linux-toradex.git/drivers/video/sh_mobile_lcdcfb.c, branch v3.12.34

fb: reorder the lock sequence to fix potential dead lock

2014-04-18T09:03:37+00:00

commit 3a41c5dbe8bc396a7fb16ca8739e945bb003342e upstream.

Following commits:

50e244cc79 fb: rework locking to fix lock ordering on takeover
e93a9a8687 fb: Yet another band-aid for fixing lockdep mess
054430e773 fbcon: fix locking harder

reworked locking to fix related lock ordering on takeover, and introduced console_lock
into fbmem, but it seems that the new lock sequence(fb_info->lock ---> console_lock)
is against with the one in console_callback(console_lock ---> fb_info->lock), and leads to
a potential dead lock as following:

[  601.079000] ======================================================
[  601.079000] [ INFO: possible circular locking dependency detected ]
[  601.079000] 3.11.0 #189 Not tainted
[  601.079000] -------------------------------------------------------
[  601.079000] kworker/0:3/619 is trying to acquire lock:
[  601.079000]  (&fb_info->lock){+.+.+.}, at: [] lock_fb_info+0x26/0x60
[  601.079000]
but task is already holding lock:
[  601.079000]  (console_lock){+.+.+.}, at: [] console_callback+0x13/0x160
[  601.079000]
which lock already depends on the new lock.

[  601.079000]
the existing dependency chain (in reverse order) is:
[  601.079000]
-> #1 (console_lock){+.+.+.}:
[  601.079000]        [] lock_acquire+0xa1/0x140
[  601.079000]        [] console_lock+0x77/0x80
[  601.079000]        [] register_framebuffer+0x1d8/0x320
[  601.079000]        [] efifb_probe+0x408/0x48f
[  601.079000]        [] platform_drv_probe+0x43/0x80
[  601.079000]        [] driver_probe_device+0x8b/0x390
[  601.079000]        [] __driver_attach+0xab/0xb0
[  601.079000]        [] bus_for_each_dev+0x5d/0xa0
[  601.079000]        [] driver_attach+0x1e/0x20
[  601.079000]        [] bus_add_driver+0x117/0x290
[  601.079000]        [] driver_register+0x7a/0x170
[  601.079000]        [] __platform_driver_register+0x4a/0x50
[  601.079000]        [] platform_driver_probe+0x1d/0xb0
[  601.079000]        [] efifb_init+0x273/0x292
[  601.079000]        [] do_one_initcall+0x102/0x1c0
[  601.079000]        [] kernel_init_freeable+0x15d/0x1ef
[  601.079000]        [] kernel_init+0xe/0xf0
[  601.079000]        [] ret_from_fork+0x7c/0xb0
[  601.079000]
-> #0 (&fb_info->lock){+.+.+.}:
[  601.079000]        [] __lock_acquire+0x1e18/0x1f10
[  601.079000]        [] lock_acquire+0xa1/0x140
[  601.079000]        [] mutex_lock_nested+0x7a/0x3b0
[  601.079000]        [] lock_fb_info+0x26/0x60
[  601.079000]        [] fbcon_blank+0x29b/0x2e0
[  601.079000]        [] do_blank_screen+0x1d8/0x280
[  601.079000]        [] console_callback+0x64/0x160
[  601.079000]        [] process_one_work+0x1f5/0x540
[  601.079000]        [] worker_thread+0x11c/0x370
[  601.079000]        [] kthread+0xed/0x100
[  601.079000]        [] ret_from_fork+0x7c/0xb0
[  601.079000]
other info that might help us debug this:

[  601.079000]  Possible unsafe locking scenario:

[  601.079000]        CPU0                    CPU1
[  601.079000]        ----                    ----
[  601.079000]   lock(console_lock);
[  601.079000]                                lock(&fb_info->lock);
[  601.079000]                                lock(console_lock);
[  601.079000]   lock(&fb_info->lock);
[  601.079000]
 *** DEADLOCK ***

so we reorder the lock sequence the same as it in console_callback() to
avoid this issue. And following Tomi's suggestion, fix these similar
issues all in fb subsystem.

Signed-off-by: Gu Zheng 
Signed-off-by: Tomi Valkeinen 
Signed-off-by: Jiri Slaby

fbdev: sh_mobile_lcdc: fixup B side hsync adjust settings

2013-03-27T13:47:24+00:00

The lcdc B side horizon output is shifted
if sh_mobile_lcdc_pan() was called.
This patch fixup this issue.
It is tested on R8A7740 Armadillo800eva HDMI output.
Special thanks to Fukushima-san, and Sano-san

Reported-by: Osamu Fukushima 
Signed-off-by: Hideyuki Sano 
Signed-off-by: Kuninori Morimoto 
Acked-by: Laurent Pinchart 
Signed-off-by: Simon Horman

Drivers: video: remove __dev* attributes.

2013-01-03T23:57:01+00:00

CONFIG_HOTPLUG is going away as an option.  As a result, the __dev*
markings need to be removed.

This change removes the use of __devinit, __devexit_p, __devinitdata,
__devinitconst, and __devexit from these drivers.

Based on patches originally written by Bill Pemberton, but redone by me
in order to handle some of the coding style issues better, by hand.

Cc: Bill Pemberton 
Signed-off-by: Greg Kroah-Hartman

fbdev: sh_mobile_lcdc: Make sh_mobile_lcdc_sys_bus_ops static

2012-11-21T14:50:02+00:00

The structure isn't used outside of its compilation unit, make it
static.

Signed-off-by: Laurent Pinchart

fbdev: sh_mobile_lcdc: Store the backlight brightness internally

2012-11-21T14:50:01+00:00

There's no need to query the hardware for the currenty brightness value
through a platform data callback when we can cache the value internally
in the LCDC driver.

Signed-off-by: Laurent Pinchart

fbdev: sh_mobile_lcdc: Remove priv argument from channel and overlay init

2012-11-21T14:50:00+00:00

The channel and overlay init functions operate on a channel and an
overlay, don't pass the priv parameter explicitly.

Signed-off-by: Laurent Pinchart

fbdev: sh_mobile_lcdc: Rename mode argument to modes

2012-11-21T14:50:00+00:00

The sh_mobile_lcdc_channel_fb_init() mode argument is used to pass a
list of modes, rename it to modes.

Signed-off-by: Laurent Pinchart

fbdev: sh_mobile_lcdc: Get display dimensions from the channel structure

2012-11-21T14:49:59+00:00

Get the display dimensions directly from the channel structure instead
of recomputing them from the monitor specs or accessing the platform
data.

Signed-off-by: Laurent Pinchart

fbdev: sh_mobile_lcdc: use dma_mmap_coherent

2012-11-21T14:31:45+00:00

fb_mmap() implemented in fbmem.c uses smem_start as the physical
address of the frame buffer.  In the sh_mobile_lcdc driver, the
smem_start is a dma_addr_t that is not a physical address when IOMMU is
enabled.  dma_mmap_coherent() maps the address correctly.

Signed-off-by: Hideki EIRAKU 
Signed-off-by: Laurent Pinchart

fbdev: sh_mobile_lcdc: Fix vertical panning step

2012-07-29T01:17:07+00:00

Commit 15dede882e564601947f2ce4b647742c0351be6d added support for
horizontal panning but accidentally computes the Y pan step value
incorrectly for NV12/21 and NV16/61 formats. Fix this.

Signed-off-by: Laurent Pinchart 
Signed-off-by: Florian Tobias Schandinat