[ Upstream commit df5db5f9ee ]
Before this patch, run_queue would demote glocks based on whether
there are any more holders. But if the glock has pending revokes that
haven't been written to the media, giving up the glock might end in
file system corruption if the revokes never get written due to
io errors, node crashes and fences, etc. In that case, another node
will replay the metadata blocks associated with the glock, but
because the revoke was never written, it could replay that block
even though the glock had since been granted to another node who
might have made changes.
This patch changes the logic in run_queue so that it never demotes
a glock until its count of pending revokes reaches zero.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9ff7828935 ]
Before this patch, if gfs2_ail_empty_gl saw there was nothing on
the ail list, it would return and not flush the log. The problem
is that there could still be a revoke for the rgrp sitting on the
sd_log_le_revoke list that's been recently taken off the ail list.
But that revoke still needs to be written, and the rgrp_go_inval
still needs to call log_flush_wait to ensure the revokes are all
properly written to the journal before we relinquish control of
the glock to another node. If we give the glock to another node
before we have this knowledge, the node might crash and its journal
replayed, in which case the missing revoke would allow the journal
replay to replay the rgrp over top of the rgrp we already gave to
another node, thus overwriting its changes and corrupting the
file system.
This patch makes gfs2_ail_empty_gl still call gfs2_log_flush rather
than returning.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1d72f7aec3 ]
If the call to scsi_add_host_with_dma() in ata_scsi_add_hosts() fails,
then we may get use-after-free KASAN warns:
==================================================================
BUG: KASAN: use-after-free in kobject_put+0x24/0x180
Read of size 1 at addr ffff0026b8c80364 by task swapper/0/1
CPU: 1 PID: 1 Comm: swapper/0 Tainted: G W 5.6.0-rc3-00004-g5a71b206ea82-dirty #1765
Hardware name: Huawei TaiShan 200 (Model 2280)/BC82AMDD, BIOS 2280-V2 CS V3.B160.01 02/24/2020
Call trace:
dump_backtrace+0x0/0x298
show_stack+0x14/0x20
dump_stack+0x118/0x190
print_address_description.isra.9+0x6c/0x3b8
__kasan_report+0x134/0x23c
kasan_report+0xc/0x18
__asan_load1+0x5c/0x68
kobject_put+0x24/0x180
put_device+0x10/0x20
scsi_host_put+0x10/0x18
ata_devres_release+0x74/0xb0
release_nodes+0x2d0/0x470
devres_release_all+0x50/0x78
really_probe+0x2d4/0x560
driver_probe_device+0x7c/0x148
device_driver_attach+0x94/0xa0
__driver_attach+0xa8/0x110
bus_for_each_dev+0xe8/0x158
driver_attach+0x30/0x40
bus_add_driver+0x220/0x2e0
driver_register+0xbc/0x1d0
__pci_register_driver+0xbc/0xd0
ahci_pci_driver_init+0x20/0x28
do_one_initcall+0xf0/0x608
kernel_init_freeable+0x31c/0x384
kernel_init+0x10/0x118
ret_from_fork+0x10/0x18
Allocated by task 5:
save_stack+0x28/0xc8
__kasan_kmalloc.isra.8+0xbc/0xd8
kasan_kmalloc+0xc/0x18
__kmalloc+0x1a8/0x280
scsi_host_alloc+0x44/0x678
ata_scsi_add_hosts+0x74/0x268
ata_host_register+0x228/0x488
ahci_host_activate+0x1c4/0x2a8
ahci_init_one+0xd18/0x1298
local_pci_probe+0x74/0xf0
work_for_cpu_fn+0x2c/0x48
process_one_work+0x488/0xc08
worker_thread+0x330/0x5d0
kthread+0x1c8/0x1d0
ret_from_fork+0x10/0x18
Freed by task 5:
save_stack+0x28/0xc8
__kasan_slab_free+0x118/0x180
kasan_slab_free+0x10/0x18
slab_free_freelist_hook+0xa4/0x1a0
kfree+0xd4/0x3a0
scsi_host_dev_release+0x100/0x148
device_release+0x7c/0xe0
kobject_put+0xb0/0x180
put_device+0x10/0x20
scsi_host_put+0x10/0x18
ata_scsi_add_hosts+0x210/0x268
ata_host_register+0x228/0x488
ahci_host_activate+0x1c4/0x2a8
ahci_init_one+0xd18/0x1298
local_pci_probe+0x74/0xf0
work_for_cpu_fn+0x2c/0x48
process_one_work+0x488/0xc08
worker_thread+0x330/0x5d0
kthread+0x1c8/0x1d0
ret_from_fork+0x10/0x18
There is also refcount issue, as well:
WARNING: CPU: 1 PID: 1 at lib/refcount.c:28 refcount_warn_saturate+0xf8/0x170
The issue is that we make an erroneous extra call to scsi_host_put()
for that host:
So in ahci_init_one()->ata_host_alloc_pinfo()->ata_host_alloc(), we setup
a device release method - ata_devres_release() - which intends to release
the SCSI hosts:
static void ata_devres_release(struct device *gendev, void *res)
{
...
for (i = 0; i < host->n_ports; i++) {
struct ata_port *ap = host->ports[i];
if (!ap)
continue;
if (ap->scsi_host)
scsi_host_put(ap->scsi_host);
}
...
}
However in the ata_scsi_add_hosts() error path, we also call
scsi_host_put() for the SCSI hosts.
Fix by removing the the scsi_host_put() calls in ata_scsi_add_hosts() and
leave this to ata_devres_release().
Fixes: f31871951b ("libata: separate out ata_host_alloc() and ata_host_register()")
Signed-off-by: John Garry <john.garry@huawei.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 64d4fc9926 ]
Fix build fault when CONFIG_HWMON is a module, and CONFIG_VIDEO_I2C
as builtin. This is due to 'imply hwmon' in the respective Kconfig.
Issue build log:
ld: drivers/media/i2c/video-i2c.o: in function `amg88xx_hwmon_init':
video-i2c.c:(.text+0x2e1): undefined reference to `devm_hwmon_device_register_with_info
Cc: rdunlap@infradead.org
Fixes: acbea67989 (media: video-i2c: add hwmon support for amg88xx)
Signed-off-by: Matt Ranostay <matt.ranostay@konsulko.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit fd1bb3ae54 ]
Commit ecedd3d7e1 ("block, bfq: get extra ref to prevent a queue
from being freed during a group move") gets an extra reference to a
bfq_queue before possibly deactivating it (temporarily), in
bfq_bfqq_move(). This prevents the bfq_queue from disappearing before
being reactivated in its new group.
Yet, the bfq_queue may also be expired (i.e., its service may be
stopped) before the bfq_queue is deactivated. And also an expiration
may lead to a premature freeing. This commit fixes this issue by
simply moving forward the getting of the extra reference already
introduced by commit ecedd3d7e1 ("block, bfq: get extra ref to
prevent a queue from being freed during a group move").
Reported-by: cki-project@redhat.com
Tested-by: cki-project@redhat.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>
[ Upstream commit efbdc76960 ]
The call to init_completion() in mrpc_queue_cmd() can theoretically
race with the call to poll_wait() in switchtec_dev_poll().
poll() write()
switchtec_dev_poll() switchtec_dev_write()
poll_wait(&s->comp.wait); mrpc_queue_cmd()
init_completion(&s->comp)
init_waitqueue_head(&s->comp.wait)
To my knowledge, no one has hit this bug.
Fix this by using reinit_completion() instead of init_completion() in
mrpc_queue_cmd().
Fixes: 080b47def5 ("MicroSemi Switchtec management interface driver")
Reported-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://lkml.kernel.org/r/20200313183608.2646-1-logang@deltatee.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 26cf52229e ]
During our testing, we found a case that shares no longer
working correctly, the cgroup topology is like:
/sys/fs/cgroup/cpu/A (shares=102400)
/sys/fs/cgroup/cpu/A/B (shares=2)
/sys/fs/cgroup/cpu/A/B/C (shares=1024)
/sys/fs/cgroup/cpu/D (shares=1024)
/sys/fs/cgroup/cpu/D/E (shares=1024)
/sys/fs/cgroup/cpu/D/E/F (shares=1024)
The same benchmark is running in group C & F, no other tasks are
running, the benchmark is capable to consumed all the CPUs.
We suppose the group C will win more CPU resources since it could
enjoy all the shares of group A, but it's F who wins much more.
The reason is because we have group B with shares as 2, since
A->cfs_rq.load.weight == B->se.load.weight == B->shares/nr_cpus,
so A->cfs_rq.load.weight become very small.
And in calc_group_shares() we calculate shares as:
load = max(scale_load_down(cfs_rq->load.weight), cfs_rq->avg.load_avg);
shares = (tg_shares * load) / tg_weight;
Since the 'cfs_rq->load.weight' is too small, the load become 0
after scale down, although 'tg_shares' is 102400, shares of the se
which stand for group A on root cfs_rq become 2.
While the se of D on root cfs_rq is far more bigger than 2, so it
wins the battle.
Thus when scale_load_down() scale real weight down to 0, it's no
longer telling the real story, the caller will have the wrong
information and the calculation will be buggy.
This patch add check in scale_load_down(), so the real weight will
be >= MIN_SHARES after scale, after applied the group C wins as
expected.
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Michael Wang <yun.wang@linux.alibaba.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Vincent Guittot <vincent.guittot@linaro.org>
Link: https://lkml.kernel.org/r/38e8e212-59a1-64b2-b247-b6d0b52d8dc1@linux.alibaba.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8277815349 ]
The gop_length field is actually only u16 and there are two more u8
fields in the message:
- the number of consecutive b-frames
- frequency of golden frames
Fix the message and thus fix the configuration of the GOP length.
Signed-off-by: Michael Tretter <m.tretter@pengutronix.de>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 2c8bd58812 ]
To minimize latency, PREEMPT_RT kernels expires hrtimers in preemptible
softirq context by default. This can be overriden by marking the timer's
expiry with HRTIMER_MODE_HARD.
sched_clock_timer is missing this annotation: if its callback is preempted
and the duration of the preemption exceeds the wrap around time of the
underlying clocksource, sched clock will get out of sync.
Mark the sched_clock_timer for expiry in hard interrupt context.
Signed-off-by: Ahmed S. Darwish <a.darwish@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lkml.kernel.org/r/20200309181529.26558-1-a.darwish@linutronix.de
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 486562da59 ]
Enclose the chained handler with chained_irq_{enter,exit}(), so that the
muxed interrupts get properly acked.
This patch also fixes a reboot bug on OX820 SoC, where the jiffies timer
interrupt is never acked. The kernel waits a clock tick forever in
calibrate_delay_converge(), which leads to a boot hang.
Fixes: c41b16f8c9 ("ARM: integrator/versatile: consolidate FPGA IRQ handling code")
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20200319023448.1479701-1-mans0n@gorani.run
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e74d93e96d ]
Field bdi->io_pages added in commit 9491ae4aad ("mm: don't cap request
size based on read-ahead setting") removes unneeded split of read requests.
Stacked drivers do not call blk_queue_max_hw_sectors(). Instead they set
limits of their devices by blk_set_stacking_limits() + disk_stack_limits().
Field bio->io_pages stays zero until user set max_sectors_kb via sysfs.
This patch updates io_pages after merging limits in disk_stack_limits().
Commit c6d6e9b0f6 ("dm: do not allow readahead to limit IO size") fixed
the same problem for device-mapper devices, this one fixes MD RAIDs.
Fixes: 9491ae4aad ("mm: don't cap request size based on read-ahead setting")
Reviewed-by: Paul Menzel <pmenzel@molgen.mpg.de>
Reviewed-by: Bob Liu <bob.liu@oracle.com>
Signed-off-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 17c4a2ae15 ]
When dma_mmap_coherent() sets up a mapping to unencrypted coherent memory
under SEV encryption and sometimes under SME encryption, it will actually
set up an encrypted mapping rather than an unencrypted, causing devices
that DMAs from that memory to read encrypted contents. Fix this.
When force_dma_unencrypted() returns true, the linear kernel map of the
coherent pages have had the encryption bit explicitly cleared and the
page content is unencrypted. Make sure that any additional PTEs we set
up to these pages also have the encryption bit cleared by having
dma_pgprot() return a protection with the encryption bit cleared in this
case.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Acked-by: Tom Lendacky <thomas.lendacky@amd.com>
Link: https://lkml.kernel.org/r/20200304114527.3636-3-thomas_os@shipmail.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6db73f17c5 ]
When SEV or SME is enabled and active, vm_get_page_prot() typically
returns with the encryption bit set. This means that users of
pgprot_modify(, vm_get_page_prot()) (mprotect_fixup(), do_mmap()) end up
with a value of vma->vm_pg_prot that is not consistent with the intended
protection of the PTEs.
This is also important for fault handlers that rely on the VMA
vm_page_prot to set the page protection. Fix this by not allowing
pgprot_modify() to change the encryption bit, similar to how it's done
for PAT bits.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Dave Hansen <dave.hansen@linux.intel.com>
Acked-by: Tom Lendacky <thomas.lendacky@amd.com>
Link: https://lkml.kernel.org/r/20200304114527.3636-2-thomas_os@shipmail.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 65a691f5f8 ]
The reason for clearing boot_ec_is_ecdt in acpi_ec_add() (if a
PNP0C09 device object matching the ECDT boot EC had been found in
the namespace) was to cause acpi_ec_ecdt_start() to return early,
but since the latter does not look at boot_ec_is_ecdt any more,
acpi_ec_add() need not clear it.
Moreover, doing that may be confusing as it may cause "DSDT" to be
printed instead of "ECDT" in the EC initialization completion
message, so stop doing it.
While at it, split the EC initialization completion message into
two messages, one regarding the boot EC and another one printed
regardless of whether or not the EC at hand is the boot one.
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 72ae194704 ]
Bail out early if the xHC host needs to be reset at resume
but driver can't access xHC PCI registers.
If xhci driver already fails to reset the controller then there
is no point in attempting to free, re-initialize, re-allocate and
re-start the host. If failure to access the host is detected later,
failing the resume, xhci interrupts will be double freed
when remove is called.
Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20200312144517.1593-2-mathias.nyman@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f7b8488bd3 ]
Commit 4791bd7d6a ("media: imx: Try colorimetry at both sink and
source pads") reworked the way that formats are set on the sink pad of
the CSI subdevice, and accidentally removed video field handling.
Restore it by defaulting to V4L2_FIELD_NONE if the field value isn't
supported, with the only two supported value being V4L2_FIELD_NONE and
V4L2_FIELD_INTERLACED.
Fixes: 4791bd7d6a ("media: imx: Try colorimetry at both sink and source pads")
Signed-off-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Reviewed-by: Rui Miguel Silva <rmfrfs@gmail.com>
Signed-off-by: Hans Verkuil <hverkuil-cisco@xs4all.nl>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit ff77042296 ]
Steps to reproduce:
BLKRESETZONE zone 0
// force EIO
pwrite(fd, buf, 4096, 4096);
[issue more IO including zone ioctls]
It will start failing randomly including IO to unrelated zones because of
->error "reuse". Trigger can be partition detection as well if test is not
run immediately which is even more entertaining.
The fix is of course to clear ->error where necessary.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alexey Dobriyan (SK hynix) <adobriyan@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9b03b71308 ]
If null_add_dev() fails then null_del_dev() is called with a NULL argument.
Make null_del_dev() handle this scenario correctly. This patch fixes the
following KASAN complaint:
null-ptr-deref in null_del_dev+0x28/0x280 [null_blk]
Read of size 8 at addr 0000000000000000 by task find/1062
Call Trace:
dump_stack+0xa5/0xe6
__kasan_report.cold+0x65/0x99
kasan_report+0x16/0x20
__asan_load8+0x58/0x90
null_del_dev+0x28/0x280 [null_blk]
nullb_group_drop_item+0x7e/0xa0 [null_blk]
client_drop_item+0x53/0x80 [configfs]
configfs_rmdir+0x395/0x4e0 [configfs]
vfs_rmdir+0xb6/0x220
do_rmdir+0x238/0x2c0
__x64_sys_unlinkat+0x75/0x90
do_syscall_64+0x6f/0x2f0
entry_SYSCALL_64_after_hwframe+0x49/0xbe
Signed-off-by: Bart Van Assche <bvanassche@acm.org>
Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Cc: Johannes Thumshirn <jth@kernel.org>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Ming Lei <ming.lei@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6ded0b61cf ]
SDEI has private events that must be registered on each CPU. When
CPUs come and go they must re-register and re-enable their private
events. Each event has flags to indicate whether this should happen
to protect against an event being registered on a CPU coming online,
while all the others are unregistering the event.
These flags are protected by the sdei_list_lock spinlock, because
the cpuhp callbacks can't take the mutex.
Hibernate needs to unregister all events, but keep the in-memory
re-register and re-enable as they are. sdei_unregister_shared()
takes the spinlock to walk the list, then calls _sdei_event_unregister()
on each shared event. _sdei_event_unregister() tries to take the
same spinlock to update re-register and re-enable. This doesn't go
so well.
Push the re-register and re-enable updates out to their callers.
sdei_unregister_shared() doesn't want these values updated, so
doesn't need to do anything.
This also fixes shared events getting lost over hibernate as this
path made them look unregistered.
Fixes: da35182724 ("firmware: arm_sdei: Add support for CPU and system power states")
Reported-by: Liguang Zhang <zhangliguang@linux.alibaba.com>
Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c50cc6dc6c ]
Some older MSM8916 Venus firmware versions also seem to indicate
support for encoding HEVC, even though they really can't.
This will lead to errors later because hfi_session_init() fails
in this case.
HEVC is already ignored for "dec_codecs", so add the same for
"enc_codecs" to make these old firmware versions work correctly.
Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
Signed-off-by: Stanimir Varbanov <stanimir.varbanov@linaro.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 36eb7dc1bd ]
imx6ul_opp_check_speed_grading is called for both i.MX6UL and i.MX6ULL.
Since the i.MX6ULL was introduced to a separate ocotp compatible node
later, it is possible that the i.MX6ULL has also dtbs with
"fsl,imx6ull-ocotp". On a system without nvmem-cell speed grade a
missing check on this node causes a driver fail without considering
the cpu speed grade.
This patch prevents unwanted cpu overclocking on i.MX6ULL with compatible
node "fsl,imx6ull-ocotp" in old dtbs without nvmem-cell speed grade.
Fixes: 2733fb0d06 ("cpufreq: imx6q: read OCOTP through nvmem for imx6ul/imx6ull")
Signed-off-by: Christoph Niedermaier <cniedermaier@dh-electronics.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 30defecb98 ]
This is an NEC remote control device shipped with the Videostrong KII Pro
tv box as well as other devices from videostrong.
Signed-off-by: Mohammad Rasim <mohammad.rasim96@gmail.com>
Signed-off-by: Sean Young <sean@mess.org>
Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 14c1fe699c ]
The interrupt is not required so use platform_irq_get_optional() to
avoid error messages like
i2c-pca-platform 22080000.i2c: IRQ index 0 not found
Signed-off-by: Chris Packham <chris.packham@alliedtelesis.co.nz>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit bcaeb886ad ]
In qlcnic_83xx_get_reset_instruction_template, the variable
of null test is bad, so correct it.
Signed-off-by: Xu Wang <vulab@iscas.ac.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 77bb2f204f ]
There is one one corner case at dma_fence_signal_locked
which will raise the NULL pointer problem just like below.
->dma_fence_signal
->dma_fence_signal_locked
->test_and_set_bit
here trigger dma_fence_release happen due to the zero of fence refcount.
->dma_fence_put
->dma_fence_release
->drm_sched_fence_release_scheduled
->call_rcu
here make the union fled “cb_list” at finished fence
to NULL because struct rcu_head contains two pointer
which is same as struct list_head cb_list
Therefore, to hold the reference of finished fence at drm_sched_process_job
to prevent the null pointer during finished fence dma_fence_signal
[ 732.912867] BUG: kernel NULL pointer dereference, address: 0000000000000008
[ 732.914815] #PF: supervisor write access in kernel mode
[ 732.915731] #PF: error_code(0x0002) - not-present page
[ 732.916621] PGD 0 P4D 0
[ 732.917072] Oops: 0002 [#1] SMP PTI
[ 732.917682] CPU: 7 PID: 0 Comm: swapper/7 Tainted: G OE 5.4.0-rc7 #1
[ 732.918980] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.8.2-0-g33fbe13 by qemu-project.org 04/01/2014
[ 732.920906] RIP: 0010:dma_fence_signal_locked+0x3e/0x100
[ 732.938569] Call Trace:
[ 732.939003] <IRQ>
[ 732.939364] dma_fence_signal+0x29/0x50
[ 732.940036] drm_sched_fence_finished+0x12/0x20 [gpu_sched]
[ 732.940996] drm_sched_process_job+0x34/0xa0 [gpu_sched]
[ 732.941910] dma_fence_signal_locked+0x85/0x100
[ 732.942692] dma_fence_signal+0x29/0x50
[ 732.943457] amdgpu_fence_process+0x99/0x120 [amdgpu]
[ 732.944393] sdma_v4_0_process_trap_irq+0x81/0xa0 [amdgpu]
v2: hold the finished fence at drm_sched_process_job instead of
amdgpu_fence_process
v3: resume the blank line
Signed-off-by: Yintian Tao <yttao@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 50e0d28d38 ]
cxgb4_ptp_fineadjtime() doesn't pass the signedness of offset delta
in FW_PTP_CMD. Fix it by passing correct sign.
Signed-off-by: Raju Rangoju <rajur@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 83a9b6f639 ]
Many systems build/test up-to-date kernels with older libcs, and
an older glibc (2.17) lacks the definition of SOL_DCCP in
/usr/include/bits/socket.h (it was added in the 4.6 timeframe).
Adding the definition to the test program avoids a compilation
failure that gets in the way of building tools/testing/selftests/net.
The test itself will work once the definition is added; either
skipping due to DCCP not being configured in the kernel under test
or passing, so there are no other more up-to-date glibc dependencies
here it seems beyond that missing definition.
Fixes: 11fb60d108 ("selftests: net: reuseport_addr_any: add DCCP")
Signed-off-by: Alan Maguire <alan.maguire@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 0da7c322f1 ]
the second input parameter of wait_for_completion_timeout should
be jiffies instead of millisecond
Signed-off-by: Luo bin <luobin9@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 33f15da216 ]
add read barrier in driver code to keep from reading other fileds
in dma memory which is writable for hw until we have verified the
memory is valid for driver
Signed-off-by: Luo bin <luobin9@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 614eaa943e ]
should disable eq irq before freeing it, must clear event queue
depth in hw before freeing relevant memory to avoid illegal
memory access and update consumer idx to avoid invalid interrupt
Signed-off-by: Luo bin <luobin9@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 96758117dc ]
it's unreliable for fw to check whether IO is stopped, so driver
wait for enough time to ensure IO process is done in hw before
freeing resources
Signed-off-by: Luo bin <luobin9@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b317538c47 ]
printk in macro vxge_debug_ll uses __VA_ARGS__ without "##" prefix,
it causes a build error when there is no variable
arguments(e.g. only fmt is specified.).
Signed-off-by: Zheng Wei <wei.zheng@vivo.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit fc191af1bb ]
Not every stmmac based platform makes use of the eth_wake_irq or eth_lpi
interrupts. Use the platform_get_irq_byname_optional variant for these
interrupts, so no error message is displayed, if they can't be found.
Rather print an information to hint something might be wrong to assist
debugging on platforms which use these interrupts.
Signed-off-by: Markus Fuchs <mklntf@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e138aa7d32 ]
Fix the interruptibility of kernel-initiated client calls so that they're
either only interruptible when they're waiting for a call slot to come
available or they're not interruptible at all. Either way, they're not
interruptible during transmission.
This should help prevent StoreData calls from being interrupted when
writeback is in progress. It doesn't, however, handle interruption during
the receive phase.
Userspace-initiated calls are still interruptable. After the signal has
been handled, sendmsg() will return the amount of data copied out of the
buffer and userspace can perform another sendmsg() call to continue
transmission.
Fixes: bc5e3a546d ("rxrpc: Use MSG_WAITALL to tell sendmsg() to temporarily ignore signals")
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 158fe66653 ]
Abstract out the calculation of there being sufficient Tx buffer space.
This is reproduced several times in the rxrpc sendmsg code.
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1bc7896e9e ]
When experimenting with bpf_send_signal() helper in our production
environment (5.2 based), we experienced a deadlock in NMI mode:
#5 [ffffc9002219f770] queued_spin_lock_slowpath at ffffffff8110be24
#6 [ffffc9002219f770] _raw_spin_lock_irqsave at ffffffff81a43012
#7 [ffffc9002219f780] try_to_wake_up at ffffffff810e7ecd
#8 [ffffc9002219f7e0] signal_wake_up_state at ffffffff810c7b55
#9 [ffffc9002219f7f0] __send_signal at ffffffff810c8602
#10 [ffffc9002219f830] do_send_sig_info at ffffffff810ca31a
#11 [ffffc9002219f868] bpf_send_signal at ffffffff8119d227
#12 [ffffc9002219f988] bpf_overflow_handler at ffffffff811d4140
#13 [ffffc9002219f9e0] __perf_event_overflow at ffffffff811d68cf
#14 [ffffc9002219fa10] perf_swevent_overflow at ffffffff811d6a09
#15 [ffffc9002219fa38] ___perf_sw_event at ffffffff811e0f47
#16 [ffffc9002219fc30] __schedule at ffffffff81a3e04d
#17 [ffffc9002219fc90] schedule at ffffffff81a3e219
#18 [ffffc9002219fca0] futex_wait_queue_me at ffffffff8113d1b9
#19 [ffffc9002219fcd8] futex_wait at ffffffff8113e529
#20 [ffffc9002219fdf0] do_futex at ffffffff8113ffbc
#21 [ffffc9002219fec0] __x64_sys_futex at ffffffff81140d1c
#22 [ffffc9002219ff38] do_syscall_64 at ffffffff81002602
#23 [ffffc9002219ff50] entry_SYSCALL_64_after_hwframe at ffffffff81c00068
The above call stack is actually very similar to an issue
reported by Commit eac9153f2b ("bpf/stackmap: Fix deadlock with
rq_lock in bpf_get_stack()") by Song Liu. The only difference is
bpf_send_signal() helper instead of bpf_get_stack() helper.
The above deadlock is triggered with a perf_sw_event.
Similar to Commit eac9153f2b, the below almost identical reproducer
used tracepoint point sched/sched_switch so the issue can be easily caught.
/* stress_test.c */
#include <stdio.h>
#include <stdlib.h>
#include <sys/mman.h>
#include <pthread.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#define THREAD_COUNT 1000
char *filename;
void *worker(void *p)
{
void *ptr;
int fd;
char *pptr;
fd = open(filename, O_RDONLY);
if (fd < 0)
return NULL;
while (1) {
struct timespec ts = {0, 1000 + rand() % 2000};
ptr = mmap(NULL, 4096 * 64, PROT_READ, MAP_PRIVATE, fd, 0);
usleep(1);
if (ptr == MAP_FAILED) {
printf("failed to mmap\n");
break;
}
munmap(ptr, 4096 * 64);
usleep(1);
pptr = malloc(1);
usleep(1);
pptr[0] = 1;
usleep(1);
free(pptr);
usleep(1);
nanosleep(&ts, NULL);
}
close(fd);
return NULL;
}
int main(int argc, char *argv[])
{
void *ptr;
int i;
pthread_t threads[THREAD_COUNT];
if (argc < 2)
return 0;
filename = argv[1];
for (i = 0; i < THREAD_COUNT; i++) {
if (pthread_create(threads + i, NULL, worker, NULL)) {
fprintf(stderr, "Error creating thread\n");
return 0;
}
}
for (i = 0; i < THREAD_COUNT; i++)
pthread_join(threads[i], NULL);
return 0;
}
and the following command:
1. run `stress_test /bin/ls` in one windown
2. hack bcc trace.py with the following change:
# --- a/tools/trace.py
# +++ b/tools/trace.py
@@ -513,6 +513,7 @@ BPF_PERF_OUTPUT(%s);
__data.tgid = __tgid;
__data.pid = __pid;
bpf_get_current_comm(&__data.comm, sizeof(__data.comm));
+ bpf_send_signal(10);
%s
%s
%s.perf_submit(%s, &__data, sizeof(__data));
3. in a different window run
./trace.py -p $(pidof stress_test) t:sched:sched_switch
The deadlock can be reproduced in our production system.
Similar to Song's fix, the fix is to delay sending signal if
irqs is disabled to avoid deadlocks involving with rq_lock.
With this change, my above stress-test in our production system
won't cause deadlock any more.
I also implemented a scale-down version of reproducer in the
selftest (a subsequent commit). With latest bpf-next,
it complains for the following potential deadlock.
[ 32.832450] -> #1 (&p->pi_lock){-.-.}:
[ 32.833100] _raw_spin_lock_irqsave+0x44/0x80
[ 32.833696] task_rq_lock+0x2c/0xa0
[ 32.834182] task_sched_runtime+0x59/0xd0
[ 32.834721] thread_group_cputime+0x250/0x270
[ 32.835304] thread_group_cputime_adjusted+0x2e/0x70
[ 32.835959] do_task_stat+0x8a7/0xb80
[ 32.836461] proc_single_show+0x51/0xb0
...
[ 32.839512] -> #0 (&(&sighand->siglock)->rlock){....}:
[ 32.840275] __lock_acquire+0x1358/0x1a20
[ 32.840826] lock_acquire+0xc7/0x1d0
[ 32.841309] _raw_spin_lock_irqsave+0x44/0x80
[ 32.841916] __lock_task_sighand+0x79/0x160
[ 32.842465] do_send_sig_info+0x35/0x90
[ 32.842977] bpf_send_signal+0xa/0x10
[ 32.843464] bpf_prog_bc13ed9e4d3163e3_send_signal_tp_sched+0x465/0x1000
[ 32.844301] trace_call_bpf+0x115/0x270
[ 32.844809] perf_trace_run_bpf_submit+0x4a/0xc0
[ 32.845411] perf_trace_sched_switch+0x10f/0x180
[ 32.846014] __schedule+0x45d/0x880
[ 32.846483] schedule+0x5f/0xd0
...
[ 32.853148] Chain exists of:
[ 32.853148] &(&sighand->siglock)->rlock --> &p->pi_lock --> &rq->lock
[ 32.853148]
[ 32.854451] Possible unsafe locking scenario:
[ 32.854451]
[ 32.855173] CPU0 CPU1
[ 32.855745] ---- ----
[ 32.856278] lock(&rq->lock);
[ 32.856671] lock(&p->pi_lock);
[ 32.857332] lock(&rq->lock);
[ 32.857999] lock(&(&sighand->siglock)->rlock);
Deadlock happens on CPU0 when it tries to acquire &sighand->siglock
but it has been held by CPU1 and CPU1 tries to grab &rq->lock
and cannot get it.
This is not exactly the callstack in our production environment,
but sympotom is similar and both locks are using spin_lock_irqsave()
to acquire the lock, and both involves rq_lock. The fix to delay
sending signal when irq is disabled also fixed this issue.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Cc: Song Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20200304191104.2796501-1-yhs@fb.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit b46b2b7ba6 ]
Commit cd28d1d6e5 ("net: phy: at803x: Disable phy delay for RGMII mode")
caused a regression for dm814x boards where NFSroot would no longer work.
Let's fix the issue by configuring "rgmii-id" mode as internal delays are
needed that is no longer the case with "rgmii" mode.
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit a43ab30dcd ]
When doing a 16-bit read that returns data in the MSB byte, the
RSB_DATA register will keep the MSB byte unchanged when doing
the following 8-bit read. sunxi_rsb_read() will then return
a result that contains high byte from 16-bit read mixed with
the 8-bit result.
The consequence is that after this happens the PMIC's regmap will
look like this: (0x33 is the high byte from the 16-bit read)
% cat /sys/kernel/debug/regmap/sunxi-rsb-3a3/registers
00: 33
01: 33
02: 33
03: 33
04: 33
05: 33
06: 33
07: 33
08: 33
09: 33
0a: 33
0b: 33
0c: 33
0d: 33
0e: 33
[snip]
Fix this by masking the result of the read with the correct mask
based on the size of the read. There are no 16-bit users in the
mainline kernel, so this doesn't need to get into the stable tree.
Signed-off-by: Ondrej Jirman <megous@megous.com>
Acked-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Signed-off-by: Sasha Levin <sashal@kernel.org>
