2003-07-21 07:48:48 +04:00
/*
* udev - add . c
*
* Userspace devfs
*
* Copyright ( C ) 2003 Greg Kroah - Hartman < greg @ kroah . com >
*
*
* This program is free software ; you can redistribute it and / or modify it
* under the terms of the GNU General Public License as published by the
* Free Software Foundation version 2 of the License .
*
* This program is distributed in the hope that it will be useful , but
* WITHOUT ANY WARRANTY ; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE . See the GNU
* General Public License for more details .
*
* You should have received a copy of the GNU General Public License along
* with this program ; if not , write to the Free Software Foundation , Inc . ,
* 675 Mass Ave , Cambridge , MA 0213 9 , USA .
*
*/
# include <stdlib.h>
# include <string.h>
# include <stdio.h>
# include <fcntl.h>
# include <unistd.h>
# include <errno.h>
2003-10-23 06:39:54 +04:00
# include <sys/stat.h>
2003-11-24 07:56:18 +03:00
# include <sys/types.h>
# include <grp.h>
[PATCH] hmm, handle net devices with udev?
Hmm, Arndt Bergmann sent a patch like this one a few weeks ago and
I want to bring the question back, if we want to handle net device
naming with udev.
With this patch it is actually possible to specify something like this
in udev.rules:
KERNEL="dummy*", SYSFS{address}="00:00:00:00:00:00", SYSFS{features}="0x0", NAME="blind%n"
KERNEL="eth*", SYSFS{address}="00:0d:60:77:30:91", NAME="private"
and you will get:
[root@pim udev.kay]# cat /proc/net/dev
Inter-| Receive | Transmit
face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
lo: 1500 30 0 0 0 0 0 0 1500 30 0 0 0 0 0 0
private: 278393 1114 0 0 0 0 0 0 153204 1468 0 0 0 0 0 0
sit0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
blind0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
The udevinfo program is also working:
[root@pim udev.kay]# ./udevinfo -a -p /sys/class/net/private
looking at class device '/sys/class/net/private':
SYSFS{addr_len}="6"
SYSFS{address}="00:0d:60:77:30:91"
SYSFS{broadcast}="ff:ff:ff:ff:ff:ff"
SYSFS{features}="0x3a9"
SYSFS{flags}="0x1003"
SYSFS{ifindex}="2"
SYSFS{iflink}="2"
SYSFS{mtu}="1500"
SYSFS{tx_queue_len}="1000"
SYSFS{type}="1"
follow the class device's "device"
looking at the device chain at '/sys/devices/pci0000:00/0000:00:1e.0/0000:02:01.0':
BUS="pci"
ID="0000:02:01.0"
SYSFS{class}="0x020000"
SYSFS{detach_state}="0"
SYSFS{device}="0x101e"
SYSFS{irq}="11"
SYSFS{subsystem_device}="0x0549"
SYSFS{subsystem_vendor}="0x1014"
SYSFS{vendor}="0x8086"
The matching device will be renamed to the given name. The device name
will not be put into the udev database, cause the kernel renames the
device and the sysfs name disappears.
I like it, cause it plugs in nicely. We have all the naming features
and sysfs queries and walks inside of udev. The sysfs timing races
are already solved and the management tools are working for net devices
too. nameif can only match the MAC address now. udev can match any sysfs
value of the device tree the net device is connected to.
But right, net devices do not have device nodes :)
2004-03-25 10:19:39 +03:00
# include <net/if.h>
# include <sys/socket.h>
# include <sys/ioctl.h>
# include <linux/sockios.h>
2003-07-21 07:48:48 +04:00
2004-02-24 06:07:25 +03:00
# include "libsysfs/sysfs/libsysfs.h"
2005-03-06 08:16:52 +03:00
# include "udev_libc_wrapper.h"
2003-07-21 07:48:48 +04:00
# include "udev.h"
2004-11-25 04:44:38 +03:00
# include "udev_utils.h"
2005-02-21 15:44:39 +03:00
# include "udev_sysfs.h"
2003-07-21 07:48:48 +04:00
# include "udev_version.h"
[PATCH] add udev logging to info log
On Thu, Jan 15, 2004 at 05:14:16AM +0100, Kay Sievers wrote:
> On Wed, Jan 14, 2004 at 01:10:43PM -0800, Greg KH wrote:
> > On Wed, Jan 14, 2004 at 02:34:26PM -0600, Clay Haapala wrote:
> > > On Wed, 14 Jan 2004, Chris Friesen spake thusly:
> > > >
> > > > Maybe for ones with a matching rule, you could print something like:
> > > >
> > > >
> > > Is the act of printing/syslogging a rule in an of itself?
> >
> > No, as currently the only way stuff ends up in the syslog is if
> > DEBUG=true is used on the build line.
> >
> > But it's sounding like we might want to change that... :)
>
> How about this in the syslog after connect/disconnect?
>
> Jan 15 05:07:45 pim udev[28007]: configured rule in '/etc/udev/udev.rules' at line 17 applied, 'video*' becomes 'video/webcam%n'
> Jan 15 05:07:45 pim udev[28007]: creating device node '/udev/video/webcam0'
> Jan 15 05:07:47 pim udev[28015]: removing device node '/udev/video/webcam0'
Here is a slightly better version. I've created a logging.h file and
moved the debug macros from udev.h in there.
If you type:
'make' - you will get a binary that prints one or two lines to syslog
if a device node is created or deleted
'make LOG=false' - you get a binary that prints asolutely nothing
'make DEBUG=true' - the same as today, it will print all debug lines
2004-01-16 08:53:20 +03:00
# include "logging.h"
2005-03-13 00:36:32 +03:00
# include "udev_rules.h"
2004-11-12 08:32:19 +03:00
# include "udev_db.h"
2005-02-04 23:38:26 +03:00
# include "udev_selinux.h"
2004-10-06 10:39:05 +04:00
2003-07-21 07:48:48 +04:00
2005-02-21 15:44:39 +03:00
int udev_make_node ( struct udevice * udev , const char * file , dev_t devt , mode_t mode , uid_t uid , gid_t gid )
2004-02-17 08:44:28 +03:00
{
2004-04-03 10:16:38 +04:00
struct stat stats ;
int retval = 0 ;
if ( stat ( file , & stats ) ! = 0 )
goto create ;
/* preserve node with already correct numbers, to not change the inode number */
if ( ( ( stats . st_mode & S_IFMT ) = = S_IFBLK | | ( stats . st_mode & S_IFMT ) = = S_IFCHR ) & &
2005-02-21 15:44:39 +03:00
( stats . st_rdev = = devt ) ) {
2004-04-03 10:16:38 +04:00
dbg ( " preserve file '%s', cause it has correct dev_t " , file ) ;
2005-02-04 23:38:26 +03:00
selinux_setfilecon ( file , udev - > kernel_name , stats . st_mode ) ;
2004-04-03 10:16:38 +04:00
goto perms ;
}
if ( unlink ( file ) ! = 0 )
dbg ( " unlink(%s) failed with error '%s' " , file , strerror ( errno ) ) ;
else
dbg ( " already present file '%s' unlinked " , file ) ;
2004-02-17 08:44:28 +03:00
2004-04-03 10:16:38 +04:00
create :
2005-02-10 21:46:50 +03:00
switch ( udev - > type ) {
2005-03-12 23:41:57 +03:00
case DEV_BLOCK :
2005-02-10 21:46:50 +03:00
mode | = S_IFBLK ;
break ;
2005-03-12 23:41:57 +03:00
case DEV_CLASS :
2005-02-10 21:46:50 +03:00
mode | = S_IFCHR ;
break ;
default :
dbg ( " unknown node type %c \n " , udev - > type ) ;
return - EINVAL ;
}
2005-02-04 23:38:26 +03:00
selinux_setfscreatecon ( file , udev - > kernel_name , mode ) ;
2005-02-21 15:44:39 +03:00
retval = mknod ( file , mode , devt ) ;
2005-04-27 10:52:14 +04:00
selinux_resetfscreatecon ( ) ;
2004-02-17 08:44:28 +03:00
if ( retval ! = 0 ) {
dbg ( " mknod(%s, %#o, %u, %u) failed with error '%s' " ,
2005-02-21 15:44:39 +03:00
file , mode , major ( devt ) , minor ( devt ) , strerror ( errno ) ) ;
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
goto exit ;
2004-02-17 08:44:28 +03:00
}
2004-04-03 10:16:38 +04:00
perms :
dbg ( " chmod(%s, %#o) " , file , mode ) ;
if ( chmod ( file , mode ) ! = 0 ) {
dbg ( " chmod(%s, %#o) failed with error '%s' " , file , mode , strerror ( errno ) ) ;
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
goto exit ;
2004-02-17 08:44:28 +03:00
}
if ( uid ! = 0 | | gid ! = 0 ) {
2004-04-03 10:16:38 +04:00
dbg ( " chown(%s, %u, %u) " , file , uid , gid ) ;
if ( chown ( file , uid , gid ) ! = 0 ) {
2004-02-17 08:44:28 +03:00
dbg ( " chown(%s, %u, %u) failed with error '%s' " ,
2004-04-03 10:16:38 +04:00
file , uid , gid , strerror ( errno ) ) ;
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
goto exit ;
2004-02-17 08:44:28 +03:00
}
}
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
exit :
return retval ;
2004-02-17 08:44:28 +03:00
}
2005-02-09 10:43:18 +03:00
static int create_node ( struct udevice * udev , struct sysfs_class_device * class_dev )
2004-03-02 10:31:06 +03:00
{
2005-03-07 06:29:43 +03:00
char filename [ PATH_SIZE ] ;
char partitionname [ PATH_SIZE ] ;
2005-03-05 07:35:31 +03:00
struct name_entry * name_loop ;
2005-03-10 04:46:26 +03:00
uid_t uid ;
gid_t gid ;
2003-12-07 20:12:07 +03:00
int tail ;
2005-03-05 07:35:31 +03:00
int i ;
2003-12-07 20:12:07 +03:00
2005-03-07 06:29:43 +03:00
snprintf ( filename , sizeof ( filename ) , " %s/%s " , udev_root , udev - > name ) ;
filename [ sizeof ( filename ) - 1 ] = ' \0 ' ;
2003-10-21 09:48:44 +04:00
2003-12-07 20:12:07 +03:00
/* create parent directories if needed */
2005-02-10 21:46:50 +03:00
if ( strchr ( udev - > name , ' / ' ) )
2003-12-07 20:12:07 +03:00
create_path ( filename ) ;
2003-11-12 14:48:01 +03:00
2005-03-10 04:46:26 +03:00
if ( strcmp ( udev - > owner , " root " ) = = 0 )
uid = 0 ;
else {
2003-11-12 14:47:57 +03:00
char * endptr ;
2005-03-10 04:46:26 +03:00
unsigned long id ;
2005-02-09 10:43:18 +03:00
2005-03-10 04:46:26 +03:00
id = strtoul ( udev - > owner , & endptr , 10 ) ;
2004-01-20 06:44:24 +03:00
if ( endptr [ 0 ] = = ' \0 ' )
2003-11-12 14:47:57 +03:00
uid = ( uid_t ) id ;
2005-03-06 08:16:52 +03:00
else
uid = lookup_user ( udev - > owner ) ;
2003-11-12 14:47:57 +03:00
}
2005-03-10 04:46:26 +03:00
if ( strcmp ( udev - > group , " root " ) = = 0 )
gid = 0 ;
else {
2003-11-12 14:47:57 +03:00
char * endptr ;
2005-03-10 04:46:26 +03:00
unsigned long id ;
2005-02-09 10:43:18 +03:00
2005-03-10 04:46:26 +03:00
id = strtoul ( udev - > group , & endptr , 10 ) ;
2004-01-20 06:44:24 +03:00
if ( endptr [ 0 ] = = ' \0 ' )
2003-11-12 14:47:57 +03:00
gid = ( gid_t ) id ;
2005-03-06 08:16:52 +03:00
else
2005-03-18 12:00:25 +03:00
gid = lookup_group ( udev - > group ) ;
2003-11-12 14:47:57 +03:00
}
2004-10-19 06:11:51 +04:00
if ( ! udev - > test_run ) {
2004-02-17 08:44:28 +03:00
info ( " creating device node '%s' " , filename ) ;
2005-02-21 15:44:39 +03:00
if ( udev_make_node ( udev , filename , udev - > devt , udev - > mode , uid , gid ) ! = 0 )
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
goto error ;
2004-02-17 08:58:25 +03:00
} else {
info ( " creating device node '%s', major = '%d', minor = '%d', "
" mode = '%#o', uid = '%d', gid = '%d' " , filename ,
2005-02-21 15:44:39 +03:00
major ( udev - > devt ) , minor ( udev - > devt ) , udev - > mode , uid , gid ) ;
2004-02-17 08:58:25 +03:00
}
2004-02-17 08:44:28 +03:00
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
/* create all_partitions if requested */
2005-02-14 08:03:06 +03:00
if ( udev - > partitions ) {
2005-02-09 10:43:18 +03:00
struct sysfs_attribute * attr ;
int range ;
/* take the maximum registered minor range */
attr = sysfs_get_classdev_attr ( class_dev , " range " ) ;
if ( attr ) {
range = atoi ( attr - > value ) ;
if ( range > 1 )
udev - > partitions = range - 1 ;
}
2004-10-19 06:11:51 +04:00
info ( " creating device partition nodes '%s[1-%i]' " , filename , udev - > partitions ) ;
if ( ! udev - > test_run ) {
for ( i = 1 ; i < = udev - > partitions ; i + + ) {
2005-02-21 15:44:39 +03:00
dev_t part_devt ;
2005-03-07 06:29:43 +03:00
snprintf ( partitionname , sizeof ( partitionname ) , " %s%d " , filename , i ) ;
partitionname [ sizeof ( partitionname ) - 1 ] = ' \0 ' ;
2005-02-21 15:44:39 +03:00
part_devt = makedev ( major ( udev - > devt ) , minor ( udev - > devt ) + 1 ) ;
udev_make_node ( udev , partitionname , part_devt , udev - > mode , uid , gid ) ;
2004-02-17 08:58:25 +03:00
}
2004-02-17 08:44:28 +03:00
}
2003-12-07 20:12:07 +03:00
}
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
/* create symlink(s) if requested */
2005-03-05 07:35:31 +03:00
list_for_each_entry ( name_loop , & udev - > symlink_list , node ) {
2005-04-27 10:52:14 +04:00
int retval ;
2005-03-07 06:29:43 +03:00
char linktarget [ PATH_SIZE ] ;
2004-11-06 16:28:01 +03:00
2005-03-07 06:29:43 +03:00
snprintf ( filename , sizeof ( filename ) , " %s/%s " , udev_root , name_loop - > name ) ;
filename [ sizeof ( filename ) - 1 ] = ' \0 ' ;
2004-11-06 16:28:01 +03:00
2004-10-19 06:11:51 +04:00
dbg ( " symlink '%s' to node '%s' requested " , filename , udev - > name ) ;
if ( ! udev - > test_run )
2005-03-05 07:35:31 +03:00
if ( strchr ( filename , ' / ' ) )
2004-03-04 05:16:35 +03:00
create_path ( filename ) ;
/* optimize relative link */
linktarget [ 0 ] = ' \0 ' ;
i = 0 ;
tail = 0 ;
2005-03-05 07:35:31 +03:00
while ( udev - > name [ i ] & & ( udev - > name [ i ] = = name_loop - > name [ i ] ) ) {
2004-10-19 06:11:51 +04:00
if ( udev - > name [ i ] = = ' / ' )
2004-03-04 05:16:35 +03:00
tail = i + 1 ;
i + + ;
}
2005-03-05 07:35:31 +03:00
while ( name_loop - > name [ i ] ! = ' \0 ' ) {
if ( name_loop - > name [ i ] = = ' / ' )
2005-03-07 06:29:43 +03:00
strlcat ( linktarget , " ../ " , sizeof ( linktarget ) ) ;
2004-03-04 05:16:35 +03:00
i + + ;
}
2003-12-07 20:12:07 +03:00
2005-03-07 06:29:43 +03:00
strlcat ( linktarget , & udev - > name [ tail ] , sizeof ( linktarget ) ) ;
2003-12-07 20:12:07 +03:00
2004-03-04 05:16:35 +03:00
dbg ( " symlink(%s, %s) " , linktarget , filename ) ;
2004-10-19 06:11:51 +04:00
if ( ! udev - > test_run ) {
2005-04-27 10:48:29 +04:00
unlink ( filename ) ;
2005-04-27 10:52:14 +04:00
selinux_setfscreatecon ( filename , udev - > kernel_name , S_IFLNK ) ;
retval = symlink ( linktarget , filename ) ;
selinux_resetfscreatecon ( ) ;
if ( retval ! = 0 )
2004-03-04 05:16:35 +03:00
dbg ( " symlink(%s, %s) failed with error '%s' " ,
linktarget , filename , strerror ( errno ) ) ;
2003-12-10 11:47:00 +03:00
}
2003-11-12 14:47:57 +03:00
}
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
return 0 ;
error :
return - 1 ;
2003-07-21 07:48:48 +04:00
}
2004-10-19 06:11:51 +04:00
static int rename_net_if ( struct udevice * udev )
[PATCH] hmm, handle net devices with udev?
Hmm, Arndt Bergmann sent a patch like this one a few weeks ago and
I want to bring the question back, if we want to handle net device
naming with udev.
With this patch it is actually possible to specify something like this
in udev.rules:
KERNEL="dummy*", SYSFS{address}="00:00:00:00:00:00", SYSFS{features}="0x0", NAME="blind%n"
KERNEL="eth*", SYSFS{address}="00:0d:60:77:30:91", NAME="private"
and you will get:
[root@pim udev.kay]# cat /proc/net/dev
Inter-| Receive | Transmit
face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
lo: 1500 30 0 0 0 0 0 0 1500 30 0 0 0 0 0 0
private: 278393 1114 0 0 0 0 0 0 153204 1468 0 0 0 0 0 0
sit0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
blind0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
The udevinfo program is also working:
[root@pim udev.kay]# ./udevinfo -a -p /sys/class/net/private
looking at class device '/sys/class/net/private':
SYSFS{addr_len}="6"
SYSFS{address}="00:0d:60:77:30:91"
SYSFS{broadcast}="ff:ff:ff:ff:ff:ff"
SYSFS{features}="0x3a9"
SYSFS{flags}="0x1003"
SYSFS{ifindex}="2"
SYSFS{iflink}="2"
SYSFS{mtu}="1500"
SYSFS{tx_queue_len}="1000"
SYSFS{type}="1"
follow the class device's "device"
looking at the device chain at '/sys/devices/pci0000:00/0000:00:1e.0/0000:02:01.0':
BUS="pci"
ID="0000:02:01.0"
SYSFS{class}="0x020000"
SYSFS{detach_state}="0"
SYSFS{device}="0x101e"
SYSFS{irq}="11"
SYSFS{subsystem_device}="0x0549"
SYSFS{subsystem_vendor}="0x1014"
SYSFS{vendor}="0x8086"
The matching device will be renamed to the given name. The device name
will not be put into the udev database, cause the kernel renames the
device and the sysfs name disappears.
I like it, cause it plugs in nicely. We have all the naming features
and sysfs queries and walks inside of udev. The sysfs timing races
are already solved and the management tools are working for net devices
too. nameif can only match the MAC address now. udev can match any sysfs
value of the device tree the net device is connected to.
But right, net devices do not have device nodes :)
2004-03-25 10:19:39 +03:00
{
int sk ;
struct ifreq ifr ;
int retval ;
2004-10-19 06:11:51 +04:00
dbg ( " changing net interface name from '%s' to '%s' " , udev - > kernel_name , udev - > name ) ;
if ( udev - > test_run )
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
return 0 ;
[PATCH] hmm, handle net devices with udev?
Hmm, Arndt Bergmann sent a patch like this one a few weeks ago and
I want to bring the question back, if we want to handle net device
naming with udev.
With this patch it is actually possible to specify something like this
in udev.rules:
KERNEL="dummy*", SYSFS{address}="00:00:00:00:00:00", SYSFS{features}="0x0", NAME="blind%n"
KERNEL="eth*", SYSFS{address}="00:0d:60:77:30:91", NAME="private"
and you will get:
[root@pim udev.kay]# cat /proc/net/dev
Inter-| Receive | Transmit
face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
lo: 1500 30 0 0 0 0 0 0 1500 30 0 0 0 0 0 0
private: 278393 1114 0 0 0 0 0 0 153204 1468 0 0 0 0 0 0
sit0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
blind0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
The udevinfo program is also working:
[root@pim udev.kay]# ./udevinfo -a -p /sys/class/net/private
looking at class device '/sys/class/net/private':
SYSFS{addr_len}="6"
SYSFS{address}="00:0d:60:77:30:91"
SYSFS{broadcast}="ff:ff:ff:ff:ff:ff"
SYSFS{features}="0x3a9"
SYSFS{flags}="0x1003"
SYSFS{ifindex}="2"
SYSFS{iflink}="2"
SYSFS{mtu}="1500"
SYSFS{tx_queue_len}="1000"
SYSFS{type}="1"
follow the class device's "device"
looking at the device chain at '/sys/devices/pci0000:00/0000:00:1e.0/0000:02:01.0':
BUS="pci"
ID="0000:02:01.0"
SYSFS{class}="0x020000"
SYSFS{detach_state}="0"
SYSFS{device}="0x101e"
SYSFS{irq}="11"
SYSFS{subsystem_device}="0x0549"
SYSFS{subsystem_vendor}="0x1014"
SYSFS{vendor}="0x8086"
The matching device will be renamed to the given name. The device name
will not be put into the udev database, cause the kernel renames the
device and the sysfs name disappears.
I like it, cause it plugs in nicely. We have all the naming features
and sysfs queries and walks inside of udev. The sysfs timing races
are already solved and the management tools are working for net devices
too. nameif can only match the MAC address now. udev can match any sysfs
value of the device tree the net device is connected to.
But right, net devices do not have device nodes :)
2004-03-25 10:19:39 +03:00
sk = socket ( PF_INET , SOCK_DGRAM , 0 ) ;
if ( sk < 0 ) {
dbg ( " error opening socket " ) ;
return - 1 ;
}
memset ( & ifr , 0x00 , sizeof ( struct ifreq ) ) ;
2005-03-07 06:29:43 +03:00
strlcpy ( ifr . ifr_name , udev - > kernel_name , IFNAMSIZ ) ;
strlcpy ( ifr . ifr_newname , udev - > name , IFNAMSIZ ) ;
[PATCH] hmm, handle net devices with udev?
Hmm, Arndt Bergmann sent a patch like this one a few weeks ago and
I want to bring the question back, if we want to handle net device
naming with udev.
With this patch it is actually possible to specify something like this
in udev.rules:
KERNEL="dummy*", SYSFS{address}="00:00:00:00:00:00", SYSFS{features}="0x0", NAME="blind%n"
KERNEL="eth*", SYSFS{address}="00:0d:60:77:30:91", NAME="private"
and you will get:
[root@pim udev.kay]# cat /proc/net/dev
Inter-| Receive | Transmit
face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
lo: 1500 30 0 0 0 0 0 0 1500 30 0 0 0 0 0 0
private: 278393 1114 0 0 0 0 0 0 153204 1468 0 0 0 0 0 0
sit0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
blind0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
The udevinfo program is also working:
[root@pim udev.kay]# ./udevinfo -a -p /sys/class/net/private
looking at class device '/sys/class/net/private':
SYSFS{addr_len}="6"
SYSFS{address}="00:0d:60:77:30:91"
SYSFS{broadcast}="ff:ff:ff:ff:ff:ff"
SYSFS{features}="0x3a9"
SYSFS{flags}="0x1003"
SYSFS{ifindex}="2"
SYSFS{iflink}="2"
SYSFS{mtu}="1500"
SYSFS{tx_queue_len}="1000"
SYSFS{type}="1"
follow the class device's "device"
looking at the device chain at '/sys/devices/pci0000:00/0000:00:1e.0/0000:02:01.0':
BUS="pci"
ID="0000:02:01.0"
SYSFS{class}="0x020000"
SYSFS{detach_state}="0"
SYSFS{device}="0x101e"
SYSFS{irq}="11"
SYSFS{subsystem_device}="0x0549"
SYSFS{subsystem_vendor}="0x1014"
SYSFS{vendor}="0x8086"
The matching device will be renamed to the given name. The device name
will not be put into the udev database, cause the kernel renames the
device and the sysfs name disappears.
I like it, cause it plugs in nicely. We have all the naming features
and sysfs queries and walks inside of udev. The sysfs timing races
are already solved and the management tools are working for net devices
too. nameif can only match the MAC address now. udev can match any sysfs
value of the device tree the net device is connected to.
But right, net devices do not have device nodes :)
2004-03-25 10:19:39 +03:00
retval = ioctl ( sk , SIOCSIFNAME , & ifr ) ;
if ( retval ! = 0 )
dbg ( " error changing net interface name " ) ;
2004-03-27 12:21:46 +03:00
close ( sk ) ;
[PATCH] hmm, handle net devices with udev?
Hmm, Arndt Bergmann sent a patch like this one a few weeks ago and
I want to bring the question back, if we want to handle net device
naming with udev.
With this patch it is actually possible to specify something like this
in udev.rules:
KERNEL="dummy*", SYSFS{address}="00:00:00:00:00:00", SYSFS{features}="0x0", NAME="blind%n"
KERNEL="eth*", SYSFS{address}="00:0d:60:77:30:91", NAME="private"
and you will get:
[root@pim udev.kay]# cat /proc/net/dev
Inter-| Receive | Transmit
face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
lo: 1500 30 0 0 0 0 0 0 1500 30 0 0 0 0 0 0
private: 278393 1114 0 0 0 0 0 0 153204 1468 0 0 0 0 0 0
sit0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
blind0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
The udevinfo program is also working:
[root@pim udev.kay]# ./udevinfo -a -p /sys/class/net/private
looking at class device '/sys/class/net/private':
SYSFS{addr_len}="6"
SYSFS{address}="00:0d:60:77:30:91"
SYSFS{broadcast}="ff:ff:ff:ff:ff:ff"
SYSFS{features}="0x3a9"
SYSFS{flags}="0x1003"
SYSFS{ifindex}="2"
SYSFS{iflink}="2"
SYSFS{mtu}="1500"
SYSFS{tx_queue_len}="1000"
SYSFS{type}="1"
follow the class device's "device"
looking at the device chain at '/sys/devices/pci0000:00/0000:00:1e.0/0000:02:01.0':
BUS="pci"
ID="0000:02:01.0"
SYSFS{class}="0x020000"
SYSFS{detach_state}="0"
SYSFS{device}="0x101e"
SYSFS{irq}="11"
SYSFS{subsystem_device}="0x0549"
SYSFS{subsystem_vendor}="0x1014"
SYSFS{vendor}="0x8086"
The matching device will be renamed to the given name. The device name
will not be put into the udev database, cause the kernel renames the
device and the sysfs name disappears.
I like it, cause it plugs in nicely. We have all the naming features
and sysfs queries and walks inside of udev. The sysfs timing races
are already solved and the management tools are working for net devices
too. nameif can only match the MAC address now. udev can match any sysfs
value of the device tree the net device is connected to.
But right, net devices do not have device nodes :)
2004-03-25 10:19:39 +03:00
return retval ;
}
2004-10-19 06:11:51 +04:00
int udev_add_device ( struct udevice * udev , struct sysfs_class_device * class_dev )
2003-07-21 07:48:48 +04:00
{
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
char * pos ;
2004-10-15 07:36:07 +04:00
int retval = 0 ;
2003-07-21 07:48:48 +04:00
2005-03-12 23:41:57 +03:00
if ( udev - > type = = DEV_BLOCK | | udev - > type = = DEV_CLASS ) {
2005-02-21 15:44:39 +03:00
udev - > devt = get_devt ( class_dev ) ;
if ( ! udev - > devt ) {
2004-10-15 07:36:07 +04:00
dbg ( " no dev-file found, do nothing " ) ;
2004-10-19 06:28:39 +04:00
return 0 ;
[PATCH] hmm, handle net devices with udev?
Hmm, Arndt Bergmann sent a patch like this one a few weeks ago and
I want to bring the question back, if we want to handle net device
naming with udev.
With this patch it is actually possible to specify something like this
in udev.rules:
KERNEL="dummy*", SYSFS{address}="00:00:00:00:00:00", SYSFS{features}="0x0", NAME="blind%n"
KERNEL="eth*", SYSFS{address}="00:0d:60:77:30:91", NAME="private"
and you will get:
[root@pim udev.kay]# cat /proc/net/dev
Inter-| Receive | Transmit
face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
lo: 1500 30 0 0 0 0 0 0 1500 30 0 0 0 0 0 0
private: 278393 1114 0 0 0 0 0 0 153204 1468 0 0 0 0 0 0
sit0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
blind0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
The udevinfo program is also working:
[root@pim udev.kay]# ./udevinfo -a -p /sys/class/net/private
looking at class device '/sys/class/net/private':
SYSFS{addr_len}="6"
SYSFS{address}="00:0d:60:77:30:91"
SYSFS{broadcast}="ff:ff:ff:ff:ff:ff"
SYSFS{features}="0x3a9"
SYSFS{flags}="0x1003"
SYSFS{ifindex}="2"
SYSFS{iflink}="2"
SYSFS{mtu}="1500"
SYSFS{tx_queue_len}="1000"
SYSFS{type}="1"
follow the class device's "device"
looking at the device chain at '/sys/devices/pci0000:00/0000:00:1e.0/0000:02:01.0':
BUS="pci"
ID="0000:02:01.0"
SYSFS{class}="0x020000"
SYSFS{detach_state}="0"
SYSFS{device}="0x101e"
SYSFS{irq}="11"
SYSFS{subsystem_device}="0x0549"
SYSFS{subsystem_vendor}="0x1014"
SYSFS{vendor}="0x8086"
The matching device will be renamed to the given name. The device name
will not be put into the udev database, cause the kernel renames the
device and the sysfs name disappears.
I like it, cause it plugs in nicely. We have all the naming features
and sysfs queries and walks inside of udev. The sysfs timing races
are already solved and the management tools are working for net devices
too. nameif can only match the MAC address now. udev can match any sysfs
value of the device tree the net device is connected to.
But right, net devices do not have device nodes :)
2004-03-25 10:19:39 +03:00
}
2003-07-21 07:48:48 +04:00
}
2005-03-13 00:36:32 +03:00
if ( udev_rules_get_name ( udev , class_dev ) ! = 0 )
2005-02-01 13:12:32 +03:00
return 0 ;
2003-11-14 07:08:10 +03:00
2004-10-19 06:11:51 +04:00
dbg ( " adding name='%s' " , udev - > name ) ;
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
2004-10-06 10:39:05 +04:00
selinux_init ( ) ;
2004-10-19 06:28:39 +04:00
2005-03-12 23:41:57 +03:00
if ( udev - > type = = DEV_BLOCK | | udev - > type = = DEV_CLASS ) {
2005-02-09 10:43:18 +03:00
retval = create_node ( udev , class_dev ) ;
2004-04-02 09:52:46 +04:00
if ( retval ! = 0 )
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
goto exit ;
2004-04-01 12:59:58 +04:00
2004-11-12 08:32:19 +03:00
if ( udev_db_add_device ( udev ) ! = 0 )
dbg ( " udev_db_add_dev failed, but we create the node anyway, "
2004-10-19 06:28:39 +04:00
" remove might not work for custom names " ) ;
[PATCH] hmm, handle net devices with udev?
Hmm, Arndt Bergmann sent a patch like this one a few weeks ago and
I want to bring the question back, if we want to handle net device
naming with udev.
With this patch it is actually possible to specify something like this
in udev.rules:
KERNEL="dummy*", SYSFS{address}="00:00:00:00:00:00", SYSFS{features}="0x0", NAME="blind%n"
KERNEL="eth*", SYSFS{address}="00:0d:60:77:30:91", NAME="private"
and you will get:
[root@pim udev.kay]# cat /proc/net/dev
Inter-| Receive | Transmit
face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
lo: 1500 30 0 0 0 0 0 0 1500 30 0 0 0 0 0 0
private: 278393 1114 0 0 0 0 0 0 153204 1468 0 0 0 0 0 0
sit0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
blind0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
The udevinfo program is also working:
[root@pim udev.kay]# ./udevinfo -a -p /sys/class/net/private
looking at class device '/sys/class/net/private':
SYSFS{addr_len}="6"
SYSFS{address}="00:0d:60:77:30:91"
SYSFS{broadcast}="ff:ff:ff:ff:ff:ff"
SYSFS{features}="0x3a9"
SYSFS{flags}="0x1003"
SYSFS{ifindex}="2"
SYSFS{iflink}="2"
SYSFS{mtu}="1500"
SYSFS{tx_queue_len}="1000"
SYSFS{type}="1"
follow the class device's "device"
looking at the device chain at '/sys/devices/pci0000:00/0000:00:1e.0/0000:02:01.0':
BUS="pci"
ID="0000:02:01.0"
SYSFS{class}="0x020000"
SYSFS{detach_state}="0"
SYSFS{device}="0x101e"
SYSFS{irq}="11"
SYSFS{subsystem_device}="0x0549"
SYSFS{subsystem_vendor}="0x1014"
SYSFS{vendor}="0x8086"
The matching device will be renamed to the given name. The device name
will not be put into the udev database, cause the kernel renames the
device and the sysfs name disappears.
I like it, cause it plugs in nicely. We have all the naming features
and sysfs queries and walks inside of udev. The sysfs timing races
are already solved and the management tools are working for net devices
too. nameif can only match the MAC address now. udev can match any sysfs
value of the device tree the net device is connected to.
But right, net devices do not have device nodes :)
2004-03-25 10:19:39 +03:00
2004-10-19 06:28:39 +04:00
/* use full path to the environment */
2005-03-07 06:29:43 +03:00
snprintf ( udev - > devname , sizeof ( udev - > devname ) , " %s/%s " , udev_root , udev - > name ) ;
udev - > devname [ sizeof ( udev - > devname ) - 1 ] = ' \0 ' ;
2004-10-19 06:28:39 +04:00
2005-03-12 23:41:57 +03:00
} else if ( udev - > type = = DEV_NET ) {
2004-10-19 06:28:39 +04:00
/* look if we want to change the name of the netif */
2004-10-19 06:11:51 +04:00
if ( strcmp ( udev - > name , udev - > kernel_name ) ! = 0 ) {
retval = rename_net_if ( udev ) ;
2004-04-02 09:52:46 +04:00
if ( retval ! = 0 )
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
goto exit ;
2004-10-19 06:28:39 +04:00
2004-12-20 10:57:31 +03:00
/* we've changed the name, now fake the devpath, cause the
* original kernel name sleeps with the fishes and we don ' t
* get an event from the kernel with the new name
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
*/
2004-10-19 06:11:51 +04:00
pos = strrchr ( udev - > devpath , ' / ' ) ;
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
if ( pos ! = NULL ) {
pos [ 1 ] = ' \0 ' ;
2005-03-07 06:29:43 +03:00
strlcat ( udev - > devpath , udev - > name , sizeof ( udev - > devpath ) ) ;
2004-10-19 06:28:39 +04:00
setenv ( " DEVPATH " , udev - > devpath , 1 ) ;
2004-12-20 10:57:31 +03:00
setenv ( " INTERFACE " , udev - > name , 1 ) ;
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
}
2004-04-01 12:59:58 +04:00
2004-10-19 06:28:39 +04:00
/* use netif name for the environment */
2005-03-07 06:29:43 +03:00
strlcpy ( udev - > devname , udev - > name , sizeof ( udev - > devname ) ) ;
2004-10-19 06:28:39 +04:00
}
[PATCH] hmm, handle net devices with udev?
Hmm, Arndt Bergmann sent a patch like this one a few weeks ago and
I want to bring the question back, if we want to handle net device
naming with udev.
With this patch it is actually possible to specify something like this
in udev.rules:
KERNEL="dummy*", SYSFS{address}="00:00:00:00:00:00", SYSFS{features}="0x0", NAME="blind%n"
KERNEL="eth*", SYSFS{address}="00:0d:60:77:30:91", NAME="private"
and you will get:
[root@pim udev.kay]# cat /proc/net/dev
Inter-| Receive | Transmit
face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
lo: 1500 30 0 0 0 0 0 0 1500 30 0 0 0 0 0 0
private: 278393 1114 0 0 0 0 0 0 153204 1468 0 0 0 0 0 0
sit0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
blind0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
The udevinfo program is also working:
[root@pim udev.kay]# ./udevinfo -a -p /sys/class/net/private
looking at class device '/sys/class/net/private':
SYSFS{addr_len}="6"
SYSFS{address}="00:0d:60:77:30:91"
SYSFS{broadcast}="ff:ff:ff:ff:ff:ff"
SYSFS{features}="0x3a9"
SYSFS{flags}="0x1003"
SYSFS{ifindex}="2"
SYSFS{iflink}="2"
SYSFS{mtu}="1500"
SYSFS{tx_queue_len}="1000"
SYSFS{type}="1"
follow the class device's "device"
looking at the device chain at '/sys/devices/pci0000:00/0000:00:1e.0/0000:02:01.0':
BUS="pci"
ID="0000:02:01.0"
SYSFS{class}="0x020000"
SYSFS{detach_state}="0"
SYSFS{device}="0x101e"
SYSFS{irq}="11"
SYSFS{subsystem_device}="0x0549"
SYSFS{subsystem_vendor}="0x1014"
SYSFS{vendor}="0x8086"
The matching device will be renamed to the given name. The device name
will not be put into the udev database, cause the kernel renames the
device and the sysfs name disappears.
I like it, cause it plugs in nicely. We have all the naming features
and sysfs queries and walks inside of udev. The sysfs timing races
are already solved and the management tools are working for net devices
too. nameif can only match the MAC address now. udev can match any sysfs
value of the device tree the net device is connected to.
But right, net devices do not have device nodes :)
2004-03-25 10:19:39 +03:00
}
2003-07-21 07:48:48 +04:00
exit :
2005-04-27 10:52:14 +04:00
selinux_exit ( ) ;
2003-11-14 07:08:10 +03:00
2003-07-21 07:48:48 +04:00
return retval ;
}