i_cnextents does not include delayed allocated extents, so switch
to the inode fork size check that we already use in other places
instead.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Streamline the conditionals so that it is more obvious which specific case
form the top of the function comments is being handled. Use gotos only
for early returns.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Switch to a single interface for flushing the log to a specific LSN, which
gives consistent trace point coverage and a less confusing interface.
The was only a single user of the previous xfs_log_force_lsn function,
which now also passes a NULL log_flushed argument.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Switch to a single interface for flushing the whole log, which gives
consistent trace point coverage, and removes the unused log_flushed
argument for the previous _xfs_log_force callers.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
The function now does something, and that something is central to our
inode logging scheme.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
The rmapbt perag metadata reservation reserves blocks for the
reverse mapping btree (rmapbt). Since the rmapbt uses blocks from
the agfl and perag accounting is updated as blocks are allocated
from the allocation btrees, the reservation actually accounts blocks
as they are allocated to (or freed from) the agfl rather than the
rmapbt itself.
While this works for blocks that are eventually used for the rmapbt,
not all agfl blocks are destined for the rmapbt. Blocks that are
allocated to the agfl (and thus "reserved" for the rmapbt) but then
used by another structure leads to a growing inconsistency over time
between the runtime tracking of rmapbt usage vs. actual rmapbt
usage. Since the runtime tracking thinks all agfl blocks are rmapbt
blocks, it essentially believes that less future reservation is
required to satisfy the rmapbt than what is actually necessary.
The inconsistency is rectified across mount cycles because the perag
reservation is initialized based on the actual rmapbt usage at mount
time. The problem, however, is that the excessive drain of the
reservation at runtime opens a window to allocate blocks for other
purposes that might be required for the rmapbt on a subsequent
mount. This problem can be demonstrated by a simple test that runs
an allocation workload to consume agfl blocks over time and then
observe the difference in the agfl reservation requirement across an
unmount/mount cycle:
mount ...: xfs_ag_resv_init: ... resv 3193 ask 3194 len 3194
...
... : xfs_ag_resv_alloc_extent: ... resv 2957 ask 3194 len 1
umount...: xfs_ag_resv_free: ... resv 2956 ask 3194 len 0
mount ...: xfs_ag_resv_init: ... resv 3052 ask 3194 len 3194
As the above tracepoints show, the reservation requirement reduces
from 3194 blocks to 2956 blocks as the workload runs. Without any
other changes in the filesystem, the same reservation requirement
jumps from 2956 to 3052 blocks over a umount/mount cycle.
To address this divergence, update the RMAPBT reservation to account
blocks used for the rmapbt only rather than all blocks filled into
the agfl. This patch makes several high-level changes toward that
end:
1.) Reintroduce an AGFL reservation type to serve as an accounting
no-op for blocks allocated to (or freed from) the AGFL.
2.) Invoke RMAPBT usage accounting from the actual rmapbt block
allocation path rather than the AGFL allocation path.
The first change is required because agfl blocks are considered free
blocks throughout their lifetime. The perag reservation subsystem is
invoked unconditionally by the allocation subsystem, so we need a
way to tell the perag subsystem (via the allocation subsystem) to
not make any accounting changes for blocks filled into the AGFL.
The second change causes the in-core RMAPBT reservation usage
accounting to remain consistent with the on-disk state at all times
and eliminates the risk of leaving the rmapbt reservation
underfilled.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
The AGFL perag reservation type accounts all allocations that feed
into (or are released from) the allocation group free list (agfl).
The purpose of the reservation is to support worst case conditions
for the reverse mapping btree (rmapbt). As such, the agfl
reservation usage accounting only considers rmapbt usage when the
in-core counters are initialized at mount time.
This implementation inconsistency leads to divergence of the in-core
and on-disk usage accounting over time. In preparation to resolve
this inconsistency and adjust the AGFL reservation into an rmapbt
specific reservation, rename the AGFL reservation type and
associated accounting fields to something more rmapbt-specific. Also
fix up a couple tracepoints that incorrectly use the AGFL
reservation type to pass the agfl state of the associated extent
where the raw reservation type is expected.
Note that this patch does not change perag reservation behavior.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
The extent swap mechanism requires a unique implementation for
rmapbt enabled filesystems. Because the rmapbt tracks extent owner
information, extent swap must individually unmap and remap each
extent between the two inodes.
The rmapbt extent swap transaction block reservation currently
accounts for the worst case bmapbt block and rmapbt block
consumption based on the extent count of each inode. There is a
corner case that exists due to the extent swap implementation that
is not covered by this reservation, however.
If one of the associated inodes is just over the max extent count
used for extent format inodes (i.e., the inode is in btree format by
a single extent), the unmap/remap cycle of the extent swap can
bounce the inode between extent and btree format multiple times,
almost as many times as there are extents in the inode (if the
opposing inode happens to have one less, for example). Each back and
forth cycle involves a block free and allocation, which isn't a
problem except for that the initial transaction reservation must
account for the total number of block allocations performed by the
chain of deferred operations. If not, a block reservation overrun
occurs and the filesystem shuts down.
Update the rmapbt extent swap block reservation to check for this
situation and add some block reservation slop to ensure the entire
operation succeeds. We'd never likely require reservation for both
inodes as fsr wouldn't defrag the file in that case, but the
additional reservation is constrained by the data fork size so be
cautious and check for both.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
The ->t_blk_res_used field tracks how many blocks have been used in
the current transaction. This should never exceed the block
reservation (->t_blk_res) for a particular transaction. We currently
assert this condition in the transaction block accounting code, but
otherwise take no additional action should this situation occur.
The overrun generally has no effect if space ends up being available
and the associated transaction commits. If the transaction is
duplicated, however, the current block usage is used to determine
the remaining block reservation to be transferred to the new
transaction. If usage exceeds reservation, this calculation
underflows and creates a transaction with an invalid and excessive
reservation. When the second transaction commits, the release of
unused blocks corrupts the in-core free space counters. With lazy
superblock accounting enabled, this inconsistency eventually
trickles to the on-disk superblock and corrupts the filesystem.
Replace the transaction block usage accounting assert with an
explicit overrun check. If the transaction overruns the reservation,
shutdown the filesystem immediately to prevent corruption. Add a new
assert to xfs_trans_dup() to catch any callers that might induce
this invalid state in the future.
Signed-off-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Noticed when looking at why cycling 600k inodes/s through the inode
cache was taking a total of 8% cpu in memset() during inode
initialisation. There is no need to zero the inode.i_data structure
twice.
This increases single threaded bulkstat throughput from ~200,000
inodes/s to ~220,000 inodes/s, so we save a substantial amount of
CPU time per inode init by doing this.
Signed-Off-By: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
The AGFL size calculation is about to get more complex, so lets turn
the macro into a function first and remove the macro.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
[darrick: forward port to newer kernel, simplify the helper]
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
There's no point in allocating a transaction and locking the inode in
preparation to clear cow blocks if there actually are any cow fork
extents. Therefore, move the xfs_reflink_cancel_cow_range hunk to
xfs_inactive and check the cow ifp first. This makes inode reclamation
run faster.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Yet another round of playing whack-a-mole with directory code that
asserts on corrupt on-disk metadata when it really should be returning
-EFSCORRUPTED instead of ASSERTing. Found by a xfs/391 crash while
lastbit fuzzing of ltail.bestcount.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
In xfs_qm_dqalloc, we join the locked quota inode to the transaction we
use to allocate blocks. If the allocation or mapping fails, we're not
allowed to unlock the inode because the transaction code is in charge of
unlocking it for us. Therefore, remove the iunlock call to avoid
blowing asserts about unbalanced locking + mount hang.
Found by corrupting the AGF and allocating space in the filesystem
(quotacheck) immediately after mount. The upcoming agfl wrapping fixup
test will trigger this scenario.
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Due to an inverted logic mistake in xfs_buftarg_isolate()
the xfs_buffers with zero b_lru_ref will take another trip
around LRU, while isolating buffers with non-zero b_lru_ref.
Additionally those isolated buffers end up right back on the LRU
once they are released, because b_lru_ref remains elevated.
Fix that circuitous route by leaving them on the LRU
as originally intended.
Signed-off-by: Vratislav Bendel <vbendel@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
xfs_trans_alloc() does GFP_KERNEL allocation, and we can call it
while holding pages locked for writeback in the ->writepages path.
The memory allocation is allowed to wait on pages under writeback,
and so can wait on pages that are tagged as writeback by the
caller.
This affects both pre-IO submission and post-IO submission paths.
Hence xfs_setsize_trans_alloc(), xfs_reflink_end_cow(),
xfs_iomap_write_unwritten() and xfs_reflink_cancel_cow_range().
xfs_iomap_write_unwritten() already does the right thing, but the
others don't. Fix them.
Signed-Off-By: Dave Chinner <dchinner@redhat.com>
Fixes: 281627df3e ("xfs: log file size updates at I/O completion time")
Fixes: 43caeb187d ("xfs: move mappings from cow fork to data fork after copy-write)"
Reviewed-by: Eric Sandeen <sandeen@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Use the VFS dirty inode tracking for lazytime inodes only, and just
log them in ->dirty_inode.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
__mark_inode_dirty already takes care of that, and for the XFS lazytime
implementation we need to know that ->dirty_inode was called because
I_DIRTY_TIME was set.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
The memcpy is guarded by a check which is performed a right before we
call xfs_log_dinode_to_disk. At this point we are sure this check will
always be false otherwise we would have errored out. So let's remove
this dead weight.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
When using large directory blocks, we regularly see memory
allocations of >64k being made for the shadow log vector buffer.
When we are under memory pressure, kmalloc() may not be able to find
contiguous memory chunks large enough to satisfy these allocations
easily, and if memory is fragmented we can potentially stall here.
TO avoid this problem, switch the log vector buffer allocation to
use kmem_alloc_large(). This will allow failed allocations to fall
back to vmalloc and so remove the dependency on large contiguous
regions of memory being available. This should prevent slowdowns
and potential stalls when memory is low and/or fragmented.
Signed-Off-By: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
Pull x86/pti updates from Thomas Gleixner:
"Yet another pile of melted spectrum related updates:
- Drop native vsyscall support finally as it causes more trouble than
benefit.
- Make microcode loading more robust. There were a few issues
especially related to late loading which are now surfacing because
late loading of the IB* microcodes addressing spectre issues has
become more widely used.
- Simplify and robustify the syscall handling in the entry code
- Prevent kprobes on the entry trampoline code which lead to kernel
crashes when the probe hits before CR3 is updated
- Don't check microcode versions when running on hypervisors as they
are considered as lying anyway.
- Fix the 32bit objtool build and a coment typo"
* 'x86-pti-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/kprobes: Fix kernel crash when probing .entry_trampoline code
x86/pti: Fix a comment typo
x86/microcode: Synchronize late microcode loading
x86/microcode: Request microcode on the BSP
x86/microcode/intel: Look into the patch cache first
x86/microcode: Do not upload microcode if CPUs are offline
x86/microcode/intel: Writeback and invalidate caches before updating microcode
x86/microcode/intel: Check microcode revision before updating sibling threads
x86/microcode: Get rid of struct apply_microcode_ctx
x86/spectre_v2: Don't check microcode versions when running under hypervisors
x86/vsyscall/64: Drop "native" vsyscalls
x86/entry/64/compat: Save one instruction in entry_INT80_compat()
x86/entry: Do not special-case clone(2) in compat entry
x86/syscalls: Use COMPAT_SYSCALL_DEFINEx() macros for x86-only compat syscalls
x86/syscalls: Use proper syscall definition for sys_ioperm()
x86/entry: Remove stale syscall prototype
x86/syscalls/32: Simplify $entry == $compat entries
objtool: Fix 32-bit build
Pull timer fix from Thomas Gleixner:
"Just a single fix which adds a missing Kconfig dependency to avoid
unmet dependency warnings"
* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
clocksource/atmel-st: Add 'depends on HAS_IOMEM' to fix unmet dependency
Pull RAS fixes from Thomas Gleixner:
"Two small fixes for RAS/MCE:
- Serialize sysfs changes to avoid concurrent modificaiton of
underlying data
- Add microcode revision to Machine Check records. This should have
been there forever, but now with the broken microcode versions in
the wild it has become important"
* 'ras-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/MCE: Serialize sysfs changes
x86/MCE: Save microcode revision in machine check records
Pull perf updates from Thomas Gleixner:
"Another set of perf updates:
- Fix a Skylake Uncore event format declaration
- Prevent perf pipe mode from crahsing which was caused by a missing
buffer allocation
- Make the perf top popup message which tells the user that it uses
fallback mode on older kernels a debug message.
- Make perf context rescheduling work correcctly
- Robustify the jump error drawing in perf browser mode so it does
not try to create references to NULL initialized offset entries
- Make trigger_on() robust so it does not enable the trigger before
everything is set up correctly to handle it
- Make perf auxtrace respect the --no-itrace option so it does not
try to queue AUX data for decoding.
- Prevent having different number of field separators in CVS output
lines when a counter is not supported.
- Make the perf kallsyms man page usage behave like it does for all
other perf commands.
- Synchronize the kernel headers"
* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/core: Fix ctx_event_type in ctx_resched()
perf tools: Fix trigger class trigger_on()
perf auxtrace: Prevent decoding when --no-itrace
perf stat: Fix CVS output format for non-supported counters
tools headers: Sync x86's cpufeatures.h
tools headers: Sync copy of kvm UAPI headers
perf record: Fix crash in pipe mode
perf annotate browser: Be more robust when drawing jump arrows
perf top: Fix annoying fallback message on older kernels
perf kallsyms: Fix the usage on the man page
perf/x86/intel/uncore: Fix Skylake UPI event format
Pull locking fix from Thomas Gleixner:
"rt_mutex_futex_unlock() grew a new irq-off call site, but the function
assumes that its always called from irq enabled context.
Use (un)lock_irqsafe() to handle the new call site correctly"
* 'locking-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
rtmutex: Make rt_mutex_futex_unlock() safe for irq-off callsites
Pull dmaengine fixes from Vinod Koul:
"Two small fixes are for this cycle:
- fix max_chunk_size for rcar-dmac for R-Car Gen3
- fix clock resource of mv_xor_v2"
* tag 'dmaengine-fix-4.16-rc5' of git://git.infradead.org/users/vkoul/slave-dma:
dmaengine: mv_xor_v2: Fix clock resource by adding a register clock
dmaengine: rcar-dmac: fix max_chunk_size for R-Car Gen3
Pull GPIO fix from Linus Walleij:
"This is a single GPIO fix for the v4.16 series affecting the Renesas
driver, and fixes wakeup from external stuff"
* tag 'gpio-v4.16-3' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio:
gpio: rcar: Use wakeup_path i.s.o. explicit clock handling
On the CP110 components which are present on the Armada 7K/8K SoC we need
to explicitly enable the clock for the registers. However it is not
needed for the AP8xx component, that's why this clock is optional.
With this patch both clock have now a name, but in order to be backward
compatible, the name of the first clock is not used. It allows to still
use this clock with a device tree using the old binding.
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Pull Kbuild fixes from Masahiro Yamada:
- make fixdep parse kconfig.h to fix missing rebuild
- replace hyphens with underscores in builtin DTB label names
- fix typos
* tag 'kbuild-fixes-v4.16-2' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kbuild: Handle builtin dtb file names containing hyphens
scripts/bloat-o-meter: fix typos in help
fixdep: do not ignore kconfig.h
fixdep: remove some false CONFIG_ matches
fixdep: remove stale references to uml-config.h
Pull block fixes from Jens Axboe:
- a xen-blkfront fix from Bhavesh with a multiqueue fix when
detaching/re-attaching
- a few important NVMe fixes, including a revert for a sysfs fix that
caused some user space confusion
- two bcache fixes by way of Michael Lyle
- a loop regression fix, fixing an issue with lost writes on DAX.
* tag 'for-linus-20180309' of git://git.kernel.dk/linux-block:
loop: Fix lost writes caused by missing flag
nvme_fc: rework sqsize handling
nvme-fabrics: Ignore nr_io_queues option for discovery controllers
xen-blkfront: move negotiate_mq to cover all cases of new VBDs
Revert "nvme: create 'slaves' and 'holders' entries for hidden controllers"
bcache: don't attach backing with duplicate UUID
bcache: fix crashes in duplicate cache device register
nvme: pci: pass max vectors as num_possible_cpus() to pci_alloc_irq_vectors
nvme-pci: Fix EEH failure on ppc
Pull device mapper fixes from Mike Snitzer:
- Fix an uninitialized variable false warning in dm bufio
- Fix DM's passthrough ioctl support to be race free against an
underlying device being removed.
- Fix corner-case of DM raid resync reporting if/when the raid becomes
degraded during resync; otherwise automated raid repair will fail.
- A few DM multipath fixes to make non-SCSI optimizations, that were
introduced during the 4.16 merge, useful for all non-SCSI devices,
rather than narrowly define this non-SCSI mode in terms of "nvme".
This allows the removal of "queue_mode nvme" that really didn't need
to be introduced. Instead DM core will internalize whether
nvme-specific IO submission optimizations are doable and DM multipath
will only do SCSI-specific device handler operations if SCSI is in
use.
* tag 'for-4.16/dm-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
dm table: allow upgrade from bio-based to specialized bio-based variant
dm mpath: remove unnecessary NVMe branching in favor of scsi_dh checks
dm table: fix "nvme" test
dm raid: fix incorrect sync_ratio when degraded
dm: use blkdev_get rather than bdgrab when issuing pass-through ioctl
dm bufio: avoid false-positive Wmaybe-uninitialized warning
Pull rdma fixes from Doug Ledford:
- Various driver bug fixes in mlx5, mlx4, bnxt_re and qedr, ranging
from bugs under load to bad error case handling
- There in one largish patch fixing the locking in bnxt_re to avoid a
machine hard lock situation
- A few core bugs on error paths
- A patch to reduce stack usage in the new CQ API
- One mlx5 regression introduced in this merge window
- There were new syzkaller scripts written for the RDMA subsystem and
we are fixing issues found by the bot
- One of the commits (aa0de36a40 “RDMA/mlx5: Fix integer overflow
while resizing CQ”) is missing part of the commit log message and one
of the SOB lines. The original patch was from Leon Romanovsky, and a
cut-n-paste separator in the commit message confused patchworks which
then put the end of message separator in the wrong place in the
downloaded patch, and I didn’t notice in time. The patch made it into
the official branch, and the only way to fix it in-place was to
rebase. Given the pain that a rebase causes, and the fact that the
patch has relevant tags for stable and syzkaller, a revert of the
munged patch and a reapplication of the original patch with the log
message intact was done.
* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (25 commits)
RDMA/mlx5: Fix integer overflow while resizing CQ
Revert "RDMA/mlx5: Fix integer overflow while resizing CQ"
RDMA/ucma: Check that user doesn't overflow QP state
RDMA/mlx5: Fix integer overflow while resizing CQ
RDMA/ucma: Limit possible option size
IB/core: Fix possible crash to access NULL netdev
RDMA/bnxt_re: Avoid Hard lockup during error CQE processing
RDMA/core: Reduce poll batch for direct cq polling
IB/mlx5: Fix an error code in __mlx5_ib_modify_qp()
IB/mlx5: When not in dual port RoCE mode, use provided port as native
IB/mlx4: Include GID type when deleting GIDs from HW table under RoCE
IB/mlx4: Fix corruption of RoCEv2 IPv4 GIDs
RDMA/qedr: Fix iWARP write and send with immediate
RDMA/qedr: Fix kernel panic when running fio over NFSoRDMA
RDMA/qedr: Fix iWARP connect with port mapper
RDMA/qedr: Fix ipv6 destination address resolution
IB/core : Add null pointer check in addr_resolve
RDMA/bnxt_re: Fix the ib_reg failure cleanup
RDMA/bnxt_re: Fix incorrect DB offset calculation
RDMA/bnxt_re: Unconditionly fence non wire memory operations
...
Pull x86 platform driver fixes from Darren Hart:
"Correct a module loading race condition between the DELL_SMBIOS
backend modules and the first user by converting them to bool features
of the DELL_SMBIOS driver. Fixup the resulting Kconfig dependency
issue with DCDBAS"
* tag 'platform-drivers-x86-v4.16-6' of git://git.infradead.org/linux-platform-drivers-x86:
platform/x86: dell-smbios: Resolve dependency error on DCDBAS
platform/x86: Allow for SMBIOS backend defaults
platform/x86: dell-smbios: Link all dell-smbios-* modules together
platform/x86: dell-smbios: Rename dell-smbios source to dell-smbios-base
platform/x86: dell-smbios: Correct some style warnings
Pull KVM fixes from Radim Krčmář:
"PPC:
- Fix guest time accounting in the host
- Fix large-page backing for radix guests on POWER9
- Fix HPT guests on POWER9 backed by 2M or 1G pages
- Compile fixes for some configs and gcc versions
s390:
- Fix random memory corruption when running as guest2 (e.g. KVM in
LPAR) and starting guest3 (e.g. nested KVM) with many CPUs
- Export forgotten io interrupt delivery statistics counter"
* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
KVM: s390: fix memory overwrites when not using SCA entries
KVM: PPC: Book3S HV: Fix guest time accounting with VIRT_CPU_ACCOUNTING_GEN
KVM: PPC: Book3S HV: Fix VRMA initialization with 2MB or 1GB memory backing
KVM: PPC: Book3S HV: Fix handling of large pages in radix page fault handler
KVM: s390: provide io interrupt kvm_stat
KVM: PPC: Book3S: Fix compile error that occurs with some gcc versions
KVM: PPC: Fix compile error that occurs when CONFIG_ALTIVEC=n
Pull xen fix from Juergen Gross:
"Just one fix for the correct error handling after a failed
device_register()"
* tag 'for-linus-4.16a-rc5-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
xen: xenbus: use put_device() instead of kfree()
Pull arm64 fixes from Catalin Marinas:
- The SMCCC firmware interface for the spectre variant 2 mitigation has
been updated to allow the discovery of whether the CPU needs the
workaround. This pull request relaxes the kernel check on the return
value from firmware.
- Fix the commit allowing changing from global to non-global page table
entries which inadvertently disallowed other safe attribute changes.
- Fix sleeping in atomic during the arm_perf_teardown_cpu() code.
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64: Relax ARM_SMCCC_ARCH_WORKAROUND_1 discovery
arm_pmu: Use disable_irq_nosync when disabling SPI in CPU teardown hook
arm64: mm: fix thinko in non-global page table attribute check
Pull Documentation build fix from Jonathan Corbet:
"The Sphinx 1.7 release broke the build process for reasons that are
mostly our fault.
This is a single fix cherry-picked from docs-next that restores docs
buildability for all supported Sphinx versions"
* tag 'docs-4.16-fix' of git://git.lwn.net/linux:
Documentation/sphinx: Fix Directive import error
Merge misc fixes from Andrew Morton:
"8 fixes"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
lib/test_kmod.c: fix limit check on number of test devices created
selftests/vm/run_vmtests: adjust hugetlb size according to nr_cpus
mm/page_alloc: fix memmap_init_zone pageblock alignment
mm/memblock.c: hardcode the end_pfn being -1
mm/gup.c: teach get_user_pages_unlocked to handle FOLL_NOWAIT
lib/bug.c: exclude non-BUG/WARN exceptions from report_bug()
bug: use %pB in BUG and stack protector failure
hugetlb: fix surplus pages accounting
As reported by Dan the parentheses is in the wrong place, and since
unlikely() call returns either 0 or 1 it's never less than zero. The
second issue is that signed integer overflows like "INT_MAX + 1" are
undefined behavior.
Since num_test_devs represents the number of devices, we want to stop
prior to hitting the max, and not rely on the wrap arround at all. So
just cap at num_test_devs + 1, prior to assigning a new device.
Link: http://lkml.kernel.org/r/20180224030046.24238-1-mcgrof@kernel.org
Fixes: d9c6a72d6f ("kmod: add test driver to stress test the module loader")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Luis R. Rodriguez <mcgrof@kernel.org>
Acked-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit b92df1de5d ("mm: page_alloc: skip over regions of invalid pfns
where possible") introduced a bug where move_freepages() triggers a
VM_BUG_ON() on uninitialized page structure due to pageblock alignment.
To fix this, simply align the skipped pfns in memmap_init_zone() the
same way as in move_freepages_block().
Seen in one of the RHEL reports:
crash> log | grep -e BUG -e RIP -e Call.Trace -e move_freepages_block -e rmqueue -e freelist -A1
kernel BUG at mm/page_alloc.c:1389!
invalid opcode: 0000 [#1] SMP
--
RIP: 0010:[<ffffffff8118833e>] [<ffffffff8118833e>] move_freepages+0x15e/0x160
RSP: 0018:ffff88054d727688 EFLAGS: 00010087
--
Call Trace:
[<ffffffff811883b3>] move_freepages_block+0x73/0x80
[<ffffffff81189e63>] __rmqueue+0x263/0x460
[<ffffffff8118c781>] get_page_from_freelist+0x7e1/0x9e0
[<ffffffff8118caf6>] __alloc_pages_nodemask+0x176/0x420
--
RIP [<ffffffff8118833e>] move_freepages+0x15e/0x160
RSP <ffff88054d727688>
crash> page_init_bug -v | grep RAM
<struct resource 0xffff88067fffd2f8> 1000 - 9bfff System RAM (620.00 KiB)
<struct resource 0xffff88067fffd3a0> 100000 - 430bffff System RAM ( 1.05 GiB = 1071.75 MiB = 1097472.00 KiB)
<struct resource 0xffff88067fffd410> 4b0c8000 - 4bf9cfff System RAM ( 14.83 MiB = 15188.00 KiB)
<struct resource 0xffff88067fffd480> 4bfac000 - 646b1fff System RAM (391.02 MiB = 400408.00 KiB)
<struct resource 0xffff88067fffd560> 7b788000 - 7b7fffff System RAM (480.00 KiB)
<struct resource 0xffff88067fffd640> 100000000 - 67fffffff System RAM ( 22.00 GiB)
crash> page_init_bug | head -6
<struct resource 0xffff88067fffd560> 7b788000 - 7b7fffff System RAM (480.00 KiB)
<struct page 0xffffea0001ede200> 1fffff00000000 0 <struct pglist_data 0xffff88047ffd9000> 1 <struct zone 0xffff88047ffd9800> DMA32 4096 1048575
<struct page 0xffffea0001ede200> 505736 505344 <struct page 0xffffea0001ed8000> 505855 <struct page 0xffffea0001edffc0>
<struct page 0xffffea0001ed8000> 0 0 <struct pglist_data 0xffff88047ffd9000> 0 <struct zone 0xffff88047ffd9000> DMA 1 4095
<struct page 0xffffea0001edffc0> 1fffff00000400 0 <struct pglist_data 0xffff88047ffd9000> 1 <struct zone 0xffff88047ffd9800> DMA32 4096 1048575
BUG, zones differ!
Note that this range follows two not populated sections
68000000-77ffffff in this zone. 7b788000-7b7fffff is the first one
after a gap. This makes memmap_init_zone() skip all the pfns up to the
beginning of this range. But this range is not pageblock (2M) aligned.
In fact no range has to be.
crash> kmem -p 77fff000 78000000 7b5ff000 7b600000 7b787000 7b788000
PAGE PHYSICAL MAPPING INDEX CNT FLAGS
ffffea0001e00000 78000000 0 0 0 0
ffffea0001ed7fc0 7b5ff000 0 0 0 0
ffffea0001ed8000 7b600000 0 0 0 0 <<<<
ffffea0001ede1c0 7b787000 0 0 0 0
ffffea0001ede200 7b788000 0 0 1 1fffff00000000
Top part of page flags should contain nodeid and zonenr, which is not
the case for page ffffea0001ed8000 here (<<<<).
crash> log | grep -o fffea0001ed[^\ ]* | sort -u
fffea0001ed8000
fffea0001eded20
fffea0001edffc0
crash> bt -r | grep -o fffea0001ed[^\ ]* | sort -u
fffea0001ed8000
fffea0001eded00
fffea0001eded20
fffea0001edffc0
Initialization of the whole beginning of the section is skipped up to
the start of the range due to the commit b92df1de5d. Now any code
calling move_freepages_block() (like reusing the page from a freelist as
in this example) with a page from the beginning of the range will get
the page rounded down to start_page ffffea0001ed8000 and passed to
move_freepages() which crashes on assertion getting wrong zonenr.
> VM_BUG_ON(page_zone(start_page) != page_zone(end_page));
Note, page_zone() derives the zone from page flags here.
From similar machine before commit b92df1de5d:
crash> kmem -p 77fff000 78000000 7b5ff000 7b600000 7b7fe000 7b7ff000
PAGE PHYSICAL MAPPING INDEX CNT FLAGS
fffff73941e00000 78000000 0 0 1 1fffff00000000
fffff73941ed7fc0 7b5ff000 0 0 1 1fffff00000000
fffff73941ed8000 7b600000 0 0 1 1fffff00000000
fffff73941edff80 7b7fe000 0 0 1 1fffff00000000
fffff73941edffc0 7b7ff000 ffff8e67e04d3ae0 ad84 1 1fffff00020068 uptodate,lru,active,mappedtodisk
All the pages since the beginning of the section are initialized.
move_freepages()' not gonna blow up.
The same machine with this fix applied:
crash> kmem -p 77fff000 78000000 7b5ff000 7b600000 7b7fe000 7b7ff000
PAGE PHYSICAL MAPPING INDEX CNT FLAGS
ffffea0001e00000 78000000 0 0 0 0
ffffea0001e00000 7b5ff000 0 0 0 0
ffffea0001ed8000 7b600000 0 0 1 1fffff00000000
ffffea0001edff80 7b7fe000 0 0 1 1fffff00000000
ffffea0001edffc0 7b7ff000 ffff88017fb13720 8 2 1fffff00020068 uptodate,lru,active,mappedtodisk
At least the bare minimum of pages is initialized preventing the crash
as well.
Customers started to report this as soon as 7.4 (where b92df1de5d was
merged in RHEL) was released. I remember reports from
September/October-ish times. It's not easily reproduced and happens on
a handful of machines only. I guess that's why. But that does not make
it less serious, I think.
Though there actually is a report here:
https://bugzilla.kernel.org/show_bug.cgi?id=196443
And there are reports for Fedora from July:
https://bugzilla.redhat.com/show_bug.cgi?id=1473242
and CentOS:
https://bugs.centos.org/view.php?id=13964
and we internally track several dozens reports for RHEL bug
https://bugzilla.redhat.com/show_bug.cgi?id=1525121
Link: http://lkml.kernel.org/r/0485727b2e82da7efbce5f6ba42524b429d0391a.1520011945.git.neelx@redhat.com
Fixes: b92df1de5d ("mm: page_alloc: skip over regions of invalid pfns where possible")
Signed-off-by: Daniel Vacek <neelx@redhat.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Paul Burton <paul.burton@imgtec.com>
Cc: Pavel Tatashin <pasha.tatashin@oracle.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
KVM is hanging during postcopy live migration with userfaultfd because
get_user_pages_unlocked is not capable to handle FOLL_NOWAIT.
Earlier FOLL_NOWAIT was only ever passed to get_user_pages.
Specifically faultin_page (the callee of get_user_pages_unlocked caller)
doesn't know that if FAULT_FLAG_RETRY_NOWAIT was set in the page fault
flags, when VM_FAULT_RETRY is returned, the mmap_sem wasn't actually
released (even if nonblocking is not NULL). So it sets *nonblocking to
zero and the caller won't release the mmap_sem thinking it was already
released, but it wasn't because of FOLL_NOWAIT.
Link: http://lkml.kernel.org/r/20180302174343.5421-2-aarcange@redhat.com
Fixes: ce53053ce3 ("kvm: switch get_user_page_nowait() to get_user_pages_unlocked()")
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reported-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Tested-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
