTHENTICATED ) +switch (key->type) { +case MGMT_LTK_UNAUTHENTICATED : authenticated =0x00 ; -else +break ; +case MGMT_LTK_AUTHENTICATED : authenticated =0x01 ; +break ; +default : +continue ; + } hci_add_ltk(hdev, &key->addr.bdaddr, addr_type, type, authenticated, key->val, key->enc_size, key->ediv,
Greg Kroah-Hartman
maintainerPATCH 134/139 lz4: fix another possible overrun
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Greg Kroah-Hartman <[email protected]>
commit 4148c1f67abf823099b2d7db6851e4aea407f5ee upstream.
There is one other possible overrun in the lz4 code as implemented by Linux at this point in time (which differs from the upstream lz4 codebase, but will get synced at in a future kernel release.) As pointed out by Don, we also need to check the overflow in the data itself.
While we are at it, replace the odd error return value with just a "simple" -1 value as the return value is never used for anything other than a basic "did this work or not" check.
--- lib/lz4/lz4_decompress.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
lib/lz4/lz4_decompress.c +3 -1
@@ -108,6 +108,8 @@ static int lz4_uncompress(const char *soif (length ==ML_MASK ) {for (; *ip ==255 ; length +=255 ) ip++; +if (unlikely (length > (size_t )(length + *ip))) + goto _output_error; length += *ip++; }
@@ -157,7 +159,7 @@ static int lz4_uncompress(const char *so/* write overflow error detected */ _output_error: -return (int ) (-(((char *)ip) - source)); +return -1 ; } static int lz4_uncompress_unknownoutputsize(const char *source, char *dest,
Greg Kroah-Hartman
maintainerPATCH 136/139 epoll: fix use-after-free in eventpoll_release_file
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Konstantin Khlebnikov <[email protected]>
commit ebe06187bf2aec10d537ce4595e416035367d703 upstream.
This fixes use-after-free of epi->fllink.next inside list loop macro. This loop actually releases elements in the body. The list is rcu-protected but here we cannot hold rcu_read_lock because we need to lock mutex inside.
The obvious solution is to use list_for_each_entry_safe(). RCU-ness isn't essential because nobody can change this list under us, it's final fput for this file.
The bug was introduced by ae10b2b4eb01 ("epoll: optimize EPOLL_CTL_DEL using rcu")
--- fs/eventpoll.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
fs/eventpoll.c +2 -2
@@ -910,7 +910,7 @@ static const struct file_operations evenvoid eventpoll_release_file (struct file *file) {struct eventpoll *ep; -struct epitem *epi; +struct epitem *epi, *next;/* * We don't want to get "file->f_lock" because it is not
@@ -926,7 +926,7 @@ void eventpoll_release_file(struct file* Besides, ep_remove() acquires the lock, so we can't hold it here. */ mutex_lock (&epmutex); -list_for_each_entry_rcu (epi, &file->f_ep_links, fllink) { +list_for_each_entry_safe (epi, next, &file->f_ep_links, fllink) { ep = epi->ep;mutex_lock_nested (&ep->mtx,0 );ep_remove (ep, epi);
Greg Kroah-Hartman
maintainerPATCH 137/139 random: fix nasty entropy accounting bug
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Theodore Ts'o <[email protected]>
commit e33ba5fa7afce1a9f159704121d4e4d110df8185 upstream.
Commit 0fb7a01af5b0 "random: simplify accounting code", introduced in v3.15, has a very nasty accounting problem when the entropy pool has has fewer bytes of entropy than the number of requested reserved bytes. In that case, "have_bytes - reserved" goes negative, and since size_t is unsigned, the expression:
ibytes = min_t(size_t, ibytes, have_bytes - reserved);
... does not do the right thing. This is rather bad, because it defeats the catastrophic reseeding feature in the xfer_secondary_pool() path.
It also can cause the "BUG: spinlock trylock failure on UP" for some kernel configurations when prandom_reseed() calls get_random_bytes() in the early init, since when the entropy count gets corrupted, credit_entropy_bits() erroneously believes that the nonblocking pool has been fully initialized (when in fact it is not), and so it calls prandom_reseed(true) recursively leading to the spinlock BUG.
The logic is *not* the same it was originally, but in the cases where it matters, the behavior is the same, and the resulting code is hopefully easier to read and understand.
--- drivers/char/random.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-)
drivers/char/random.c +9 -8
@@ -979,7 +979,6 @@ static void push_to_pool(struct work_strstatic size_t account (struct entropy_store *r,size_t nbytes,int min,int reserved) { -int have_bytes;int entropy_count, orig;size_t ibytes;
@@ -988,17 +987,19 @@ static size_t account(struct entropy_sto/* Can we pull enough? */ retry: entropy_count = orig =ACCESS_ONCE (r->entropy_count); - have_bytes = entropy_count >> (ENTROPY_SHIFT +3 ); ibytes = nbytes;/* If limited, never pull more than available */ -if (r->limit) - ibytes =min_t (size_t, ibytes, have_bytes - reserved); +if (r->limit) { +int have_bytes = entropy_count >> (ENTROPY_SHIFT +3 ); + +if ((have_bytes -= reserved) <0 ) + have_bytes =0 ; + ibytes =min_t (size_t, ibytes, have_bytes); + }if (ibytes < min) ibytes =0 ; -if (have_bytes >= ibytes + reserved) - entropy_count -= ibytes << (ENTROPY_SHIFT +3 ); -else - entropy_count = reserved << (ENTROPY_SHIFT +3 ); +if ((entropy_count -= ibytes << (ENTROPY_SHIFT +3 )) <0 ) + entropy_count =0 ;if (cmpxchg (&r->entropy_count, orig, entropy_count) != orig) goto retry;
Greg Kroah-Hartman
maintainerPATCH 138/139 builddeb: use $OBJCOPY variable instead of objcopy
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Fathi Boudra <[email protected]>
commit 6b4a144a92ab81a1f45fb9b12aebaaaee0d08120 upstream.
In cross-build environment, we expect to use the cross-compiler objcopy instead of the host objcopy.
It fixes following build failures:
objcopy --only-keep-debug lib/modules/3.14/kernel/net/ipv6/xfrm6_mode_tunnel.ko /srv/build/linux/debian/dbgtmp/usr/lib/debug/lib/modules/3.14/kernel/net/ipv6/xfrm6_mode_tunnel.ko
objcopy: Unable to recognise the format of the input file `lib/modules/3.14/kernel/net/ipv6/xfrm6_mode_tunnel.ko'
--- scripts/package/builddeb | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
scripts/package/builddeb +3 -3
@@ -155,11 +155,11 @@ if grep -q '^CONFIG_MODULES=y' $KCONFIG_ for module in $(find lib/modules/ -name *.ko); do mkdir -p $(dirname $dbg_dir/usr/lib/debug/$module) # only keep debug symbols in the debug file - objcopy --only-keep-debug $module $dbg_dir/usr/lib/debug/$module + $OBJCOPY --only-keep-debug $module $dbg_dir/usr/lib/debug/$module # strip original module from debug symbols - objcopy --strip-debug $module + $OBJCOPY --strip-debug $module # then add a link to those - objcopy --add-gnu-debuglink=$dbg_dir/usr/lib/debug/$module $module + $OBJCOPY --add-gnu-debuglink=$dbg_dir/usr/lib/debug/$module $module done ) fi
Greg Kroah-Hartman
maintainerPATCH 139/139 efi-pstore: Fix an overflow on 32-bit builds
In reply to Greg Kroah-Hartman3.15-stable review patch. If anyone has any objections, please let me know.
------------------
From: Andrzej Zaborowski <[email protected]>
commit 783ee43118dc773bc8b0342c5b230e017d5a04d0 upstream.
In generic_id the long int timestamp is multiplied by 100000 and needs an explicit cast to u64.
Without that the id in the resulting pstore filename is wrong and userspace may have problems parsing it, but more importantly files in pstore can never be deleted and may fill the EFI flash (brick device?). This happens because when generic pstore code wants to delete a file, it passes the id to the EFI backend which reinterpretes it and a wrong variable name is attempted to be deleted. There's no error message but after remounting pstore, deleted files would reappear.
--- drivers/firmware/efi/efi-pstore.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/firmware/efi/efi-pstore.c +1 -1
@@ -40,7 +40,7 @@ struct pstore_read_data {static inline u64 generic_id (unsigned long timestamp,unsigned int part,int count) { -return (timestamp *100 + part) *1000 + count; +return ((u64 ) timestamp *100 + part) *1000 + count; }static int efi_pstore_read_func (struct efivar_entry *entry,void *data)
Quoted 8 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 3.15.3 release. There are 139 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.
Responses should be made by Mon Jun 30 17:45:39 UTC 2014. Anything received after that time might be too late.
Build results:
total: 144 pass: 136 skipped: 5 fail: 3
Failed builds:
powerpc:allmodconfig
powerpc:allmodconfig
unicore32:defconfig
Qemu tests all passed.
Results are as expected. Details are available at http://server.roeck-us.net:8010/builders.
Guenter
At Sat, 28 Jun 2014 15:32:53 -0700,
Quoted 18 lines · Guenter Roeck
Quoted 8 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 3.15.3 release. There are 139 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.
Responses should be made by Mon Jun 30 17:45:39 UTC 2014. Anything received after that time might be too late.
Build results:
total: 144 pass: 136 skipped: 5 fail: 3Failed builds:
powerpc:allmodconfig
powerpc:allmodconfig
unicore32:defconfig
Quoted 6 lines
Qemu tests all passed.
Results are as expected. Details are available at http://server.roeck-us.net:8010/builders.
Guenter
3.15.3-rc1, 3.14.10-rc1, 3.10.46-r1, and 3.4.96-rc1 passed my test.
- Test Cases:
- Build this kernel.
- Boot this kernel.
- Build the latest mainline kernel with this kernel.
- Test Tool:
https://github.com/satoru-takeuchi/test-linux-stable
- Test Result (kernel .config, ktest config and test log):
http://satoru-takeuchi.org/test-linux-stable/results/<version>-<test datetime>.tar.xz
- Build Environment:
- OS: Debian Jessy x86_64
- CPU: Intel(R) Core(TM) i5-2400 CPU @ 3.10GHz x 4
- memory: 8GB
- Test Target Environment:
- Debian Jessy x86_64 (KVM guest on the Build Environment)
- # of vCPU: 2
- memory: 2GB
Thanks,
Satoru
Greg Kroah-Hartman
maintainerQuoted 45 lines · Satoru Takeuchi
At Sat, 28 Jun 2014 15:32:53 -0700,
Quoted 18 lines · Guenter Roeck
Quoted 8 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 3.15.3 release. There are 139 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.
Responses should be made by Mon Jun 30 17:45:39 UTC 2014. Anything received after that time might be too late.
Build results:
total: 144 pass: 136 skipped: 5 fail: 3Failed builds:
powerpc:allmodconfig
powerpc:allmodconfig
unicore32:defconfigQuoted 6 lines
Qemu tests all passed.
Results are as expected. Details are available at http://server.roeck-us.net:8010/builders.
Guenter
3.15.3-rc1, 3.14.10-rc1, 3.10.46-r1, and 3.4.96-rc1 passed my test.
- Test Cases:
- Build this kernel.
- Boot this kernel.
- Build the latest mainline kernel with this kernel.- Test Tool:
https://github.com/satoru-takeuchi/test-linux-stable- Test Result (kernel .config, ktest config and test log):
http://satoru-takeuchi.org/test-linux-stable/results/<version>-<test datetime>.tar.xz- Build Environment:
- OS: Debian Jessyx86_64
- CPU: Intel(R) Core(TM) i5-2400 CPU @ 3.10GHz x 4
- memory: 8GB
Quoted 4 lines
- Test Target Environment:
- Debian Jessyx86_64(KVM guest on the Build Environment)
- # of vCPU: 2
- memory: 2GB
Thanks for testing and letting me know.
greg k-h
Quoted 8 lines · Greg Kroah-Hartman
This is the start of the stable review cycle for the 3.15.3 release. There are 139 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.
Responses should be made by Mon Jun 30 17:45:39 UTC 2014. Anything received after that time might be too late.
Quoted 8 lines
The whole patch series can be found in one patch at:
kernel.org/pub/linux/kernel/v3.0/stable-review/patch-3.15.3-rc1.gzand the diffstat can be found below.
thanks,
greg k-h
Compiled and booted on my test system. No dmesg regressions.
-- Shuah
Signature · 4 lines
Shuah Khan Senior Linux Kernel Developer - Open Source Group Samsung Research America(Silicon Valley) [email protected] | (970) 672-0658