On Wed, 1 Apr 2015, Sudip Mukherjee wrote:
> > don't get smug. :-) i had sort of concluded there was no actual
> > udev rule i could point at, so it's just the case that udev
> > processes those events internally? is this written down anywhere?
> > everyone happily explains how misc devices get their /dev file
> > automatically, but no one i've seen goes that extra step to
> > explain how that happens.
>
> misc_register() is ultimately calling device_add() which is again
> calling device_create_file() , and this one is actually creating the
> node in /dev . (if i am not wrong).
close, and now i can come across as an expert since i checked the
source file. :-) from drivers/base/core.c, in device_add():
if (MAJOR(dev->devt)) {
error = device_create_file(dev, &dev_attr_dev);
if (error)
goto DevAttrError;
error = device_create_sys_dev_entry(dev);
if (error)
goto SysEntryError;
devtmpfs_create_node(dev);
}
so device_add() creates device_create_file(), then calls
devtmpfs_create_node() which, based on whether you selected that
devtmpfs CONFIG option, either creates the /dev file or does nothing.
it's all clear to me now ...
rday
--
========================================================================
Robert P. J. Day Ottawa, Ontario, CANADA
http://crashcourse.ca
Twitter: http://twitter.com/rpjday
LinkedIn: http://ca.linkedin.com/in/rpjday
========================================================================
_______________________________________________
Kernelnewbies mailing list
Kernelnewbies@xxxxxxxxxxxxxxxxx
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
