All of lore.kernel.org
 help / color / mirror / Atom feed
From: Shuah Khan <shuahkh@osg.samsung.com>
To: Mauro Carvalho Chehab <mchehab@osg.samsung.com>,
	Shuah Khan <shuah.kh@samsung.com>
Cc: "Laurent Pinchart" <laurent.pinchart@ideasonboard.com>,
	"Linux Media Mailing List" <linux-media@vger.kernel.org>,
	"Mauro Carvalho Chehab" <mchehab@infradead.org>,
	"Javier Martinez Canillas" <javier@osg.samsung.com>,
	"Rafael Lourenço de Lima Chehab" <chehabrafael@gmail.com>,
	"Lars-Peter Clausen" <lars@metafoo.de>,
	"Shuah Khan" <shuahkh@osg.samsung.com>,
	"Linux Media Mailing List" <linux-media@vger.kernel.org>
Subject: Re: [PATCH 4/4] [meida] media-device: dynamically allocate struct media_devnode
Date: Thu, 28 Apr 2016 08:50:59 -0600	[thread overview]
Message-ID: <57222353.6090107@osg.samsung.com> (raw)
In-Reply-To: <20160428084155.65c812b1@recife.lan>

On 04/28/2016 05:41 AM, Mauro Carvalho Chehab wrote:
> Em Wed, 27 Apr 2016 16:20:56 -0600
> Shuah Khan <shuah.kh@samsung.com> escreveu:
> 
>> On 03/24/2016 05:37 AM, Mauro Carvalho Chehab wrote:
>>> Em Thu, 24 Mar 2016 10:24:44 +0200
>>> Laurent Pinchart <laurent.pinchart@ideasonboard.com> escreveu:
>>>   
>>>> On Wednesday 23 Mar 2016 16:27:46 Mauro Carvalho Chehab wrote:  
>>>>> struct media_devnode is currently embedded at struct media_device.
>>>>>
>>>>> While this works fine during normal usage, it leads to a race
>>>>> condition during devnode unregister. the problem is that drivers
>>>>> assume that, after calling media_device_unregister(), the struct
>>>>> that contains media_device can be freed. This is not true, as it
>>>>> can't be freed until userspace closes all opened /dev/media devnodes.
>>>>>
>>>>> In other words, if the media devnode is still open, and media_device
>>>>> gets freed, any call to an ioctl will make the core to try to access
>>>>> struct media_device, with will cause an use-after-free and even GPF.
>>>>>
>>>>> Fix this by dynamically allocating the struct media_devnode and only
>>>>> freeing it when it is safe.    
>>
>> Hi Mauro,
>>
>> I think this is the patch you were referring to in response to the patch
>> I sent out. Looks like this is still under review and some outstanding
>> issues. This patch itself doesn't ensure media_devnode sticks around
>> until the last app. closes the cdev. More work is needed such as adding
>> cdev parent and providing kobject release function that can be called
>> from cdev-core which will free media_devnode when the last cdev ref
>> is gone.
>>
>> Anyway, since you asked me to do the fix on top of your patch, I am asking
>> to see if this patch is in a good shape for me to apply. As such, we no
>> longer have sound/us/media.c in the mix. Hence this patch needs work before
>> I can base my work on it.
> 
> Well, the only issue on it is that it needed to be rebased on the top of
> the current tree. I did such rebase for you, at:
> 	https://git.linuxtv.org/mchehab/experimental.git/log/?h=au0828-unbind-fixes-v5
> 
> As said before, it doesn't touch cdev, nor adds any kref.
> 
> I'm running it today with the stress test. So far (~100 unbind loops, with 5
> concurrent accesses via mc_nextgen_test), the only issue it got so
> far seems to be at V4L2 cdev stuff (not at the media side, but at the
> V4L2 API side):

Are you planning to debug this further to isolate the problem?

thanks,
-- Shuah

> 
> 
> [  445.428212] kasan: CONFIG_KASAN_INLINE enabled
> [  445.428346] kasan: GPF could be caused by NULL-ptr deref or user memory access
> [  445.428750] general protection fault: 0000 [#2] SMP KASAN
> [  445.429417] Modules linked in: ir_lirc_codec lirc_dev au8522_dig xc5000 tuner au8522_decoder au8522_common au0828 videobuf2_vmalloc videobuf2_memops videobuf2_v4l2 videobuf2_core tveeprom dvb_core rc_core v4l2_common videodev media cpufreq_powersave cpufreq_conservative cpufreq_userspace cpufreq_stats parport_pc ppdev lp parport snd_hda_codec_hdmi intel_rapl x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm iTCO_wdt iTCO_vendor_support irqbypass crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel i915 snd_hda_codec_realtek snd_hda_codec_generic sha256_ssse3 sha256_generic hmac drbg btusb btrtl btbcm aesni_intel evdev snd_hda_intel aes_x86_64 lrw btintel i2c_algo_bit snd_hda_codec gf128mul bluetooth glue_helper drm_kms_helper snd_hwdep ablk_helper cryptd snd_hda_core drm serio_raw
> [  445.431431]  rfkill sg snd_pcm pcspkr snd_timer mei_me snd mei lpc_ich i2c_i801 mfd_core soundcore battery dw_dmac video dw_dmac_core i2c_designware_platform i2c_designware_core acpi_pad button tpm_tis tpm ext4 crc16 jbd2 mbcache dm_mod sd_mod ahci libahci psmouse libata e1000e ehci_pci xhci_pci ptp scsi_mod ehci_hcd pps_core xhci_hcd fan thermal sdhci_acpi sdhci mmc_core i2c_hid hid
> [  445.431907] CPU: 0 PID: 32248 Comm: v4l_id Tainted: G      D         4.6.0-rc2+ #68
> [  445.431984] Hardware name:                  /NUC5i7RYB, BIOS RYBDWi35.86A.0350.2015.0812.1722 08/12/2015
> [  445.436336] task: ffff8803bde0b000 ti: ffff8803a22d8000 task.ti: ffff8803a22d8000
> [  445.440571] RIP: 0010:[<ffffffff81d77a61>]  [<ffffffff81d77a61>] usb_ifnum_to_if+0x31/0x270
> [  445.444745] RSP: 0018:ffff8803a22df8d0  EFLAGS: 00010282
> [  445.448765] RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 1ffff10074699c5e
> [  445.452648] RDX: 000000000000009e RSI: 0000000000000000 RDI: 00000000000004f0
> [  445.456339] RBP: ffff8803a22df908 R08: 0000000000000001 R09: 0000000000000001
> [  445.459885] R10: ffff8803a2254580 R11: 0000000000000000 R12: ffffffffa12be0c0
> [  445.463473] R13: ffff8803a34ceb3c R14: ffff8803a34cc080 R15: ffffffffa12a89b0
> [  445.466938] FS:  00007f4629888700(0000) GS:ffff8803c6800000(0000) knlGS:0000000000000000
> [  445.470457] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [  445.473960] CR2: 00007f0e00d19000 CR3: 00000003b0bca000 CR4: 00000000003406f0
> [  445.477441] Stack:
> [  445.480959]  ffff8803a3480ba0 ffff8803a34ce2f0 ffff8803a34cc000 ffffffffa12be0c0
> [  445.484563]  ffff8803a34ceb3c ffff8803a34cc080 ffffffffa12a89b0 ffff8803a22df940
> [  445.488178]  ffffffffa12a4a45 ffff8803a22df940 ffff8803a34cc000 0000000000000000
> [  445.491780] Call Trace:
> [  445.495317]  [<ffffffffa12a89b0>] ? au0828_v4l2_close+0x5a0/0x5a0 [au0828]
> [  445.498886]  [<ffffffffa12a4a45>] au0828_analog_stream_enable+0x85/0x330 [au0828]
> [  445.502461]  [<ffffffffa12a8b11>] au0828_v4l2_open+0x161/0x350 [au0828]
> [  445.506012]  [<ffffffffa12a89b0>] ? au0828_v4l2_close+0x5a0/0x5a0 [au0828]
> [  445.509531]  [<ffffffffa1169561>] v4l2_open+0x1d1/0x350 [videodev]
> [  445.513097]  [<ffffffff815cc071>] chrdev_open+0x1f1/0x580
> [  445.516643]  [<ffffffff815cbe80>] ? cdev_put+0x50/0x50
> [  445.520129]  [<ffffffff815b98a7>] do_dentry_open+0x5d7/0xac0
> [  445.523582]  [<ffffffff815cbe80>] ? cdev_put+0x50/0x50
> [  445.526994]  [<ffffffff815bc05b>] vfs_open+0x16b/0x1e0
> [  445.530442]  [<ffffffff815e1c0b>] ? may_open+0x14b/0x260
> [  445.533882]  [<ffffffff815eb3f7>] path_openat+0x4f7/0x3a00
> [  445.537253]  [<ffffffff8156cc95>] ? alloc_debug_processing+0x75/0x1c0
> [  445.540731]  [<ffffffff815eaf00>] ? vfs_create+0x390/0x390
> [  445.544198]  [<ffffffff811ad88e>] ? __kernel_text_address+0x7e/0xa0
> [  445.547650]  [<ffffffff8109154f>] ? print_context_stack+0x7f/0xf0
> [  445.551089]  [<ffffffff8124b110>] ? debug_check_no_locks_freed+0x290/0x290
> [  445.554512]  [<ffffffff815b105b>] ? create_object+0x3eb/0x940
> [  445.558004]  [<ffffffff8124a5f6>] ? trace_hardirqs_on_caller+0x16/0x590
> [  445.561511]  [<ffffffff815f1cd5>] do_filp_open+0x175/0x230
> [  445.565031]  [<ffffffff815f1b60>] ? user_path_mountpoint_at+0x40/0x40
> [  445.568540]  [<ffffffff822d8567>] ? _raw_spin_unlock+0x27/0x40
> [  445.572056]  [<ffffffff81615b1a>] ? __alloc_fd+0x19a/0x4b0
> [  445.575491]  [<ffffffff8156d653>] ? kmem_cache_alloc+0x233/0x300
> [  445.579023]  [<ffffffff815bc615>] do_sys_open+0x195/0x340
> [  445.582510]  [<ffffffff8123eb5f>] ? up_read+0x1f/0x40
> [  445.585079]  [<ffffffff815bc480>] ? filp_open+0x60/0x60
> [  445.588457]  [<ffffffff81242681>] ? trace_hardirqs_off_caller+0x21/0x290
> [  445.591960]  [<ffffffff8100401b>] ? trace_hardirqs_on_thunk+0x1b/0x1d
> [  445.595413]  [<ffffffff815bc7de>] SyS_open+0x1e/0x20
> [  445.598904]  [<ffffffff822d8dc0>] entry_SYSCALL_64_fastpath+0x23/0xc1
> [  445.602394]  [<ffffffff81242681>] ? trace_hardirqs_off_caller+0x21/0x290
> [  445.605902] Code: 48 b8 00 00 00 00 00 fc ff df 48 89 e5 41 57 41 56 41 55 41 54 53 48 89 fb 48 81 c7 f0 04 00 00 48 89 fa 48 c1 ea 03 48 83 ec 10 <80> 3c 02 00 0f 85 c6 01 00 00 4c 8b bb f0 04 00 00 4d 85 ff 0f 
> [  445.609826] RIP  [<ffffffff81d77a61>] usb_ifnum_to_if+0x31/0x270
> [  445.613516]  RSP <ffff8803a22df8d0>
> [  445.617247] ---[ end trace 9127ab975e0f4104 ]---
> 
> I got two of those errors already.
> 
> 
>> Lars gave a few comments on the patch I sent out in the code that makes
>> devnode dynamic which are relevant to be folded into your patch. Added
>> Lars to this thread.
> 
> Yeah, I saw Lars comments. I'll answer to his emails in a few.
> 
>>
>> P.S: removed alsa folks and alsa list and added linux-media
>>
>> thanks,
>> -- Shuah
>>>>
>>>> We have the exact same issue with video_device, and there we've solved the 
>>>> problem by passing the release call all the way up to the driver. I'm open to 
>>>> discuss what the best solution is, but I don't think we should special-case 
>>>> media_device unless there are compelling arguments regarding why different 
>>>> solutions are needed for media_device and video_device.  
>>>
>>> The relationship between a video driver and  video_device/v4l2_dev is
>>> different. On V4L2 we have:
>>> 	- One driver using video_device resources;
>>> 	- multiple video_device devnodes.
>>>
>>> For media devices, the relationship is the opposite:
>>> 	- multiple independent drivers using media_devnode.
>>> 	- One media device node;
>>>
>>> On media devices, when multiple drivers are sharing the same devnode, the
>>> .probe() order can be different than the .release() order.
>>>
>>> So, we don't need to use the same solution as we did for video_device
>>> on media controller. Actually, the V4L2 solution won't work.
>>>
>>> On V4L2, a video device is typically initialized with:
>>>
>>>         video-dev->release = video_device_release;
>>>         err = video_register_device(video_dev,VFL_TYPE_GRABBER,
>>>                                     video_nr[dev->nr]);
>>>
>>> And video_device_release is simply a kfree:
>>>
>>> void video_device_release(struct video_device *vdev)
>>> {
>>>         kfree(vdev);
>>> }
>>>
>>> The caller driver may opt to use its own code to free the resources
>>> instead of the core one, but it needs to free vdev in the end
>>> (or some struct that embedds it).
>>>
>>> In the specific case of media, drivers don't need to touch or even
>>> be aware of media_devnode, as the creation of the media devnode is
>>> handled internally by the core. Also, there's no good reason to
>>> make the caller drivers to be aware of that.
>>>
>>> So, the approach taken by this patch is actually simpler, as the
>>> kfree() is internal to the core, and it doesn't require
>>> any callbacks. This patch provides all that it is needed to make devnode
>>> destroy safe. 
>>>
>>> On the common case where one driver allocates one /dev/media devnode,
>>> using the standard media_device_register()/media_device_unregister(),
>>> grants that a media_devnode instance will only be freed after all uses
>>> have gone, including open() descriptors. It also grants that the caller
>>> can free its own resources after media_device_unregister(), because
>>> media_devnode won't use media_device anymore.
>>>
>>> This happens because media_devnode_is_registered() will return
>>> false after media_device_unregister(), and the media_ioctl logic
>>> will return an error in this case:
>>> __media_ioctl(struct file *filp, unsigned int cmd, unsigned long arg,
>>> 	      long (*ioctl_func)(struct file *filp, unsigned int cmd,
>>> 				 unsigned long arg))
>>> {
>>> 	struct media_devnode *devnode = media_devnode_data(filp);
>>>
>>> 	if (!ioctl_func)
>>> 		return -ENOTTY;
>>>
>>> 	if (!media_devnode_is_registered(devnode))
>>> 		return -EIO;
>>> 		/* IMHO, it should be -ENODEV here */
>>>
>>> 	return ioctl_func(filp, cmd, arg);
>>> }
>>>
>>> all other syscalls have a similar test.
>>>
>>> When more than one driver shares the same media devnode - e. g. the
>>> case that it is currently using media_device_*_devres(), the V4L2
>>> solution of exposing the .release() callback to the caller driver
>>> won't work, as the unbind order can be different than the binding
>>> one. So, it is not possible to have .release() callbacks.
>>>
>>> On the multiple drivers scenario, a kref is used to identify when
>>> all drivers called media_device_unregister_devres(). Only when the
>>> last driver called it, it will do the actual media_device cleanups
>>> and will wait for userspace to close all opened file descriptors,
>>> calling kfree(media_devnode) only after that. It is also safe for
>>> a device driver to cleanup its own resources after
>>> media_device_release_devres(), as, if the driver is not the last
>>> one, media_device and media_devnode will still be allocated, and,
>>> if it is the last one, this will fallback on the case of a single
>>> driver.
>>>
>>> I can't think on any other race-free solution than the one implemented
>>> by this patch, and still being simple.
>>>   
>>>> I also suspect we will need to consider dynamic pipeline management sooner 
>>>> than later to solve the problem properly if we don't want to create code today 
>>>> that will need to be completely reworked tomorrow.  
>>>
>>> On the stress testing we're doing, we're removing/recreating part of the
>>> graph, by unbinding/rebinding one one of the drivers, while keep calling
>>> G_TOPOLOGY on an endless loop.
>>>
>>> It is working quite well. The change from semaphore->mutex, suggested
>>> by Sakari seemed to solve all the locking issues we had before.
>>>
>>> Ok, I didn't test SETUP_LINK, but, as it is now protected by the same
>>> mutex, except for some hidden bug, I guess it will work just fine.
>>>
>>> So, I don't see any need to change the locking schema at the core,
>>> to avoid race issues.
>>>   
>>>>  
>>>>> Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
>>>>> ---
>>>>>  drivers/media/media-device.c           | 38 +++++++++++++++++++------------
>>>>>  drivers/media/media-devnode.c          |  7 ++++++-
>>>>>  drivers/media/usb/au0828/au0828-core.c |  4 ++--
>>>>>  drivers/media/usb/uvc/uvc_driver.c     |  2 +-
>>>>>  include/media/media-device.h           |  5 +----
>>>>>  include/media/media-devnode.h          | 15 ++++++++++++--
>>>>>  sound/usb/media.c                      |  8 +++----
>>>>>  7 files changed, 52 insertions(+), 27 deletions(-)
>>>>>
>>>>> diff --git a/drivers/media/media-device.c b/drivers/media/media-device.c
>>>>> index 10cc4766de10..d10dc615e7a8 100644
>>>>> --- a/drivers/media/media-device.c
>>>>> +++ b/drivers/media/media-device.c
>>>>> @@ -428,7 +428,7 @@ static long media_device_ioctl(struct file *filp,
>>>>> unsigned int cmd, unsigned long arg)
>>>>>  {
>>>>>  	struct media_devnode *devnode = media_devnode_data(filp);
>>>>> -	struct media_device *dev = to_media_device(devnode);
>>>>> +	struct media_device *dev = devnode->media_dev;
>>>>>  	long ret;
>>>>>
>>>>>  	switch (cmd) {
>>>>> @@ -504,7 +504,7 @@ static long media_device_compat_ioctl(struct file *filp,
>>>>> unsigned int cmd, unsigned long arg)
>>>>>  {
>>>>>  	struct media_devnode *devnode = media_devnode_data(filp);
>>>>> -	struct media_device *dev = to_media_device(devnode);
>>>>> +	struct media_device *dev = devnode->media_dev;
>>>>>  	long ret;
>>>>>
>>>>>  	switch (cmd) {
>>>>> @@ -540,7 +540,8 @@ static const struct media_file_operations
>>>>> media_device_fops = { static ssize_t show_model(struct device *cd,
>>>>>  			  struct device_attribute *attr, char *buf)
>>>>>  {
>>>>> -	struct media_device *mdev = to_media_device(to_media_devnode(cd));
>>>>> +	struct media_devnode *devnode = to_media_devnode(cd);
>>>>> +	struct media_device *mdev = devnode->media_dev;
>>>>>
>>>>>  	return sprintf(buf, "%.*s\n", (int)sizeof(mdev->model), mdev->model);
>>>>>  }
>>>>> @@ -718,25 +719,36 @@ EXPORT_SYMBOL_GPL(media_device_cleanup);
>>>>>  int __must_check __media_device_register(struct media_device *mdev,
>>>>>  					 struct module *owner)
>>>>>  {
>>>>> +	struct media_devnode *devnode;
>>>>>  	int ret;
>>>>>
>>>>>  	mutex_lock(&mdev->graph_mutex);
>>>>>
>>>>> +	devnode = kzalloc(sizeof(*devnode), GFP_KERNEL);
>>>>> +	if (!devnode)
>>>>> +		return -ENOMEM;
>>>>> +
>>>>>  	/* Register the device node. */
>>>>> -	mdev->devnode.fops = &media_device_fops;
>>>>> -	mdev->devnode.parent = mdev->dev;
>>>>> -	mdev->devnode.release = media_device_release;
>>>>> +	mdev->devnode = devnode;
>>>>> +	devnode->fops = &media_device_fops;
>>>>> +	devnode->parent = mdev->dev;
>>>>> +	devnode->release = media_device_release;
>>>>>
>>>>>  	/* Set version 0 to indicate user-space that the graph is static */
>>>>>  	mdev->topology_version = 0;
>>>>>
>>>>> -	ret = media_devnode_register(&mdev->devnode, owner);
>>>>> -	if (ret < 0)
>>>>> +	ret = media_devnode_register(mdev, devnode, owner);
>>>>> +	if (ret < 0) {
>>>>> +		mdev->devnode = NULL;
>>>>> +		kfree(devnode);
>>>>>  		goto err;
>>>>> +	}
>>>>>
>>>>> -	ret = device_create_file(&mdev->devnode.dev, &dev_attr_model);
>>>>> +	ret = device_create_file(&devnode->dev, &dev_attr_model);
>>>>>  	if (ret < 0) {
>>>>> -		media_devnode_unregister(&mdev->devnode);
>>>>> +		mdev->devnode = NULL;
>>>>> +		media_devnode_unregister(devnode);
>>>>> +		kfree(devnode);
>>>>>  		goto err;
>>>>>  	}
>>>>>
>>>>> @@ -800,9 +812,9 @@ static void __media_device_unregister(struct
>>>>> media_device *mdev) }
>>>>>
>>>>>  	/* Check if mdev devnode was registered */
>>>>> -	if (media_devnode_is_registered(&mdev->devnode)) {
>>>>> -		device_remove_file(&mdev->devnode.dev, &dev_attr_model);
>>>>> -		media_devnode_unregister(&mdev->devnode);
>>>>> +	if (media_devnode_is_registered(mdev->devnode)) {
>>>>> +		device_remove_file(&mdev->devnode->dev, &dev_attr_model);
>>>>> +		media_devnode_unregister(mdev->devnode);
>>>>>  	}
>>>>>
>>>>>  	dev_dbg(mdev->dev, "Media device unregistered\n");
>>>>> diff --git a/drivers/media/media-devnode.c b/drivers/media/media-devnode.c
>>>>> index ae2bc0b7a368..db47063d8801 100644
>>>>> --- a/drivers/media/media-devnode.c
>>>>> +++ b/drivers/media/media-devnode.c
>>>>> @@ -44,6 +44,7 @@
>>>>>  #include <linux/uaccess.h>
>>>>>
>>>>>  #include <media/media-devnode.h>
>>>>> +#include <media/media-device.h>
>>>>>
>>>>>  #define MEDIA_NUM_DEVICES	256
>>>>>  #define MEDIA_NAME		"media"
>>>>> @@ -74,6 +75,8 @@ static void media_devnode_release(struct device *cd)
>>>>>  	/* Release media_devnode and perform other cleanups as needed. */
>>>>>  	if (devnode->release)
>>>>>  		devnode->release(devnode);
>>>>> +
>>>>> +	kfree(devnode);
>>>>>  }
>>>>>
>>>>>  static struct bus_type media_bus_type = {
>>>>> @@ -218,7 +221,8 @@ static const struct file_operations media_devnode_fops =
>>>>> { .llseek = no_llseek,
>>>>>  };
>>>>>
>>>>> -int __must_check media_devnode_register(struct media_devnode *devnode,
>>>>> +int __must_check media_devnode_register(struct media_device *mdev,
>>>>> +					struct media_devnode *devnode,
>>>>>  					struct module *owner)
>>>>>  {
>>>>>  	int minor;
>>>>> @@ -237,6 +241,7 @@ int __must_check media_devnode_register(struct
>>>>> media_devnode *devnode, mutex_unlock(&media_devnode_lock);
>>>>>
>>>>>  	devnode->minor = minor;
>>>>> +	devnode->media_dev = mdev;
>>>>>
>>>>>  	/* Part 2: Initialize and register the character device */
>>>>>  	cdev_init(&devnode->cdev, &media_devnode_fops);
>>>>> diff --git a/drivers/media/usb/au0828/au0828-core.c
>>>>> b/drivers/media/usb/au0828/au0828-core.c index 85c13ca5178f..598a85388d77
>>>>> 100644
>>>>> --- a/drivers/media/usb/au0828/au0828-core.c
>>>>> +++ b/drivers/media/usb/au0828/au0828-core.c
>>>>> @@ -142,7 +142,7 @@ static void au0828_unregister_media_device(struct
>>>>> au0828_dev *dev) struct media_device *mdev = dev->media_dev;
>>>>>  	struct media_entity_notify *notify, *nextp;
>>>>>
>>>>> -	if (!mdev || !media_devnode_is_registered(&mdev->devnode))
>>>>> +	if (!mdev || !media_devnode_is_registered(mdev->devnode))
>>>>>  		return;
>>>>>
>>>>>  	/* Remove au0828 entity_notify callbacks */
>>>>> @@ -480,7 +480,7 @@ static int au0828_media_device_register(struct
>>>>> au0828_dev *dev, if (!dev->media_dev)
>>>>>  		return 0;
>>>>>
>>>>> -	if (!media_devnode_is_registered(&dev->media_dev->devnode)) {
>>>>> +	if (!media_devnode_is_registered(dev->media_dev->devnode)) {
>>>>>
>>>>>  		/* register media device */
>>>>>  		ret = media_device_register(dev->media_dev);
>>>>> diff --git a/drivers/media/usb/uvc/uvc_driver.c
>>>>> b/drivers/media/usb/uvc/uvc_driver.c index 451e84e962e2..302e284a95eb
>>>>> 100644
>>>>> --- a/drivers/media/usb/uvc/uvc_driver.c
>>>>> +++ b/drivers/media/usb/uvc/uvc_driver.c
>>>>> @@ -1674,7 +1674,7 @@ static void uvc_delete(struct uvc_device *dev)
>>>>>  	if (dev->vdev.dev)
>>>>>  		v4l2_device_unregister(&dev->vdev);
>>>>>  #ifdef CONFIG_MEDIA_CONTROLLER
>>>>> -	if (media_devnode_is_registered(&dev->mdev.devnode))
>>>>> +	if (media_devnode_is_registered(dev->mdev.devnode))
>>>>>  		media_device_unregister(&dev->mdev);
>>>>>  	media_device_cleanup(&dev->mdev);
>>>>>  #endif
>>>>> diff --git a/include/media/media-device.h b/include/media/media-device.h
>>>>> index e59772ed8494..b04cfa907350 100644
>>>>> --- a/include/media/media-device.h
>>>>> +++ b/include/media/media-device.h
>>>>> @@ -347,7 +347,7 @@ struct media_entity_notify {
>>>>>  struct media_device {
>>>>>  	/* dev->driver_data points to this struct. */
>>>>>  	struct device *dev;
>>>>> -	struct media_devnode devnode;
>>>>> +	struct media_devnode *devnode;
>>>>>
>>>>>  	char model[32];
>>>>>  	char driver_name[32];
>>>>> @@ -403,9 +403,6 @@ struct usb_device;
>>>>>  #define MEDIA_DEV_NOTIFY_PRE_LINK_CH	0
>>>>>  #define MEDIA_DEV_NOTIFY_POST_LINK_CH	1
>>>>>
>>>>> -/* media_devnode to media_device */
>>>>> -#define to_media_device(node) container_of(node, struct media_device,
>>>>> devnode) -
>>>>>  /**
>>>>>   * media_entity_enum_init - Initialise an entity enumeration
>>>>>   *
>>>>> diff --git a/include/media/media-devnode.h b/include/media/media-devnode.h
>>>>> index e1d5af077adb..cc2b3155593c 100644
>>>>> --- a/include/media/media-devnode.h
>>>>> +++ b/include/media/media-devnode.h
>>>>> @@ -33,6 +33,8 @@
>>>>>  #include <linux/device.h>
>>>>>  #include <linux/cdev.h>
>>>>>
>>>>> +struct media_device;
>>>>> +
>>>>>  /*
>>>>>   * Flag to mark the media_devnode struct as registered. Drivers must not
>>>>> touch * this flag directly, it will be set and cleared by
>>>>> media_devnode_register and @@ -81,6 +83,8 @@ struct media_file_operations {
>>>>>   * before registering the node.
>>>>>   */
>>>>>  struct media_devnode {
>>>>> +	struct media_device *media_dev;
>>>>> +
>>>>>  	/* device ops */
>>>>>  	const struct media_file_operations *fops;
>>>>>
>>>>> @@ -103,7 +107,8 @@ struct media_devnode {
>>>>>  /**
>>>>>   * media_devnode_register - register a media device node
>>>>>   *
>>>>> - * @devnode: media device node structure we want to register
>>>>> + * @media_dev: struct media_device we want to register a device node
>>>>> + * @devnode: the device node to unregister
>>>>>   * @owner: should be filled with %THIS_MODULE
>>>>>   *
>>>>>   * The registration code assigns minor numbers and registers the new device
>>>>> node @@ -116,7 +121,8 @@ struct media_devnode {
>>>>>   * the media_devnode structure is *not* called, so the caller is
>>>>> responsible for * freeing any data.
>>>>>   */
>>>>> -int __must_check media_devnode_register(struct media_devnode *devnode,
>>>>> +int __must_check media_devnode_register(struct media_device *mdev,
>>>>> +					struct media_devnode *devnode,
>>>>>  					struct module *owner);
>>>>>
>>>>>  /**
>>>>> @@ -146,9 +152,14 @@ static inline struct media_devnode
>>>>> *media_devnode_data(struct file *filp) *	false otherwise.
>>>>>   *
>>>>>   * @devnode: pointer to struct &media_devnode.
>>>>> + *
>>>>> + * Note: If mdev is NULL, it also returns false.
>>>>>   */
>>>>>  static inline int media_devnode_is_registered(struct media_devnode
>>>>> *devnode) {
>>>>> +	if (!devnode)
>>>>> +		return false;
>>>>> +
>>>>>  	return test_bit(MEDIA_FLAG_REGISTERED, &devnode->flags);
>>>>>  }
>>>>>
>>>>> diff --git a/sound/usb/media.c b/sound/usb/media.c
>>>>> index 6db4878045e5..8fed0adec9e1 100644
>>>>> --- a/sound/usb/media.c
>>>>> +++ b/sound/usb/media.c
>>>>> @@ -136,7 +136,7 @@ void media_snd_stream_delete(struct snd_usb_substream
>>>>> *subs) struct media_device *mdev;
>>>>>
>>>>>  		mdev = mctl->media_dev;
>>>>> -		if (mdev && media_devnode_is_registered(&mdev->devnode)) {
>>>>> +		if (mdev && media_devnode_is_registered(mdev->devnode)) {
>>>>>  			media_devnode_remove(mctl->intf_devnode);
>>>>>  			media_device_unregister_entity(&mctl->media_entity);
>>>>>  			media_entity_cleanup(&mctl->media_entity);
>>>>> @@ -241,14 +241,14 @@ static void media_snd_mixer_delete(struct
>>>>> snd_usb_audio *chip) if (!mixer->media_mixer_ctl)
>>>>>  			continue;
>>>>>
>>>>> -		if (media_devnode_is_registered(&mdev->devnode)) {
>>>>> +		if (media_devnode_is_registered(mdev->devnode)) {
>>>>>  			media_device_unregister_entity(&mctl->media_entity);
>>>>>  			media_entity_cleanup(&mctl->media_entity);
>>>>>  		}
>>>>>  		kfree(mctl);
>>>>>  		mixer->media_mixer_ctl = NULL;
>>>>>  	}
>>>>> -	if (media_devnode_is_registered(&mdev->devnode))
>>>>> +	if (media_devnode_is_registered(mdev->devnode))
>>>>>  		media_devnode_remove(chip->ctl_intf_media_devnode);
>>>>>  	chip->ctl_intf_media_devnode = NULL;
>>>>>  }
>>>>> @@ -268,7 +268,7 @@ int media_snd_device_create(struct snd_usb_audio *chip,
>>>>>  	if (!mdev->dev)
>>>>>  		media_device_usb_init(mdev, usbdev, NULL);
>>>>>
>>>>> -	if (!media_devnode_is_registered(&mdev->devnode)) {
>>>>> +	if (!media_devnode_is_registered(mdev->devnode)) {
>>>>>  		ret = media_device_register(mdev);
>>>>>  		if (ret) {
>>>>>  			dev_err(&usbdev->dev,    
>>>>  
>>>
>>>   
>>
>>
> 
> 


  reply	other threads:[~2016-04-28 14:51 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-03-23 19:27 [PATCH 0/4] Some fixes and cleanups for the Media Controller code Mauro Carvalho Chehab
2016-03-23 19:27 ` [PATCH 1/4] [media] media-device: Simplify compat32 logic Mauro Carvalho Chehab
2016-03-24  8:40   ` Laurent Pinchart
2016-03-23 19:27 ` [PATCH 2/4] [media] media-devnode: fix namespace mess Mauro Carvalho Chehab
2016-03-23 19:27 ` [PATCH 3/4] [media] media-device: get rid of a leftover comment Mauro Carvalho Chehab
2016-03-24  8:41   ` Laurent Pinchart
2016-03-23 19:27 ` [PATCH 4/4] [meida] media-device: dynamically allocate struct media_devnode Mauro Carvalho Chehab
2016-03-23 20:28   ` kbuild test robot
2016-03-23 20:28     ` kbuild test robot
2016-03-24  8:24   ` Laurent Pinchart
2016-03-24 11:37     ` Mauro Carvalho Chehab
2016-04-27 22:20       ` Shuah Khan
2016-04-28 11:41         ` Mauro Carvalho Chehab
2016-04-28 14:50           ` Shuah Khan [this message]
2016-04-28 15:04             ` Mauro Carvalho Chehab
2016-04-28 16:23               ` Shuah Khan
2016-04-28 16:43                 ` Mauro Carvalho Chehab
2016-03-24 10:03 ` [PATCH 0/4] Some fixes and cleanups for the Media Controller code Sakari Ailus

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=57222353.6090107@osg.samsung.com \
    --to=shuahkh@osg.samsung.com \
    --cc=chehabrafael@gmail.com \
    --cc=javier@osg.samsung.com \
    --cc=lars@metafoo.de \
    --cc=laurent.pinchart@ideasonboard.com \
    --cc=linux-media@vger.kernel.org \
    --cc=mchehab@infradead.org \
    --cc=mchehab@osg.samsung.com \
    --cc=shuah.kh@samsung.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.