[PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
The i2c-dev calls i2c_get_adapter() from the .open() callback, which doesn't let the adapter device unregister unless the .close() callback is called. On some platforms (like Google ARA), this doesn't let the modules (hardware attached to the phone) eject from the phone as the cleanup path for the module hasn't finished yet (i2c adapter not removed). We can't let the userspace block the kernel forever in such cases. Fix this by calling i2c_get_adapter() from all other file operations, i.e. read/write/ioctl, to make sure the adapter doesn't get away while we are in the middle of a operation, but not otherwise. In .open() we will release the adapter device before returning and so if there is no data transfer in progress, then the i2c-dev doesn't block the adapter from unregistering. Signed-off-by: Viresh Kumar --- drivers/i2c/i2c-dev.c | 72 ++- include/linux/i2c.h | 1 + 2 files changed, 66 insertions(+), 7 deletions(-) diff --git a/drivers/i2c/i2c-dev.c b/drivers/i2c/i2c-dev.c index 66f323fd3982..b2562603daa9 100644 --- a/drivers/i2c/i2c-dev.c +++ b/drivers/i2c/i2c-dev.c @@ -142,13 +142,25 @@ static ssize_t i2cdev_read(struct file *file, char __user *buf, size_t count, int ret; struct i2c_client *client = file->private_data; + struct i2c_adapter *adap; + + adap = i2c_get_adapter(client->adapter_nr); + if (!adap) + return -ENODEV; + + if (adap != client->adapter) { + ret = -EINVAL; + goto put_adapter; + } if (count > 8192) count = 8192; tmp = kmalloc(count, GFP_KERNEL); - if (tmp == NULL) - return -ENOMEM; + if (tmp == NULL) { + ret = -ENOMEM; + goto put_adapter; + } pr_debug("i2c-dev: i2c-%d reading %zu bytes.\n", iminor(file_inode(file)), count); @@ -157,6 +169,9 @@ static ssize_t i2cdev_read(struct file *file, char __user *buf, size_t count, if (ret >= 0) ret = copy_to_user(buf, tmp, count) ? -EFAULT : ret; kfree(tmp); + +put_adapter: + i2c_put_adapter(adap); return ret; } @@ -166,19 +181,34 @@ static ssize_t i2cdev_write(struct file *file, const char __user *buf, int ret; char *tmp; struct i2c_client *client = file->private_data; + struct i2c_adapter *adap; + + adap = i2c_get_adapter(client->adapter_nr); + if (!adap) + return -ENODEV; + + if (adap != client->adapter) { + ret = -EINVAL; + goto put_adapter; + } if (count > 8192) count = 8192; tmp = memdup_user(buf, count); - if (IS_ERR(tmp)) - return PTR_ERR(tmp); + if (IS_ERR(tmp)) { + ret = PTR_ERR(tmp); + goto put_adapter; + } pr_debug("i2c-dev: i2c-%d writing %zu bytes.\n", iminor(file_inode(file)), count); ret = i2c_master_send(client, tmp, count); kfree(tmp); + +put_adapter: + i2c_put_adapter(adap); return ret; } @@ -412,9 +442,9 @@ static noinline int i2cdev_ioctl_smbus(struct i2c_client *client, return res; } -static long i2cdev_ioctl(struct file *file, unsigned int cmd, unsigned long arg) +static long __i2cdev_ioctl(struct i2c_client *client, unsigned int cmd, + unsigned long arg) { - struct i2c_client *client = file->private_data; unsigned long funcs; dev_dbg(&client->adapter->dev, "ioctl, cmd=0x%02x, arg=0x%02lx\n", @@ -480,6 +510,28 @@ static long i2cdev_ioctl(struct file *file, unsigned int cmd, unsigned long arg) return 0; } +static long i2cdev_ioctl(struct file *file, unsigned int cmd, unsigned long arg) +{ + struct i2c_client *client = file->private_data; + struct i2c_adapter *adap; + unsigned long ret; + + adap = i2c_get_adapter(client->adapter_nr); + if (!adap) + return -ENODEV; + + if (adap != client->adapter) { + ret = -EINVAL; + goto put_adapter; + } + + ret = __i2cdev_ioctl(client, cmd, arg); + +put_adapter: + i2c_put_adapter(adap); + return ret; +} + static int i2cdev_open(struct inode *inode, struct file *file) { unsigned int minor = iminor(inode); @@ -504,9 +556,16 @@ static int i2cdev_open(struct inode *inode, struct file *file) } snprintf(client->name, I2C_NAME_SIZE, "i2c-dev %d", adap->nr); + client->adapter_nr = minor; client->adapter = adap; file->private_data = client; + /* +* Allow the adapter to unregister while userspace has opened the i2c +* device. +*/ + i2c_put_adapter(client->adapter); + return 0; } @@ -514,7 +573,6 @@ static int i2cdev_release(struct inode *inode, struct file *f
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On 11-07-16, 14:50, Greg KH wrote: > Ideally, yes, userspace will have closed that device node. But again, > hardware isn't kind and sometimes decides to be yanked out by users > before they tell the kernel about it. We handle this for almost all > other device subsystems, i2c is one of the last to be fixed up in this > manner. Sorry it's taken us well over a decade to get here :) > > hope that explains things better, @Jean: Ping :) -- viresh
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
Hi Viresh, On Wed, 6 Jul 2016 13:55:40 -0700, Viresh Kumar wrote: > On 06-07-16, 19:12, Jean Delvare wrote: > > Well well... I don't like this patch at all to be honest. > > Sure, I didn't like it much as well. I just wanted people to comment on what > else we can do here. We don't really want to add out-of-mainline stuff here. > > > My first question would be: what is keeping /dev/i2c-* open all the > > time? Originally i2c-dev was developed with development and debugging > > tools in mind (the i2c-tools suite.) The device nodes were never meant > > to be kept open for more than a few seconds. > > We thought that buggy userspace shouldn't be allowed to get kernel into > trouble. > Isn't that the case ? Buggy user-space run by root can cause any kind of trouble. And the point being discussed here is amongst the most minor of these. But I even disagree with calling it buggy. > In our case this is what happens: > - userspace opens the file descriptor > - we try to forcefully remove the module from phone (that doesn't talk to > userspace to stop using the device). > - The module doesn't get ejected unless the app closes the fd. So you are basically building a test case to cause the problem. It's artificial. The adapter reference being held while the device node is open isn't a bug, it is a design decision. I would consider revisiting that design if there was a real world case where it causes trouble, but not for an artificially created test case. I don't see anything fundamentally wrong in the design anyway. I do not expect to be allowed to remove a hard disk drive from my system while its partitions are mounted, and I don't expect to be able to unmount partitions while users have files opened on them. You always have to tear things down in the right order. Same here. > > Do you have user-space i2c device drivers on your system? Which ones, > > No. Its probably an app written by some of our module app developers. > > > and why (I would expect all useful i2c devices to have a kernel > > driver.) > > That's what we have. Still no details here. What app, what is it doing, to what device is it talking, why is it not a kernel driver, and why do they keep the device node opened all the time? > > Requesting and freeing the i2c adapter for every transaction is going > > Well, we are just finding it (taking a reference of it) and the dropping its > reference. Yes, that's what I meant, sorry for using the wrong terms. > > to add a lot of overhead to all existing tools :-( > > :( i2cdump typically runs 258 ioctls on the device node, i2cdetect 235. i2c_get_adapter isn't cheap. Multiple function calls, mutex locking/unlocking, preemption disabling/enabling... You don't want do to that repeatedly if it can be avoided. So, nack from me. > > It's not like every user can open i2c device nodes and block the > > system. Only selected users should be able to open i2c device nodes > > (only root by default) so they should be responsible for not > > misbehaving. > > Hmmm. The problem is that they weren't told when the module tries to go away > and > so they don't know that they need to close the fd. See my previous questions. We still don't know why they are doing what they are doing in user-space, nor why they think they have to keep the device node opened. They could always kill the application in question with: # fuser -k /dev/i2c-* before removing the module. Or find a more polite way to tell the application to quit. If they want to do it in user-space, they have to do it right. -- Jean Delvare SUSE L3 Support
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
Hi, [auto build test WARNING on wsa/i2c/for-next] [also build test WARNING on v4.7-rc6 next-20160705] [if your patch is applied to the wrong git tree, please drop us a note to help improve the system] url: https://github.com/0day-ci/linux/commits/Viresh-Kumar/i2c-dev-Don-t-let-userspace-block-adapter/20160706-110245 base: https://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux.git i2c/for-next reproduce: make htmldocs All warnings (new ones prefixed by >>): include/linux/init.h:1: warning: no structured comments found kernel/sched/core.c:2079: warning: No description found for parameter 'cookie' kernel/sys.c:1: warning: no structured comments found drivers/dma-buf/seqno-fence.c:1: warning: no structured comments found >> include/linux/i2c.h:242: warning: No description found for parameter >> 'adapter_nr' vim +/adapter_nr +242 include/linux/i2c.h d64f73be David Brownell 2007-07-12 226 * managing the device. ^1da177e Linus Torvalds 2005-04-16 227 */ ^1da177e Linus Torvalds 2005-04-16 228 struct i2c_client { 2096b956 David Brownell 2007-05-01 229unsigned short flags; /* div., see below */ 5071860a Jean Delvare2005-07-20 230unsigned short addr; /* chip address - NOTE: 7bit*/ ^1da177e Linus Torvalds 2005-04-16 231 /* addresses are stored in the */ 5071860a Jean Delvare2005-07-20 232 /* _LOWER_ 7 bits */ 2096b956 David Brownell 2007-05-01 233char name[I2C_NAME_SIZE]; ^1da177e Linus Torvalds 2005-04-16 234struct i2c_adapter *adapter; /* the adapter we sit on*/ ^1da177e Linus Torvalds 2005-04-16 235struct device dev; /* the device structure */ 8e29da9e Wolfram Sang2008-07-01 236int irq; /* irq issued by device */ 42fc20b8 Viresh Kumar2016-07-05 237int adapter_nr; 4735c98f Jean Delvare2008-07-14 238struct list_head detected; d5fd120e Jean Delvare2015-01-26 239 #if IS_ENABLED(CONFIG_I2C_SLAVE) 4b1acc43 Wolfram Sang2014-11-18 240i2c_slave_cb_t slave_cb; /* callback for slave mode */ d5fd120e Jean Delvare2015-01-26 241 #endif ^1da177e Linus Torvalds 2005-04-16 @242 }; ^1da177e Linus Torvalds 2005-04-16 243 #define to_i2c_client(d) container_of(d, struct i2c_client, dev) ^1da177e Linus Torvalds 2005-04-16 244 9b766b81 David Brownell 2008-01-27 245 extern struct i2c_client *i2c_verify_client(struct device *dev); 643dd09e Stephen Warren 2012-04-17 246 extern struct i2c_adapter *i2c_verify_adapter(struct device *dev); 9b766b81 David Brownell 2008-01-27 247 a61fc683 Ben Gardner 2005-07-27 248 static inline struct i2c_client *kobj_to_i2c_client(struct kobject *kobj) a61fc683 Ben Gardner 2005-07-27 249 { d75d53cd Mark M. Hoffman 2007-07-12 250struct device * const dev = container_of(kobj, struct device, kobj); :: The code at line 242 was first introduced by commit :: 1da177e4c3f41524e886b7f1b8a0c1fc7321cac2 Linux-2.6.12-rc2 :: TO: Linus Torvalds :: CC: Linus Torvalds --- 0-DAY kernel test infrastructureOpen Source Technology Center https://lists.01.org/pipermail/kbuild-all Intel Corporation .config.gz Description: Binary data
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On Tue, Jul 05, 2016 at 07:57:07PM -0700, Viresh Kumar wrote: > The i2c-dev calls i2c_get_adapter() from the .open() callback, which > doesn't let the adapter device unregister unless the .close() callback > is called. > > On some platforms (like Google ARA), this doesn't let the modules > (hardware attached to the phone) eject from the phone as the cleanup > path for the module hasn't finished yet (i2c adapter not removed). > > We can't let the userspace block the kernel forever in such cases. > > Fix this by calling i2c_get_adapter() from all other file operations, > i.e. read/write/ioctl, to make sure the adapter doesn't get away while > we are in the middle of a operation, but not otherwise. In .open() we > will release the adapter device before returning and so if there is no > data transfer in progress, then the i2c-dev doesn't block the adapter > from unregistering. > > Signed-off-by: Viresh Kumar I'd think Jean has more experience with I2C hotplugging approaches and difficulties, so I'd be interested in his high level review. However: > @@ -234,6 +234,7 @@ struct i2c_client { > struct i2c_adapter *adapter;/* the adapter we sit on*/ > struct device dev; /* the device structure */ > int irq;/* irq issued by device */ > + int adapter_nr; > struct list_head detected; Adding something to *every* i2c_client for this corner case sounds pretty expensive to me. Regards, Wolfram signature.asc Description: PGP signature
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On 2016-07-06 04:57, Viresh Kumar wrote: > The i2c-dev calls i2c_get_adapter() from the .open() callback, which > doesn't let the adapter device unregister unless the .close() callback > is called. > > On some platforms (like Google ARA), this doesn't let the modules > (hardware attached to the phone) eject from the phone as the cleanup > path for the module hasn't finished yet (i2c adapter not removed). > > We can't let the userspace block the kernel forever in such cases. > > Fix this by calling i2c_get_adapter() from all other file operations, > i.e. read/write/ioctl, to make sure the adapter doesn't get away while > we are in the middle of a operation, but not otherwise. In .open() we > will release the adapter device before returning and so if there is no > data transfer in progress, then the i2c-dev doesn't block the adapter > from unregistering. > > Signed-off-by: Viresh Kumar > --- > drivers/i2c/i2c-dev.c | 72 > ++- > include/linux/i2c.h | 1 + > 2 files changed, 66 insertions(+), 7 deletions(-) > > diff --git a/drivers/i2c/i2c-dev.c b/drivers/i2c/i2c-dev.c > index 66f323fd3982..b2562603daa9 100644 > --- a/drivers/i2c/i2c-dev.c > +++ b/drivers/i2c/i2c-dev.c > @@ -142,13 +142,25 @@ static ssize_t i2cdev_read(struct file *file, char > __user *buf, size_t count, > int ret; > > struct i2c_client *client = file->private_data; > + struct i2c_adapter *adap; > + > + adap = i2c_get_adapter(client->adapter_nr); > + if (!adap) > + return -ENODEV; > + > + if (adap != client->adapter) { > + ret = -EINVAL; > + goto put_adapter; > + } I don't see how this can work with the i2c-demux-pinctrl driver. I also wonder if/how other muxes handle this relaxed adapter lifetime thingy? Out of curiosity, why would client->adapter change anyway? (that is, if not because of a demux-pinctrl op) Or maybe I'm just paranoid, but this is not obvious... Cheers, Peter
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On 06-07-16, 15:55, Wolfram Sang wrote: > Adding something to *every* i2c_client for this corner case sounds > pretty expensive to me. I agree with you on that. I wanted to avoid it, but I couldn't :( Lets see how Jean suggests to handle it. -- viresh
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On 06-07-16, 10:22, Peter Rosin wrote: > On 2016-07-06 04:57, Viresh Kumar wrote: > > The i2c-dev calls i2c_get_adapter() from the .open() callback, which > > doesn't let the adapter device unregister unless the .close() callback > > is called. > > > > On some platforms (like Google ARA), this doesn't let the modules > > (hardware attached to the phone) eject from the phone as the cleanup > > path for the module hasn't finished yet (i2c adapter not removed). > > > > We can't let the userspace block the kernel forever in such cases. > > > > Fix this by calling i2c_get_adapter() from all other file operations, > > i.e. read/write/ioctl, to make sure the adapter doesn't get away while > > we are in the middle of a operation, but not otherwise. In .open() we > > will release the adapter device before returning and so if there is no > > data transfer in progress, then the i2c-dev doesn't block the adapter > > from unregistering. > > > > Signed-off-by: Viresh Kumar > > --- > > drivers/i2c/i2c-dev.c | 72 > > ++- > > include/linux/i2c.h | 1 + > > 2 files changed, 66 insertions(+), 7 deletions(-) > > > > diff --git a/drivers/i2c/i2c-dev.c b/drivers/i2c/i2c-dev.c > > index 66f323fd3982..b2562603daa9 100644 > > --- a/drivers/i2c/i2c-dev.c > > +++ b/drivers/i2c/i2c-dev.c > > @@ -142,13 +142,25 @@ static ssize_t i2cdev_read(struct file *file, char > > __user *buf, size_t count, > > int ret; > > > > struct i2c_client *client = file->private_data; > > + struct i2c_adapter *adap; > > + > > + adap = i2c_get_adapter(client->adapter_nr); > > + if (!adap) > > + return -ENODEV; > > + > > + if (adap != client->adapter) { > > + ret = -EINVAL; > > + goto put_adapter; > > + } > > I don't see how this can work with the i2c-demux-pinctrl driver. > I also wonder if/how other muxes handle this relaxed adapter > lifetime thingy? I would like to mention here that I am no I2C expert and have limited knowledge of it :) I haven't had a look at the muxes implementation earlier, now that I looked at them, I see that they unregister/register the adapter, perhaps while switching functionality. I am not sure though, if this patch will break it or not. And I don't have a way of testing it out. > Out of curiosity, why would client->adapter change anyway? > (that is, if not because of a demux-pinctrl op) I didn't mean that it will change, and perhaps we can add a WARN_ON(adap != client->adapter). But, thinking about it again now, I think it is possible. What about this sequence: - i2c-adap-register (address P1) - .open(), client->adapter = P1; - .read/write/ioctl().. - i2c-adap-unregister (adapter freed) - i2c-adap-register with same adapter_nr (address P2); - .read/write/ioctl(). Wouldn't the address differ here ? -- viresh
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On 07/06/2016 04:33 PM, Viresh Kumar wrote: > On 06-07-16, 10:22, Peter Rosin wrote: >> On 2016-07-06 04:57, Viresh Kumar wrote: >>> The i2c-dev calls i2c_get_adapter() from the .open() callback, which >>> doesn't let the adapter device unregister unless the .close() callback >>> is called. >>> >>> On some platforms (like Google ARA), this doesn't let the modules >>> (hardware attached to the phone) eject from the phone as the cleanup >>> path for the module hasn't finished yet (i2c adapter not removed). >>> >>> We can't let the userspace block the kernel forever in such cases. >>> >>> Fix this by calling i2c_get_adapter() from all other file operations, >>> i.e. read/write/ioctl, to make sure the adapter doesn't get away while >>> we are in the middle of a operation, but not otherwise. In .open() we >>> will release the adapter device before returning and so if there is no >>> data transfer in progress, then the i2c-dev doesn't block the adapter >>> from unregistering. >>> >>> Signed-off-by: Viresh Kumar >>> --- >>> drivers/i2c/i2c-dev.c | 72 >>> ++- >>> include/linux/i2c.h | 1 + >>> 2 files changed, 66 insertions(+), 7 deletions(-) >>> >>> diff --git a/drivers/i2c/i2c-dev.c b/drivers/i2c/i2c-dev.c >>> index 66f323fd3982..b2562603daa9 100644 >>> --- a/drivers/i2c/i2c-dev.c >>> +++ b/drivers/i2c/i2c-dev.c >>> @@ -142,13 +142,25 @@ static ssize_t i2cdev_read(struct file *file, char >>> __user *buf, size_t count, >>> int ret; >>> >>> struct i2c_client *client = file->private_data; >>> + struct i2c_adapter *adap; >>> + >>> + adap = i2c_get_adapter(client->adapter_nr); >>> + if (!adap) >>> + return -ENODEV; >>> + >>> + if (adap != client->adapter) { >>> + ret = -EINVAL; >>> + goto put_adapter; >>> + } >> >> I don't see how this can work with the i2c-demux-pinctrl driver. >> I also wonder if/how other muxes handle this relaxed adapter >> lifetime thingy? > > I would like to mention here that I am no I2C expert and have limited > knowledge of it :) > > I haven't had a look at the muxes implementation earlier, now that I > looked at them, I see that they unregister/register the adapter, > perhaps while switching functionality. > > I am not sure though, if this patch will break it or not. And I don't > have a way of testing it out. > >> Out of curiosity, why would client->adapter change anyway? >> (that is, if not because of a demux-pinctrl op) > > I didn't mean that it will change, and perhaps we can add a > WARN_ON(adap != client->adapter). > > But, thinking about it again now, I think it is possible. > > What about this sequence: > > - i2c-adap-register (address P1) > - .open(), client->adapter = P1; > - .read/write/ioctl().. > - i2c-adap-unregister (adapter freed) > - i2c-adap-register with same adapter_nr (address P2); > - .read/write/ioctl(). > > Wouldn't the address differ here ? There is no guarantee that the address will be different. While it is unlikely the memory allocated might give out the same address for the second adapter if the first one has been freed.
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On 2016-07-06 16:43, Lars-Peter Clausen wrote: > On 07/06/2016 04:33 PM, Viresh Kumar wrote: >> On 06-07-16, 10:22, Peter Rosin wrote: >>> On 2016-07-06 04:57, Viresh Kumar wrote: The i2c-dev calls i2c_get_adapter() from the .open() callback, which doesn't let the adapter device unregister unless the .close() callback is called. On some platforms (like Google ARA), this doesn't let the modules (hardware attached to the phone) eject from the phone as the cleanup path for the module hasn't finished yet (i2c adapter not removed). We can't let the userspace block the kernel forever in such cases. Fix this by calling i2c_get_adapter() from all other file operations, i.e. read/write/ioctl, to make sure the adapter doesn't get away while we are in the middle of a operation, but not otherwise. In .open() we will release the adapter device before returning and so if there is no data transfer in progress, then the i2c-dev doesn't block the adapter from unregistering. Signed-off-by: Viresh Kumar --- drivers/i2c/i2c-dev.c | 72 ++- include/linux/i2c.h | 1 + 2 files changed, 66 insertions(+), 7 deletions(-) diff --git a/drivers/i2c/i2c-dev.c b/drivers/i2c/i2c-dev.c index 66f323fd3982..b2562603daa9 100644 --- a/drivers/i2c/i2c-dev.c +++ b/drivers/i2c/i2c-dev.c @@ -142,13 +142,25 @@ static ssize_t i2cdev_read(struct file *file, char __user *buf, size_t count, int ret; struct i2c_client *client = file->private_data; + struct i2c_adapter *adap; + + adap = i2c_get_adapter(client->adapter_nr); + if (!adap) + return -ENODEV; + + if (adap != client->adapter) { + ret = -EINVAL; + goto put_adapter; + } >>> >>> I don't see how this can work with the i2c-demux-pinctrl driver. >>> I also wonder if/how other muxes handle this relaxed adapter >>> lifetime thingy? >> >> I would like to mention here that I am no I2C expert and have limited >> knowledge of it :) >> >> I haven't had a look at the muxes implementation earlier, now that I >> looked at them, I see that they unregister/register the adapter, >> perhaps while switching functionality. >> >> I am not sure though, if this patch will break it or not. And I don't >> have a way of testing it out. >> >>> Out of curiosity, why would client->adapter change anyway? >>> (that is, if not because of a demux-pinctrl op) >> >> I didn't mean that it will change, and perhaps we can add a >> WARN_ON(adap != client->adapter). >> >> But, thinking about it again now, I think it is possible. >> >> What about this sequence: >> >> - i2c-adap-register (address P1) >> - .open(), client->adapter = P1; >> - .read/write/ioctl().. >> - i2c-adap-unregister (adapter freed) >> - i2c-adap-register with same adapter_nr (address P2); >> - .read/write/ioctl(). >> >> Wouldn't the address differ here ? > > There is no guarantee that the address will be different. While it is > unlikely the memory allocated might give out the same address for the second > adapter if the first one has been freed. Exactly, so the stored address had better be correct, and in that case there is no need for the new adapter_nr in every client, you could just go with client->adapter->nr instead. Which just shows that the whole thing is fishy and that the adapter has to remain alive. BTW, is there any guarantee that adapter numbers will not get reused? Cheers, Peter
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On 06-07-16, 16:43, Lars-Peter Clausen wrote: > There is no guarantee that the address will be different. While it is > unlikely the memory allocated might give out the same address for the second > adapter if the first one has been freed. Oh yeah, thanks for correcting me :) -- viresh
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On 06-07-16, 17:04, Peter Rosin wrote: > Exactly, so the stored address had better be correct, and in No. > that case there is no need for the new adapter_nr in every > client, you could just go with client->adapter->nr instead. client->adapter may be a dangling pointer at this point if the adapter is freed, so we can't use that blindly for sure. > Which just shows that the whole thing is fishy and that the > adapter has to remain alive. BTW, is there any guarantee that > adapter numbers will not get reused? We are allocating them from idr and that will reuse them once they get freed. -- viresh
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On Wed, 6 Jul 2016 15:55:24 +0900, Wolfram Sang wrote: > On Tue, Jul 05, 2016 at 07:57:07PM -0700, Viresh Kumar wrote: > > The i2c-dev calls i2c_get_adapter() from the .open() callback, which > > doesn't let the adapter device unregister unless the .close() callback > > is called. > > > > On some platforms (like Google ARA), this doesn't let the modules > > (hardware attached to the phone) eject from the phone as the cleanup > > path for the module hasn't finished yet (i2c adapter not removed). > > > > We can't let the userspace block the kernel forever in such cases. > > > > Fix this by calling i2c_get_adapter() from all other file operations, > > i.e. read/write/ioctl, to make sure the adapter doesn't get away while > > we are in the middle of a operation, but not otherwise. In .open() we > > will release the adapter device before returning and so if there is no > > data transfer in progress, then the i2c-dev doesn't block the adapter > > from unregistering. > > > > Signed-off-by: Viresh Kumar > > I'd think Jean has more experience with I2C hotplugging approaches and > difficulties, so I'd be interested in his high level review. Well well... I don't like this patch at all to be honest. My first question would be: what is keeping /dev/i2c-* open all the time? Originally i2c-dev was developed with development and debugging tools in mind (the i2c-tools suite.) The device nodes were never meant to be kept open for more than a few seconds. Do you have user-space i2c device drivers on your system? Which ones, and why (I would expect all useful i2c devices to have a kernel driver.) And why do they keep their i2c device node opened all the time? Requesting and freeing the i2c adapter for every transaction is going to add a lot of overhead to all existing tools :-( It's not like every user can open i2c device nodes and block the system. Only selected users should be able to open i2c device nodes (only root by default) so they should be responsible for not misbehaving. -- Jean Delvare SUSE L3 Support
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On 06-07-16, 19:12, Jean Delvare wrote: > Well well... I don't like this patch at all to be honest. Sure, I didn't like it much as well. I just wanted people to comment on what else we can do here. We don't really want to add out-of-mainline stuff here. > My first question would be: what is keeping /dev/i2c-* open all the > time? Originally i2c-dev was developed with development and debugging > tools in mind (the i2c-tools suite.) The device nodes were never meant > to be kept open for more than a few seconds. We thought that buggy userspace shouldn't be allowed to get kernel into trouble. Isn't that the case ? In our case this is what happens: - userspace opens the file descriptor - we try to forcefully remove the module from phone (that doesn't talk to userspace to stop using the device). - The module doesn't get ejected unless the app closes the fd. > Do you have user-space i2c device drivers on your system? Which ones, No. Its probably an app written by some of our module app developers. > and why (I would expect all useful i2c devices to have a kernel > driver.) That's what we have. > Requesting and freeing the i2c adapter for every transaction is going Well, we are just finding it (taking a reference of it) and the dropping its reference. > to add a lot of overhead to all existing tools :-( :( > It's not like every user can open i2c device nodes and block the > system. Only selected users should be able to open i2c device nodes > (only root by default) so they should be responsible for not > misbehaving. Hmmm. The problem is that they weren't told when the module tries to go away and so they don't know that they need to close the fd. Also coming to the earlier thing, I though even the buggy userspace thing shouldn't be allowed to block kernel device unregisteration. -- viresh
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
Hi Greg, Viresh, On Mon, 11 Jul 2016 14:50:58 -0700, Greg KH wrote: > On Mon, Jul 11, 2016 at 02:22:09PM +0200, Jean Delvare wrote: > > So you are basically building a test case to cause the problem. It's > > artificial. The adapter reference being held while the device node is > > open isn't a bug, it is a design decision. I would consider revisiting > > that design if there was a real world case where it causes trouble, but > > not for an artificially created test case. > > > > I don't see anything fundamentally wrong in the design anyway. I do not > > expect to be allowed to remove a hard disk drive from my system while > > its partitions are mounted, and I don't expect to be able to unmount > > partitions while users have files opened on them. You always have to > > tear things down in the right order. Same here. > > But that's exactly what you do today when your USB disk falls out of > it's connection. The kernel recovers and you move on. Hopefully it > wasn't your root partition :) > > Same for a serial device that has open userspace descriptors that is > removed from the system, or almost any other type of device that can be > physically removed from a system while it is currently running (PCI, > firewire, thunderbolt, etc.) What happens if you have an i2c device on > a PCI card that is removed while userspace has that device descriptor > open? You need to "invalidate" it and not oops if userspace keeps > reading/writing to it. Honestly I have no idea what would happen in this case. I would expect the PCI card to be offlined by software first, before it is physically removed. Hot-unplug doesn't mean hot-tearoff ;-) In case unprepared hardware tear-off actually happens (more realistically on USB rather than PCI I suppose) I agree we want to avoid oopses and other horrible consequences and behave as smoothly as possible. The code was not written with this scenario in mind, so additional work is certainly needed. > > (...) > > Still no details here. What app, what is it doing, to what device is it > > talking, why is it not a kernel driver, and why do they keep the device > > node opened all the time? > > Any random application can write to an i2c device in this system, as > long as it has "permission" to do so. But, it's hardware, and on a bus > that sometimes can be yanked out of the system. When that happens, > userspace will be notified of the removal, and "should" be nice and I don't think there is any such notification on i2c device nodes at the moment. Unless it's something generic. But I'm certain i2cdump etc. aren't listening anyway. > clean up after itself. But there will always be some lag between the > actual removal when the kernel figures it out, and when userspace > finally closes that device node. > > So not crashing the kernel is a nice thing to prevent from happening > during that window of when the device is removed and userspace hasn't > quite noticed it yet. I agree. > > (...) > > See my previous questions. We still don't know why they are doing what > > they are doing in user-space, nor why they think they have to keep the > > device node opened. > > > > They could always kill the application in question with: > > # fuser -k /dev/i2c-* > > before removing the module. Or find a more polite way to tell the > > application to quit. If they want to do it in user-space, they have to > > do it right. > > Ideally, yes, userspace will have closed that device node. But again, > hardware isn't kind and sometimes decides to be yanked out by users > before they tell the kernel about it. We handle this for almost all > other device subsystems, i2c is one of the last to be fixed up in this > manner. Sorry it's taken us well over a decade to get here :) The problem is that the patch proposed by Viresh has nothing to do with this. It's not adding notifications, just changing the time frame during which user-space holds a reference to the i2c (bus) device. The goal as I understand it is to allow *prepared* hot-unplug (in the form of "rmmod i2c-bus-device-driver" or sysfs-based offlining?) while user-space processes have i2c device nodes open. Unprepared hot-unplug will still go wrong exactly as it goes now. My point is that prepared hot-unplug can already be achieved today without any patch. Or possibly improved by adding a notification mechanism. But not by changing the reference holding design. Not only the proposed patch does not help and degrades the performance, but it breaks assumptions. For example, it would allow an application to open an i2c bus, then you remove its driver and load another i2c bus driver, which gets the same bus number, and now the application writes to a completely different I2C bus segment. The current reference model prevents that, on purpose. So, again, nack from me. -- Jean Delvare SUSE L3 Support
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
Hi Jean, On 25-07-16, 11:39, Jean Delvare wrote: > The problem is that the patch proposed by Viresh has nothing to do with > this. It's not adding notifications, just changing the time frame during > which user-space holds a reference to the i2c (bus) device. The goal as > I understand it is to allow *prepared* hot-unplug (in the form of > "rmmod i2c-bus-device-driver" or sysfs-based offlining?) while Not really. We are concerned about both prepared and Unprepared cases. This *hacky* patch was useful in case of *unprepared* hot-unplug as well. Here is the sequence of events: - open() i2c device from userspace - do some operations on the device read/write/ioctls() .. - Module hot-unplugged (*unprepared*) - Some of the ongoing i2c transactions may just fail, that is fine .. - Kernel detected the interrupt about module removal and tries to cleanup the devices.. - Now, kernel can not remove the i2c device, unless user application has closed the file descriptor. And so kernel is waiting in the driver's ->remove() callback forever. Also, there is no way to co-ordinate (in Android) with the Applications using the device. They can crash or fail out if they want to, but the kernel shouldn't stop removal of a hardware module in that case. > user-space processes have i2c device nodes open. Unprepared hot-unplug > will still go wrong exactly as it goes now. > My point is that prepared hot-unplug can already be achieved today > without any patch. Yeah, if we have the option of stopping the applications before the device is gone. > Or possibly improved by adding a notification > mechanism. But not by changing the reference holding design. > > Not only the proposed patch does not help and degrades the performance, > but it breaks assumptions. For example, it would allow an application > to open an i2c bus, then you remove its driver and load another i2c bus > driver, which gets the same bus number, and now the application writes > to a completely different I2C bus segment. The current reference model > prevents that, on purpose. > > So, again, nack from me. Yeah, the patch wasn't great and I knew it from the beginning. But we are looking for a solution that can be accepted and so need advice from you guys :) -- viresh
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
Hi Viresh, On Mon, 25 Jul 2016 15:31:40 -0700, Viresh Kumar wrote: > On 25-07-16, 11:39, Jean Delvare wrote: > > The problem is that the patch proposed by Viresh has nothing to do with > > this. It's not adding notifications, just changing the time frame during > > which user-space holds a reference to the i2c (bus) device. The goal as > > I understand it is to allow *prepared* hot-unplug (in the form of > > "rmmod i2c-bus-device-driver" or sysfs-based offlining?) while > > Not really. We are concerned about both prepared and Unprepared cases. > > This *hacky* patch was useful in case of *unprepared* hot-unplug as well. > > Here is the sequence of events: > - open() i2c device from userspace > - do some operations on the device read/write/ioctls() .. > - Module hot-unplugged (*unprepared*) > - Some of the ongoing i2c transactions may just fail, that is fine .. > - Kernel detected the interrupt about module removal and tries to > cleanup the devices.. > - Now, kernel can not remove the i2c device, unless user application > has closed the file descriptor. Well, what is the application doing at that point? What error codes is it getting? Should it not close the device node and bail out? > And so kernel is waiting in the driver's ->remove() callback forever. > > Also, there is no way to co-ordinate (in Android) with the > Applications using the device. Why? Looks like a serious limitation. At this point I still don't know what application we are talking about, why it has to be in user-space when I2C device drivers are supposed to be on the kernel side, nor why they have to keep the i2c device node opened all the time. Why do you insist on relying on i2c-dev when it is so clearly no compatible with your requirements? > They can crash or fail out if they > want to, but the kernel shouldn't stop removal of a hardware module in > that case. If they crash or bail out, that would close the device, so no problem. The problem would be if they stay alive and misbehave. If the kernel should be completely independent from what user-space is doing, this is simply not compatible with reference counting. The whole point of counting references is to know when a device is still in use, and prevent it from being removed while this is the case. If you say that devices can be removed at any time and this is OK, then you should not be counting references at all, this is pointless. But I doubt the i2c code is currently ready for this. > > user-space processes have i2c device nodes open. Unprepared hot-unplug > > will still go wrong exactly as it goes now. > > > My point is that prepared hot-unplug can already be achieved today > > without any patch. > > Yeah, if we have the option of stopping the applications before the > device is gone. > > > Or possibly improved by adding a notification > > mechanism. But not by changing the reference holding design. > > > > Not only the proposed patch does not help and degrades the performance, > > but it breaks assumptions. For example, it would allow an application > > to open an i2c bus, then you remove its driver and load another i2c bus > > driver, which gets the same bus number, and now the application writes > > to a completely different I2C bus segment. The current reference model > > prevents that, on purpose. > > > > So, again, nack from me. > > Yeah, the patch wasn't great and I knew it from the beginning. But we > are looking for a solution that can be accepted and so need advice > from you guys :) I have no idea, sorry. Hopefully Greg or Wolfram know more? Check what other subsystems are doing? -- Jean Delvare SUSE L3 Support
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
Hi Jean, On Tue, Jul 26, 2016 at 12:41 AM, Jean Delvare wrote: > Hi Viresh, > > On Mon, 25 Jul 2016 15:31:40 -0700, Viresh Kumar wrote: >> On 25-07-16, 11:39, Jean Delvare wrote: >> > The problem is that the patch proposed by Viresh has nothing to do with >> > this. It's not adding notifications, just changing the time frame during >> > which user-space holds a reference to the i2c (bus) device. The goal as >> > I understand it is to allow *prepared* hot-unplug (in the form of >> > "rmmod i2c-bus-device-driver" or sysfs-based offlining?) while >> >> Not really. We are concerned about both prepared and Unprepared cases. >> >> This *hacky* patch was useful in case of *unprepared* hot-unplug as well. >> >> Here is the sequence of events: >> - open() i2c device from userspace >> - do some operations on the device read/write/ioctls() .. >> - Module hot-unplugged (*unprepared*) >> - Some of the ongoing i2c transactions may just fail, that is fine .. >> - Kernel detected the interrupt about module removal and tries to >> cleanup the devices.. >> - Now, kernel can not remove the i2c device, unless user application >> has closed the file descriptor. > > Well, what is the application doing at that point? What error codes is > it getting? Should it not close the device node and bail out? It should. Will it? It is up to the application. It may decide to keep that file descriptor open forever. > >> And so kernel is waiting in the driver's ->remove() callback forever. >> >> Also, there is no way to co-ordinate (in Android) with the >> Applications using the device. > > Why? Looks like a serious limitation. > > At this point I still don't know what application we are talking about, > why it has to be in user-space when I2C device drivers are supposed to > be on the kernel side, nor why they have to keep the i2c device node > opened all the time. > > Why do you insist on relying on i2c-dev when it is so clearly no > compatible with your requirements? If you provide an interface expect it to be used, maybe even in the ways you did not anticipated. Kernel's (and out task) is to make the interface behave properly, not police users. > >> They can crash or fail out if they >> want to, but the kernel shouldn't stop removal of a hardware module in >> that case. > > If they crash or bail out, that would close the device, so no problem. > The problem would be if they stay alive and misbehave. > > If the kernel should be completely independent from what user-space is > doing, this is simply not compatible with reference counting. The whole > point of counting references is to know when a device is still in use, > and prevent it from being removed while this is the case. If you say > that devices can be removed at any time and this is OK, then you should > not be counting references at all, this is pointless. But I doubt the > i2c code is currently ready for this. There are multitude of reference counts, all counting different things, not one refcount for the entire thing. You have refcount for the file object, you have memory refcounts, you have device object refcounts. And note that device and driver bond is not refcounted at all. It is supposed to be allowed to be broken at [pretty much] any moment. You just need to be careful when doing so ;) > >> > user-space processes have i2c device nodes open. Unprepared hot-unplug >> > will still go wrong exactly as it goes now. >> >> > My point is that prepared hot-unplug can already be achieved today >> > without any patch. >> >> Yeah, if we have the option of stopping the applications before the >> device is gone. >> >> > Or possibly improved by adding a notification >> > mechanism. But not by changing the reference holding design. >> > >> > Not only the proposed patch does not help and degrades the performance, >> > but it breaks assumptions. For example, it would allow an application >> > to open an i2c bus, then you remove its driver and load another i2c bus >> > driver, which gets the same bus number, and now the application writes >> > to a completely different I2C bus segment. The current reference model >> > prevents that, on purpose. >> > >> > So, again, nack from me. >> >> Yeah, the patch wasn't great and I knew it from the beginning. But we >> are looking for a solution that can be accepted and so need advice >> from you guys :) > > I have no idea, sorry. Hopefully Greg or Wolfram know more? Check what > other subsystems are doing? One option is to, upon device removal, to mark it as "dead" to allow accesses through i2c-dev to return error, release all hardware resources, but keep the object in memory in zombie state, waiting for the last user to drop the reference. Once that happens (maybe years later) you finally can release last bits of memory. You can check how we do that for input_dev/evdev pair. Thanks. -- Dmitry
Re: [PATCH 2/2] i2c-dev: Don't block the adapter from unregistering
On Mon, Jul 11, 2016 at 02:22:09PM +0200, Jean Delvare wrote: > Hi Viresh, > > On Wed, 6 Jul 2016 13:55:40 -0700, Viresh Kumar wrote: > > On 06-07-16, 19:12, Jean Delvare wrote: > > > Well well... I don't like this patch at all to be honest. > > > > Sure, I didn't like it much as well. I just wanted people to comment on what > > else we can do here. We don't really want to add out-of-mainline stuff here. > > > > > My first question would be: what is keeping /dev/i2c-* open all the > > > time? Originally i2c-dev was developed with development and debugging > > > tools in mind (the i2c-tools suite.) The device nodes were never meant > > > to be kept open for more than a few seconds. > > > > We thought that buggy userspace shouldn't be allowed to get kernel into > > trouble. > > Isn't that the case ? > > Buggy user-space run by root can cause any kind of trouble. And the > point being discussed here is amongst the most minor of these. But I > even disagree with calling it buggy. > > > In our case this is what happens: > > - userspace opens the file descriptor > > - we try to forcefully remove the module from phone (that doesn't talk to > > userspace to stop using the device). > > - The module doesn't get ejected unless the app closes the fd. > > So you are basically building a test case to cause the problem. It's > artificial. The adapter reference being held while the device node is > open isn't a bug, it is a design decision. I would consider revisiting > that design if there was a real world case where it causes trouble, but > not for an artificially created test case. > > I don't see anything fundamentally wrong in the design anyway. I do not > expect to be allowed to remove a hard disk drive from my system while > its partitions are mounted, and I don't expect to be able to unmount > partitions while users have files opened on them. You always have to > tear things down in the right order. Same here. But that's exactly what you do today when your USB disk falls out of it's connection. The kernel recovers and you move on. Hopefully it wasn't your root partition :) Same for a serial device that has open userspace descriptors that is removed from the system, or almost any other type of device that can be physically removed from a system while it is currently running (PCI, firewire, thunderbolt, etc.) What happens if you have an i2c device on a PCI card that is removed while userspace has that device descriptor open? You need to "invalidate" it and not oops if userspace keeps reading/writing to it. > > > Do you have user-space i2c device drivers on your system? Which ones, > > > > No. Its probably an app written by some of our module app developers. > > > > > and why (I would expect all useful i2c devices to have a kernel > > > driver.) > > > > That's what we have. > > Still no details here. What app, what is it doing, to what device is it > talking, why is it not a kernel driver, and why do they keep the device > node opened all the time? Any random application can write to an i2c device in this system, as long as it has "permission" to do so. But, it's hardware, and on a bus that sometimes can be yanked out of the system. When that happens, userspace will be notified of the removal, and "should" be nice and clean up after itself. But there will always be some lag between the actual removal when the kernel figures it out, and when userspace finally closes that device node. So not crashing the kernel is a nice thing to prevent from happening during that window of when the device is removed and userspace hasn't quite noticed it yet. > > > Requesting and freeing the i2c adapter for every transaction is going > > > > Well, we are just finding it (taking a reference of it) and the dropping its > > reference. > > Yes, that's what I meant, sorry for using the wrong terms. > > > > to add a lot of overhead to all existing tools :-( > > > > :( > > i2cdump typically runs 258 ioctls on the device node, i2cdetect 235. > i2c_get_adapter isn't cheap. Multiple function calls, mutex > locking/unlocking, preemption disabling/enabling... You don't want do > to that repeatedly if it can be avoided. So, nack from me. > > > > It's not like every user can open i2c device nodes and block the > > > system. Only selected users should be able to open i2c device nodes > > > (only root by default) so they should be responsible for not > > > misbehaving. > > > > Hmmm. The problem is that they weren't told when the module tries to go > > away and > > so they don't know that they need to close the fd. > > See my previous questions. We still don't know why they are doing what > they are doing in user-space, nor why they think they have to keep the > device node opened. > > They could always kill the application in question with: > # fuser -k /dev/i2c-* > before removing the module. Or find a more polite way to tell the > application to quit. If they want to do it in user-space, they have to > do it ri