Newer GCC warns about possible truncations of two generated path names as
we're concatenating the configurable sysfs and debugfs path prefixes
with a filename and placing the results in buffers of the same size as
the maximum length of the prefixes.
Newer GCC warns about a possible truncation of a generated sysfs path
name as we're concatenating a directory path with a file name and
placing the result in a buffer that is half the size of the maximum
length of the directory path (which is user controlled).
loopback_test.c: In function 'open_poll_files':
loopback_test.c:651:31: warning: '%s' directive output may be truncated writing up to 511 bytes into a region of size 255 [-Wformat-truncation=]
651 | snprintf(buf, sizeof(buf), "%s%s", dev->sysfs_entry, "iteration_count");
| ^~
loopback_test.c:651:3: note: 'snprintf' output between 16 and 527 bytes into a destination of size 255
651 | snprintf(buf, sizeof(buf), "%s%s", dev->sysfs_entry, "iteration_count");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Fix this by making sure the buffer is large enough the concatenated
strings.
Fixes: 6b0658f68786 ("greybus: tools: Add tools directory to greybus repo and add loopback") Fixes: 9250c0ee2626 ("greybus: Loopback_test: use poll instead of inotify") Signed-off-by: Johan Hovold <johan@kernel.org> Link: https://lore.kernel.org/r/20200312110151.22028-3-johan@kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On a system configured to trigger a crash_kexec() reboot, when only one CPU
is online and another CPU panics while starting-up, crash_smp_send_stop()
will fail to send any STOP message to the other already online core,
resulting in fail to freeze and registers not properly saved.
Moreover even if the proper messages are sent (case CPUs > 2)
it will similarly fail to account for the booting CPU when executing
the final stop wait-loop, so potentially resulting in some CPU not
been waited for shutdown before rebooting.
A tangible effect of this behaviour can be observed when, after a panic
with kexec enabled and loaded, on the following reboot triggered by kexec,
the cpu that could not be successfully stopped fails to come back online:
Make crash_smp_send_stop() account also for the online status of the
calling CPU while evaluating how many CPUs are effectively online: this way
the right number of STOPs is sent and all other stopped-cores's registers
are properly saved.
On a system with only one CPU online, when another one CPU panics while
starting-up, smp_send_stop() will fail to send any STOP message to the
other already online core, resulting in a system still responsive and
alive at the end of the panic procedure.
Make smp_send_stop() account also for the online status of the calling CPU
while evaluating how many CPUs are effectively online: this way, the right
number of STOPs is sent, so enforcing a proper freeze of the system at the
end of panic even under the above conditions.
Fixes: 08e875c16a16c ("arm64: SMP support") Reported-by: Dave Martin <Dave.Martin@arm.com> Acked-by: Mark Rutland <mark.rutland@arm.com> Signed-off-by: Cristian Marussi <cristian.marussi@arm.com> Signed-off-by: Will Deacon <will@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The recent futex inode life time fix changed the ordering of the futex key
union struct members, but forgot to adjust the hash function accordingly,
As a result the hashing omits the leading 64bit and even hashes beyond the
futex key causing a bad hash distribution which led to a ~100% performance
regression.
Hand in the futex key pointer instead of a random struct member and make
the size calculation based of the struct offset.
Fixes: 8019ad13ef7f ("futex: Fix inode life-time issue") Reported-by: Rong Chen <rong.a.chen@intel.com> Decoded-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Tested-by: Rong Chen <rong.a.chen@intel.com> Link: https://lkml.kernel.org/r/87h7yy90ve.fsf@nanos.tec.linutronix.de Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
As reported by Jann, ihold() does not in fact guarantee inode
persistence. And instead of making it so, replace the usage of inode
pointers with a per boot, machine wide, unique inode identifier.
This sequence number is global, but shared (file backed) futexes are
rare enough that this should not become a performance issue.
Clang's -Wpointer-to-int-cast deviates from GCC in that it warns when
casting to enums. The kernel does this in certain places, such as device
tree matches to set the version of the device being used, which allows
the kernel to avoid using a gigantic union.
To avoid a ton of false positive warnings, disable this particular part
of the warning, which has been split off into a separate diagnostic so
that the entire warning does not need to be turned off for clang. It
will be visible under W=1 in case people want to go about fixing these
easily and enabling the warning treewide.
Vishay has published a new version of "Designing the VCNL4200 Into an
Application" application note in October 2019. The new version specifies
that there is +-20% of part to part tolerance. This explains the drift
seen during experiments. The proximity pulse width is also changed from
32us to 30us. According to the support, the tolerance also applies to
ambient light.
So update the sampling periods. As the reading is blocking, current
users may notice slightly longer response time.
Fixes: be38866fbb97 ("iio: vcnl4000: add support for VCNL4200") Reviewed-by: Guido Günther <agx@sigxcpu.org> Signed-off-by: Tomas Novotny <tomas@novotny.cz> Cc: <Stable@vger.kernel.org> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
By default, tty_port_init() initializes those parameters to a multiple
of HZ. For instance in line 69 of tty_port.c:
port->close_delay = (50 * HZ) / 100;
https://github.com/torvalds/linux/blob/master/drivers/tty/tty_port.c#L69
With e.g. CONFIG_HZ = 250 (as this is the case for Ubuntu 18.04
linux-image-4.15.0-37-generic), the default setting for close_delay is
thus 125.
When ioctl(fd, TIOCGSERIAL, &s) is executed, the setting returned in
user space is '12' (125/10). When ioctl(fd, TIOCSSERIAL, &s) is then
executed with the same setting '12', the value is interpreted as '120'
which is different from the current setting and a EPERM error may be
raised by set_serial_info() if !CAP_SYS_ADMIN.
https://github.com/torvalds/linux/blob/master/drivers/usb/class/cdc-acm.c#L919
close_delay and closing_wait are specified in hundredth of a second but stored
internally in jiffies. Use the jiffies_to_msecs() and msecs_to_jiffies()
functions to convert from each other.
Commit 3f8fd02b1bf1 ("mm/vmalloc: Sync unmappings in
__purge_vmap_area_lazy()") introduced a call to vmalloc_sync_all() in
the vunmap() code-path. While this change was necessary to maintain
correctness on x86-32-pae kernels, it also adds additional cycles for
architectures that don't need it.
Specifically on x86-64 with CONFIG_VMAP_STACK=y some people reported
severe performance regressions in micro-benchmarks because it now also
calls the x86-64 implementation of vmalloc_sync_all() on vunmap(). But
the vmalloc_sync_all() implementation on x86-64 is only needed for newly
created mappings.
To avoid the unnecessary work on x86-64 and to gain the performance
back, split up vmalloc_sync_all() into two functions:
* vmalloc_sync_mappings(), and
* vmalloc_sync_unmappings()
Most call-sites to vmalloc_sync_all() only care about new mappings being
synchronized. The only exception is the new call-site added in the
above mentioned commit.
Shile Zhang directed us to a report of an 80% regression in reaim
throughput.
Fixes: 3f8fd02b1bf1 ("mm/vmalloc: Sync unmappings in __purge_vmap_area_lazy()") Reported-by: kernel test robot <oliver.sang@intel.com> Reported-by: Shile Zhang <shile.zhang@linux.alibaba.com> Signed-off-by: Joerg Roedel <jroedel@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Tested-by: Borislav Petkov <bp@suse.de> Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> [GHES] Cc: Dave Hansen <dave.hansen@linux.intel.com> Cc: Andy Lutomirski <luto@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@redhat.com> Cc: <stable@vger.kernel.org> Link: http://lkml.kernel.org/r/20191009124418.8286-1-joro@8bytes.org Link: https://lists.01.org/hyperkitty/list/lkp@lists.01.org/thread/4D3JPPHBNOSPFK2KEPC6KGKS6J25AIDB/ Link: http://lkml.kernel.org/r/20191113095530.228959-1-shile.zhang@linux.alibaba.com Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit bd4c82c22c36 ("mm, THP, swap: delay splitting THP after swapped
out") supported writing THP to a swap device but forgot to upgrade an
older commit df8c94d13c7e ("page-flags: define behavior of FS/IO-related
flags on compound pages") which could trigger a crash during THP
swapping out with DEBUG_VM_PGFLAGS=y,
This only happens with a mmotm patch "mm/memcontrol.c: allocate
shrinker_map on appropriate NUMA node" [2] which effectively calls
kmalloc_node for each possible node. SLUB however only allocates
kmem_cache_node on online N_NORMAL_MEMORY nodes, and relies on
node_to_mem_node to return such valid node for other nodes since commit a561ce00b09e ("slub: fall back to node_to_mem_node() node if allocating
on memoryless node"). This is however not true in this configuration
where the _node_numa_mem_ array is not initialized for nodes 0 and 2-31,
thus it contains zeroes and get_partial() ends up accessing
non-allocated kmem_cache_node.
A related issue was reported by Bharata (originally by Ramachandran) [3]
where a similar PowerPC configuration, but with mainline kernel without
patch [2] ends up allocating large amounts of pages by kmalloc-1k
kmalloc-512. This seems to have the same underlying issue with
node_to_mem_node() not behaving as expected, and might probably also
lead to an infinite loop with CONFIG_SLUB_CPU_PARTIAL [4].
This patch should fix both issues by not relying on node_to_mem_node()
anymore and instead simply falling back to NUMA_NO_NODE, when
kmalloc_node(node) is attempted for a node that's not online, or has no
usable memory. The "usable memory" condition is also changed from
node_present_pages() to N_NORMAL_MEMORY node state, as that is exactly
the condition that SLUB uses to allocate kmem_cache_node structures.
The check in get_partial() is removed completely, as the checks in
___slab_alloc() are now sufficient to prevent get_partial() being
reached with an invalid node.
This is just a cleanup addition to Jann's fix to properly update the
transaction ID for the slub slowpath in commit fd4d9c7d0c71 ("mm: slub:
add missing TID bump..").
The transaction ID is what protects us against any concurrent accesses,
but we should really also make sure to make the 'freelist' comparison
itself always use the same freelist value that we then used as the new
next free pointer.
Jann points out that if we do all of this carefully, we could skip the
transaction ID update for all the paths that only remove entries from
the lists, and only update the TID when adding entries (to avoid the ABA
issue with cmpxchg and list handling re-adding a previously seen value).
But this patch just does the "make sure to cmpxchg the same value we
used" rather than then try to be clever.
An eventfd monitors multiple memory thresholds of the cgroup, closes them,
the kernel deletes all events related to this eventfd. Before all events
are deleted, another eventfd monitors the memory threshold of this cgroup,
leading to a crash:
We can reproduce this problem in the following ways:
1. We create a new cgroup subdirectory and a new eventfd, and then we
monitor multiple memory thresholds of the cgroup through this eventfd.
2. closing this eventfd, and __mem_cgroup_usage_unregister_event ()
will be called multiple times to delete all events related to this
eventfd.
The first time __mem_cgroup_usage_unregister_event() is called, the
kernel will clear all items related to this eventfd in thresholds->
primary.
Since there is currently only one eventfd, thresholds-> primary becomes
empty, so the kernel will set thresholds-> primary and hresholds-> spare
to NULL. If at this time, the user creates a new eventfd and monitor
the memory threshold of this cgroup, kernel will re-initialize
thresholds-> primary.
Then when __mem_cgroup_usage_unregister_event () is called for the
second time, because thresholds-> primary is not empty, the system will
access thresholds-> spare, but thresholds-> spare is NULL, which will
trigger a crash.
In general, the longer it takes to delete all events related to this
eventfd, the easier it is to trigger this problem.
The solution is to check whether the thresholds associated with the
eventfd has been cleared when deleting the event. If so, we do nothing.
[akpm@linux-foundation.org: fix comment, per Kirill] Fixes: 907860ed381a ("cgroups: make cftype.unregister_event() void-returning") Signed-off-by: Chunguang Xu <brookxu@tencent.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Acked-by: Michal Hocko <mhocko@suse.com> Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Vladimir Davydov <vdavydov.dev@gmail.com> Cc: <stable@vger.kernel.org> Link: http://lkml.kernel.org/r/077a6f67-aefa-4591-efec-f2f3af2b0b02@gmail.com Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The offset into the array was specified in bytes but should
be in terms of 32-bit words. Also prevent large reads that
would also cause a buffer overread.
v2: Read from correct offset from internal storage buffer.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com> Acked-by: Christian König <christian.koenig@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Cc: stable@vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
During a rename whiteout, if btrfs_whiteout_for_rename() returns an error
we can end up returning from btrfs_rename() with the log context object
still in the root's log context list - this happens if 'sync_log' was
set to true before we called btrfs_whiteout_for_rename() and it is
dangerous because we end up with a corrupt linked list (root->log_ctxs)
as the log context object was allocated on the stack.
After btrfs_rename() returns, any task that is running btrfs_sync_log()
concurrently can end up crashing because that linked list is traversed by
btrfs_sync_log() (through btrfs_remove_all_log_ctxs()). That results in
the same issue that commit e6c617102c7e4 ("Btrfs: fix log context list
corruption after rename exchange operation") fixed.
Fixes: d4682ba03ef618 ("Btrfs: sync log after logging new name") CC: stable@vger.kernel.org # 4.19+ Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: David Sterba <dsterba@suse.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
libtraceevent (used by perf and trace-cmd) failed to parse the
xhci_urb_dequeue trace event. This is because the user space trace
event format parsing is not a full C compiler. It can handle some basic
logic, but is not meant to be able to handle everything C can do.
In cases where a trace event field needs to be converted from a number
to a string, there's the __print_symbolic() macro that should be used:
See samples/trace_events/trace-events-sample.h
Some xhci trace events open coded the __print_symbolic() causing the
user spaces tools to fail to parse it. This has to be replaced with
__print_symbolic() instead.
armv7a-hardfloat-linux-gnueabi-ld: drivers/rtc/rtc-max8907.o: in function `max8907_rtc_probe':
rtc-max8907.c:(.text+0x400): undefined reference to `regmap_irq_get_virq'
get_char was erroneously given the address of the pointer to the text
instead of the address of the text, thus leading to random crashes when
the user requests speaking a word while the current position is on a space
character and say_word_ctl is not enabled.
A scripted conversion from userland POLL* to kernel EPOLL* constants
mistakingly replaced the poll flags in the loopback_test tool, which
therefore no longer builds.
SAMA5D2x doesn't drive CMD line if GPIO is used as CD line (at least
SAMA5D27 doesn't). Fix this by forcing card-detect in the module
if module-controlled CD is not used.
Fixed commit addresses the problem only for non-removable cards. This
amends it to also cover gpio-cd case.
Cc: stable@vger.kernel.org Fixes: 7a1e3f143176 ("mmc: sdhci-of-at91: force card detect value for non removable devices") Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl> Acked-by: Adrian Hunter <adrian.hunter@intel.com> Link: https://lore.kernel.org/r/8d10950d9940468577daef4772b82a071b204716.1584290561.git.mirq-linux@rere.qmqm.pl Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The differential channels require writing the channel offset register (COR).
Otherwise they do not work in differential mode.
The configuration of COR is missing in triggered mode.
Fixes: 5e1a1da0f8c9 ("iio: adc: at91-sama5d2_adc: add hw trigger and buffer support") Signed-off-by: Eugen Hristev <eugen.hristev@microchip.com> Cc: <Stable@vger.kernel.org> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
At the moment, reading from in_magn_*_raw in sysfs tends to return
large values around 65000, even though the output of ak8974 is actually
limited to ±32768. This happens because the value is never converted
to the signed 16-bit integer variant.
Add an explicit cast to s16 to fix this.
Fixes: 7c94a8b2ee8c ("iio: magn: add a driver for AK8974") Signed-off-by: Stephan Gerhold <stephan@gerhold.net> Reviewed-by: Linus Waleij <linus.walleij@linaro.org> Cc: <Stable@vger.kernel.org> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Master mode should be disabled when stopping. This mainly impacts
possible other use-case after timer has been stopped. Currently,
master mode remains set (from start routine).
Fixes: 6fb34812c2a2 ("iio: stm32 trigger: Add support for TRGO2 triggers") Signed-off-by: Fabrice Gasnier <fabrice.gasnier@st.com> Cc: <Stable@vger.kernel.org> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fixes: 3d56e19815b3 ("iio: accel: st_accel: Add support for the SMO8840 ACPI id") Signed-off-by: Wen-chien Jesse Sung <jesse.sung@canonical.com> Tested-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Hans de Goede <hdegoede@redhat.com> Cc: <Stable@vger.kernel.org> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The return value checks in snd_pcm_plug_alloc() are covered with
snd_BUG_ON() macro that may trigger a kernel WARNING depending on the
kconfig. But since the error condition can be triggered by a weird
user space parameter passed to OSS layer, we shouldn't give the kernel
stack trace just for that. As it's a normal error condition, let's
remove snd_BUG_ON() macro usage there.
Each OSS PCM plugins allocate its internal buffer per pre-calculation
of the max buffer size through the chain of plugins (calling
src_frames and dst_frames callbacks). This works for most plugins,
but the rate plugin might behave incorrectly. The calculation in the
rate plugin involves with the fractional position, i.e. it may vary
depending on the input position. Since the buffer size
pre-calculation is always done with the offset zero, it may return a
shorter size than it might be; this may result in the out-of-bound
access as spotted by fuzzer.
This patch addresses those possible buffer overflow accesses by simply
setting the upper limit per the given buffer size for each plugin
before src_frames() and after dst_frames() calls.
The virmidi driver handles sysex event exceptionally in a short-cut
snd_seq_dump_var_event() call, but this missed the reset of the
running status. As a result, it may lead to an incomplete command
right after the sysex when an event with the same running status was
queued.
Fix it by clearing the running status properly via alling
snd_midi_event_reset_decode() for that code path.
The MIDI input event parser of the LINE6 driver may enter into an
endless loop when the unexpected data sequence is given, as it tries
to continue the secondary bytes without termination. Also, when the
input data is too short, the parser returns a negative error, while
the caller doesn't handle it properly. This would lead to the
unexpected behavior as well.
This patch addresses those issues by checking the return value
correctly and handling the one-byte event in the parser properly.
This controller timeouts during suspend (S3) with
[ 240.521724] xhci_hcd 0000:30:00.3: WARN: xHC save state timeout
[ 240.521729] xhci_hcd 0000:30:00.3: ERROR mismatched command completion event
thus preventing the system from entering S3.
Moreover it remains in an undefined state where some connected devices stop
working until a reboot.
Apply the XHCI_SUSPEND_DELAY quirk to make it suspend properly.
We have been receiving bug reports that ethernet connections over
RTL8153 based ethernet adapters stops working after a while with
errors like these showing up in dmesg when the ethernet stops working:
This has been reported on Dell WD15 docks, Belkin USB-C Express Dock 3.1
docks and with generic USB to ethernet dongles using the RTL8153
chipsets. Some users have tried adding usbcore.quirks=0bda:8153:k to
the kernel commandline and all users who have tried this report that
this fixes this.
Also note that we already have an existing NO_LPM quirk for the RTL8153
used in the Microsoft Surface Dock (where it uses a different usb-id).
This commit adds a NO_LPM quirk for the generic Realtek RTL8153
0bda:8153 usb-id, fixing the Tx timeout errors on these devices.
The bfq_find_set_group() function takes as input a blkcg (which represents
a cgroup) and retrieves the corresponding bfq_group, then it updates the
bfq internal group hierarchy (see comments inside the function for why
this is needed) and finally it returns the bfq_group.
In the hierarchy update cycle, the pointer holding the correct bfq_group
that has to be returned is mistakenly used to traverse the hierarchy
bottom to top, meaning that in each iteration it gets overwritten with the
parent of the current group. Since the update cycle stops at root's
children (depth = 2), the overwrite becomes a problem only if the blkcg
describes a cgroup at a hierarchy level deeper than that (depth > 2). In
this case the root's child that happens to be also an ancestor of the
correct bfq_group is returned. The main consequence is that processes
contained in a cgroup at depth greater than 2 are wrongly placed in the
group described above by BFQ.
This commits fixes this problem by using a different bfq_group pointer in
the update cycle in order to avoid the overwrite of the variable holding
the original group reference.
Reported-by: Kwon Je Oh <kwonje.oh2@gmail.com> Signed-off-by: Carlo Nonato <carlo.nonato95@gmail.com> Signed-off-by: Paolo Valente <paolo.valente@linaro.org> Signed-off-by: Jens Axboe <axboe@kernel.dk> Signed-off-by: Sasha Levin <sashal@kernel.org>
Below is sample callstack when the issue is reproduced on purpose by
reordering the updates of req->body and req->state and adding delay in
code between updates of req->state and req->body.
The virt_rmb() is added in the 'true' path of test_reply(). The "while"
is changed to "do while" so that test_reply() is used as a read memory
barrier.
[Why]
Swath sizes are being calculated incorrectly. The horizontal swath size
should be the product of block height, viewport width, and bytes per
element, but the calculation uses viewport height instead of width. The
vertical swath size is similarly incorrectly calculated. The effect of
this is that we report the wrong DCC caps.
[How]
Use viewport width in the horizontal swath size calculation and viewport
height in the vertical swath size calculation.
Signed-off-by: Josip Pavic <Josip.Pavic@amd.com> Reviewed-by: Aric Cyr <Aric.Cyr@amd.com> Acked-by: Rodrigo Siqueira <Rodrigo.Siqueira@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
[How]
when we disable MST we should clear the cur link settings (lane_count=0 is
good enough). This will cause us to not reallocate payloads earlier than
expected and not throw the warning
The compiler uses the PIC-relative method to access static variables
instead of GOT when the code model is PIC. Therefore, the limitation of
the access range from the instruction to the symbol address is +-2GB.
Under this circumstance, the kernel cannot load a kernel module if this
module has static per-CPU symbols declared by DEFINE_PER_CPU(). The reason
is that kernel relocates the .data..percpu section of the kernel module to
the end of kernel's .data..percpu. Hence, the distance between the per-CPU
symbols and the instruction will exceed the 2GB limits. To solve this
problem, the kernel should place the loaded module in the memory area
[&_end-2G, VMALLOC_END].
Signed-off-by: Vincent Chen <vincent.chen@sifive.com> Suggested-by: Alexandre Ghiti <alex@ghiti.fr> Suggested-by: Anup Patel <anup@brainfault.org> Tested-by: Alexandre Ghiti <alex@ghiti.fr> Tested-by: Carlos de Paula <me@carlosedp.com> Signed-off-by: Palmer Dabbelt <palmerdabbelt@google.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
In cases where dec_in_flight() has to requeue the integrity_bio_wait
work to transfer the rest of the data, the bio's __bi_remaining might
already have been decremented to 0, e.g.: if bio passed to underlying
data device was split via blk_queue_split().
Use dm_bio_{record,restore} rather than effectively open-coding them in
dm-integrity -- these methods now manage __bi_remaining too.
Depends-on: f7f0b057a9c1 ("dm bio record: save/restore bi_end_io and bi_integrity") Reported-by: Daniel Glöckner <dg@emlix.com> Suggested-by: Mikulas Patocka <mpatocka@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
altera_get_note is called from altera_init, where key is kzalloc(33).
When the allocation functions are annotated to allow the compiler to see
the sizes of objects, and with FORTIFY_SOURCE, we see:
In file included from drivers/misc/altera-stapl/altera.c:14:0:
In function ‘strlcpy’,
inlined from ‘altera_init’ at drivers/misc/altera-stapl/altera.c:2189:5:
include/linux/string.h:378:4: error: call to ‘__write_overflow’ declared with attribute error: detected write beyond size of object passed as 1st parameter
__write_overflow();
^~~~~~~~~~~~~~~~~~
The error triggers because the length of 'key' is 33, but the copy
uses length supplied as the 'length' parameter, which is always
256. Split the size parameter into key_len and val_len, and use the
appropriate length depending on what is being copied.
Writing to the built-in strings arrays doesn't work if driver is loaded
as kernel module. This is also considered as a bad pattern. Fix this by
adding a call to clk_get() with legacy clock name. This fixes following
kernel oops if driver is loaded as module:
Unable to handle kernel paging request at virtual address bf047978
pgd = (ptrval)
[bf047978] *pgd=59344811, *pte=5903c6df, *ppte=5903c65f
Internal error: Oops: 80f [#1] SMP ARM
Modules linked in: mc exynosdrm(+) analogix_dp rtc_s3c exynos_ppmu i2c_gpio
CPU: 1 PID: 212 Comm: systemd-udevd Not tainted 5.6.0-rc2-next-20200219 #326
videodev: Linux video capture interface: v2.00
Hardware name: Samsung Exynos (Flattened Device Tree)
PC is at exynos_dsi_probe+0x1f0/0x384 [exynosdrm]
LR is at exynos_dsi_probe+0x1dc/0x384 [exynosdrm]
...
Process systemd-udevd (pid: 212, stack limit = 0x(ptrval))
...
[<bf03cf14>] (exynos_dsi_probe [exynosdrm]) from [<c09b1ca0>] (platform_drv_probe+0x6c/0xa4)
[<c09b1ca0>] (platform_drv_probe) from [<c09afcb8>] (really_probe+0x210/0x350)
[<c09afcb8>] (really_probe) from [<c09aff74>] (driver_probe_device+0x60/0x1a0)
[<c09aff74>] (driver_probe_device) from [<c09b0254>] (device_driver_attach+0x58/0x60)
[<c09b0254>] (device_driver_attach) from [<c09b02dc>] (__driver_attach+0x80/0xbc)
[<c09b02dc>] (__driver_attach) from [<c09ade00>] (bus_for_each_dev+0x68/0xb4)
[<c09ade00>] (bus_for_each_dev) from [<c09aefd8>] (bus_add_driver+0x130/0x1e8)
[<c09aefd8>] (bus_add_driver) from [<c09b0d64>] (driver_register+0x78/0x110)
[<c09b0d64>] (driver_register) from [<bf038558>] (exynos_drm_init+0xe8/0x11c [exynosdrm])
[<bf038558>] (exynos_drm_init [exynosdrm]) from [<c0302fa8>] (do_one_initcall+0x50/0x220)
[<c0302fa8>] (do_one_initcall) from [<c03dd02c>] (do_init_module+0x60/0x210)
[<c03dd02c>] (do_init_module) from [<c03dbf44>] (load_module+0x1c0c/0x2310)
[<c03dbf44>] (load_module) from [<c03dc85c>] (sys_finit_module+0xac/0xbc)
[<c03dc85c>] (sys_finit_module) from [<c0301000>] (ret_fast_syscall+0x0/0x54)
Exception stack(0xd979bfa8 to 0xd979bff0)
...
---[ end trace db16efe05faab470 ]---
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com> Reviewed-by: Andrzej Hajda <a.hajda@samsung.com> Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Properly propagate error value from devm_regulator_bulk_get() and don't
confuse user with meaningless warning about failure in getting regulators
in case of deferred probe.
Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com> Reviewed-by: Krzysztof Kozlowski <krzk@kernel.org> Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
In the public interface for chipselect, there is always an entry
commented as "Dummy generic FIFO entry" pushed down to the fifo right
after the activate/deactivate command. The dummy entry is 0x0,
irregardless if the intention was to activate or deactive the cs. This
causes the cs line to glitch rather than beeing activated in the case
when there was an activate command.
This has been observed on oscilloscope, and have caused problems for at
least one specific flash device type connected to the qspi port. After
the change the glitch is gone and cs goes active when intended.
The reason why this worked before (except for the glitch) was because
when sending the actual data, the CS bits are once again set. Since
most flashes uses mode 0, there is always a half clk period anyway for
cs to clk active setup time. If someone would rely on timing from a
chip_select call to a transfer_one, it would fail though.
It is unknown why the dummy entry was there in the first place, git log
seems to be of no help in this case. The reference manual gives no
indication of the necessity of this. In fact the lower 8 bits are a
setup (or hold in case of deactivate) time expressed in cycles. So this
should not be needed to fulfill any setup/hold timings.
Signed-off-by: Thommy Jakobsson <thommyj@gmail.com> Reviewed-by: Naga Sureshkumar Relli <naga.sureshkumar.relli@xilinx.com> Link: https://lore.kernel.org/r/20200224162643.29102-1-thommyj@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
In some circumstances on Intel LPSS controllers, toggling the LPSS
CS control register doesn't actually cause the CS line to toggle.
This seems to be failure of dynamic clock gating that occurs after
going through a suspend/resume transition, where the controller
is sent through a reset transition. This ruins SPI transactions
that either rely on delay_usecs, or toggle the CS line without
sending data.
Whenever CS is toggled, momentarily set the clock gating register
to "Force On" to poke the controller into acting on CS.
'dma-ranges' in a PCI bridge node does correctly set dma masks for PCI
devices not described in the DT. Certain DRA7 platforms (e.g., DRA76)
has RAM above 32-bit boundary (accessible with LPAE config) though the
PCIe bridge will be able to access only 32-bits. Add 'dma-ranges'
property in PCIe RC DT nodes to indicate the host bridge can access
only 32 bits.
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> Signed-off-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Selecting CONFIG_DEBUG_INFO_BTF results in the below warning from ld:
ld: warning: orphan section `.BTF' from `.btf.vmlinux.bin.o' being placed in section `.BTF'
Include .BTF section in vmlinux explicitly to fix the same.
spi_qup_suspend() will cause synchronous external abort when
runtime suspend is enabled and applied, as it tries to
access SPI controller register while clock is already disabled
in spi_qup_pm_suspend_runtime().
The cursor and primary planes were hard coded.
Now search for them for passing to drm_crtc_init_with_planes
Signed-off-by: Evan Benn <evanbenn@chromium.org> Reviewed-by: Sean Paul <seanpaul@chromium.org> Signed-off-by: CK Hu <ck.hu@mediatek.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Similarly to commit c543cb4a5f07 ("ipv4: ensure rcu_read_lock() in
ipv4_link_failure()"), __ip_options_compile() must be called under rcu
protection.
Fixes: 3da1ed7ac398 ("net: avoid use IPCB in cipso_v4_error") Suggested-by: Guillaume Nault <gnault@redhat.com> Signed-off-by: Matteo Croce <mcroce@redhat.com> Acked-by: Paul Moore <paul@paul-moore.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
9dbbedaa6171 ("efi: Make efi_rts_work accessible to efi page fault handler")
converted 'efi_rts_work' from an auto variable to a global variable.
However, when submitting the work, INIT_WORK_ONSTACK() was still used,
causing the following complaint from debugobjects:
When kmem_cache_alloc_bulk() attempts to allocate N objects from a percpu
freelist of length M, and N > M > 0, it will first remove the M elements
from the percpu freelist, then call ___slab_alloc() to allocate the next
element and repopulate the percpu freelist. ___slab_alloc() can re-enable
IRQs via allocate_slab(), so the TID must be bumped before ___slab_alloc()
to properly commit the freelist head change.
Fix it by unconditionally bumping c->tid when entering the slowpath.
When the uaccess .fixup section was renamed to .text.fixup, one case was
missed. Under ld.bfd, the orphaned section was moved close to .text
(since they share the "ax" bits), so things would work normally on
uaccess faults. Under ld.lld, the orphaned section was placed outside
the .text section, making it unreachable.
It is possible for a system with an ARMv8 timer to run a 32-bit kernel.
When this happens we will unconditionally have the vDSO code remove the
__vdso_gettimeofday and __vdso_clock_gettime symbols because
cntvct_functional() returns false since it does not match that
compatibility string.
Fixes: ecf99a439105 ("ARM: 8331/1: VDSO initialization, mapping, and synchronization") Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The len used for skb_put_padto is wrong, it need to add len of hdr.
In qrtr_node_enqueue, local variable size_t len is assign with
skb->len, then skb_push(skb, sizeof(*hdr)) will add skb->len with
sizeof(*hdr), so local variable size_t len is not same with skb->len
after skb_push(skb, sizeof(*hdr)).
Then the purpose of skb_put_padto(skb, ALIGN(len, 4)) is to add add
pad to the end of the skb's data if skb->len is not aligned to 4, but
unfortunately it use len instead of skb->len, at this line, skb->len
is 32 bytes(sizeof(*hdr)) more than len, for example, len is 3 bytes,
then skb->len is 35 bytes(3 + 32), and ALIGN(len, 4) is 4 bytes, so
__skb_put_padto will do nothing after check size(35) < len(4), the
correct value should be 36(sizeof(*hdr) + ALIGN(len, 4) = 32 + 4),
then __skb_put_padto will pass check size(35) < len(36) and add 1 byte
to the end of skb's data, then logic is correct.
function of skb_push:
void *skb_push(struct sk_buff *skb, unsigned int len)
{
skb->data -= len;
skb->len += len;
if (unlikely(skb->data < skb->head))
skb_under_panic(skb, len, __builtin_return_address(0));
return skb->data;
}
function of skb_put_padto
static inline int skb_put_padto(struct sk_buff *skb, unsigned int len)
{
return __skb_put_padto(skb, len, true);
}
function of __skb_put_padto
static inline int __skb_put_padto(struct sk_buff *skb, unsigned int len,
bool free_on_error)
{
unsigned int size = skb->len;
if (unlikely(size < len)) {
len -= size;
if (__skb_pad(skb, len, free_on_error))
return -ENOMEM;
__skb_put(skb, len);
}
return 0;
}
Signed-off-by: Carl Huang <cjhuang@codeaurora.org> Signed-off-by: Wen Gong <wgong@codeaurora.org> Cc: Doug Anderson <dianders@chromium.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After commit 515db266a9da ("driver core: Remove device link creation
limitation"), if PM-runtime flags are passed to device_link_add(), it
will fail (returning NULL) due to an overly restrictive flags check
introduced by that commit.
Fix this issue by extending the check in question to cover the
PM-runtime flags too.
Fixes: 515db266a9da ("driver core: Remove device link creation limitation") Reported-by: Dmitry Osipenko <digetx@gmail.com> Tested-by: Jon Hunter <jonathanh@nvidia.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Tested-by: Dmitry Osipenko <digetx@gmail.com> Tested-by: Marek Szyprowski <m.szyprowski@samsung.com> Link: https://lore.kernel.org/r/7674989.cD04D8YV3U@kreacher Signed-off-by: Saravana Kannan <saravanak@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If device_link_add() is called for a consumer/supplier pair with an
existing device link between them and the existing link's type is
not in agreement with the flags passed to that function by its
caller, NULL will be returned. That is seriously inconvenient,
because it forces the callers of device_link_add() to worry about
what others may or may not do even if that is not relevant to them
for any other reasons.
It turns out, however, that this limitation can be made go away
relatively easily.
The underlying observation is that if DL_FLAG_STATELESS has been
passed to device_link_add() in flags for the given consumer/supplier
pair at least once, calling either device_link_del() or
device_link_remove() to release the link returned by it should work,
but there are no other requirements associated with that flag. In
turn, if at least one of the callers of device_link_add() for the
given consumer/supplier pair has not passed DL_FLAG_STATELESS to it
in flags, the driver core should track the status of the link and act
on it as appropriate (ie. the link should be treated as "managed").
This means that DL_FLAG_STATELESS needs to be set for managed device
links and it should be valid to call device_link_del() or
device_link_remove() to drop references to them in certain
sutiations.
To allow that to happen, introduce a new (internal) device link flag
called DL_FLAG_MANAGED and make device_link_add() set it automatically
whenever DL_FLAG_STATELESS is not passed to it. Also make it take
additional references to existing device links that were previously
stateless (that is, with DL_FLAG_STATELESS set and DL_FLAG_MANAGED
unset) and will need to be managed going forward and initialize
their status (which has been DL_STATE_NONE so far).
Accordingly, when a managed device link is dropped automatically
by the driver core, make it clear DL_FLAG_MANAGED, reset the link's
status back to DL_STATE_NONE and drop the reference to it associated
with DL_FLAG_MANAGED instead of just deleting it right away (to
allow it to stay around in case it still needs to be released
explicitly by someone).
With that, since setting DL_FLAG_STATELESS doesn't mean that the
device link in question is not managed any more, replace all of the
status-tracking checks against DL_FLAG_STATELESS with analogous
checks against DL_FLAG_MANAGED and update the documentation to
reflect these changes.
While at it, make device_link_add() reject flags that it does not
recognize, including DL_FLAG_MANAGED.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Reviewed-by: Saravana Kannan <saravanak@google.com> Tested-by: Marek Szyprowski <m.szyprowski@samsung.com> Review-by: Saravana Kannan <saravanak@google.com> Link: https://lore.kernel.org/r/2305283.AStDPdUUnE@kreacher Signed-off-by: Saravana Kannan <saravanak@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add a new device link flag, DL_FLAG_AUTOPROBE_CONSUMER, to request the
driver core to probe for a consumer driver automatically after binding
a driver to the supplier device on a persistent managed device link.
As unbinding the supplier driver on a managed device link causes the
consumer driver to be detached from its device automatically, this
flag provides a complementary mechanism which is needed to address
some "composite device" use cases.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Saravana Kannan <saravanak@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Even though stateful device links are managed by the driver core in
principle, their creators are allowed and sometimes even expected
to drop references to them via device_link_del() or
device_link_remove(), but that doesn't really play well with the
"persistent" link concept.
If "persistent" managed device links are created from driver
probe callbacks, device_link_add() called to do that will take a
new reference on the link each time the callback runs and those
references will never be dropped, which kind of isn't nice.
This issues arises because of the link reference counting carried
out by device_link_add() for existing links, but that is only done to
avoid deleting device links that may still be necessary, which
shouldn't be a concern for managed (stateful) links. These device
links are managed by the driver core and whoever creates one of them
will need it at least as long as until the consumer driver is detached
from its device and deleting it may be left to the driver core just
fine.
For this reason, rework device_link_add() to apply the reference
counting to stateless links only and make device_link_del() and
device_link_remove() drop references to stateless links only too.
After this change, if called to add a stateful device link for
a consumer-supplier pair for which a stateful device link is
present already, device_link_add() will return the existing link
without incrementing its reference counter. Accordingly,
device_link_del() and device_link_remove() will WARN() and do
nothing when called to drop a reference to a stateful link. Thus,
effectively, all stateful device links will be owned by the driver
core.
In addition, clean up the handling of the link management flags,
DL_FLAG_AUTOREMOVE_CONSUMER and DL_FLAG_AUTOREMOVE_SUPPLIER, so that
(a) they are never set at the same time and (b) if device_link_add()
is called for a consumer-supplier pair with an existing stateful link
between them, the flags of that link will be combined with the flags
passed to device_link_add() to ensure that the life time of the link
is sufficient for all of the callers of device_link_add() for the
same consumer-supplier pair.
Update the device_link_add() kerneldoc comment to reflect the
above changes.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Saravana Kannan <saravanak@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently, it is not valid to add a device link from a consumer
driver ->probe callback to a supplier that is still probing too, but
generally this is a valid use case. For example, if the consumer has
just acquired a resource that can only be available if the supplier
is functional, adding a device link to that supplier right away
should be safe (and even desirable arguably), but device_link_add()
doesn't handle that case correctly and the initial state of the link
created by it is wrong then.
To address this problem, change the initial state of device links
added between a probing supplier and a probing consumer to
DL_STATE_CONSUMER_PROBE and update device_links_driver_bound() to
skip such links on the supplier side.
With this change, if the supplier probe completes first,
device_links_driver_bound() called for it will skip the link state
update and when it is called for the consumer, the link state will
be updated to "active". In turn, if the consumer probe completes
first, device_links_driver_bound() called for it will change the
state of the link to "active" and when it is called for the
supplier, the link status update will be skipped.
However, in principle the supplier or consumer probe may still fail
after the link has been added, so modify device_links_no_driver() to
change device links in the "active" or "consumer probe" state to
"dormant" on the supplier side and update __device_links_no_driver()
to change the link state to "available" only if it is "consumer
probe" or "active".
Then, if the supplier probe fails first, the leftover link to the
probing consumer will become "dormant" and device_links_no_driver()
called for the consumer (when its probe fails) will clean it up.
In turn, if the consumer probe fails first, it will either drop the
link, or change its state to "available" and, in the latter case,
when device_links_no_driver() is called for the supplier, it will
update the link state to "dormant". [If the supplier probe fails,
but the consumer probe succeeds, which should not happen as long as
the consumer driver is correct, the link still will be around, but
it will be "dormant" until the supplier is probed again.]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Saravana Kannan <saravanak@google.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
DL_FLAG_AUTOREMOVE_CONSUMER/SUPPLIER means "Remove the link
automatically on consumer/supplier driver unbind", that means we should
remove whole the device_link when there is no this driver no matter what
the ref_count of the link is.
According to the App note[1] detailing the tuning algorithm, for
temperatures < -20C, the initial tuning value should be min(largest value
in LPW - 24, ceil(13/16 ratio of LPW)). The largest value in LPW is
(max_window + 4 * (max_len - 1)) and not (max_window + 4 * max_len) itself.
Fix this implementation.
commit 5b0d62108b46 ("mmc: sdhci-omap: Add platform specific reset
callback") skips data resets during tuning operation. Because of this,
a data error or data finish interrupt might still arrive after a command
error has been handled and the mrq ended. This ends up with a "mmc0: Got
data interrupt 0x00000002 even though no data operation was in progress"
error message.
Fix this by adding a platform specific callback for sdhci_irq. Mark the
mrq as a failure but wait for a data interrupt instead of calling
finish_mrq().
Fixes: 5b0d62108b46 ("mmc: sdhci-omap: Add platform specific reset
callback") Signed-off-by: Faiz Abbas <faiz_abbas@ti.com> Acked-by: Adrian Hunter <adrian.hunter@intel.com> Cc: stable@vger.kernel.org Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
In the implementation of i2400m_op_rfkill_sw_toggle() the allocated
buffer for cmd should be released before returning. The
documentation for i2400m_msg_to_dev() says when it returns the buffer
can be reused. Meaning cmd should be released in either case. Move
kfree(cmd) before return to be reached by all execution paths.
Fixes: 2507e6ab7a9a ("wimax: i2400: fix memory leak") Signed-off-by: Navid Emamdoost <navid.emamdoost@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
In i2400m_op_rfkill_sw_toggle cmd buffer should be released along with
skb response.
Signed-off-by: Navid Emamdoost <navid.emamdoost@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
journal_head::b_transaction and journal_head::b_next_transaction could
be accessed concurrently as noticed by KCSAN,
LTP: starting fsync04
/dev/zero: Can't open blockdev
EXT4-fs (loop0): mounting ext3 file system using the ext4 subsystem
EXT4-fs (loop0): mounted filesystem with ordered data mode. Opts: (null)
==================================================================
BUG: KCSAN: data-race in __jbd2_journal_refile_buffer [jbd2] / jbd2_write_access_granted [jbd2]
write to 0xffff99f9b1bd0e30 of 8 bytes by task 25721 on cpu 70:
__jbd2_journal_refile_buffer+0xdd/0x210 [jbd2]
__jbd2_journal_refile_buffer at fs/jbd2/transaction.c:2569
jbd2_journal_commit_transaction+0x2d15/0x3f20 [jbd2]
(inlined by) jbd2_journal_commit_transaction at fs/jbd2/commit.c:1034
kjournald2+0x13b/0x450 [jbd2]
kthread+0x1cd/0x1f0
ret_from_fork+0x27/0x50
read to 0xffff99f9b1bd0e30 of 8 bytes by task 25724 on cpu 68:
jbd2_write_access_granted+0x1b2/0x250 [jbd2]
jbd2_write_access_granted at fs/jbd2/transaction.c:1155
jbd2_journal_get_write_access+0x2c/0x60 [jbd2]
__ext4_journal_get_write_access+0x50/0x90 [ext4]
ext4_mb_mark_diskspace_used+0x158/0x620 [ext4]
ext4_mb_new_blocks+0x54f/0xca0 [ext4]
ext4_ind_map_blocks+0xc79/0x1b40 [ext4]
ext4_map_blocks+0x3b4/0x950 [ext4]
_ext4_get_block+0xfc/0x270 [ext4]
ext4_get_block+0x3b/0x50 [ext4]
__block_write_begin_int+0x22e/0xae0
__block_write_begin+0x39/0x50
ext4_write_begin+0x388/0xb50 [ext4]
generic_perform_write+0x15d/0x290
ext4_buffered_write_iter+0x11f/0x210 [ext4]
ext4_file_write_iter+0xce/0x9e0 [ext4]
new_sync_write+0x29c/0x3b0
__vfs_write+0x92/0xa0
vfs_write+0x103/0x260
ksys_write+0x9d/0x130
__x64_sys_write+0x4c/0x60
do_syscall_64+0x91/0xb05
entry_SYSCALL_64_after_hwframe+0x49/0xbe
5 locks held by fsync04/25724:
#0: ffff99f9911093f8 (sb_writers#13){.+.+}, at: vfs_write+0x21c/0x260
#1: ffff99f9db4c0348 (&sb->s_type->i_mutex_key#15){+.+.}, at: ext4_buffered_write_iter+0x65/0x210 [ext4]
#2: ffff99f5e7dfcf58 (jbd2_handle){++++}, at: start_this_handle+0x1c1/0x9d0 [jbd2]
#3: ffff99f9db4c0168 (&ei->i_data_sem){++++}, at: ext4_map_blocks+0x176/0x950 [ext4]
#4: ffffffff99086b40 (rcu_read_lock){....}, at: jbd2_write_access_granted+0x4e/0x250 [jbd2]
irq event stamp: 1407125
hardirqs last enabled at (1407125): [<ffffffff980da9b7>] __find_get_block+0x107/0x790
hardirqs last disabled at (1407124): [<ffffffff980da8f9>] __find_get_block+0x49/0x790
softirqs last enabled at (1405528): [<ffffffff98a0034c>] __do_softirq+0x34c/0x57c
softirqs last disabled at (1405521): [<ffffffff97cc67a2>] irq_exit+0xa2/0xc0
Reported by Kernel Concurrency Sanitizer on:
CPU: 68 PID: 25724 Comm: fsync04 Tainted: G L 5.6.0-rc2-next-20200221+ #7
Hardware name: HPE ProLiant DL385 Gen10/ProLiant DL385 Gen10, BIOS A40 07/10/2019
The plain reads are outside of jh->b_state_lock critical section which result
in data races. Fix them by adding pairs of READ|WRITE_ONCE().
We can't just use the top bits of the last sync event as they could be
off-by-one every 65,536 seconds, giving an error in reconstruction of
65,536 seconds.
This patch uses the difference in the bottom 16 bits (mod 2^16) to
calculate an offset that needs to be applied to the last sync event to
get to the current time.
Signed-off-by: Alexandru-Mihai Maftei <amaftei@solarflare.com> Acked-by: Martin Habets <mhabets@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
Packet forwarding is not working in rmnet bridge mode.
Because when a packet is forwarded, skb_push() for an ethernet header
is needed. But it doesn't call skb_push().
So, the ethernet header will be lost.
Test commands:
modprobe rmnet
ip netns add nst
ip netns add nst2
ip link add veth0 type veth peer name veth1
ip link add veth2 type veth peer name veth3
ip link set veth1 netns nst
ip link set veth3 netns nst2
ip link add rmnet0 link veth0 type rmnet mux_id 1
ip link set veth2 master rmnet0
ip link set veth0 up
ip link set veth2 up
ip link set rmnet0 up
ip a a 192.168.100.1/24 dev rmnet0
ip netns exec nst ip link set veth1 up
ip netns exec nst ip a a 192.168.100.2/24 dev veth1
ip netns exec nst2 ip link set veth3 up
ip netns exec nst2 ip a a 192.168.100.3/24 dev veth3
ip netns exec nst2 ping 192.168.100.2
Fixes: 60d58f971c10 ("net: qualcomm: rmnet: Implement bridge mode") Signed-off-by: Taehee Yoo <ap420073@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
In order to attach a bridge interface to the rmnet interface,
"master" operation is used.
(e.g. ip link set dummy1 master rmnet0)
But, in the rmnet_add_bridge(), which is a callback of ->ndo_add_slave()
doesn't register lower interface.
So, ->ndo_del_slave() doesn't work.
There are other problems too.
1. It couldn't detect circular upper/lower interface relationship.
2. It couldn't prevent stack overflow because of too deep depth
of upper/lower interface
3. It doesn't check the number of lower interfaces.
4. Panics because of several reasons.
The root problem of these issues is actually the same.
So, in this patch, these all problems will be fixed.
Test commands:
modprobe rmnet
ip link add dummy0 type dummy
ip link add rmnet0 link dummy0 type rmnet mux_id 1
ip link add dummy1 master rmnet0 type dummy
ip link add dummy2 master rmnet0 type dummy
ip link del rmnet0
ip link del dummy2
ip link del dummy1
netdev_upper_dev_link() is useful to manage lower/upper interfaces.
And this function internally validates looping, maximum depth.
All or most virtual interfaces that could have a real interface
(e.g. macsec, macvlan, ipvlan etc.) use lower/upper infrastructure.
Test commands:
modprobe rmnet
ip link add dummy0 type dummy
ip link add rmnet1 link dummy0 type rmnet mux_id 1
for i in {2..100}
do
let A=$i-1
ip link add rmnet$i link rmnet$A type rmnet mux_id $i
done
ip link del dummy0
The purpose of the test commands is to make stack overflow.
Fixes: b37f78f234bf ("net: qualcomm: rmnet: Fix crash on real dev unregistration") Signed-off-by: Taehee Yoo <ap420073@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
Basically, duplicate mux id isn't be allowed.
So, the creation of rmnet will be failed if there is duplicate mux id
is existing.
But, changelink routine doesn't check duplicate mux id.
Test commands:
modprobe rmnet
ip link add dummy0 type dummy
ip link add rmnet0 link dummy0 type rmnet mux_id 1
ip link add rmnet1 link dummy0 type rmnet mux_id 2
ip link set rmnet1 type rmnet mux_id 1
Fixes: 23790ef12082 ("net: qualcomm: rmnet: Allow to configure flags for existing devices") Signed-off-by: Taehee Yoo <ap420073@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
The notifier_call() of the slave interface removes rmnet interface with
unregister_netdevice_queue().
But, before calling unregister_netdevice_queue(), it acquires
rcu readlock.
In the RCU critical section, sleeping isn't be allowed.
But, unregister_netdevice_queue() internally calls synchronize_net(),
which would sleep.
So, suspicious RCU usage warning occurs.
Test commands:
modprobe rmnet
ip link add dummy0 type dummy
ip link add dummy1 type dummy
ip link add rmnet0 link dummy0 type rmnet mux_id 1
ip link set dummy1 master rmnet0
ip link del dummy0
rmnet_get_port() internally calls rcu_dereference_rtnl(),
which checks RTNL.
But rmnet_get_port() could be called by packet path.
The packet path is not protected by RTNL.
So, the suspicious RCU usage problem occurs.
Test commands:
modprobe rmnet
ip netns add nst
ip link add veth0 type veth peer name veth1
ip link set veth1 netns nst
ip link add rmnet0 link veth0 type rmnet mux_id 1
ip netns exec nst ip link add rmnet1 link veth1 type rmnet mux_id 1
ip netns exec nst ip link set veth1 up
ip netns exec nst ip link set rmnet1 up
ip netns exec nst ip a a 192.168.100.2/24 dev rmnet1
ip link set veth0 up
ip link set rmnet0 up
ip a a 192.168.100.1/24 dev rmnet0
ping 192.168.100.2
Fixes: 23790ef12082 ("net: qualcomm: rmnet: Allow to configure flags for existing devices") Signed-off-by: Taehee Yoo <ap420073@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
rmnet registers IFLA_LINK interface as a lower interface.
But, IFLA_LINK could be NULL.
In the current code, rmnet doesn't check IFLA_LINK.
So, panic would occur.
Test commands:
modprobe rmnet
ip link add rmnet0 type rmnet mux_id 1
As the description before netdev_run_todo, we cannot call free_netdev
before rtnl_unlock, fix it by reorder the code.
Signed-off-by: yangerkun <yangerkun@huawei.com> Reviewed-by: Oliver Hartkopp <socketcan@hartkopp.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
When queueing a signal, we increment both the users count of pending
signals (for RLIMIT_SIGPENDING tracking) and we increment the refcount
of the user struct itself (because we keep a reference to the user in
the signal structure in order to correctly account for it when freeing).
That turns out to be fairly expensive, because both of them are atomic
updates, and particularly under extreme signal handling pressure on big
machines, you can get a lot of cache contention on the user struct.
That can then cause horrid cacheline ping-pong when you do these
multiple accesses.
So change the reference counting to only pin the user for the _first_
pending signal, and to unpin it when the last pending signal is
dequeued. That means that when a user sees a lot of concurrent signal
queuing - which is the only situation when this matters - the only
atomic access needed is generally the 'sigpending' count update.
This was noticed because of a particularly odd timing artifact on a
dual-socket 96C/192T Cascade Lake platform: when you get into bad
contention, on that machine for some reason seems to be much worse when
the contention happens in the upper 32-byte half of the cacheline.
As a result, the kernel test robot will-it-scale 'signal1' benchmark had
an odd performance regression simply due to random alignment of the
'struct user_struct' (and pointed to a completely unrelated and
apparently nonsensical commit for the regression).
Avoiding the double increments (and decrements on the dequeueing side,
of course) makes for much less contention and hugely improved
performance on that will-it-scale microbenchmark.
Quoting Feng Tang:
"It makes a big difference, that the performance score is tripled! bump
from original 17000 to 54000. Also the gap between 5.0-rc6 and
5.0-rc6+Jiri's patch is reduced to around 2%"
[ The "2% gap" is the odd cacheline placement difference on that
platform: under the extreme contention case, the effect of which half
of the cacheline was hot was 5%, so with the reduced contention the
odd timing artifact is reduced too ]
It does help in the non-contended case too, but is not nearly as
noticeable.
Reported-and-tested-by: Feng Tang <feng.tang@intel.com> Cc: Eric W. Biederman <ebiederm@xmission.com> Cc: Huang, Ying <ying.huang@intel.com> Cc: Philip Li <philip.li@intel.com> Cc: Andi Kleen <andi.kleen@intel.com> Cc: Jiri Olsa <jolsa@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
local->sta_mtx is held in __ieee80211_check_fast_rx_iface().
No need to use list_for_each_entry_rcu() as it also requires
a cond argument to avoid false lockdep warnings when not used in
RCU read-side section (with CONFIG_PROVE_RCU_LIST).
Therefore use list_for_each_entry();
The KS8851 requires that packet RX and TX are mutually exclusive.
Currently, the driver hopes to achieve this by disabling interrupt
from the card by writing the card registers and by disabling the
interrupt on the interrupt controller. This however is racy on SMP.
Replace this approach by expanding the spinlock used around the
ks_start_xmit() TX path to ks_irq() RX path to assure true mutual
exclusion and remove the interrupt enabling/disabling, which is
now not needed anymore. Furthermore, disable interrupts also in
ks_net_stop(), which was missing before.
Note that a massive improvement here would be to re-use the KS8851
driver approach, which is to move the TX path into a worker thread,
interrupt handling to threaded interrupt, and synchronize everything
with mutexes, but that would be a much bigger rework, for a separate
patch.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: David S. Miller <davem@davemloft.net> Cc: Lukas Wunner <lukas@wunner.de> Cc: Petr Stetiar <ynezz@true.cz> Cc: YueHaibing <yuehaibing@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
projects / linux-stable / log