Can't cleanly unload driver
I am trying to load/unload a basic driver skeleton. I did an insmod ./mydriver.ko , then a mknod /dev/mydriver0 c <major #> 0 , then a ln -sf mydriver0 /dev/mydriver All is well. Now to unload rmmod mydriver rm -f /dev/mydriver rm -f /dev/mydriver0 But rmmod reports Module mydriver is not currently loaded. But it is in /proc/devices: 248 mydriver Trying to reload it gives: insmod ./mydriver.ko Could not insert module mydriver.ko: File exists. I could reboot .... again ... but what is wrong? I have source for the driver, as I am developing it. WS
On Thu, 26 Dec 2013 17:50:52 -0800, Eric Fowler said:
I did an insmod ./mydriver.ko
But rmmod reports Module mydriver is not currently loaded.
But it is in /proc/devices: 248 mydriver
Trying to reload it gives: insmod ./mydriver.ko
Could not insert module mydriver.ko: File exists.
Out of curiosity, what name is reported if you do an 'lsmod' after loading it? And what's the reference count on it? I wonder if you're ever-so-slightly out of sync someplace (a 'mydriver' in one place, and 'Mydriver' in another, or something equally misbegotten).
I have checked that and the problem persists. [eric@localhost parprt]$ ls -la /lib/modules/`uname -r`/build lrwxrwxrwx. 1 root root 23 Dec 23 17:06 /lib/modules/3.12.6/build -> /home/eric/build/kernel [eric@localhost parprt]$ pwd /home/eric/drivers/parprt [eric@localhost parprt]$ ls -la /lib/modules/`uname -r`/build lrwxrwxrwx. 1 root root 23 Dec 23 17:06 /lib/modules/3.12.6/build -> /home/eric/build/kernel [eric@localhost parprt]$ ls -la *.ko -rw-rw-r--. 1 eric eric 86187 Dec 27 15:18 foobar.ko [eric@localhost parprt]$ su Password:**** [root@localhost parprt]# /sbin/insmod ./foobar.ko insmod: ERROR: could not insert module ./foobar.ko: File exists [root@localhost parprt]# rmmod foobar rmmod: ERROR: Module foobar is not currently loaded [root@localhost parprt]# rmmod foobar.ko rmmod: ERROR: Module foobar is not currently loaded [root@localhost parprt]# modinfo foobar modinfo: ERROR: Module foobar not found. On Thu, Dec 26, 2013 at 9:55 PM, <Valdis.Kletnieks@vt.edu> wrote:
On Thu, 26 Dec 2013 17:50:52 -0800, Eric Fowler said:
I did an insmod ./mydriver.ko
But rmmod reports Module mydriver is not currently loaded.
But it is in /proc/devices: 248 mydriver
Trying to reload it gives: insmod ./mydriver.ko
Could not insert module mydriver.ko: File exists.
Out of curiosity, what name is reported if you do an 'lsmod' after loading it? And what's the reference count on it? I wonder if you're ever-so-slightly out of sync someplace (a 'mydriver' in one place, and 'Mydriver' in another, or something equally misbegotten).
-- cc:NSA
lsmod shows no driver. On Fri, Dec 27, 2013 at 6:59 PM, <Valdis.Kletnieks@vt.edu> wrote:
On Fri, 27 Dec 2013 16:10:43 -0800, Eric Fowler said:
I have checked that and the problem persists.
I see insmod and rmmod - but no lsmod output. Not ls, lsmod. Different command.
-- cc:NSA
I suspect I am doing something wrong in the code with register/unregister_chrdev(), but I have been over that code a million times. It looks fine. Now: insmod the device, OK rmmod the device, OK Check /proc/devices , device # is present insmod the device again, fails with ERROR: could not insert module ./foobar.ko: Device or resource busy On Fri, Dec 27, 2013 at 6:59 PM, <Valdis.Kletnieks@vt.edu> wrote:
On Fri, 27 Dec 2013 16:10:43 -0800, Eric Fowler said:
I have checked that and the problem persists.
I see insmod and rmmod - but no lsmod output. Not ls, lsmod. Different command.
-- cc:NSA
On Fri, 27 Dec 2013 19:33:50 -0800, Eric Fowler said:
I suspect I am doing something wrong in the code with register/unregister_chrdev(), but I have been over that code a million times. It looks fine.
Now: insmod the device, OK rmmod the device, OK Check /proc/devices , device # is present insmod the device again, fails with ERROR: could not insert module ./foobar.ko: Device or resource busy
It does smell like an unregister issue. You may want to try adding printk() calls to print out the return code from register and unregister. I'm willing to bet that (a) the unegister is failing because somebody still has a reference on the device, and (b) the second register call fails because the device already exists, causing your module_init() to bail out. The fun is that you may not have taken a reference on the device directly yourself - you may have called some other get_foo() that ends up taking an implicit reference under the covers, causing issues when you fail to call put_foo() at the right place...
Still working on this. Here is some dmesg spew: [ 514.245846] foobar: module verification failed: signature and/or required key missing - tainting kernel [ 514.245937] kobject: 'foobar' (f7f060c8): kobject_add_internal: parent: 'module', set: 'module' [ 514.245951] kobject: 'holders' (f5ff3d40): kobject_add_internal: parent: 'foobar', set: '<NULL>' [ 514.245981] kobject: 'notes' (f2d25f80): kobject_add_internal: parent: 'foobar', set: '<NULL>' [ 514.245987] kobject: 'foobar' (f7f060c8): kobject_uevent_env [ 514.245998] kobject: 'foobar' (f7f060c8): fill_kobj_path: path = '/module/foobar' So it looks like kernel validation is failing. I have printk's in my init fxn that are never turning up in /var/log/messages, until, weirdly, AFTER I remove the device: <insmod device> Dec 30 09:43:03 localhost kernel: [ 514.245846] foobar: module verification failed: signature and/or required key missing - tainting kernel Dec 30 09:43:16 localhost fprintd[1085]: ** Message: No devices in use, exit <rmmod device> Dec 30 09:45:53 localhost kernel: [ 514.249323] foobar: got device number 248, minor is 0 <<<<----THIS IS IN init() fxn Dec 30 09:45:53 localhost kernel: [ 684.102912] unregister_chrdev(248) called for foobar<7>[ 684.102927] kobject: '(null)' (f7f06220): kobject_cleanup, parent (null) <insmod> insmod: ERROR: could not insert module ./foobar.ko: Device or resource busy On Fri, Dec 27, 2013 at 9:13 PM, <Valdis.Kletnieks@vt.edu> wrote:
On Fri, 27 Dec 2013 19:33:50 -0800, Eric Fowler said:
I suspect I am doing something wrong in the code with register/unregister_chrdev(), but I have been over that code a million times. It looks fine.
Now: insmod the device, OK rmmod the device, OK Check /proc/devices , device # is present insmod the device again, fails with ERROR: could not insert module ./foobar.ko: Device or resource busy
It does smell like an unregister issue. You may want to try adding printk() calls to print out the return code from register and unregister. I'm willing to bet that (a) the unegister is failing because somebody still has a reference on the device, and (b) the second register call fails because the device already exists, causing your module_init() to bail out.
The fun is that you may not have taken a reference on the device directly yourself - you may have called some other get_foo() that ends up taking an implicit reference under the covers, causing issues when you fail to call put_foo() at the right place...
-- cc:NSA
Hi Eric, I have seen some errors with module reference counting with a nicely written code, but culprit for my case was a missing compilation flag -DMODULE which gives definition of THIS_MODULE, otherwise it is null e.g. for modules which are compiled in kernel, so they are never unloaded. Unless you have some customization done to Makefiles, this definition should be included, but its anyways good to double check and rule out this possibility. -Rajat On Mon, Dec 30, 2013 at 9:48 AM, Eric Fowler <eric.fowler@gmail.com> wrote:
Still working on this. Here is some dmesg spew:
[ 514.245846] foobar: module verification failed: signature and/or required key missing - tainting kernel [ 514.245937] kobject: 'foobar' (f7f060c8): kobject_add_internal: parent: 'module', set: 'module' [ 514.245951] kobject: 'holders' (f5ff3d40): kobject_add_internal: parent: 'foobar', set: '<NULL>' [ 514.245981] kobject: 'notes' (f2d25f80): kobject_add_internal: parent: 'foobar', set: '<NULL>' [ 514.245987] kobject: 'foobar' (f7f060c8): kobject_uevent_env [ 514.245998] kobject: 'foobar' (f7f060c8): fill_kobj_path: path = '/module/foobar'
So it looks like kernel validation is failing. I have printk's in my init fxn that are never turning up in /var/log/messages, until, weirdly, AFTER I remove the device:
<insmod device> Dec 30 09:43:03 localhost kernel: [ 514.245846] foobar: module verification failed: signature and/or required key missing - tainting kernel Dec 30 09:43:16 localhost fprintd[1085]: ** Message: No devices in use, exit
<rmmod device> Dec 30 09:45:53 localhost kernel: [ 514.249323] foobar: got device number 248, minor is 0 <<<<----THIS IS IN init() fxn Dec 30 09:45:53 localhost kernel: [ 684.102912] unregister_chrdev(248) called for foobar<7>[ 684.102927] kobject: '(null)' (f7f06220): kobject_cleanup, parent (null)
<insmod> insmod: ERROR: could not insert module ./foobar.ko: Device or resource busy
On Fri, Dec 27, 2013 at 9:13 PM, <Valdis.Kletnieks@vt.edu> wrote:
On Fri, 27 Dec 2013 19:33:50 -0800, Eric Fowler said:
I suspect I am doing something wrong in the code with register/unregister_chrdev(), but I have been over that code a million times. It looks fine.
Now: insmod the device, OK rmmod the device, OK Check /proc/devices , device # is present insmod the device again, fails with ERROR: could not insert module ./foobar.ko: Device or resource busy
It does smell like an unregister issue. You may want to try adding printk() calls to print out the return code from register and unregister. I'm willing to bet that (a) the unegister is failing because somebody still has a reference on the device, and (b) the second register call fails because the device already exists, causing your module_init() to bail out.
The fun is that you may not have taken a reference on the device directly yourself - you may have called some other get_foo() that ends up taking an implicit reference under the covers, causing issues when you fail to call put_foo() at the right place...
-- cc:NSA
_______________________________________________ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Actually, I was working with a homegrown makefile. I made the change you recommended and am still having the problem. I am thinking now about something else as being the problem. Is this code legitimate? //// file scope struct cdev my_cdev; ///inside init function cdev_init(&my_cdev, ....); cdev_add(&my_cdev, ....); ///exit fxn cdev_del(&my_cdev); In other words, the memory for the cdev struct comes from static memory for the driver, not a call to cdev_alloc() or kmalloc(). On Mon, Dec 30, 2013 at 10:18 AM, Rajat Sharma <fs.rajat@gmail.com> wrote:
Hi Eric,
I have seen some errors with module reference counting with a nicely written code, but culprit for my case was a missing compilation flag -DMODULE which gives definition of THIS_MODULE, otherwise it is null e.g. for modules which are compiled in kernel, so they are never unloaded. Unless you have some customization done to Makefiles, this definition should be included, but its anyways good to double check and rule out this possibility.
-Rajat
On Mon, Dec 30, 2013 at 9:48 AM, Eric Fowler <eric.fowler@gmail.com>wrote:
Still working on this. Here is some dmesg spew:
[ 514.245846] foobar: module verification failed: signature and/or required key missing - tainting kernel [ 514.245937] kobject: 'foobar' (f7f060c8): kobject_add_internal: parent: 'module', set: 'module' [ 514.245951] kobject: 'holders' (f5ff3d40): kobject_add_internal: parent: 'foobar', set: '<NULL>' [ 514.245981] kobject: 'notes' (f2d25f80): kobject_add_internal: parent: 'foobar', set: '<NULL>' [ 514.245987] kobject: 'foobar' (f7f060c8): kobject_uevent_env [ 514.245998] kobject: 'foobar' (f7f060c8): fill_kobj_path: path = '/module/foobar'
So it looks like kernel validation is failing. I have printk's in my init fxn that are never turning up in /var/log/messages, until, weirdly, AFTER I remove the device:
<insmod device> Dec 30 09:43:03 localhost kernel: [ 514.245846] foobar: module verification failed: signature and/or required key missing - tainting kernel Dec 30 09:43:16 localhost fprintd[1085]: ** Message: No devices in use, exit
<rmmod device> Dec 30 09:45:53 localhost kernel: [ 514.249323] foobar: got device number 248, minor is 0 <<<<----THIS IS IN init() fxn Dec 30 09:45:53 localhost kernel: [ 684.102912] unregister_chrdev(248) called for foobar<7>[ 684.102927] kobject: '(null)' (f7f06220): kobject_cleanup, parent (null)
<insmod> insmod: ERROR: could not insert module ./foobar.ko: Device or resource busy
On Fri, Dec 27, 2013 at 9:13 PM, <Valdis.Kletnieks@vt.edu> wrote:
On Fri, 27 Dec 2013 19:33:50 -0800, Eric Fowler said:
I suspect I am doing something wrong in the code with register/unregister_chrdev(), but I have been over that code a million times. It looks fine.
Now: insmod the device, OK rmmod the device, OK Check /proc/devices , device # is present insmod the device again, fails with ERROR: could not insert module ./foobar.ko: Device or resource busy
It does smell like an unregister issue. You may want to try adding printk() calls to print out the return code from register and unregister. I'm willing to bet that (a) the unegister is failing because somebody still has a reference on the device, and (b) the second register call fails because the device already exists, causing your module_init() to bail out.
The fun is that you may not have taken a reference on the device directly yourself - you may have called some other get_foo() that ends up taking an implicit reference under the covers, causing issues when you fail to call put_foo() at the right place...
-- cc:NSA
_______________________________________________ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
-- cc:NSA
So, I finally figured this out, and for the benefit of web posterity and the search engines, I'm gonna lay it out: I was using *both* the register_chrdev() and the cdev_*() functions on the same device. I guess you can't do that. I can't, anyway. Rubini's scull code uses both, and it looks like he uses both on the same device, although I bet he doesn't (haven't looked closely at the code). I switched to using register_chrdev()/unregister_chrdev() and all is happy now. Thanks to all who helped. On Mon, Dec 30, 2013 at 10:57 AM, Eric Fowler <eric.fowler@gmail.com> wrote:
Actually, I was working with a homegrown makefile. I made the change you recommended and am still having the problem.
I am thinking now about something else as being the problem.
Is this code legitimate?
//// file scope
struct cdev my_cdev;
///inside init function cdev_init(&my_cdev, ....); cdev_add(&my_cdev, ....);
///exit fxn cdev_del(&my_cdev);
In other words, the memory for the cdev struct comes from static memory for the driver, not a call to cdev_alloc() or kmalloc().
On Mon, Dec 30, 2013 at 10:18 AM, Rajat Sharma <fs.rajat@gmail.com> wrote:
Hi Eric,
I have seen some errors with module reference counting with a nicely written code, but culprit for my case was a missing compilation flag -DMODULE which gives definition of THIS_MODULE, otherwise it is null e.g. for modules which are compiled in kernel, so they are never unloaded. Unless you have some customization done to Makefiles, this definition should be included, but its anyways good to double check and rule out this possibility.
-Rajat
On Mon, Dec 30, 2013 at 9:48 AM, Eric Fowler <eric.fowler@gmail.com>wrote:
Still working on this. Here is some dmesg spew:
[ 514.245846] foobar: module verification failed: signature and/or required key missing - tainting kernel [ 514.245937] kobject: 'foobar' (f7f060c8): kobject_add_internal: parent: 'module', set: 'module' [ 514.245951] kobject: 'holders' (f5ff3d40): kobject_add_internal: parent: 'foobar', set: '<NULL>' [ 514.245981] kobject: 'notes' (f2d25f80): kobject_add_internal: parent: 'foobar', set: '<NULL>' [ 514.245987] kobject: 'foobar' (f7f060c8): kobject_uevent_env [ 514.245998] kobject: 'foobar' (f7f060c8): fill_kobj_path: path = '/module/foobar'
So it looks like kernel validation is failing. I have printk's in my init fxn that are never turning up in /var/log/messages, until, weirdly, AFTER I remove the device:
<insmod device> Dec 30 09:43:03 localhost kernel: [ 514.245846] foobar: module verification failed: signature and/or required key missing - tainting kernel Dec 30 09:43:16 localhost fprintd[1085]: ** Message: No devices in use, exit
<rmmod device> Dec 30 09:45:53 localhost kernel: [ 514.249323] foobar: got device number 248, minor is 0 <<<<----THIS IS IN init() fxn Dec 30 09:45:53 localhost kernel: [ 684.102912] unregister_chrdev(248) called for foobar<7>[ 684.102927] kobject: '(null)' (f7f06220): kobject_cleanup, parent (null)
<insmod> insmod: ERROR: could not insert module ./foobar.ko: Device or resource busy
On Fri, Dec 27, 2013 at 9:13 PM, <Valdis.Kletnieks@vt.edu> wrote:
On Fri, 27 Dec 2013 19:33:50 -0800, Eric Fowler said:
I suspect I am doing something wrong in the code with register/unregister_chrdev(), but I have been over that code a million times. It looks fine.
Now: insmod the device, OK rmmod the device, OK Check /proc/devices , device # is present insmod the device again, fails with ERROR: could not insert module ./foobar.ko: Device or resource busy
It does smell like an unregister issue. You may want to try adding printk() calls to print out the return code from register and unregister. I'm willing to bet that (a) the unegister is failing because somebody still has a reference on the device, and (b) the second register call fails because the device already exists, causing your module_init() to bail out.
The fun is that you may not have taken a reference on the device directly yourself - you may have called some other get_foo() that ends up taking an implicit reference under the covers, causing issues when you fail to call put_foo() at the right place...
-- cc:NSA
_______________________________________________ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
-- cc:NSA
-- cc:NSA
participants (3)
-
Eric Fowler -
Rajat Sharma -
Valdis.Kletnieks＠vt.edu