This reverts commit da7ee60b03.
The current code is not mature enough, the API should allow a single
protocol to be specified. Also, the current code contains heuristics
that will depend on module load order.
Signed-off-by: David Härdeman <david@hardeman.nu>
Acked-by: Antti Seppälä <a.seppala@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Fix a couple of warnings like this:
[linux-4.2-rc7/drivers/s390/block/dcssblk.c:553]:
(style) Array index 'j' is used before limits check.
Reported-by: David Binderman <dcb314@hotmail.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
With the removal of the dynamic reallocation of page tables for
KVM (see git commit 0b46e0a3ec)
the page table allocation / freeing code can be simplified.
The page table free code can now use the alloc_pgste bit in the
mm context to decide if a page table is 2K or 4K, there is no mix
of different sized page tables anymore. This eliminates the need
to use "page->_mapcount == 0" to check for 4K page table.
Use the lower two bits in page->_mapcount to indicate which
2K fragments of the 4K page are in use.
As 31-bit support is gone, remove the two defines ALLOC_ORDER
and FRAG_MASK and use the constants directly where appropriate.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The error messages in pci_insn are for debug purposes only. Move
them to the debugfs.
Reviewed-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Sebastian Ott <sebott@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Change machine_check_init() to an early_initcall(). This makes sure it will
be called before all other cpus are online and therfore saves us a lot of
pointless smp_call_function() calls.
The control register settings will be forwarded to the other cpus when they
will be brought online.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Reviewed-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
There is no need to busy loop and monopolize a cpu for up to ~2 seconds.
The code in question that calls mdelay() is preemptible anyway, so better
let the kernel schedule different processes than just looping and causing
unnecessary delays.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The z/VM driver sets bit "63-22" in control register zero to one in order
to enable the CP Service interrupt (0x2603). However the irq subclass mask
that normally corresponds to the CP Service interrupt is
"63-54" (== "63-22-32").
So it looks like the author read the documentation with the 32 bit sized
cr0 register bit positions (== 22), but didn't realize that bit numbers
change, if applied to a 64 bit register (== 54) due to the numbering
scheme.
Also use irq_subclass_register() instead if ctl_set_bit() since multiple
services depend on the service signal subclass mask, which is the correct
bit. This also explains why nobody noticed the bug, since the bit is always
enabled anyway (e.g. pfault).
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Reviewed-by: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
The novx parameter disables the vector facility but the HWCAP_S390_VXRS
bit in the ELf hardware capabilies is always set if the machine has
the vector facility. If the user space program uses the "vx" string
in the features field of /proc/cpuinfo to utilize vector instruction
it will crash if the novx kernel paramter is set.
Convert setup_hwcaps to an arch_initcall and use MACHINE_HAS_VX to
decide if the HWCAPS_S390_VXRS bit needs to be set.
Cc: stable@vger.kernel.org # 3.18+
Reported-by: Ulrich Weigand <uweigand@de.ibm.com>
Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
get_user() and put_user() are inline functions in the meantime
again. Both will generate the mvcos instruction if compiled
with -march=z10 (or greater).
The kernel parameter "uaccess_primary" can only change the behavior
of out-of-line uaccess functions like copy_from_user() to not use
the mvcos instruction, but not for the above named inlined functions.
Therefore it is quite useless and the parameter can be removed.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
This patch removes TP3 support on CHV since there is no support
for HBR2 on this platform.
v2: rename the function to indicate it checks source rates (Jani)
v3: update comment to indicate TP3 dependency on HBR2 supported
hardware (Jani)
Cc: stable@vger.kernel.org # v4.1+
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com>
[Jani: fixed a couple of checkpatch warnings.]
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This patch removes 5.4Gbps from supported link rate for CHV since
it is not supported in it.
v2: change the ordering for better readability (Ville)
Cc: stable@vger.kernel.org # v4.1+
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
This reverts
commit fe51bfb95c.
Author: Ville Syrjälä <ville.syrjala@linux.intel.com>
Date: Thu Mar 12 17:10:38 2015 +0200
CHV does not support intermediate frequencies so reverting the
patch that added it in the first place
Cc: stable@vger.kernel.org # v4.1+
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Sivakumar Thulasimani <sivakumar.thulasimani@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Remove two more statements which always evaluate to 'false'.
These are more leftovers from the 31 bit era.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Reviewed-by: David Hildenbrand <dahi@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Remove the two facility bits
50 - constrained transactional-execution facility
74 - transactional-execution facility
from the required facilities if the kernel is built with -march=zEC12.
E.g. z/VM 6.3 doesn't virtualize the TX facility yet. Therefore a kernel
built with -march=zEC12 and ipl'ed on a zEC12 machine as a z/VM 6.3 guest
will emit a message about the missing facilities and stop working.
The kernel however doesn't make use of the TX facility, therefore remove
the two TX related facility bits and fix this unpleasant behavior.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Reviewed-by: David Hildenbrand <dahi@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
By accident this level has been removed by the NUMA infrastructure patch.
For non-NUMA systems with CPUs that span more than one book, this makes
the scheduler only use one of the books and the other books remain idle.
Fix this and re-add the missing level.
For NUMA and non-NUMA we have the following scheduling domains and groups:
- SMT (Groups: CPU threads)
- MC (Groups: Cores)
- BOOK (Groups: Books)
For the non-NUMA case we have one last level scheduling domain:
- DIE (Groups: Whole system, has all CPUs -> cpu_cpu_mask)
For the NUMA case we have the following two last level scheduling domains:
- DIE (Groups: NUMA nodes -> cpu_cpu_mask -> returns node siblings)
- NUMA (Groups: Whole system, has all CPUs -> created in sched_init_numa())
Fixes: e8054b654b ("s390/numa: add topology tree infrastructure")
Reported-and-tested-by: Evgeny Cherkashin <Eugene.Crosser@ru.ibm.com>
Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
make payload expression aware of the fact that VLAN offload may have
removed a vlan header.
When we encounter tagged skb, transparently insert the tag into the
register so that vlan header matching can work without userspace being
aware of offload features.
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This reverts
commit 047fe6e6db
Author: David Weinehall <david.weinehall@linux.intel.com>
Date: Tue Aug 4 16:55:52 2015 +0300
drm/i915: Allow parsing of variable size child device entries from VBT
That commit is not valid for v4.2, however it will be valid for v4.3. It
was simply queued too early.
The referenced regressing commit is just fine until the size of struct
common_child_dev_config changes, and that won't happen until
v4.3. Indeed, the expected size checks here rely on the increased size
of the struct, breaking new platforms.
Fixes: 047fe6e6db ("drm/i915: Allow parsing of variable size child device entries from VBT")
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: David Weinehall <david.weinehall@linux.intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
the PPC_MPC512x config automatically selected USB_EHCI_BIG_ENDIAN_*
switches, which made Kconfig warn about "unmet direct dependencies":
scripts/kconfig/conf --silentoldconfig Kconfig
warning: (PPC_MPC512x && 440EPX) selects USB_EHCI_BIG_ENDIAN_DESC which has unmet direct dependencies (USB_SUPPORT && USB && USB_EHCI_HCD)
warning: (PPC_MPC512x && PPC_PS3 && PPC_CELLEB && 440EPX) selects USB_EHCI_BIG_ENDIAN_MMIO which has unmet direct dependencies (USB_SUPPORT && USB && USB_EHCI_HCD)
warning: (PPC_MPC512x && 440EPX) selects USB_EHCI_BIG_ENDIAN_DESC which has unmet direct dependencies (USB_SUPPORT && USB && USB_EHCI_HCD)
warning: (PPC_MPC512x && PPC_PS3 && PPC_CELLEB && 440EPX) selects USB_EHCI_BIG_ENDIAN_MMIO which has unmet direct dependencies (USB_SUPPORT && USB && USB_EHCI_HCD)
make the selected entries additionally depend on USB_EHCI_HCD which
silences the warning
Signed-off-by: Gerhard Sittig <gsi@denx.de>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Pstore only supports one backend at a time. The preferred
pstore backend is set by passing the pstore.backend=<name>
argument to the kernel at boot time. Currently, while trying
to register with pstore, nvram throws an error message even
when "pstore.backend != nvram", which is unnecessary. This
patch removes the error message in case "pstore.backend != nvram".
Signed-off-by: Hari Bathini <hbathini@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
This reverts commit 4e379d36c0.
This commit opens up a race between the recovery code and the open code.
Reported-by: Olga Kornievskaia <aglo@umich.edu>
Cc: stable@vger.kernel # v4.0+
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
REPORT LUNS should not fail just because the allocation length is less
than 16. The relevant section of SPC-4 is:
4.2.5.6 Allocation length
The ALLOCATION LENGTH field specifies the maximum number of bytes or
blocks that an application client has allocated in the Data-In
Buffer. The ALLOCATION LENGTH field specifies bytes unless a
different requirement is stated in the command definition.
An allocation length of zero specifies that no data shall be
transferred. This condition shall not be considered an error.
So we should just truncate our response rather than return an error.
Signed-off-by: Roland Dreier <roland@purestorage.com>
Signed-off-by: Spencer Baugh <sbaugh@catern.com>
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org>
If we have an OPEN_DOWNGRADE and CLOSE race with one another, we want
to ensure that the layout is forgotten by the client, so that we
start afresh with a new layoutget.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
We currently register a platform device for e820 type-12 memory and
register a nvdimm bus beneath it. Registering the platform device
triggers the device-core machinery to probe for a driver, but that
search currently comes up empty. Building the nvdimm-bus registration
into the e820_pmem platform device registration in this way forces
libnvdimm to be built-in. Instead, convert the built-in portion of
CONFIG_X86_PMEM_LEGACY to simply register a platform device and move the
rest of the logic to the driver for e820_pmem, for the following
reasons:
1/ Letting e820_pmem support be a module allows building and testing
libnvdimm.ko changes without rebooting
2/ All the normal policy around modules can be applied to e820_pmem
(unbind to disable and/or blacklisting the module from loading by
default)
3/ Moving the driver to a generic location and converting it to scan
"iomem_resource" rather than "e820.map" means any other architecture can
take advantage of this simple nvdimm resource discovery mechanism by
registering a resource named "Persistent Memory (legacy)"
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
The helper pnfs_roc() has already verified that we have no delegations,
and no further open files, hence no outstanding I/O and it has marked
all the return-on-close lsegs as being invalid.
Furthermore, it sets the NFS_LAYOUT_RETURN bit, thus serialising the
close/delegreturn with all future layoutget calls on this inode.
The checks in pnfs_roc_drain() for valid layout segments are therefore
redundant: those cannot exist until another layoutget completes.
The other check for whether or not NFS_LAYOUT_RETURN is set, actually
causes a hang, since we already know that we hold that flag.
To fix, we therefore strip out all the functionality in pnfs_roc_drain()
except the retrieval of the barrier state, and then rename the function
accordingly.
Reported-by: Christoph Hellwig <hch@infradead.org>
Fixes: 5c4a79fb2b ("Don't prevent layoutgets when doing return-on-close")
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Jeff Kirsher says:
====================
Intel Wired LAN Driver Updates 2015-08-18
This series contains updates to igb, e100, e1000e and ixgbe.
Shota Suzuki provides a fix for a possible overflow in
igb_set_interrupt_capability() which leads to an oops. When changing the
number of queues by "ethtool -L", set IGB_FLAG_QUEUE_PAIRS in the same
manner as when initializing the igb driver.
Vasily Averin provides a fix for a missing rtnl_unlock() for when we
error out due to not being able to allocate memory for our queues.
Stefan Assman provides a couple of fixes for igb/igbvf. First changes
the igb driver in probe to simply call igb_enable_sriov() instead of
igb_sriov_reinit() since we are starting from scratch. Then in igbvf,
fix the driver where it does not clear the buffer_info->dma in all
cases after calling dma_unmap_single(), which was found by changing the
MTU twice.
Richard Cochran implements the periodic output function using the
programmable clock outputs available in i210 when possible, falling
back to the target time for longer periods.
Todd adds support for the Marvell PHY 1512 which is required for i354
devices. Then updates igb to make sure SR-IOV init uses the correct
number of queues, since recent changes could result in the PF holding
onto all of the queues.
Alex Williamson provides a fix in the case where a guest OS does not
support hot-unplug, so disable SR-IOV prior to unregister_netdev() to
avoid the problem.
Jia-Ju Bai provides several patches, first knocks some collecting dust
off an old e100 driver to add a check to avoid a null pointer
dereference. Then cleans up a possible resource leak by releasing the
skb buffer allocated when the e100_xmit_prepare() runs into an issue
in the DMA mapping. In igb, add a missing rtnl_unlock() for when we
error out due to igb_sriov_reinit() in the igb_init_interrupt_scheme().
Provides a e1000e fix, based on suggestions from Alex Duyck to move
head/tail register writing to e1000_configure_tx/rx() to avoid a
possible null pointer dereference (similar to igb driver). Lastly,
fix a possible memory leak in igb_probe(), where the memory shadow_vfta
allocated by kcalloc in igb_sw_init() is not freed.
Mark simplifies port-specific macros for ixgbe by eliminating explicit
comparisons with 0 and enclose formal parameters in parens to eliminate
the risk of an operator precedence issue.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Nikolay Aleksandrov says:
====================
vrf: a few simplifications and cleanups
These patches remove some unnecessary checks (patches 3, 4), unnecessary
num_slaves member and refcnt manipulations which are already done by the
upper functions.
====================
Acked-by: David Ahern <dsa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We can drop the check because if vrf_ptr is present then we must have
the vrf device as a master and since we're running with rtnl it can't go
away.
Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
dstats and rth are always present because we fail the device registration
if they can't be allocated in vrf_init() (ndo_init) so drop the unnecessary
checks.
Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
slave_queue has a num_slaves member which is unused, drop it.
Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
netdev_master_upper_dev_link/unlink already do a dev_hold/put on the
devices being linked, so no need to take another reference.
Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Removed references to ib_reg_phsy_mr() and PMR which was added
to deal with some Chelsio specific scenario, but no longer needed
now.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Amir Shehata <amir.shehata@intel.com>
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, two routes going through the same tunnel interface are considered
the same even when they are routed to a different host after encapsulation.
This causes all routes added after the first one to have incorrect
encapsulation parameters.
This is nicely visible by doing:
# ip r a 192.168.1.2/32 dev vxlan0 tunnel dst 10.0.0.2
# ip r a 192.168.1.3/32 dev vxlan0 tunnel dst 10.0.0.3
# ip r
[...]
192.168.1.2/32 tunnel id 0 src 0.0.0.0 dst 10.0.0.2 [...]
192.168.1.3/32 tunnel id 0 src 0.0.0.0 dst 10.0.0.2 [...]
Implement the missing comparison function.
Fixes: 3093fbe7ff ("route: Per route IP tunnel metadata via lightweight tunnel")
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The built lwtunnel_state struct has to be freed after comparison.
Fixes: 571e722676 ("ipv4: support for fib route lwtunnel encap attributes")
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Acked-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The object tt_local is allocated with kmalloc and not initialized when the
function batadv_tt_local_add checks for the vlan. But this function can
only cleanup the object when the (not yet initialized) reference counter of
the object is 1. This is unlikely and thus the object would leak when the
vlan could not be found.
Instead the uninitialized object tt_local has to be freed manually and the
pointer has to set to NULL to avoid calling the function which would try to
decrement the reference counter of the not existing object.
CID: 1316518
Fixes: 354136bcc3 ("batman-adv: fix kernel crash due to missing NULL checks")
Signed-off-by: Sven Eckelmann <sven@narfation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
You can't use kstrtoul() with an int or it causes memory corruption.
Also j should be unsigned or we have underflow bugs.
I considered changing "j" to unsigned long but everything fits in a u32.
Fixes: 8e3d04fd7d ('cxgb4: Add MPS tracing support')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When keeping the configuration set by the bootloader (by using
the marvell,nand-keep-config property), the pxa3xx_nand_detect_config()
function is called and set the chunk size to 512 as a default value if
NDCR_PAGE_SZ is not set.
In the other case, when not keeping the bootloader configuration, no
chunk size is set. Fix this by adding a default chunk size of 512.
Fixes: 70ed85232a ("mtd: nand: pxa3xx: Introduce multiple page I/O
support")
Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Acked-by: Robert Jarzmik <robert.jarzmik@free>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Remove extra space after the "platform:" prefix and make the alias matches
driver name.
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Acked-by: Roger Quadros <rogerq@ti.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
static analysis by smatch caught the following error:
drivers/mtd/maps/physmap_of.c:135 of_get_probes()
error: potential null dereference 'res'. (kzalloc returns null)
Check for failed kzalloc and return -ENOMEM in of_flash_probe if
this occurs.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Filesystems are responsible to manage file coherency between the page
cache and direct I/O. The generic dio code flushes dirty pages over the
range of a dio to ensure that the dio read or a future buffered read
returns the correct data. XFS has generally followed this pattern,
though traditionally has flushed and invalidated the range from the
start of the I/O all the way to the end of the file. This changed after
the following commit:
7d4ea3ce xfs: use ranged writeback and invalidation for direct IO
... as the full file flush was no longer necessary to deal with the
strange post-eof delalloc issues that were since fixed. Unfortunately,
we have since received complaints about performance degradation due to
the increased exclusive iolock cycles (which locks out parallel dio
submission) that occur when a file has cached pages. This does not occur
on filesystems that use the generic code as it also does not incorporate
locking.
The exclusive iolock is acquired any time the inode mapping has cached
pages, regardless of whether they reside in the range of the I/O or not.
If not, the flush/inval calls do no work and the lock was cycled for no
reason.
Under consideration of the cost of the exclusive iolock, update the dio
read and write handlers to flush and invalidate the entire mapping when
cached pages exist. In most cases, this increases the cost of the
initial flush sequence but eliminates the need for further lock cycles
and flushes so long as the workload does not actively mix direct and
buffered I/O. This also more closely matches historical behavior and
performance characteristics that users have come to expect.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <david@fromorbit.com>
struct xfs_attr_leafblock contains 'entries' array which is declared
with size 1 altough it can in fact contain much more entries. Since this
array is followed by further struct members, gcc (at least in version
4.8.3) thinks that the array has the fixed size of 1 element and thus
may optimize away all accesses beyond the end of array resulting in
non-working code. This problem was only observed with userspace code in
xfsprogs, however it's better to be safe in kernel as well and have
matching kernel and xfsprogs definitions.
cc: <stable@vger.kernel.org>
Signed-off-by: Jan Kara <jack@suse.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
In the dir3 data block readahead function, use the regular read
verifier to check the block's CRC and spot-check the block contents
instead of directly calling only the spot-checking routine. This
prevents corrupted directory data blocks from being read into the
kernel, which can lead to garbage ls output and directory loops (if
say one of the entries contains slashes and other junk).
cc: <stable@vger.kernel.org> # 3.12 - 4.2
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
If mtd_device_register() fails in nettel_init(), iomap left mapped.
The patch adds failure handling for mtd_device_register().
Found by Linux Driver Verification project (linuxtesting.org).
Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
The recent change to the readdir locking made in 40194ec ("xfs:
reinstate the ilock in xfs_readdir") for CXFS directory sanity was
probably the wrong thing to do. Deep in the readdir code we
can take page faults in the filldir callback, and so taking a page
fault while holding an inode ilock creates a new set of locking
issues that lockdep warns all over the place about.
The locking order for regular inodes w.r.t. page faults is io_lock
-> pagefault -> mmap_sem -> ilock. The directory readdir code now
triggers ilock -> page fault -> mmap_sem. While we cannot deadlock
at this point, it inverts all the locking patterns that lockdep
normally sees on XFS inodes, and so triggers lockdep. We worked
around this with commit 93a8614 ("xfs: fix directory inode iolock
lockdep false positive"), but that then just moved the lockdep
warning to deeper in the page fault path and triggered on security
inode locks. Fixing the shmem issue there just moved the lockdep
reports somewhere else, and now we are getting false positives from
filesystem freezing annotations getting confused.
Further, if we enter memory reclaim in a readdir path, we now get
lockdep warning about potential deadlocks because the ilock is held
when we enter reclaim. This, again, is different to a regular file
in that we never allow memory reclaim to run while holding the ilock
for regular files. Hence lockdep now throws
ilock->kmalloc->reclaim->ilock warnings.
Basically, the problem is that the ilock is being used to protect
the directory data and the inode metadata, whereas for a regular
file the iolock protects the data and the ilock protects the
metadata. From the VFS perspective, the i_mutex serialises all
accesses to the directory data, and so not holding the ilock for
readdir doesn't matter. The issue is that CXFS doesn't access
directory data via the VFS, so it has no "data serialisaton"
mechanism. Hence we need to hold the IOLOCK in the correct places to
provide this low level directory data access serialisation.
The ilock can then be used just when the extent list needs to be
read, just like we do for regular files. The directory modification
code can take the iolock exclusive when the ilock is also taken,
and this then ensures that readdir is correct excluded while
modifications are in progress.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Lockdep annotations are a maintenance nightmare. Locking has to be
modified to suit the limitations of the annotations, and we're
always having to fix the annotations because they are unable to
express the complexity of locking heirarchies correctly.
So, next up, we've got more issues with lockdep annotations for
inode locking w.r.t. XFS_LOCK_PARENT:
- lockdep classes are exclusive and can't be ORed together
to form new classes.
- IOLOCK needs multiple PARENT subclasses to express the
changes needed for the readdir locking rework needed to
stop the endless flow of lockdep false positives involving
readdir calling filldir under the ILOCK.
- there are only 8 unique lockdep subclasses available,
so we can't create a generic solution.
IOWs we need to treat the 3-bit space available to each lock type
differently:
- IOLOCK uses xfs_lock_two_inodes(), so needs:
- at least 2 IOLOCK subclasses
- at least 2 IOLOCK_PARENT subclasses
- MMAPLOCK uses xfs_lock_two_inodes(), so needs:
- at least 2 MMAPLOCK subclasses
- ILOCK uses xfs_lock_inodes with up to 5 inodes, so needs:
- at least 5 ILOCK subclasses
- one ILOCK_PARENT subclass
- one RTBITMAP subclass
- one RTSUM subclass
For the IOLOCK, split the space into two sets of subclasses.
For the MMAPLOCK, just use half the space for the one subclass to
match the non-parent lock classes of the IOLOCK.
For the ILOCK, use 0-4 as the ILOCK subclasses, 5-7 for the
remaining individual subclasses.
Because they are now all different, modify xfs_lock_inumorder() to
handle the nested subclasses, and to assert fail if passed an
invalid subclass. Further, annotate xfs_lock_inodes() to assert fail
if an invalid combination of lock primitives and inode counts are
passed that would result in a lockdep subclass annotation overflow.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
The node directory lookup code uses a state structure that tracks the
path of buffers used to search for the hash of a filename through the
leaf blocks. When the lookup encounters a block that ends with the
requested hash, but the entry has not yet been found, it must shift over
to the next block and continue looking for the entry (i.e., duplicate
hashes could continue over into the next block). This shift mechanism
involves walking back up and down the state structure, replacing buffers
at the appropriate btree levels as necessary.
When a buffer is replaced, the old buffer is released and the new buffer
read into the active slot in the path structure. Because the buffer is
read directly into the path slot, a buffer read failure can result in
setting a NULL buffer pointer in an active slot. This throws off the
state cleanup code in xfs_dir2_node_lookup(), which expects to release a
buffer from each active slot. Instead, a BUG occurs due to a NULL
pointer dereference:
BUG: unable to handle kernel NULL pointer dereference at 00000000000001e8
IP: [<ffffffffa0585063>] xfs_trans_brelse+0x2a3/0x3c0 [xfs]
...
RIP: 0010:[<ffffffffa0585063>] [<ffffffffa0585063>] xfs_trans_brelse+0x2a3/0x3c0 [xfs]
...
Call Trace:
[<ffffffffa05250c6>] xfs_dir2_node_lookup+0xa6/0x2c0 [xfs]
[<ffffffffa0519f7c>] xfs_dir_lookup+0x1ac/0x1c0 [xfs]
[<ffffffffa055d0e1>] xfs_lookup+0x91/0x290 [xfs]
[<ffffffffa05580b3>] xfs_vn_lookup+0x73/0xb0 [xfs]
[<ffffffff8122de8d>] lookup_real+0x1d/0x50
[<ffffffff8123330e>] path_openat+0x91e/0x1490
[<ffffffff81235079>] do_filp_open+0x89/0x100
...
This has been reproduced via a parallel fsstress and filesystem shutdown
workload in a loop. The shutdown triggers the read error in the
aforementioned codepath and causes the BUG in xfs_dir2_node_lookup().
Update xfs_da3_path_shift() to update the active path slot atomically
with respect to the caller when a buffer is replaced. This ensures that
the caller always sees the old or new buffer in the slot and prevents
the NULL pointer dereference.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
The sparse inodes feature is currently considered experimental. We warn
at mount time from xfs_mount_validate_sb(). This function is part of the
superblock verifier codepath, however, which means it could be invoked
repeatedly on superblock reads or writes. This is currently only
noticeable from userspace, where mkfs produces multiple warnings at
format time.
As mkfs warnings were not the intent of this change, relocate the mount
time warning to xfs_fs_fill_super(), which is only invoked once and only
in kernel space.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
