greg@kroah.com [Thu, 25 Dec 2003 03:44:19 +0000 (19:44 -0800)]
[PATCH] fix long delay for all devices in namedev
Now we only sleep if we can't find the device file, and we have
a hack to sleep for 1 second if we are on a partition. This will be
removed when the libsysfs change gets made...
In udev-009 and previous releases, for klibc compatibility, the pgsize var
in sysfs_read_attribute() (file sysfs_dir.c under libsysfs) is handcoded
to 0x4000.
Should it not be 4096 bytes (0x1000 in hex) instead of 0x4000 (16k bytes)?
2003-12-22 multipath-010
* tweak the install target in Makefile
* stop passing fds as argument : this change enable a strict
segregation of ugly 2.4 code
* sysfs version of get_lun_strings()
* be careful about the return of get_unique_id() since errors
formerly caught up by if(open()) in the caller fn are now returned
by get_unique_id()
* send get_serial() in unused.c
2003-12-22 multipath-010
* don't print .sg_dev if equal to .dev (2.6) in print_path()
* since the kernel code handles defective paths, remove all
code to cope with them :
* move do_tur() to unused.c
* remove .state from path struct
* remove .state settings & conditionals
* add a cmdline switch to force maps to failover mode,
ie 1 path per priority group
* add default policies to the whitelist array (spread io ==
MULTIBUS / io forced to 1 path == FAILOVER)
* move get_disk_size() call out of add_map() to coalesce()
* comment tricky coalesce() fn
* bogus unsused.c file renamed to unused.c
2003-12-20 multipath-010
* big ChangeLog update
* start to give a little control over target params :
introduce cmdline arg -i to control polling interval
* cope with hotplug-style calling convention :
ie "multipath scsi $DEVPATH" ... to avoid messing with
online maps not concerned by an event
* example hotplug agent to drop in /etc/hotplug.d/scsi
* revert the run & resched patch : unless someone proves me
wrong, this was overdesigned
* move commented out functions in unused.c
* update multipath target params to "udm[23] style"
* mp target now supports nr_path == 1, so do we
* add gratuitous free()
* push version forward
[PATCH] experimental CALLOUT script for devfs ide node creation with cd, disc, part
Here is a experimental CALLOUT script for udev to create devfs nodes
for IDE-devices. Not that I need these, I just wanted to see if it works :)
The script is really stupid, no error handling, nothing more than
absolutely needed. The rule uses the 'k' format char of the previous
patch. The %D is not used, so the user can have disc or disk :)
this single line:
CALLOUT, BUS="ide", PROGRAM="/etc/udev/ide-devfs.sh %k %b %n", ID="hd*", NAME="%1c", SYMLINK="%2c"
creates the following on my machine with two hard disks, one DVD and a PCMCIA-compact-flash inserted:
/udev
|-- hda
|-- hda1
|-- hda2
|-- hda4
|-- hdb
|-- hdb1
|-- hdc
|-- hde
|-- hde1
`-- ide
|-- host0
| |-- bus0
| | |-- target0
| | | `-- lun0
| | | |-- disc -> ../../../../../hda
| | | |-- part1 -> ../../../../../hda1
| | | |-- part2 -> ../../../../../hda2
| | | `-- part4 -> ../../../../../hda4
| | `-- target1
| | `-- lun0
| | |-- disc -> ../../../../../hdb
| | `-- part1 -> ../../../../../hdb1
| `-- bus1
| `-- target0
| `-- lun0
| `-- cd -> ../../../../../hdc
`-- host2
`-- bus0
`-- target0
`-- lun0
|-- disc -> ../../../../../hde
`-- part1 -> ../../../../../hde1
Ananth released sysfsutils 0.4.0 last night, I'm sure you saw the email.
Here's a patch with the latest changes from the pre-patch I already
gave you. It includes sysfs_get_device_parent(), which you said you
needed. I've run your test scripts and I've built scsi_id. Please
play around with this and check it out.
There are quite a few changes. Please do not access
structure pointers, like sysfs_device's parent, directly like
dev->parent. Please use the "get" function to retrieve. The functions
load things on demand and refresh views under the covers.
I've moved the malloc out of the udevdb into udev-remove to free the
struct after use and not to allocate a different struct in the case the
device is not in the data base. I seems a bit easier to read.
udev segfaults with NULL-pointer, if the device is not in the database:
./test.block: line 29: 4844 Segmentation fault $BIN block
Dec 17 22:47:42 pim udev[4882]: udev_remove_device: '/block/sdy' not found in database, falling back on default name
Dec 17 22:47:42 pim udev[4882]: udev_remove_device: name is '(null)'
Scsi_id hasn't been changed to use the latest libsysfs changes. The
"directory" in the sysfs_class_device is now considered "private" and only
should be accessed using functions. Treating the structures as handles lets
us only load information when it's needed, reducing caching or stale
information and also helping performance.
[PATCH] mark config files as such in the rpm spec file
Got a nasty surprise after upgrading my RPM. The config files weren't
marked as configs and were subsequently replaced! DOH! Fortunately, I
didn't have much in there yet but that certainly isn't something others
will like going forward. Attached is a simple patch to the spec file so
they aren't replaced in the future. The config files included in the
RPM should wind up as udev.conf.rpmnew, etc if they are different from
what is installed.
olh@suse.de [Wed, 17 Dec 2003 07:41:59 +0000 (23:41 -0800)]
[PATCH] dump latest klibc into the udev build tree
KLIBC is used as an internal makefile variable, it expands to either
true or false right now. udev should use something else than KLIBC to
allow build against the latest and greatest klibc version.
[PATCH] don't rely on field order in namedev_parse
o change the parsing to get a key from the rule and sort it
into our list of known keys instead of expecting a special order
o the key to match a sysfs file must be prependend by 'SYSFS_' now
to match with the new parsing.
(The config must be changed, but it's a bit more descriptive too.)
o put names of fields in define's, like the name of the methods
o update all tests and the man page
rml@ximian.com [Wed, 17 Dec 2003 07:30:15 +0000 (23:30 -0800)]
[PATCH] install initscript in udev rpm
Attached patch installs the initscript via 'make install' and adds it to
the RPM package. The RPM script then runs chkconfig(8) to setup the
initscript to run at the appropriate runlevels.
* Make the HW-specific get_unique_id switch pretty
* Prepare to field-test by whitelisting all known fibre array, try to
fetch WWID from the standard EVPD 0x83 off 8 for everyone ... we will
learn from feedback :)
Could you drop a note with the udev-009 release-notes asking for testing
this WWID fetching thing ?
Try this patch if you like, to get special parts of the callout output.
This beast works now:
CALLOUT, BUS="scsi", PROGRAM="/bin/echo -n node link1 link2", ID="node *", NAME="%1c", SYMLINK="%2c %3c"
The callout returned string is separated by spaces and is
addressed by the "len" value of the 'c' format char.
Since we support symlinks, this my be useful for other uses of callout too.
introduce 'len number' for format chars
the first use is 'c'-the callout return to select a part of the output string like:
CALLOUT, BUS="scsi", PROGRAM="/bin/echo -n node link1 link2", ID="node *", NAME="%1c", SYMLINK="%2c %3c"
(note: first part is requested by len=1, len=0 will return the whole string)
add a test to udev-test.pl
I am sending you a pre-release patch. It's everything that's in our
current CVS tree. It adds the functionality you've been looking for. Please
play with this before checking it into your tree, I'd like to know if
it's ok with you or if you find problems. I have tested this out with
test.all and the perl regression test. Let me know what you think.
Still need to do more testing for our work and add some more functions
related to the changes.
I've gone into namedev.c and udev-add.c to make the necessary changes
in line with the library. I have not gone and edited any of the "extras".
Changes:
1) Libsysfs object structures work more as handles now, their included
directories or devices are labeled private. If you need attributes
from a sysfs_class_device, call the available function and don't access
the directory directly. Same holds true for a sysfs_class_device
sysfs_device. Do not access the link directly but call the function
sysfs_get_classdev_device() instead. We only populate entries upon
request, makes things faster and uses less memory.
2) Added sysfs_get_classdev_parent() as requested.
3) Changed getpagesize to sysconf.
4) Added sysfs_refresh_attributes function for refreshing views of
attribute lists. We still need to add refresh for links and subdirs. All
udev needs to do is keep calling sysfs_get_classdev_attr() and that will
internally call the refresh routine.
As usual, when the stuff gets in the tree - here is a small
man page update and a TODO line removal.
mention multiple symlinks
update example with silly symlink rule :)
shorten example line to have less than 80 chars
remove "want symlink support" from TODO
* configure the multipath target with round-robin path selector and
conservative default for a start : yes it makes this release the first
really useful one.
* temporarily disable map creation for single path device due to current
restrictive defaults in the kernel target. Sistina should work it out.
* correct the strncmp logic in blacklist function.
* update the Makefiles to autodetect libgcc.a & gcc includes
"ulibc-style". Factorisation of udevdirs & others niceties
* drop a hint about absent /dev/sd? on failed open() for poor Debian
users who don't imagine their favorite distro with only 16 preconfigured
SCSI device nodes :)
david@fubar.dk [Mon, 8 Dec 2003 17:19:19 +0000 (09:19 -0800)]
[PATCH] D-BUS patch for udev-008
Attached is a patch against udev-008 to send out a D-BUS message when a
device node is added or removed.
Using D-BUS lingo, udev acquires the org.kernel.udev service and sends
out a NodeCreated or NodeDeleted signal on the
org.kernel.udev.NodeMonitor interface. Each signal carries two
parameters: the node in question and the corresponding sysfs path.
[Note: the D-BUS concepts of service, interface, object can be a bit
confusing at first glance]
An example program listening for these messages looks like this
#!/usr/bin/python
import dbus
import gtk
def udev_signal_received(dbus_iface, member, service, object_path, message):
[filename, sysfs_path] = message.get_args_list()
if member=='NodeCreated':
print 'Node %s created for %s'%(filename, sysfs_path)
elif member=='NodeDeleted':
print 'Node %s deleted for %s'%(filename, sysfs_path)
def main():
bus = dbus.Bus(dbus.Bus.TYPE_SYSTEM)
bus.add_signal_receiver(udev_signal_received,
'org.kernel.udev.NodeMonitor', # interface
'org.kernel.udev', # service
'/org/kernel/udev/NodeMonitor') # object
gtk.mainloop()
if __name__ == '__main__':
main()
and this is the output when hot-plugging some usb-storage.
[david@laptop udev-008]$ ~/node_monitor.py
Node /udev/sda created for /block/sda
Node /udev/sda1 created for /block/sda/sda1
Node /udev/sda1 deleted for /block/sda/sda1
Node /udev/sda deleted for /block/sda
The patch requires D-BUS 0.20 or later while the python example program
requires D-BUS from CVS as I only recently applied a patch against the
python bindings.
> > here is a experimental symlink creation patch - for discussion,
> > in which direction we should go.
> > It is possible now to define SYMLINK= after the NAME= in udev.rules.
> > The link is relative to the node, but the path is not optimized now
> > if the node and the link are in the same nested directory.
> > Only one link is supported, cause i need to sleep now :)
> >
> > 06-simple-symlink-creation.diff
> > simple symlink creation
> > reorganized udev-remove to have access to the symlink field
> > subdir creation/removal are functions now
> > udev-test.pl tests for link creation/removal
Here is a new version with relative link target path optimization
an better tests in udev-test.pl:
[PATCH] better allow builds of extras programs under udev
Here is an improved version of the patch that enables builds of the extras
progams for the targets all, clean, install, and uninstall, and passes
down the "prefix" for use by install and uninstall.
This patch enables building of the "extras" programs using the same build
environment as udev (i.e. build with udev's versions of klibc and
sysfsutils).
* implement a reschedule flag in /var/run. Last thing the prog do before
exit is check if a call to multipath was done (but canceled by
/var/run/multipath.run check) during its execution. If so restart the
main loop.
* implement a blacklist of sysfs bdev to not bother with for now (hd,
md, dm, sr, scd, ram, raw). This avoid sending SG_IO to unappropiate
devices.
here is the next update which brings this multipath to the state i'm not
ashamed of it being in udev :)
* drop a libdevmapper copy in extras/multipath; maybe discussions w/
Sistina folks will bring a better solution in the future.
* drop a putchar usage in libdevmapper to compile cleanly with klibc
* drop another such usage of my own in main.c
* massage the Makefile to compile libdevmapper against klibc
* use "ld" to produce the binary rather than "gcc -static"
* stop being stupid w/ uneeded major, minor & dev in main.c:dm_mk_node()
* reverse to creating striped target for now because the multipath
target is more hairy than expected initialy
* push the version code to 009 to be in synch w/ udev
builds & run here.
binary size is 43ko, which is fairly gratifying after all the efforts
I've put to compiling it with klibc :)
I've never seen any project where the documentation is in sync with the
code at this early stage. Nice! So here is the patch to make it extra perfect :)
remove random indent to be consistent
style is "shell style"
s/wildcard/pattern/
mention negation char in character class
While I was adding pattern match to the LABEL method i hit a bug.
We modify a string returned from libsysfs, so with every iteration is is
truncated by one char:
Dec 4 02:27:16 pim udev[23307]: do_label: dev->bus='scsi' sysfs_device->bus='scsi'
Dec 4 02:27:16 pim udev[23307]: do_label: look for device attribute 'vendor'
Dec 4 02:27:16 pim udev[23307]: do_label: xxx 'IBM-ESXS '
Dec 4 02:27:16 pim udev[23307]: do_label: compare attribute 'vendor' value 'IBM-ESX' with '?IBM-ESXS'
Dec 4 02:27:16 pim udev[23307]: do_label: dev->bus='scsi' sysfs_device->bus='scsi'
Dec 4 02:27:16 pim udev[23307]: do_label: look for device attribute 'vendor'
Dec 4 02:27:16 pim udev[23307]: do_label: xxx 'IBM-ESX'
Dec 4 02:27:16 pim udev[23307]: do_label: compare attribute 'vendor' value 'IBM-ES' with 'IBM-ESXS?'
Dec 4 02:27:16 pim udev[23307]: do_label: dev->bus='scsi' sysfs_device->bus='scsi'
Dec 4 02:27:16 pim udev[23307]: do_label: look for device attribute 'vendor'
Dec 4 02:27:16 pim udev[23307]: do_label: xxx 'IBM-ES'
Dec 4 02:27:16 pim udev[23307]: do_label: compare attribute 'vendor' value 'IBM-E' with 'IBM-ES??'
Dec 4 02:27:16 pim udev[23307]: do_label: dev->bus='scsi' sysfs_device->bus='scsi'
Dec 4 02:27:16 pim udev[23307]: do_label: look for device attribute 'vendor'
Dec 4 02:27:16 pim udev[23307]: do_label: xxx 'IBM-E'
Dec 4 02:27:16 pim udev[23307]: do_label: compare attribute 'vendor' value 'IBM-' with 'IBM-ESXSS'
I changed the behavior to remove only the line feed.
03-bug-in-linefeed-removal.diff
remove only the line feed from string not every last char
As promised yesterday, here is a patch to implement a more advanced
pattern matching instead of the simple '*'.
We can remove the "tty"="tty" line from udev.rules now and
replace "tty*" by "tty[0-9]*" to catch only the vc's.
implement pattern matching in namedev
'*' - to match zero or more chars
'?' - to match exactly one char
'[]' - character classes with ranges '[0-9]'and negation [!A]
greg@kroah.com [Wed, 3 Dec 2003 09:08:46 +0000 (01:08 -0800)]
[PATCH] add support for a main udev config file, udev.conf.
the older udev.config file is now called udev.rules.
This allows us to better control configuration values, and move away from
the environment variables.