... into returning a positive to path_openat(), which would interpret that
as "symlink had been encountered" and proceed to corrupt memory, etc.
It can only happen due to a bug in some ->open() instance or in some LSM
hook, etc., so we report any such event *and* make sure it doesn't trick
us into further unpleasantness.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In __request_region, if a conflict with a BUSY and MUXED resource is
detected, then the caller goes to sleep and waits for the resource to be
released. A pointer on the conflicting resource is kept. At wake-up
this pointer is used as a parent to retry to request the region.
A first problem is that this pointer might well be invalid (if for
example the conflicting resource have already been freed). Another
problem is that the next call to __request_region() fails to detect a
remaining conflict. The previously conflicting resource is passed as a
parameter and __request_region() will look for a conflict among the
children of this resource and not at the resource itself. It is likely
to succeed anyway, even if there is still a conflict.
Instead, the parent of the conflicting resource should be passed to
__request_region().
As a fix, this patch doesn't update the parent resource pointer in the
case we have to wait for a muxed region right after.
Reported-and-tested-by: Vincent Pelletier <plr.vincent@gmail.com> Signed-off-by: Simon Guinot <simon.guinot@sequanux.org> Tested-by: Vincent Donnefort <vdonnefort@gmail.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The qword_get() function NUL-terminates its output buffer. If the input
string is in hex format \xXXXX... and the same length as the output
buffer, there is an off-by-one:
This patch ensures the NUL terminator doesn't fall outside the output
buffer.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: J. Bruce Fields <bfields@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The ftrace:function event is only displayed for parsing the function tracer
data. It is not used to enable function tracing, and does not include an
"enable" file in its event directory.
Originally, this event was kept separate from other events because it did
not have a ->reg parameter. But perf added a "reg" parameter for its use
which caused issues, because it made the event available to functions where
it was not compatible for.
Commit 9b63776fa3ca9 "tracing: Do not enable function event with enable"
added a TRACE_EVENT_FL_IGNORE_ENABLE flag that prevented the function event
from being enabled by normal trace events. But this commit missed keeping
the function event from being displayed by the "available_events" directory,
which is used to show what events can be enabled by set_event.
One documented way to enable all events is to:
cat available_events > set_event
But because the function event is displayed in the available_events, this
now causes an INVALID error:
cat: write error: Invalid argument
Reported-by: Chunyu Hu <chuhu@redhat.com> Fixes: 9b63776fa3ca9 "tracing: Do not enable function event with enable" Signed-off-by: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
During error recovery, the device could be removed as part of the
partial hotplug. The criterion used to come with partial hotplug
is: if the device driver provides error_detected(), slot_reset()
and resume() callbacks, it's immune from hotplug. Otherwise,
it's going to experience partial hotplug during EEH recovery. But
the criterion isn't correct enough: mlx4_core driver for Mellanox
adapters provides error_detected(), slot_reset() callbacks, but
resume() isn't there. Those Mellanox adapters won't be to involved
in the partial hotplug.
This fixes the criterion to a practical one: adpater with driver
that provides error_detected(), slot_reset() will be immune from
partial hotplug. resume() isn't mandatory.
Fixes: f2da4ccf ("powerpc/eeh: More relaxed hotplug criterion") Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit e8dd2d2d641c ("Silence compiler warning in arch/x86/kvm/emulate.c",
2015-09-06) broke boot of the Hurd. The bug is that the "default:"
case actually could modify "la", but after the patch this change is
not reflected in *linear.
The bug is visible whenever a non-zero segment base causes the linear
address to wrap around the 4GB mark.
Sometimes when setting a breakpoint a process doesn't stop on it.
This is because the debug registers are not loaded correctly on
VCPU load.
The following simple reproducer from Oleg Nesterov tries using debug
registers in two threads. To see the bug, run a 2-VCPU guest with
"taskset -c 0" and run "./bp 0 1" inside the guest.
When we allocate bitmaps in vgic_vcpu_init_maps, we divide the number of
bits we need by 8 to figure out how many bytes to allocate. However,
bitmap elements are always accessed as unsigned longs, and if we didn't
happen to allocate a size such that size % sizeof(unsigned long) == 0,
bitmap accesses may go past the end of the allocation.
When using KASAN (which does byte-granular access checks), this results
in a continuous stream of BUGs whenever these bitmaps are accessed:
=============================================================================
BUG kmalloc-128 (Tainted: G B ): kasan: bad access detected
-----------------------------------------------------------------------------
In async_pf we try to allocate with NOWAIT to get an element quickly
or fail. This code also handle failures gracefully. Lets silence
potential page allocation failures under load.
The existing msi-map code is fine for shifting the entire RID space
upwards, but attempting finer-grained remapping reveals a bug. It turns
out that we are mistakenly treating the msi-base part as an offset, not
as a new base to remap onto, so things get squiffy when rid-base is
nonzero. Fix this, and at the same time add a sanity check against
having msi-map-mask clash with a nonzero rid-base, as that's another
thing one can easily get wrong.
Signed-off-by: Robin Murphy <robin.murphy@arm.com> Reviewed-by: Marc Zyngier <marc.zyngier@arm.com> Tested-by: Stuart Yoder <stuart.yoder@nxp.com> Acked-by: Mark Rutland <mark.rutland@arm.com> Acked-by: David Daney <david.daney@cavium.com> Signed-off-by: Rob Herring <robh@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In the case where d_add_unique() finds an appropriate alias to use it will
have already incremented the reference count. An additional dget() to swap
the open context's dentry is unnecessary and will leak a reference.
Signed-off-by: Benjamin Coddington <bcodding@redhat.com> Fixes: 275bb307865a3 ("NFSv4: Move dentry instantiation into the NFSv4-...") Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 35dc248383bbab0a7203fca4d722875bc81ef091 introduced a check for
current->mm to see if we have a user space context and only copies data
if we do. Now if an IO gets interrupted by a signal data isn't copied
into user space any more (as we don't have a user space context) but
user space isn't notified about it.
This patch modifies the behaviour to return -EINTR from bio_uncopy_user()
to notify userland that a signal has interrupted the syscall, otherwise
it could lead to a situation where the caller may get a buffer with
no data returned.
This can be reproduced by issuing SG_IO ioctl()s in one thread while
constantly sending signals to it.
Fixes: 35dc248 [SCSI] sg: Fix user memory corruption when SG_IO is interrupted by a signal Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de> Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: Jens Axboe <axboe@fb.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Starting from Intel Sunrisepoint (Skylake PCH) the iTCO watchdog
resources have been moved to reside under the i801 SMBus host
controller whereas previously they were under the LPC device.
This patch adds Intel lewisburg SMBus support for iTCO device.
It allows to load watchdog dynamically when the hardware is
present.
Fixes: cdc5a3110e7c ("i2c: i801: add Intel Lewisburg device IDs") Reviewed-by: Jean Delvare <jdelvare@suse.de> Signed-off-by: Alexandra Yates <alexandra.yates@linux.intel.com> Signed-off-by: Wolfram Sang <wsa@the-dreams.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If phy_pm_runtime_get_sync failed but we already
enable regulator, current code return directly without
doing regulator_disable. This patch fix this problem
and cleanup err handle of phy_power_on to be more readable.
Fixes: 3be88125d85d ("phy: core: Support regulator ...") Cc: Roger Quadros <rogerq@ti.com> Cc: Axel Lin <axel.lin@ingics.com> Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com> Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If cgroup writeback is in use, an inode is associated with a cgroup
for writeback. If the inode's main dirtier changes to another cgroup,
the association gets updated asynchronously. Nothing was pinning the
superblock while such switches are in progress and superblock could go
away while async switching is pending or in progress leading to
crashes like the following.
There are three subsystem callbacks in css shutdown path -
css_offline(), css_released() and css_free(). Except for
css_released(), cgroup core didn't guarantee the order of invocation.
css_offline() or css_free() could be called on a parent css before its
children. This behavior is unexpected and led to bugs in cpu and
memory controller.
This patch updates offline path so that a parent css is never offlined
before its children. Each css keeps online_cnt which reaches zero iff
itself and all its children are offline and offline_css() is invoked
only after online_cnt reaches zero.
This fixes the memory controller bug and allows the fix for cpu
controller.
If "cpuset.memory_migrate" is set, when a process is moved from one
cpuset to another with a different memory node mask, pages in used by
the process are migrated to the new set of nodes. This was performed
synchronously in the ->attach() callback, which is synchronized
against process management. Recently, the synchronization was changed
from per-process rwsem to global percpu rwsem for simplicity and
optimization.
Combined with the synchronous mm migration, this led to deadlocks
because mm migration could schedule a work item which may in turn try
to create a new worker blocking on the process management lock held
from cgroup process migration path.
This heavy an operation shouldn't be performed synchronously from that
deep inside cgroup migration in the first place. This patch punts the
actual migration to an ordered workqueue and updates cgroup process
migration and cpuset config update paths to flush the workqueue after
all locks are released. This way, the operations still seem
synchronous to userland without entangling mm migration with process
management synchronization. CPU hotplug can also invoke mm migration
but there's no reason for it to wait for mm migrations and thus
doesn't synchronize against their completions.
A Root Port's AER structure (rpc) contains a queue of events. aer_irq()
enqueues AER status information and schedules aer_isr() to dequeue and
process it. When we remove a device, aer_remove() waits for the queue to
be empty, then frees the rpc struct.
But aer_isr() references the rpc struct after dequeueing and possibly
emptying the queue, which can cause a use-after-free error as in the
following scenario with two threads, aer_isr() on the left and a
concurrent aer_remove() on the right:
Thread A Thread B
-------- --------
aer_irq():
rpc->prod_idx++
aer_remove():
wait_event(rpc->prod_idx == rpc->cons_idx)
# now blocked until queue becomes empty
aer_isr(): # ...
rpc->cons_idx++ # unblocked because queue is now empty
... kfree(rpc)
mutex_unlock(&rpc->rpc_mutex)
To prevent this problem, use flush_work() to wait until the last scheduled
instance of aer_isr() has completed before freeing the rpc struct in
aer_remove().
I reproduced this use-after-free by flashing a device FPGA and
re-enumerating the bus to find the new device. With SLUB debug, this
crashes with 0x6b bytes (POISON_FREE, the use-after-free magic number) in
GPR25:
ARConnect/MCIP Inter-Core-Interrupt module can't send interrupt to
local core. So use core intc capability to trigger software
interrupt to self, using an unsued IRQ #21.
This showed up as csd deadlock with LTP trace_sched on a dual core
system. This test acts as scheduler fuzzer, triggering all sorts of
schedulting activity. Trouble starts with IPI to self, which doesn't get
delivered (effectively lost due to H/w capability), but the msg intended
to be sent remain enqueued in per-cpu @ipi_data.
All subsequent IPIs to this core from other cores get elided due to the
IPI coalescing optimization in ipi_send_msg_one() where a pending msg
implies an IPI already sent and assumes other core is yet to ack it.
After the elided IPI, other core simply goes into csd_lock_wait()
but never comes out as this core never sees the interrupt.
Returning to delay slot, riding an interrupti, had one loose end.
AUX_USER_SP used for restoring user mode SP upon RTIE was not being
setup from orig task's saved value, causing task to use wrong SP,
leading to ProtV errors.
The reason being:
- INTERRUPT_EPILOGUE returns to a kernel trampoline, thus not expected to restore it
- EXCEPTION_EPILOGUE is not used at all
Fix that by restoring AUX_USER_SP explicitly in the trampoline.
This was broken in the original workaround, but the error scenarios got
reduced considerably since v3.14 due to following:
1. The Linuxthreads.old based userspace at the time caused many more
exceptions in delay slot than the current NPTL based one.
Infact with current userspace the error doesn't happen at all.
2. Return from interrupt (delay slot or otherwise) doesn't get exercised much
after commit 4de0e52867d8 ("Really Re-enable interrupts to avoid deadlocks")
since IRQ_ACTIVE.active being clear means most returns are as if from pure
kernel (even for active interrupts)
Infact the issue only happened in an experimental branch where I was tinkering with
reverted 4de0e52867d8
Fixes: 4255b07f2c9c ("ARCv2: STAR 9000793984: Handle return from intr to Delay Slot") Signed-off-by: Vineet Gupta <vgupta@synopsys.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The bulk of ATA host state machine is implemented by
ata_sff_hsm_move(). The function is called from either the interrupt
handler or, if polling, a work item. Unlike from the interrupt path,
the polling path calls the function without holding the host lock and
ata_sff_hsm_move() selectively grabs the lock.
This is completely broken. If an IRQ triggers while polling is in
progress, the two can easily race and end up accessing the hardware
and updating state machine state at the same time. This can put the
state machine in an illegal state and lead to a crash like the
following.
Fix it by ensuring that the polling path is holding the host lock
before entering ata_sff_hsm_move() so that all hardware accesses and
state updates are performed under the host lock.
Commit 84e32a0 ("qla2xxx: Use pci_enable_msix_range() instead of
pci_enable_msix()") introduced a regression when target mode is enabled.
In qla24xx_enable_msix(), ha->max_rsp_queues was incorrectly set
to a value higher than the number of response queues allocated causing
an invalid dereference. Specifically here in qla2x00_init_rings():
*rsp->in_ptr = 0;
Add additional check to make sure the pointer is valid. following
call stack will be seen
The non-DT platform that uses this driver (actually the AVR32) was taking a bad
branch for determining if the IP would use gpio for CS.
Adding the presence of DT as a condition fixes this issue.
Fixes: 4820303480a1 ("spi: atmel: add support for the internal chip-select of the spi controller") Reported-by: Mans Rullgard <mans@mansr.com> Signed-off-by: Cyrille Pitchen <cyrille.pitchen@atmel.com>
[nicolas.ferre@atmel.com: extract from ml discussion] Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com> Tested-by: Mans Rullgard <mans@mansr.com> Signed-off-by: Mark Brown <broonie@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes a race between setting of SCF_SEND_DELAYED_TAS
in transport_send_task_abort(), and check of the same bit in
transport_check_aborted_status().
It adds a __transport_check_aborted_status() version that is
used by target_execute_cmd() when se_cmd->t_state_lock is
held, and a transport_check_aborted_status() wrapper for
all other existing callers.
Also, it handles the case where the check happens before
transport_send_task_abort() gets called. For this, go
ahead and set SCF_SEND_DELAYED_TAS early when necessary,
and have transport_send_task_abort() send the abort.
Cc: Quinn Tran <quinn.tran@qlogic.com> Cc: Himanshu Madhani <himanshu.madhani@qlogic.com> Cc: Sagi Grimberg <sagig@mellanox.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.de> Cc: Andy Grover <agrover@redhat.com> Cc: Mike Christie <mchristi@redhat.com> Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To address the bug where fabric driver level shutdown
of se_cmd occurs at the same time when TMR CMD_T_ABORTED
is happening resulting in a -1 ->cmd_kref, this patch
adds a CMD_T_FABRIC_STOP bit that is used to determine
when TMR + driver I_T nexus shutdown is happening
concurrently.
It changes target_sess_cmd_list_set_waiting() to obtain
se_cmd->cmd_kref + set CMD_T_FABRIC_STOP, and drop local
reference in target_wait_for_sess_cmds() and invoke extra
target_put_sess_cmd() during Task Aborted Status (TAS)
when necessary.
Also, it adds a new target_wait_free_cmd() wrapper around
transport_wait_for_tasks() for the special case within
transport_generic_free_cmd() to set CMD_T_FABRIC_STOP,
and is now aware of CMD_T_ABORTED + CMD_T_TAS status
bits to know when an extra transport_put_cmd() during
TAS is required.
Note transport_generic_free_cmd() is expected to block on
cmd->cmd_wait_comp in order to follow what iscsi-target
expects during iscsi_conn context se_cmd shutdown.
Cc: Quinn Tran <quinn.tran@qlogic.com> Cc: Himanshu Madhani <himanshu.madhani@qlogic.com> Cc: Sagi Grimberg <sagig@mellanox.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.de> Cc: Andy Grover <agrover@redhat.com> Cc: Mike Christie <mchristi@redhat.com> Signed-off-by: Nicholas Bellinger <nab@daterainc.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes a bug in TMR task aborted status (TAS)
handling when multiple sessions are connected to the
same target WWPN endpoint and se_node_acl descriptor,
resulting in TASK_ABORTED status to not be generated
for aborted se_cmds on the remote port.
This is due to core_tmr_handle_tas_abort() incorrectly
comparing se_node_acl instead of se_session, for which
the multi-session case is expected to be sharing the
same se_node_acl.
Instead, go ahead and update core_tmr_handle_tas_abort()
to compare tmr_sess + cmd->se_sess in order to determine
if the LUN_RESET was received on a different I_T nexus,
and TASK_ABORTED status response needs to be generated.
Reviewed-by: Christoph Hellwig <hch@lst.de> Cc: Quinn Tran <quinn.tran@qlogic.com> Cc: Himanshu Madhani <himanshu.madhani@qlogic.com> Cc: Sagi Grimberg <sagig@mellanox.com> Cc: Hannes Reinecke <hare@suse.de> Cc: Andy Grover <agrover@redhat.com> Cc: Mike Christie <mchristi@redhat.com> Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes a NULL pointer se_cmd->cmd_kref < 0
refcount bug during TMR LUN_RESET with active TMRs,
triggered during se_cmd + se_tmr_req descriptor
shutdown + release via core_tmr_drain_tmr_list().
To address this bug, go ahead and obtain a local
kref_get_unless_zero(&se_cmd->cmd_kref) for active I/O
to set CMD_T_ABORTED, and transport_wait_for_tasks()
followed by the final target_put_sess_cmd() to drop
the local ->cmd_kref.
Also add two new checks within target_tmr_work() to
avoid CMD_T_ABORTED -> TFO->queue_tm_rsp() callbacks
ahead of invoking the backend -> fabric put in
transport_cmd_check_stop_to_fabric().
For good measure, also change core_tmr_release_req()
to use list_del_init() ahead of se_tmr_req memory
free.
Reviewed-by: Quinn Tran <quinn.tran@qlogic.com> Cc: Himanshu Madhani <himanshu.madhani@qlogic.com> Cc: Sagi Grimberg <sagig@mellanox.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.de> Cc: Andy Grover <agrover@redhat.com> Cc: Mike Christie <mchristi@redhat.com> Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch fixes a NULL pointer se_cmd->cmd_kref < 0
refcount bug during TMR LUN_RESET with active se_cmd
I/O, that can be triggered during se_cmd descriptor
shutdown + release via core_tmr_drain_state_list() code.
To address this bug, add common __target_check_io_state()
helper for ABORT_TASK + LUN_RESET w/ CMD_T_COMPLETE
checking, and set CMD_T_ABORTED + obtain ->cmd_kref for
both cases ahead of last target_put_sess_cmd() after
TFO->aborted_task() -> transport_cmd_finish_abort()
callback has completed.
It also introduces SCF_ACK_KREF to determine when
transport_cmd_finish_abort() needs to drop the second
extra reference, ahead of calling target_put_sess_cmd()
for the final kref_put(&se_cmd->cmd_kref).
It also updates transport_cmd_check_stop() to avoid
holding se_cmd->t_state_lock while dropping se_cmd
device state via target_remove_from_state_list(), now
that core_tmr_drain_state_list() is holding the
se_device lock while checking se_cmd state from
within TMR logic.
Finally, move transport_put_cmd() release of SGL +
TMR + extended CDB memory into target_free_cmd_mem()
in order to avoid potential resource leaks in TMR
ABORT_TASK + LUN_RESET code-paths. Also update
target_release_cmd_kref() accordingly.
Reviewed-by: Quinn Tran <quinn.tran@qlogic.com> Cc: Himanshu Madhani <himanshu.madhani@qlogic.com> Cc: Sagi Grimberg <sagig@mellanox.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.de> Cc: Andy Grover <agrover@redhat.com> Cc: Mike Christie <mchristi@redhat.com> Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
After login to the desktop on Dell Inspiron 3162,
there's a very loud background noise comes from the builtin speaker.
The noise does not go away even if the speaker is muted.
The noise disappears after using the aamix fixup.
Codec: Realtek ALC3234
Address: 0
AFG Function Id: 0x1 (unsol 1)
Vendor Id: 0x10ec0255
Subsystem Id: 0x10280725
Revision Id: 0x100002
No Modem Function Group found
Some Skylake machines show the codec probe errors in certain
situations, e.g. HP Z240 desktop fails to probe the onboard Realtek
codec at reloading the snd-hda-intel module like:
snd_hda_intel 0000:00:1f.3: spurious response 0x200:0x2, last cmd=0x000000
snd_hda_intel 0000:00:1f.3: azx_get_response timeout, switching to polling mode: lastcmd=0x000f0000
snd_hda_intel 0000:00:1f.3: No response from codec, disabling MSI: last cmd=0x000f0000
snd_hda_intel 0000:00:1f.3: Codec #0 probe error; disabling it...
hdaudio hdaudioC0D2: no AFG or MFG node found
snd_hda_intel 0000:00:1f.3: no codecs initialized
Also, HP G470 G3 suffers from the similar problem, as reported in
bugzilla below. On this machine, the codec probe error appears even
at a fresh boot.
As Libin suggested, the same workaround used for Broxton in the commit
[6639484ddaf6: ALSA: hda - disable dynamic clock gating on Broxton
before reset] can be applied for Skylake in order to fix this problem.
The Intel HW team also confirmed that this is needed for SKL.
This patch makes the workaround applied to both SKL and BXT
platforms. The referred macros are moved and one superfluous macro
(IS_BROXTON()) is another one (IS_BXT()) as well.
Workqueue used to implicity guarantee that work items queued without
explicit CPU specified are put on the local CPU. Recent changes in
timer broke the guarantee and led to vmstat breakage which was fixed
by 176bed1de5bf ("vmstat: explicitly schedule per-cpu work on the CPU
we need it to run on").
vmstat is the most likely to expose the issue and it's quite possible
that there are other similar problems which are a lot more difficult
to trigger. As a preventive measure, 874bbfe600a6 ("workqueue: make
sure delayed work run in local cpu") was applied to restore the local
CPU guarnatee. Unfortunately, the change exposed a bug in timer code
which got fixed by 22b886dd1018 ("timers: Use proper base migration in
add_timer_on()"). Due to code restructuring, the commit couldn't be
backported beyond certain point and stable kernels which only had 874bbfe600a6 started crashing.
The local CPU guarantee was accidental more than anything else and we
want to get rid of it anyway. As, with the vmstat case fixed, 874bbfe600a6 is causing more problems than it's fixing, it has been
decided to take the chance and officially break the guarantee by
reverting the commit. A debug feature will be added to force foreign
CPU assignment to expose cases relying on the guarantee and fixes for
the individual cases will be backported to stable as necessary.
Signed-off-by: Tejun Heo <tj@kernel.org> Fixes: 874bbfe600a6 ("workqueue: make sure delayed work run in local cpu") Link: http://lkml.kernel.org/g/20160120211926.GJ10810@quack.suse.cz Cc: Mike Galbraith <umgwanakikbuti@gmail.com> Cc: Henrique de Moraes Holschuh <hmh@hmh.eng.br> Cc: Daniel Bilik <daniel.bilik@neosystem.cz> Cc: Jan Kara <jack@suse.cz> Cc: Shaohua Li <shli@fb.com> Cc: Sasha Levin <sasha.levin@oracle.com> Cc: Ben Hutchings <ben@decadent.org.uk> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Daniel Bilik <daniel.bilik@neosystem.cz> Cc: Jiri Slaby <jslaby@suse.cz> Cc: Michal Hocko <mhocko@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When looking up the pool_workqueue to use for an unbound workqueue,
workqueue assumes that the target CPU is always bound to a valid NUMA
node. However, currently, when a CPU goes offline, the mapping is
destroyed and cpu_to_node() returns NUMA_NO_NODE.
This has always been broken but hasn't triggered often enough before 874bbfe600a6 ("workqueue: make sure delayed work run in local cpu").
After the commit, workqueue forcifully assigns the local CPU for
delayed work items without explicit target CPU to fix a different
issue. This widens the window where CPU can go offline while a
delayed work item is pending causing delayed work items dispatched
with target CPU set to an already offlined CPU. The resulting
NUMA_NO_NODE mapping makes workqueue try to queue the work item on a
NULL pool_workqueue and thus crash.
While 874bbfe600a6 has been reverted for a different reason making the
bug less visible again, it can still happen. Fix it by mapping
NUMA_NO_NODE to the default pool_workqueue from unbound_pwq_by_node().
This is a temporary workaround. The long term solution is keeping CPU
-> NODE mapping stable across CPU off/online cycles which is being
worked on.
During a sw scan ieee80211_iface_work ignores work items for all vifs.
However after the scan complete work is requeued only for STA, ADHOC
and MESH iftypes.
This occasionally results in event processing getting delayed/not
processed for iftype AP when it coexists with a STA. This can result
in data halt and eventually disconnection on the AP interface.
The code within wait_event_interruptible() is called with
!TASK_RUNNING, so mustn't call any functions that can sleep,
like mutex_lock().
Since we re-check the list_empty() in a loop after the wait,
it's safe to simply use list_empty() without locking.
This bug has existed forever, but was only discovered now
because all userspace implementations, including the default
'rfkill' tool, use poll() or select() to get a readable fd
before attempting to read.
commit 0ff53d096422 sets the next tick interrupt to the last jiffies update,
i.e. in the past, because the forward operation is invoked before the set
operation. There is no resulting damage (yet), but we get an extra pointless
tick interrupt.
Revert the order so we get the next tick interrupt in the future.
Fixes: commit 0ff53d096422 "tick: sched: Force tick interrupt and get rid of softirq magic" Signed-off-by: Wanpeng Li <wanpeng.li@hotmail.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Link: http://lkml.kernel.org/r/1453893967-3458-1-git-send-email-wanpeng.li@hotmail.com Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This commit merged interval and overall counters reading into single
read_counters function.
The old interval code cleaned the stddev data for some reason (it's
never displayed in interval mode) and the mentioned commit kept on
cleaning the stddev data in merged function, which resulted in the
stddev not being displayed.
Removing the wrong stddev data cleanup init_stats call.
Reported-and-Tested-by: Mel Gorman <mgorman@techsingularity.net> Signed-off-by: Jiri Olsa <jolsa@kernel.org> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Fixes: 106a94a0f8c2 ("perf stat: Introduce read_counters function") Link: http://lkml.kernel.org/r/1453290995-18485-4-git-send-email-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit f79b0d9c223c ("staging: speakup: Fixed warning <linux/serial.h>
instead of <asm/serial.h>") broke the port information in the speakup
driver: SERIAL_PORT_DFNS only gets defined if asm/serial.h is included,
and no other header includes asm/serial.h.
We here make sure serialio.c does get the arch-specific definition of
SERIAL_PORT_DFNS from asm/serial.h, if any.
Along the way, this makes sure that we do have information for the
requested serial port number (index)
Fixes: f79b0d9c223c ("staging: speakup: Fixed warning <linux/serial.h> instead of <asm/serial.h>") Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit ca369d51b3e1 ("block/sd: Fix device-imposed transfer length
limits") accidentally switched optimal I/O size reporting from bytes to
block layer sectors.
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Reported-by: Christian Borntraeger <borntraeger@de.ibm.com> Tested-by: Christian Borntraeger <borntraeger@de.ibm.com> Fixes: ca369d51b3e1649be4a72addd6d6a168cfb3f537 Reviewed-by: James E.J. Bottomley <James.Bottomley@HansenPartnership.com> Reviewed-by: Ewan D. Milne <emilne@redhat.com> Reviewed-by: Matthew R. Ochs <mrochs@linux.vnet.ibm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit d15f9d694b77 ("libceph: check data_len in ->alloc_msg()")
mistakenly bumped the log level on the "tid %llu unknown, skipping"
message. Turn it back into a dout() - stray replies are perfectly
normal when OSDs flap, crash, get killed for testing purposes, etc.
The contract between try_read() and try_write() is that when called
each processes as much data as possible. When instructed by osd_client
to skip a message, try_read() is violating this contract by returning
after receiving and discarding a single message instead of checking for
more. try_write() then gets a chance to write out more requests,
generating more replies/skips for try_read() to handle, forcing the
messenger into a starvation loop.
Reported-by: Varada Kari <Varada.Kari@sandisk.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com> Tested-by: Varada Kari <Varada.Kari@sandisk.com> Reviewed-by: Alex Elder <elder@linaro.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There are a number of problems with revoking a "was sending" message:
(1) We never make any attempt to revoke data - only kvecs contibute to
con->out_skip. However, once the header (envelope) is written to the
socket, our peer learns data_len and sets itself to expect at least
data_len bytes to follow front or front+middle. If ceph_msg_revoke()
is called while the messenger is sending message's data portion,
anything we send after that call is counted by the OSD towards the now
revoked message's data portion. The effects vary, the most common one
is the eventual hang - higher layers get stuck waiting for the reply to
the message that was sent out after ceph_msg_revoke() returned and
treated by the OSD as a bunch of data bytes. This is what Matt ran
into.
(2) Flat out zeroing con->out_kvec_bytes worth of bytes to handle kvecs
is wrong. If ceph_msg_revoke() is called before the tag is sent out or
while the messenger is sending the header, we will get a connection
reset, either due to a bad tag (0 is not a valid tag) or a bad header
CRC, which kind of defeats the purpose of revoke. Currently the kernel
client refuses to work with header CRCs disabled, but that will likely
change in the future, making this even worse.
(3) con->out_skip is not reset on connection reset, leading to one or
more spurious connection resets if we happen to get a real one between
con->out_skip is set in ceph_msg_revoke() and before it's cleared in
write_partial_skip().
Fixing (1) and (3) is trivial. The idea behind fixing (2) is to never
zero the tag or the header, i.e. send out tag+header regardless of when
ceph_msg_revoke() is called. That way the header is always correct, no
unnecessary resets are induced and revoke stands ready for disabled
CRCs. Since ceph_msg_revoke() rips out con->out_msg, introduce a new
"message out temp" and copy the header into it before sending.
Reported-by: Matt Conner <matt.conner@keepertech.com> Signed-off-by: Ilya Dryomov <idryomov@gmail.com> Tested-by: Matt Conner <matt.conner@keepertech.com> Reviewed-by: Sage Weil <sage@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Before this patch, a process with some permissive seccomp filter
that was applied by root without NO_NEW_PRIVS was able to add
more filters to itself without setting NO_NEW_PRIVS by setting
the new filter from a throwaway thread with NO_NEW_PRIVS.
There is a race discovered by Juri, where we are able to:
- create and read a sysfs file before policy->governor_data is being set
to a non NULL value.
OR
- set policy->governor_data to NULL, and reading a file before being
destroyed.
There are two definitions of pxa_cpufreq_change_voltage, with slightly
different prototypes after one of them had its argument marked 'const'.
Now the other one (for !CONFIG_REGULATOR) produces a harmless warning:
drivers/cpufreq/pxa2xx-cpufreq.c: In function 'pxa_set_target':
drivers/cpufreq/pxa2xx-cpufreq.c:291:36: warning: passing argument 1 of 'pxa_cpufreq_change_voltage' discards 'const' qualifier from pointer target type [-Wdiscarded-qualifiers]
ret = pxa_cpufreq_change_voltage(&pxa_freq_settings[idx]);
^
drivers/cpufreq/pxa2xx-cpufreq.c:205:12: note: expected 'struct pxa_freqs *' but argument is of type 'const struct pxa_freqs *'
static int pxa_cpufreq_change_voltage(struct pxa_freqs *pxa_freq)
^
This changes the prototype in the same way as the other, which
avoids the warning.
Fixes: 03c229906311 (cpufreq: pxa: make pxa_freqs arrays const) Signed-off-by: Arnd Bergmann <arnd@arndb.de> Acked-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Thermal hook gpio_fan_get_cur_state is only interested in knowing
the current speed index that was setup in the system, this is
already available as part of fan_data->speed_index which is always
set by set_fan_speed. Using get_fan_speed_index is useful when we
have no idea about the fan speed configuration (for example during
fan_ctrl_init).
When thermal framework invokes
gpio_fan_get_cur_state=>get_fan_speed_index via gpio_fan_get_cur_state
especially in a polled configuration for thermal governor, we
basically hog the i2c interface to the extent that other functions
fail to get any traffic out :(.
Instead, just provide the last state set in the driver - since the gpio
fan driver is responsible for the fan state immaterial of override, the
fan_data->speed_index should accurately reflect the state.
Since Linux 4.0 the CPU fan speed is going up and down on Dell Studio
XPS 8000 and 8100 for unknown reasons. The 8100 was already
blacklisted in commit a4b45b25f18d ("hwmon: (dell-smm) Blacklist
Dell Studio XPS 8100"). This patch blacklists the XPS 8000.
Without further debugging on the affected machine, it is not possible
to find the problem. For more details see
https://bugzilla.kernel.org/show_bug.cgi?id=100121
When a new cooling device is registered, we need to update the
thermal zone to set the new registered cooling device to a proper
state.
This fixes a problem that the system is cool, while the fan devices
are left running on full speed after boot, if fan device is registered
after thermal zone device.
Here is the history of why current patch looks like this:
https://patchwork.kernel.org/patch/7273041/
Current thermal code does not handle system sleep well because
1. the cooling device cooling state may be changed during suspend
2. the previous temperature reading becomes invalid after resumed because
it is got before system sleep
3. updating thermal zone device during suspending/resuming
is wrong because some devices may have already been suspended
or may have not been resumed.
Thus, the proper way to do this is to cancel all thermal zone
device update requirements during suspend/resume, and after all
the devices have been resumed, reset and update every registered
thermal zone devices.
This also fixes a regression introduced by:
Commit 19593a1fb1f6 ("ACPI / fan: convert to platform driver")
Because, with above commit applied, all the fan devices are attached
to the acpi_general_pm_domain, and they are turned on by the pm_domain
automatically after resume, without the awareness of thermal core.
After thermal zone device registered, as we have not read any
temperature before, thus tz->temperature should not be 0,
which actually means 0C, and thermal trend is not available.
In this case, we need specially handling for the first
thermal_zone_device_update().
Both thermal core framework and step_wise governor is
enhanced to handle this. And since the step_wise governor
is the only one that uses trends, so it's the only thermal
governor that needs to be updated.
The current code is problematic when the QP creation and ipoib is used to
support NFS and NFS desires to do IO for paging purposes. In that case, the
GFP_KERNEL allocation in qib_qp.c causes a deadlock in tight memory
situations.
This fix adds support to create queue pair with GFP_NOIO flag for connected
mode only to cleanly fail the create queue pair in those situations.
ib_send_cm_drep() calls cm_enter_timewait() while holding a spinlock
that can be locked from inside an interrupt handler. Hence do not
enable interrupts inside cm_enter_timewait() if called with interrupts
disabled.
This patch fixes e.g. the following deadlock: Acked-by: Erez Shitrit <erezsh@mellanox.com>
=================================
[ INFO: inconsistent lock state ]
4.4.0-rc7+ #1 Tainted: G E
---------------------------------
inconsistent {HARDIRQ-ON-W} -> {IN-HARDIRQ-W} usage.
swapper/8/0 [HC1[1]:SC0[0]:HE0:SE1] takes:
(&(&cm_id_priv->lock)->rlock){?.+...}, at: [<ffffffffa036eec4>] cm_establish+0x
74/0x1b0 [ib_cm]
{HARDIRQ-ON-W} state was registered at:
[<ffffffff810a3c11>] mark_held_locks+0x71/0x90
[<ffffffff810a3e87>] trace_hardirqs_on_caller+0xa7/0x1c0
[<ffffffff810a3fad>] trace_hardirqs_on+0xd/0x10
[<ffffffff8151c40b>] _raw_spin_unlock_irq+0x2b/0x40
[<ffffffffa036ea8e>] cm_enter_timewait+0xae/0x100 [ib_cm]
[<ffffffffa036ff76>] ib_send_cm_drep+0xb6/0x190 [ib_cm]
[<ffffffffa052ed08>] srp_cm_handler+0x128/0x1a0 [ib_srp]
[<ffffffffa0370340>] cm_process_work+0x20/0xf0 [ib_cm]
[<ffffffffa0371335>] cm_dreq_handler+0x135/0x2c0 [ib_cm]
[<ffffffffa03733c5>] cm_work_handler+0x75/0xd0 [ib_cm]
[<ffffffff8107184d>] process_one_work+0x1bd/0x460
[<ffffffff81073148>] worker_thread+0x118/0x420
[<ffffffff81078454>] kthread+0xe4/0x100
[<ffffffff8151cbbf>] ret_from_fork+0x3f/0x70
irq event stamp: 1672286
hardirqs last enabled at (1672283): [<ffffffff81408ec0>] poll_idle+0x10/0x80
hardirqs last disabled at (1672284): [<ffffffff8151d304>] common_interrupt+0x84/0x89
softirqs last enabled at (1672286): [<ffffffff8105b4dc>] _local_bh_enable+0x1c/0x50
softirqs last disabled at (1672285): [<ffffffff8105b697>] irq_enter+0x47/0x70
other info that might help us debug this:
Possible unsafe locking scenario:
The commit 2895b2cad6e7 ("dmaengine: dw: fix cyclic transfer callbacks")
re-enabled BLOCK interrupts with regard to make cyclic transfers work. However,
this change becomes a regression for non-cyclic transfers as interrupt counters
under stress test had been grown enormously (approximately per 4-5 bytes in the
UART loop back test).
Taking into consideration above enable BLOCK interrupts if and only if channel
is programmed to perform cyclic transfer.
Fixes: 2895b2cad6e7 ("dmaengine: dw: fix cyclic transfer callbacks") Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Mans Rullgard <mans@mansr.com> Tested-by: Mans Rullgard <mans@mansr.com> Acked-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Vinod Koul <vinod.koul@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cyclic transfer callbacks rely on block completion interrupts which were
disabled in commit ff7b05f29fd4 ("dmaengine/dw_dmac: Don't handle block
interrupts"). This re-enables block interrupts so the cyclic callbacks
can work. Other transfer types are not affected as they set the INT_EN
bit only on the last block.
Commit 61e183f83069 ("dmaengine/dw_dmac: Reconfigure interrupt and
chan_cfg register on resume") moved some channel initialisation to
a new function which must be called before starting a transfer.
This updates dw_dma_cyclic_start() to use dwc_dostart() like the other
modes, thus ensuring dwc_initialize() gets called and removing some code
duplication.
Fixes: 61e183f83069 ("dmaengine/dw_dmac: Reconfigure interrupt and chan_cfg register on resume") Signed-off-by: Mans Rullgard <mans@mansr.com> Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by: Vinod Koul <vinod.koul@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ACPI 6.1 clarified that multi-interface dimms require multiple control
region entries (DCRs) per dimm. Previously we were assuming that a
control region is only present when block-data-windows are present.
This implementation was done with an eye to be compatibility with the
looser ACPI 6.0 interpretation of this table.
1/ When coalescing the memory device (MEMDEV) tables for a single dimm,
coalesce on device_handle rather than control region index.
2/ Whenever we disocver a control region with non-zero block windows
re-scan for block-data-window (BDW) entries.
We may need to revisit this if a DIMM ever implements a format interface
outside of blk or pmem, but that is not on the foreseeable horizon.
Signed-off-by: Dan Williams <dan.j.williams@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The quirk to get "acpi_backlight=vendor" behavior by default on the
Dell Inspiron 5737 was added before we started doing
"acpi_backlight=native" by default on Win8 ready machines.
Since we now avoid using acpi-video as backlight driver on these machines
by default (using the native driver instead) we no longer need this quirk.
Moreover the vendor driver does not work after a suspend/resume where
as the native driver does.
This reverts commit 08a56226d847 (ACPI / video: Add Dell Inspiron 5737
to the blacklist).
Link: https://bugzilla.kernel.org/show_bug.cgi?id=111061 Reported-and-tested-by: erusan@gmail.com Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The Toshiba Satellite R830 needs disable_backlight_sysfs_if=1, just like
the Toshiba Portege R830. Add a quirk for this.
Link: https://bugzilla.kernel.org/show_bug.cgi?id=21012 Tested-by: To Do <entodoays@gmail.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The Toshiba Portege R700 needs disable_backlight_sysfs_if=1, just like
the Toshiba Portege R830. Add a quirk for this.
Link: https://bugzilla.kernel.org/show_bug.cgi?id=21012 Tested-by: Emma Reisz <emmareisz@outlook.com> Signed-off-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The sw842 library code was merged in linux-4.1 and causes a very rare randconfig
failure when CONFIG_CRC32 is not set:
lib/built-in.o: In function `sw842_compress':
oid_registry.c:(.text+0x12ddc): undefined reference to `crc32_be'
lib/built-in.o: In function `sw842_decompress':
oid_registry.c:(.text+0x137e4): undefined reference to `crc32_be'
This adds an explict 'select CRC32' statement, similar to what the other users
of the crc32 code have. In practice, CRC32 is always enabled anyway because
over 100 other symbols select it.
Commit 9d99a8dda154 ("nvme: move hardware structures out of the uapi
version of nvme.h") renamed nvme.h to nvme_ioctl.h, but the uapi list
still refers to nvme.h. People trying to install the headers hit a
failure as the header no longer exists.
Signed-off-by: Mike Frysinger <vapier@gentoo.org> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jens Axboe <axboe@fb.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Like the Yoga 900 models the Lenovo Yoga 700 does not have a
hw rfkill switch, and trying to read the hw rfkill switch through the
ideapad module causes it to always reported blocking breaking wifi.
This commit adds the Lenovo Yoga 700 to the no_hw_rfkill dmi list, fixing
the wifi breakage.
One of the newest ideapad models also lacks a physical hw rfkill switch,
and trying to read the hw rfkill switch through the ideapad module
causes it to always reported blocking breaking wifi.
If transflective backlight is supported and the brightness is zero
(lowest brightness level), the set_lcd_brightness function will activate
the transflective backlight, making the LCD appear to be turned off.
This patch fixes the issue by incrementing the brightness level, and
by doing so, avoiding the activation of the tranflective backlight.
set_power_state defaults to no displays, so we need to update
the display configuration after setting up the powerstate on the
first call. In most cases this is not an issue since ends up
getting called multiple times at any given modeset and the proper
order is achieved in the display changed handling at the top of
the function.
Reviewed-by: Christian König <christian.koenig@amd.com> Acked-by: Jordan Lazare <Jordan.Lazare@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Limit the amount of time radeon_flip_work_func can
delay programming a page flip, by both limiting the
maximum amount of time per wait cycle and the maximum
number of wait cycles. Continue the flip if the limit
is exceeded, even if that may result in a visual or
timing glitch.
This is to prevent a hang of page flips, as reported
in fdo bug #93746: Disconnecting a DisplayPort display
in parallel to a kms pageflip getting queued can cause
the following hang of page flips and thereby an unusable
desktop:
2. Hotunplug of display causes the driver to DPMS OFF
the unplugged display. Display engine shuts down,
scanout no longer moves, but stays at its resting
position at start line of vblank.
3. radeon_flip_work_func executes while crtc is off, and
due to the non-moving scanout position, the new flip
delay code introduced into Linux 4.4 by
commit 5b5561b3660d ("drm/radeon: Fixup hw vblank counter/ts..")
enters an infinite wait loop.
4. After reconnecting the display, the pageflip continues
to hang in 3. and the display doesn't update its view
of the desktop.
This patch fixes the Linux 4.4 regression from fdo bug #93746
drm_vblank_offdelay can have three different types of values:
< 0 is to be always treated the same as dev->vblank_disable_immediate
= 0 is to be treated as "never disable vblanks"
> 0 is to be treated as disable immediate if kms driver wants it
that way via dev->vblank_disable_immediate. Otherwise it is
a disable timeout in msecs.
This got broken in Linux 3.18+ for the implementation of
drm_vblank_on. If the user specified a value of zero which should
always reenable vblank irqs in this function, a kms driver could
override the users choice by setting vblank_disable_immediate
to true. This patch fixes the regression and keeps the user in
control.
v2: Only reenable vblank if there are clients left or the user
requested to "never disable vblanks" via offdelay 0. Enabling
vblanks even in the "delayed disable" case (offdelay > 0) was
specifically added by Ville in commit cd19e52aee922
("drm: Kick start vblank interrupts at drm_vblank_on()"),
but after discussion it turns out that this was done by accident.
Citing Ville: "I think it just ended up as a mess due to changing
some of the semantics of offdelay<0 vs. offdelay==0 vs.
disable_immediate during the review of the series. So yeah, given
how drm_vblank_put() works now, I'd just make this check for
offdelay==0."
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: michel@daenzer.net Cc: vbabka@suse.cz Cc: ville.syrjala@linux.intel.com Cc: daniel.vetter@ffwll.ch Cc: dri-devel@lists.freedesktop.org Cc: alexander.deucher@amd.com Cc: christian.koenig@amd.com Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Changes to drm_update_vblank_count() in Linux 4.4 broke the
behaviour of the pre/post modeset functions as the new update
code doesn't deal with hw vblank counter resets inbetween calls
to drm_vblank_pre_modeset an drm_vblank_post_modeset, as it
should.
This causes mistreatment of such hw counter resets as counter
wraparound, and thereby large forward jumps of the software
vblank counter which in turn cause vblank event dispatching
and vblank waits to fail/hang --> userspace clients hang.
This symptom was reported on radeon-kms to cause a infinite
hang of KDE Plasma 5 shell's login procedure, preventing users
from logging in.
Fix this by detecting when drm_update_vblank_count() is called
inside a pre->post modeset interval. If so, clamp valid vblank
increments to the safe values 0 and 1, pretty much restoring
the update behavior of the old update code of Linux 4.3 and
earlier. Also reset the last recorded hw vblank count at call
to drm_vblank_post_modeset() to be safe against hw that after
modesetting, dpms on etc. only fires its first vblank irq after
drm_vblank_post_modeset() was already called.
This fixes a regression introduced by the new drm_update_vblank_count()
implementation in Linux 4.4:
Restrict the bump of the software vblank counter in drm_update_vblank_count()
to a safe maximum value of +1 whenever there is the possibility that
concurrent readers of vblank timestamps could be active at the moment,
as the current implementation of the timestamp caching and updating is
not safe against concurrent readers for calls to store_vblank() with a
bump of anything but +1. A bump != 1 would very likely return corrupted
timestamps to userspace, because the same slot in the cache could
be concurrently written by store_vblank() and read by one of those
readers in a non-atomic fashion and without the read-retry logic
detecting this collision.
Concurrent readers can exist while drm_update_vblank_count() is called
from the drm_vblank_off() or drm_vblank_on() functions or other non-vblank-
irq callers. However, all those calls are happening with the vbl_lock
locked thereby preventing a drm_vblank_get(), so the vblank refcount
can't increase while drm_update_vblank_count() is executing. Therefore
a zero vblank refcount during execution of that function signals that
is safe for arbitrary counter bumps if called from outside vblank irq,
whereas a non-zero count is not safe.
Whenever the function is called from vblank irq, we have to assume concurrent
readers could show up any time during its execution, even if the refcount
is currently zero, as vblank irqs are usually only enabled due to the
presence of readers, and because when it is called from vblank irq it
can't hold the vbl_lock to protect it from sudden bumps in vblank refcount.
Therefore also restrict bumps to +1 when the function is called from vblank
irq.
Such bumps of more than +1 can happen at other times than reenabling
vblank irqs, e.g., when regular vblank interrupts get delayed by more
than 1 frame due to long held locks, long irq off periods, realtime
preemption on RT kernels, or system management interrupts.
A better solution would be to rewrite the timestamp caching to use
full seqlocks to allow concurrent writes and reads for arbitrary
vblank counter increments.
v2: Add code comment that this is essentially a hack and should
be replaced by a full seqlock implementation for caching of
timestamps.
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: michel@daenzer.net Cc: vbabka@suse.cz Cc: ville.syrjala@linux.intel.com Cc: daniel.vetter@ffwll.ch Cc: dri-devel@lists.freedesktop.org Cc: alexander.deucher@amd.com Cc: christian.koenig@amd.com Signed-off-by: Dave Airlie <airlied@redhat.com>
Otherwise if a kms driver calls into drm_vblank_off() more than once
before calling drm_vblank_on() again, the redundant calls to
vblank_disable_and_save() will call drm_update_vblank_count()
while hw vblank counters and vblank timestamping are in a undefined
state during modesets, dpms off etc.
At least with the legacy drm helpers it is not unusual to
get multiple calls to drm_vblank_off and drm_vblank_on, e.g.,
half a dozen calls to drm_vblank_off and two calls to drm_vblank_on
were observed on radeon-kms during dpms-off -> dpms-on transition.
We don't no-op calls from atomic modesetting drivers, as they
should do a proper job of tracking hw state.
Fixes large jumps of the software maintained vblank counter due to
the hardware vblank counter resetting to zero during dpms off or
modeset, e.g., if radeon-kms is modified to use drm_vblank_off/on
instead of drm_vblank_pre/post_modeset().
This fixes a regression caused by the changes made to
drm_update_vblank_count() in Linux 4.4.
v2: Don't no-op on atomic modesetting drivers, per suggestion
of Daniel Vetter.
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: michel@daenzer.net Cc: vbabka@suse.cz Cc: ville.syrjala@linux.intel.com Cc: alexander.deucher@amd.com Cc: christian.koenig@amd.com Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We need to use post-decrement to get the pci_map_page undone also for
i==0, and to avoid some very unpleasant behaviour if pci_map_page
failed already at i==0.
Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Rasmus Villemoes <linux@rasmusvillemoes.dk> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We don't actually check for INTEL_OUTPUT_DP_MST at all in here, as a
result we skip assigning a DPLL to any DP MST ports, which makes link
training fail:
[ 1442.933896] [drm:intel_power_well_enable] enabling DDI D power well
[ 1442.933905] [drm:skl_set_power_well] Enabling DDI D power well
[ 1442.933957] [drm:intel_mst_pre_enable_dp] 0
[ 1442.935474] [drm:intel_dp_set_signal_levels] Using signal levels 00000000
[ 1442.935477] [drm:intel_dp_set_signal_levels] Using vswing level 0
[ 1442.935480] [drm:intel_dp_set_signal_levels] Using pre-emphasis level 0
[ 1442.936190] [drm:intel_dp_set_signal_levels] Using signal levels 05000000
[ 1442.936193] [drm:intel_dp_set_signal_levels] Using vswing level 1
[ 1442.936195] [drm:intel_dp_set_signal_levels] Using pre-emphasis level 1
[ 1442.936858] [drm:intel_dp_set_signal_levels] Using signal levels 08000000
[ 1442.936862] [drm:intel_dp_set_signal_levels] Using vswing level 2
…
[ 1442.998253] [drm:intel_dp_link_training_clock_recovery [i915]] *ERROR* too many full retries, give up
[ 1442.998512] [drm:intel_dp_start_link_train [i915]] *ERROR* failed to train DP, aborting
After which the pipe state goes completely out of sync:
[ 70.075596] [drm:check_crtc_state] [CRTC:25]
[ 70.075696] [drm:intel_pipe_config_compare [i915]] *ERROR* mismatch in ddi_pll_sel (expected 0x00000000, found 0x00000001)
[ 70.075747] [drm:intel_pipe_config_compare [i915]] *ERROR* mismatch in shared_dpll (expected -1, found 0)
[ 70.075798] [drm:intel_pipe_config_compare [i915]] *ERROR* mismatch in dpll_hw_state.ctrl1 (expected 0x00000000, found 0x00000021)
[ 70.075840] [drm:intel_pipe_config_compare [i915]] *ERROR* mismatch in dpll_hw_state.cfgcr1 (expected 0x00000000, found 0x80400173)
[ 70.075884] [drm:intel_pipe_config_compare [i915]] *ERROR* mismatch in dpll_hw_state.cfgcr2 (expected 0x00000000, found 0x000003a5)
[ 70.075954] [drm:intel_pipe_config_compare [i915]] *ERROR* mismatch in base.adjusted_mode.crtc_clock (expected 262750, found 72256)
[ 70.075999] [drm:intel_pipe_config_compare [i915]] *ERROR* mismatch in port_clock (expected 540000, found 148500)
And if you're especially lucky, it keeps going downhill:
On SKL+ plane scaling is mutually exclusive with color keying. The code
check for this, but during some refactoring the code got changed to
also reject primary plane windowing when color keying is used. There is
no such restriction in the hardware, so restore the original logic.
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> Fixes: 061e4b8d650a ("drm/i915: clean up atomic plane check functions, v2.") Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1452883613-28549-1-git-send-email-ville.syrjala@linux.intel.com Reviewed-by: Matt Roper <matthew.d.roper@intel.com> Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
(cherry picked from commit 693bdc28a733dba68b86af295e7509812fec35d9) Signed-off-by: Jani Nikula <jani.nikula@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Per DP spec, the source device should fall back to 18 bpp, VESA range
RGB when the sink capability is unknown. Fix the color depth
clamping. 18 bpp color depth should ensure full color range in automatic
mode.
The clamping has been HDMI specific since its introduction in
We need to set the DC FLUSH PIPE_CONTROL bit on Gen7+ to guarantee
that writes performed via the HDC are visible in memory. Fixes an
intermittent failure in a Piglit test that writes to a BO from a
shader using GL atomic counters (implemented as HDC untyped atomics)
and then expects the memory to read back the same value after mapping
it on the CPU.
gmbus also needs the power domain infrastructure right from the start,
since as soon as we register the i2c controllers someone can use them.
v2: Adjust cleanup paths too (Chris).
v3: Rebase onto -nightly (totally bogus tree I had lying around) and
also move dpio init head (Ville).
v4: Ville instead suggested to move gmbus setup later in the sequence,
since it's only needed by the modeset code.
v5: Move even close to the actual user, right next to the comment that
states where we really need gmbus (and interrupts!).
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> Cc: Patrik Jakobsson <patrik.jakobsson@linux.intel.com> Cc: Imre Deak <imre.deak@intel.com> Cc: Jani Nikula <jani.nikula@intel.com> Cc: Meelis Roos <mroos@linux.ee> Cc: Chris Wilson <chris@chris-wilson.co.uk> Fixes: ac9b8236551d ("drm/i915: Introduce a gmbus power domain")
References: http://www.spinics.net/lists/intel-gfx/msg83075.html Signed-off-by: Daniel Vetter <daniel.vetter@intel.com> Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> Link: http://patchwork.freedesktop.org/patch/msgid/1452682528-19437-1-git-send-email-daniel.vetter@ffwll.ch Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This fixes reprobing of display connectors on resume. After some
talking with danvet on IRC, I learned that calling
drm_helper_hpd_irq_event() does actually trigger a full reprobe of each
connector's status. It turns out this is the actual reason reprobing on
resume hasn't been working (this was observed on a T440s):
- We call hpd_init()
- We check each connector for a couple of things before marking
connector->polled with DRM_CONNECTOR_POLL_HPD, one of which is an
active encoder. Of course, a disconnected port won't have an
active encoder, so we don't add the flag to any of the
connectors.
- We call drm_helper_hpd_irq_event()
- drm_helper_irq_event() checks each connector for the
DRM_CONNECTOR_POLL_HPD flag. The only one that has it is eDP-1,
so we skip reprobing each connector except that one.
In addition, we also now avoid setting connector->polled to
DRM_CONNECTOR_POLL_HPD for MST connectors, since their reprobing is
handled by the mst helpers. This is probably what was originally
intended to happen here.
Changes since V1:
* Use the explanation of the issue as the commit message instead
* Change the title of the commit, since this does more then just stop a
check for an encoder now
* Add "Fixes" line for the patch that introduced this regression
* Don't enable DRM_CONNECTOR_POLL_HPD for mst connectors
Changes since V2:
* Put patch changelog above Signed-off-by
* Follow Daniel Vetter's suggestion for making the code here a bit more
legible
Following a GPU reset, we may leave the context in a poorly defined
state, and reloading from that context will leave the GPU flummoxed. For
secondary contexts, this will lead to that context being banned - but
currently it is also causing the default context to become banned,
leading to turmoil in the shared state.
which quietly introduced the removal of the MI_RESTORE_INHIBIT on the
default context.
v2: Mark the global default context as uninitialized on GPU reset so
that the context-local workarounds are reloaded upon re-enabling.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Michel Thierry <michel.thierry@intel.com> Cc: Mika Kuoppala <mika.kuoppala@intel.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Link: http://patchwork.freedesktop.org/patch/msgid/1448630935-27377-1-git-send-email-chris@chris-wilson.co.uk Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
[danvet: This seems to fix a gpu hand on after the first resume,
resulting in any future suspend operation failing with -EIO because
the gpu seems to be in a funky state. Somehow this patch fixes that.] Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
(cherry picked from commit 42f1cae8c079bcceb3cff079fddc3ff8852c788f) Signed-off-by: Jani Nikula <jani.nikula@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Christian König <christian.koenig@amd.com> Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Properly setup the DFS divider for DP audio for DCE4.1.
Signed-off-by: Slava Grigorev <slava.grigorev@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Move encoding of DFS (digital frequency synthesizer) divider into a
separate function and improve calculation precision.
Signed-off-by: Slava Grigorev <slava.grigorev@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This is preparation for the fixes in the following patches.
Signed-off-by: Slava Grigorev <slava.grigorev@amd.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> Reviewed-by: Christian König <christian.koenig@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
projects / karo-tx-linux.git / log