linux.git/drivers/pinctrl/pinctrl-single.c, branch v4.2

pinctrl: single: ensure pcs irq will not be forced threaded

2015-07-20T09:01:52+00:00

The PSC IRQ is requested using request_irq() API and as result it can
be forced to be threaded IRQ in RT-Kernel if PCS_QUIRK_HAS_SHARED_IRQ
is enabled for pinctrl domain.

As result, following 'possible irq lock inversion dependency' report
can be seen:
=========================================================
[ INFO: possible irq lock inversion dependency detected ]
3.14.43-rt42-00360-g96ff499-dirty #24 Not tainted
---------------------------------------------------------
irq/369-pinctrl/927 just changed the state of lock:
 (&pcs->lock){+.....}, at: [] pcs_irq_handle+0x48/0x9c
but this lock was taken by another, HARDIRQ-safe lock in the past:
 (&irq_desc_lock_class){-.....}

and interrupts could create inverse lock ordering between them.

other info that might help us debug this:
 Possible interrupt unsafe locking scenario:

       CPU0                    CPU1
       ----                    ----
  lock(&pcs->lock);
                               local_irq_disable();
                               lock(&irq_desc_lock_class);
                               lock(&pcs->lock);
  
    lock(&irq_desc_lock_class);

 *** DEADLOCK ***

no locks held by irq/369-pinctrl/927.

the shortest dependencies between 2nd lock and 1st lock:
  -> (&irq_desc_lock_class){-.....} ops: 58724 {
     IN-HARDIRQ-W at:
                       [] lock_acquire+0x9c/0x158
                       [] _raw_spin_lock+0x48/0x58
                       [] handle_fasteoi_irq+0x24/0x15c
                       [] generic_handle_irq+0x3c/0x4c
                       [] handle_IRQ+0x50/0xa0
                       [] gic_handle_irq+0x3c/0x6c
                       [] __irq_svc+0x44/0x8c
                       [] arch_cpu_idle+0x40/0x4c
                       [] cpu_startup_entry+0x270/0x2e0
                       [] rest_init+0xd4/0xe4
                       [] start_kernel+0x3d0/0x3dc
                       [<80008084>] 0x80008084
     INITIAL USE at:
                      [] lock_acquire+0x9c/0x158
                      [] _raw_spin_lock_irqsave+0x54/0x68
                      [] __irq_get_desc_lock+0x64/0xa4
                      [] irq_set_chip+0x30/0x78
                      [] irq_set_chip_and_handler_name+0x24/0x3c
                      [] gic_irq_domain_map+0x48/0xb4
                      [] irq_domain_associate+0x84/0x1d4
                      [] irq_create_mapping+0x80/0x11c
                      [] irq_create_of_mapping+0x80/0x120
                      [] irq_of_parse_and_map+0x34/0x3c
                      [] omap_dm_timer_init_one+0x90/0x30c
                      [] omap5_realtime_timer_init+0x8c/0x48c
                      [] time_init+0x28/0x38
                      [] start_kernel+0x240/0x3dc
                      [<80008084>] 0x80008084
   }
   ... key      at: [] irq_desc_lock_class+0x0/0x8
   ... acquired at:
   [] _raw_spin_lock+0x48/0x58
   [] pcs_irq_unmask+0x58/0xa0
   [] irq_enable+0x38/0x48
   [] irq_startup+0x78/0x7c
   [] __setup_irq+0x4a8/0x4f4
   [] request_threaded_irq+0xb8/0x138
   [] omap_8250_startup+0x4c/0x148
   [] serial8250_startup+0x24/0x30
   [] uart_startup.part.9+0x5c/0x1b4
   [] uart_open+0xf4/0x16c
   [] tty_open+0x170/0x61c
   [] chrdev_open+0xbc/0x1b4
   [] do_dentry_open+0x1e8/0x2bc
   [] finish_open+0x44/0x5c
   [] do_last.isra.47+0x710/0xca0
   [] path_openat+0xc4/0x640
   [] do_filp_open+0x3c/0x98
   [] do_sys_open+0x114/0x1d8
   [] SyS_open+0x28/0x2c
   [] kernel_init_freeable+0x168/0x1e4
   [] kernel_init+0x1c/0xf8
   [] ret_from_fork+0x14/0x20

-> (&pcs->lock){+.....} ops: 65 {
   HARDIRQ-ON-W at:
                    [] lock_acquire+0x9c/0x158
                    [] _raw_spin_lock+0x48/0x58
                    [] pcs_irq_handle+0x48/0x9c
                    [] pcs_irq_handler+0x1c/0x28
                    [] irq_forced_thread_fn+0x30/0x74
                    [] irq_thread+0x158/0x1c4
                    [] kthread+0xd4/0xe8
                    [] ret_from_fork+0x14/0x20
   INITIAL USE at:
                   [] lock_acquire+0x9c/0x158
                   [] _raw_spin_lock_irqsave+0x54/0x68
                   [] pcs_enable+0x7c/0xe8
                   [] pinmux_enable_setting+0x178/0x220
                   [] pinctrl_select_state+0x110/0x194
                   [] pinctrl_bind_pins+0x7c/0x108
                   [] driver_probe_device+0x70/0x254
                   [] __driver_attach+0x9c/0xa0
                   [] bus_for_each_dev+0x78/0xac
                   [] driver_attach+0x2c/0x30
                   [] bus_add_driver+0x15c/0x204
                   [] driver_register+0x88/0x108
                   [] __platform_driver_register+0x64/0x6c
                   [] omap_hsmmc_driver_init+0x1c/0x20
                   [] do_one_initcall+0x110/0x170
                   [] kernel_init_freeable+0x140/0x1e4
                   [] kernel_init+0x1c/0xf8
                   [] ret_from_fork+0x14/0x20
 }
 ... key      at: [] __key.18572+0x0/0x8
 ... acquired at:
   [] mark_lock+0x388/0x76c
   [] __lock_acquire+0x6d0/0x1f98
   [] lock_acquire+0x9c/0x158
   [] _raw_spin_lock+0x48/0x58
   [] pcs_irq_handle+0x48/0x9c
   [] pcs_irq_handler+0x1c/0x28
   [] irq_forced_thread_fn+0x30/0x74
   [] irq_thread+0x158/0x1c4
   [] kthread+0xd4/0xe8
   [] ret_from_fork+0x14/0x20

stack backtrace:
CPU: 1 PID: 927 Comm: irq/369-pinctrl Not tainted 3.14.43-rt42-00360-g96ff499-dirty #24
[] (unwind_backtrace) from [] (show_stack+0x20/0x24)
[] (show_stack) from [] (dump_stack+0x84/0xd0)
[] (dump_stack) from [] (print_irq_inversion_bug+0x1d0/0x21c)
[] (print_irq_inversion_bug) from [] (check_usage_backwards+0xb4/0x11c)
[] (check_usage_backwards) from [] (mark_lock+0x388/0x76c)
[] (mark_lock) from [] (__lock_acquire+0x6d0/0x1f98)
[] (__lock_acquire) from [] (lock_acquire+0x9c/0x158)
[] (lock_acquire) from [] (_raw_spin_lock+0x48/0x58)
[] (_raw_spin_lock) from [] (pcs_irq_handle+0x48/0x9c)
[] (pcs_irq_handle) from [] (pcs_irq_handler+0x1c/0x28)
[] (pcs_irq_handler) from [] (irq_forced_thread_fn+0x30/0x74)
[] (irq_forced_thread_fn) from [] (irq_thread+0x158/0x1c4)
[] (irq_thread) from [] (kthread+0xd4/0xe8)
[] (kthread) from [] (ret_from_fork+0x14/0x20)

To fix it use IRQF_NO_THREAD to ensure that pcs irq will not be forced threaded.

Cc: Tony Lindgren 
Cc: Sebastian Andrzej Siewior 
Signed-off-by: Grygorii Strashko 
Acked-by: Tony Lindgren 
Signed-off-by: Linus Walleij

pinctrl: make pinctrl_register() return proper error code

2015-06-10T12:49:52+00:00

Currently, pinctrl_register() just returns NULL on error, so the
callers can not know the exact reason of the failure.

Some of the pinctrl drivers return -EINVAL, some -ENODEV, and some
-ENOMEM on error of pinctrl_register(), although the error code
might be different from the real cause of the error.

This commit reworks pinctrl_register() to return the appropriate
error code and modifies all of the pinctrl drivers to use IS_ERR()
for the error checking and PTR_ERR() for getting the error code.

Signed-off-by: Masahiro Yamada 
Acked-by: Patrice Chotard 
Acked-by: Thierry Reding 
Acked-by: Heiko Stuebner 
Tested-by: Mika Westerberg 
Acked-by: Mika Westerberg 
Acked-by: Lee Jones 
Acked-by: Sören Brinkmann 
Acked-by: Laurent Pinchart 
Acked-by: Ray Jui 
Acked-by: Antoine Tenart 
Acked-by: Hongzhou Yang 
Acked-by: Wei Chen 
Signed-off-by: Linus Walleij

pinctrl: single: Constify irq_domain_ops

2015-05-06T13:32:45+00:00

The irq_domain_ops are not modified by the driver and the irqdomain core
code accepts pointer to a const data.

Signed-off-by: Krzysztof Kozlowski 
Signed-off-by: Linus Walleij

pinctrl: constify of_device_id array

2015-03-27T08:58:35+00:00

of_device_id is always used as const.
(See driver.of_match_table and open firmware functions)

Signed-off-by: Fabian Frederick 
Acked-by: Jean-Christophe PLAGNIOL-VILLARD 
Acked-by: Patrice Chotard 
Acked-by: Maxime Coquelin 
Acked-by: Hongzhou Yang 
Acked-by: Lee Jones 
Signed-off-by: Linus Walleij

pinctrl: drop owner assignment from platform_drivers

2014-10-20T14:21:21+00:00

A platform_driver does not need to set an owner, it will be populated by the
driver core.

Signed-off-by: Wolfram Sang

pinctrl: single: fix freudian slip

2014-09-05T08:33:25+00:00

commit 03e9f0cac5da6af85758276cb4624caf5911f2b9
"pinctrl: clean up after enable refactoring"
renamed the vtable callback .enable to .set_mux. The
renaming was done manually, and one of the alterations
contained a freudian slip. I confess, I am human.

Signed-off-by: Linus Walleij

pinctrl: clean up after enable refactoring

2014-09-04T08:05:07+00:00

commit 2243a87d90b42eb38bc281957df3e57c712b5e56
"pinctrl: avoid duplicated calling enable_pinmux_setting for a pin"
removed the .disable callback from the struct pinmux_ops,
making the .enable() callback the only remaining callback.

However .enable() is a bad name as it seems to imply that a
muxing can also be disabled. Rename the callback to .set_mux()
and also take this opportunity to clean out any remaining
mentions of .disable() from the documentation.

Acked-by: Stephen Warren 
Acked-by: Bjorn Andersson 
Acked-by: Fan Wu 
Signed-off-by: Linus Walleij

pinctrl: single: AM437x: Add pinctrl compatibility

2014-08-29T07:28:35+00:00

AM437x pinctrl definitions now differ from traditional 16 bit OMAP pin
ctrl definitions, in that all 32 bits are used to describe a single pin

Also the location of wakeupenable and event bits have changed.

Signed-off-by: Keerthy 
[nm@ti.com: minor updates]
Signed-off-by: Nishanth Menon 
Acked-by: Tony Lindgren 
Signed-off-by: Linus Walleij

pinctrl: single: Add DRA7 pinctrl compatibility

2014-08-29T07:27:44+00:00

DRA7 pinctrl definitions now differ from traditional 16 bit OMAP pin
ctrl definitions, in that all 32 bits are used to describe a single pin

Also the location of wakeupenable and event bits have changed.

Signed-off-by: Nishanth Menon 
Acked-by: Tony Lindgren 
Signed-off-by: Linus Walleij

pinctrl: pinctrl-single.c: Cleaning up wrong format string usage

2014-07-11T12:08:35+00:00

%d in format string used, but the type is unsigned int

This was found using a static code analysis program called cppcheck

Signed-off-by: Rickard Strandqvist 
Signed-off-by: Linus Walleij