Commit Graph

7 Commits

Author SHA1 Message Date
Stephan Gerhold f44714b4eb soc: qcom: smsm: Fix missed interrupts if state changes while masked
[ Upstream commit e3d4571955050736bbf3eda0a9538a09d9fcfce8 ]

The SMSM driver detects interrupt edges by tracking the last state
it has seen (and has triggered the interrupt handler for). This works
fine, but only if the interrupt does not change state while masked.

For example, if an interrupt is unmasked while the state is HIGH,
the stored last_value for that interrupt might still be LOW. Then,
when the remote processor triggers smsm_intr() we assume that nothing
has changed, even though the state might have changed from HIGH to LOW.

Attempt to fix this by checking the current remote state before
unmasking an IRQ. Use atomic operations to avoid the interrupt handler
from interfering with the unmask function.

This fixes modem crashes in some edge cases with the BAM-DMUX driver.
Specifically, the BAM-DMUX interrupt handler is not called for the
HIGH -> LOW smsm state transition if the BAM-DMUX driver is loaded
(and therefore unmasks the interrupt) after the modem was already started:

qcom-q6v5-mss 4080000.remoteproc: fatal error received: a2_task.c:3188:
  Assert FALSE failed: A2 DL PER deadlock timer expired waiting for Apps ACK

Fixes: c97c4090ff ("soc: qcom: smsm: Add driver for Qualcomm SMSM")
Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
Link: https://lore.kernel.org/r/20210712135703.324748-2-stephan@gerhold.net
Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
2021-09-15 09:47:33 +02:00
Thomas Gleixner 97fb5e8d9b treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 284
Based on 1 normalized pattern(s):

  this program is free software you can redistribute it and or modify
  it under the terms of the gnu general public license version 2 and
  only version 2 as published by the free software foundation this
  program is distributed in the hope that it will be useful but
  without any warranty without even the implied warranty of
  merchantability or fitness for a particular purpose see the gnu
  general public license for more details

extracted by the scancode license scanner the SPDX license identifier

  GPL-2.0-only

has been chosen to replace the boilerplate/reference in 294 file(s).

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Allison Randal <allison@lohutok.net>
Reviewed-by: Alexios Zavras <alexios.zavras@intel.com>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190529141900.825281744@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-06-05 17:36:37 +02:00
Johan Hovold 8804517e9f soc: qcom: smsm: fix child-node lookup
Fix child-node lookup during probe, which ended up searching the whole
device tree depth-first starting at the parent rather than just matching
on its children.

Note that the original premature free of the parent node has already
been fixed separately.

Also note that this pattern of looking up the first child node with a
given property is rare enough that a generic helper is probably not
warranted.

Fixes: c97c4090ff ("soc: qcom: smsm: Add driver for Qualcomm SMSM")
Fixes: 3e8b554114 ("soc: qcom: smsm: fix of_node refcnting problem")
Cc: Bjorn Andersson <bjorn.andersson@linaro.org>
Cc: Rob Clark <robdclark@gmail.com>
Signed-off-by: Johan Hovold <johan@kernel.org>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
2017-12-20 15:38:34 -06:00
Rob Clark 3e8b554114 soc: qcom: smsm: fix of_node refcnting problem
of_find_node_with_property() drops the reference to the 'from' node,
which eventually (after enough -EPROBE_DEFERs) drops the last reference
to the node causing all sorts of fun problems, and this nice splat.

  BUG: sleeping function called from invalid context at ../kernel/locking/mutex.c:747
  in_atomic(): 1, irqs_disabled(): 128, pid: 33, name: kworker/0:1
  4 locks held by kworker/0:1/33:
   #0:  ("events"){.+.+.+}, at: [<ffff0000080fa91c>] process_one_work+0x1a4/0x728
   #1:  (deferred_probe_work){+.+.+.}, at: [<ffff0000080fa91c>] process_one_work+0x1a4/0x728
   #2:  (&dev->mutex){......}, at: [<ffff000008676078>] __device_attach+0x30/0x168
   #3:  (devtree_lock){......}, at: [<ffff000008828fd0>] of_find_node_with_property+0x30/0xe0
  irq event stamp: 18976
  hardirqs last  enabled at (18975): [<ffff00000815794c>] __down_trylock_console_sem+0x74/0xb8
  hardirqs last disabled at (18976): [<ffff0000089e26d4>] _raw_spin_lock_irqsave+0x2c/0x78
  softirqs last  enabled at (16880): [<ffff0000080e0f00>] __do_softirq+0x580/0x640
  softirqs last disabled at (16871): [<ffff0000080e13a4>] irq_exit+0xe4/0x138
  CPU: 0 PID: 33 Comm: kworker/0:1 Tainted: G            E   4.12.0-rc5+ #1455
  Hardware name: qualcomm dragonboard410c/dragonboard410c, BIOS 2017.07-rc1-00234-g22fa70a-dirty 06/26/2017
  Workqueue: events deferred_probe_work_func
  Call trace:
  [<ffff000008089ee0>] dump_backtrace+0x0/0x230
  [<ffff00000808a134>] show_stack+0x24/0x30
  [<ffff0000084e1944>] dump_stack+0xac/0xe8
  [<ffff00000810d7e0>] ___might_sleep+0x150/0x230
  [<ffff00000810d918>] __might_sleep+0x58/0x90
  [<ffff0000089dde18>] __mutex_lock+0x50/0x870
  [<ffff0000089de674>] mutex_lock_nested+0x3c/0x50
  [<ffff000008388ae0>] kernfs_remove+0x30/0x50
  [<ffff00000838b720>] sysfs_remove_dir+0x58/0x70
  [<ffff0000084e393c>] kobject_del+0x1c/0x58
  [<ffff0000084e374c>] kobject_put+0xb4/0x208
  [<ffff00000882c364>] of_node_put+0x24/0x30
  [<ffff000008829018>] of_find_node_with_property+0x78/0xe0
  [<ffff000000aff5f4>] qcom_smsm_probe+0x194/0x720 [smsm]
  [<ffff0000086793b4>] platform_drv_probe+0x74/0x110
  [<ffff0000086765bc>] driver_probe_device+0x2b4/0x420
  [<ffff000008676920>] __device_attach_driver+0xd0/0x150
  [<ffff000008673e78>] bus_for_each_drv+0x68/0xa8
  [<ffff00000867611c>] __device_attach+0xd4/0x168
  [<ffff000008676a1c>] device_initial_probe+0x24/0x30
  [<ffff000008675380>] bus_probe_device+0xa0/0xa8
  [<ffff000008675948>] deferred_probe_work_func+0xb8/0xf8
  [<ffff0000080fa9d4>] process_one_work+0x25c/0x728
  [<ffff0000080faef4>] worker_thread+0x54/0x3d8
  [<ffff0000081031d8>] kthread+0x110/0x140
  [<ffff000008082d90>] ret_from_fork+0x10/0x40
  OF: ERROR: Bad of_node_put() on /smsm
  CPU: 0 PID: 33 Comm: kworker/0:1 Tainted: G        W   E   4.12.0-rc5+ #1455
  Hardware name: qualcomm dragonboard410c/dragonboard410c, BIOS 2017.07-rc1-00234-g22fa70a-dirty 06/26/2017
  Workqueue: events deferred_probe_work_func

Signed-off-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
2017-08-08 14:37:42 -05:00
Jonathan Neuschäfer 35dfa3efea soc: qcom: smsm: Improve error handling, quiesce probe deferral
Don't use size if info indicates an error condition. Previously a
non-ENOENT error (such as -EPROBE_DEFER) would lead to size being used
even though it hadn't necessarily been initialized in qcom_smem_get.

Don't print an error message in the -EPROBE_DEFER case.

Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Reviewed-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
2017-06-05 21:50:44 -05:00
Bjorn Andersson 3680a4a974 soc: qcom: Update properties for smem state referencing
Update the property names to match device tree bindings, the correct
values should be qcom,smem-states and qcom,smem-state-names.

Also update the #qcom,smem-state-cells for consistency, before we merge
any users of these properties.

Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Andy Gross <andy.gross@linaro.org>
2016-06-24 22:53:41 -05:00
Bjorn Andersson c97c4090ff soc: qcom: smsm: Add driver for Qualcomm SMSM
This driver exposed the Qualcomm Shared Memory State Machine bits.

Signed-off-by: Bjorn Andersson <bjorn.andersson@sonymobile.com>
Signed-off-by: Andy Gross <agross@codeaurora.org>
2015-12-08 13:01:01 -06:00