It seems the following patch resolves the problem... moreover it should
resolve also the other issues regarding missing cciss, ida and rd block
files in /dev for some particular hw RAID controllers...
Cheers,
-Andrea
-------- Original Message --------
Subject: [systemimager-commits] r3578 - in trunk/initrd_source: make.d
patches skel/etc/init.d
Date: Sun, 21 May 2006 10:09:38 -0500
From: Andrea Righi <[EMAIL PROTECTED]>
Reply-To: sisuite-devel@lists.sourceforge.net
To: [EMAIL PROTECTED]
Author: arighi
Date: 2006-05-21 10:09:37 -0500 (Sun, 21 May 2006)
New Revision: 3578
Added:
trunk/initrd_source/patches/udev.names_rules.patch
Modified:
trunk/initrd_source/make.d/udev.rul
trunk/initrd_source/skel/etc/init.d/functions
Log:
fix: added udev rules to populate /dev with files for RAID controllers
and other useful block devices
Modified: trunk/initrd_source/make.d/udev.rul
===================================================================
--- trunk/initrd_source/make.d/udev.rul 2006-05-20 19:56:20 UTC (rev 3577)
+++ trunk/initrd_source/make.d/udev.rul 2006-05-21 15:09:37 UTC (rev 3578)
@@ -31,6 +31,7 @@
$(INITRD_BUILD_DIR).prep
rsync -av --exclude 'usr/share/man/' $(UDEV_DIR)/debian/udev/
$(INITRD_BUILD_DIR)/
install -m 755 $(UDEV_DIR)/debian/udev.init
$(INITRD_BUILD_DIR)/etc/init.d/udev
+ install -m 755 $(UDEV_DIR)/udeb/extras/raid-devfs.sh
$(INITRD_BUILD_DIR)/lib/udev/raid-devfs.sh
$(UDEV_DIR).unpack: $(INITRD_DIR)/make.d/udev.rul \
Added: trunk/initrd_source/patches/udev.names_rules.patch
===================================================================
--- trunk/initrd_source/patches/udev.names_rules.patch 2006-05-20
19:56:20 UTC (rev 3577)
+++ trunk/initrd_source/patches/udev.names_rules.patch 2006-05-21
15:09:37 UTC (rev 3578)
@@ -0,0 +1,17 @@
+--- udev-079.orig/debian/rules.d/20-names.rules 2006-05-21
16:10:16.000000000 +0200
++++ udev-079/debian/rules.d/20-names.rules 2006-05-21
16:10:19.000000000 +0200
+@@ -14,6 +14,14 @@
+ KERNEL=="device-mapper", NAME="mapper/control"
+ KERNEL=="dm-[0-9]*", NAME="dm/%n"
+
++# RAID controllers
++KERNEL=="cciss!*|ida!*|rd!*", PROGRAM="/lib/udev/raid-devfs.sh %k", \
++ NAME="%c{1}", SYMLINK+="%k"
++# other block devices
++KERNEL=="md[0-9]*", NAME="md/%n",
++KERNEL=="loop[0-9]*", NAME="loop/%n"
++KERNEL=="ram[0-9]*", NAME="rd/%n"
++
+ # Input devices, group under /dev/input
+ KERNEL=="event[0-9]*", NAME="input/%k"
+ KERNEL=="mice", NAME="input/%k"
Modified: trunk/initrd_source/skel/etc/init.d/functions
===================================================================
--- trunk/initrd_source/skel/etc/init.d/functions 2006-05-20 19:56:20
UTC (rev 3577)
+++ trunk/initrd_source/skel/etc/init.d/functions 2006-05-21 15:09:37
UTC (rev 3578)
@@ -737,9 +737,9 @@
#
# workaround chicken/egg bug in mdadm and raidautorun with udev
# they do the ioctl on the not yet existing device node...
- for i in 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
23 24 25 26 27 28 29 30 31 ; do
- [ -e /dev/md$i ] || mknod -m 660 /dev/md$i b 9 $i
- done
+ #for i in 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
22 23 24 25 26 27 28 29 30 31 ; do
+ # [ -e /dev/md$i ] || mknod -m 660 /dev/md$i b 9 $i
+ #done
}
#
################################################################################
dann frazier wrote:
> On Tue, May 16, 2006 at 03:18:03PM -0500, finley wrote:
>> Author: finley
>> Date: 2006-05-16 15:18:02 -0500 (Tue, 16 May 2006)
>> New Revision: 3571
>>
>> Modified:
>> trunk/initrd_source/skel/etc/init.d/functions
>> Log:
>> * test for md device files before creating
>>
>> Modified: trunk/initrd_source/skel/etc/init.d/functions
>> ===================================================================
>> --- trunk/initrd_source/skel/etc/init.d/functions 2006-05-16 20:16:59 UTC
>> (rev 3570)
>> +++ trunk/initrd_source/skel/etc/init.d/functions 2006-05-16 20:18:02 UTC
>> (rev 3571)
>> @@ -731,10 +731,14 @@
>> logmsg start_udevd
>> /etc/init.d/udev start
>>
>> + #XXX We should figure out the root cause for this, at some point,
>> + # and fix the actual problem, rather than just treating the
>> + # symptom as we are doing here. -BEF-
>> + #
>> # workaround chicken/egg bug in mdadm and raidautorun with udev
>> # they do the ioctl on the not yet existing device node...
>> for i in 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
>> 24 25 26 27 28 29 30 31 ; do
>> - mknod -m 660 /dev/md$i b 9 $i
>> + [ -e /dev/md$i ] || mknod -m 660 /dev/md$i b 9 $i
>> done
>> }
>> #
>
> Is that what you're missing? From /etc/init.d/udev:
>
> # wait for the udevd childs to finish
> log_action_begin_msg "Waiting for /dev to be fully populated"
> while [ -d /dev/.udev/queue/ ]; do
> sleep 1
> udevd_timeout=$(($udevd_timeout - 1))
> if [ $udevd_timeout -eq 0 ]; then
> # ps axf
> break
> fi
> done
>
-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Sisuite-devel mailing list
Sisuite-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-devel
