gmane.linux.hotplug.devel

Re: [ANNOUNCE] udev 130 release

Marco d'Itri — 2008-10-07T14:32:53

Users or scripts? Maybe an hidden directory would be more appropriate.

Re: [ANNOUNCE] udev 130 release

Kay Sievers — 2008-10-07T14:29:00

Because it's not for udev itself, it does not use it. It's a public interface, and the counterpart to the /sys/dev/ directory. Users can retrieve the device name of a device without the need to use any library, or run a program like udevadm. Thanks, Kay -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo< at >vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html

Re: [ANNOUNCE] udev 130 release

Marco d'Itri — 2008-10-07T13:15:32

Do we really need to expose this in /dev/? Why not put this in /dev/.udev/ or some other hidden directory?

[ANNOUNCE] udev 130 release

Kay Sievers — 2008-10-07T13:11:16

Here comes a new udev version. Thanks to all who have contributed to this release. The tarball can be found here: ftp://ftp.kernel.org/pub/linux/utils/kernel/hotplug/ The development repository can be found here: http://www.kernel.org/git/?p=linux/hotplug/udev.git;a=summary The ChangeLog can be found here: http://www.kernel.org/git/?p=linux/hotplug/udev.git;a=blob;hb=HEAD;f=ChangeLog udev 130 ======== Bugfixes. Kernel devices and device nodes are connected now by reverse indices in /sys and /dev. A device number retrieved by a stat() or similar, the kernel device directory can be found by looking up: /sys/dev/{block,char}/: and the device node of the same device by looking up: /dev/{block,char}/: -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo< at >vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html

Re: ide dvd burner detected as ide-tape

Thomas Drillich — 2008-10-04T12:12:43

Am Freitag, 3. Oktober 2008 schrieb Bartlomiej Zolnierkiewicz: Seems that can happen Bart ;-) it's a 2.6.26.5 kernel running on debian/testing. I moved ide now to module alim15x3 7504 0 [permanent] ide_pci_generic 5252 0 [permanent] ide_core 140572 2 ide_pci_generic,alim15x3 but it didn't changed anything .. as expected. 00:1f.0 IDE interface: ALi Corporation M5229 IDE (rev c7) I added /proc/config.gz to the dmesg output

Re: udev 129 says it supports linux 2.6.19 but prints a warning

Kay Sievers — 2008-10-04T08:21:29

Yeah, we kind of already "depend" on it. It just depends on what features you are looking at. We can not know which udev features other services use, they usually just depend on a specific udev version, not a kernel/sysfs version, but udev can not offer some features on older kernels. Yeah, it's hard to tell. It will not stop creating device nodes, but the rest we can not be sure about. :) I'll do this. Thanks, Kay -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo< at >vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html

Re: udev 129 says it supports linux 2.6.19 but prints a warning

Mikael Magnusson — 2008-10-04T08:13:21

2008/10/4 Kay Sievers vrfy.org>: Ah, okay. In the commit message you used the phrasing "Future udev versions will depend on the current sysfs layout". Okay, I interpreted the "depend on" as "require" in the commit message. That is probably a little less confusing, thanks :).

Re: udev 129 says it supports linux 2.6.19 but prints a warning

Kay Sievers — 2008-10-04T07:53:36

There is no specific plan to remove basic support for the deprecated sysfs layout. It will likely continue to work fine for basic udev operations. But people working on current udev versions usually don't test this anymore, so it might be that there are bugs introduced which are not handled, hence the "is unsupported" not "it will stop working". The deprecated sysfs layout misses information, userspace starts to rely on, it misses devices, which are not exported by the kernel, because they could not be inserted into the sysfs device tree without breaking things. Also some more specialized udev rule features, which are used, do not work with the old layout. Storage volume handling and asynchronous media change events in DeviceKit also not work with the old sysfs layout. So it's not strictly a udev requirement, it's that system services will increasingly depend on new features udev offers, which will fail in in subtle ways with older kernels or the deprecated sysfs layout. I can change the message to: d

Re: udev 129 says it supports linux 2.6.19 but prints a warning

Marco d'Itri — 2008-10-03T21:50:52

Agreed. Every time kernel vs. udev requirements change it should be stated prominently in the release notes, or else it will be hell again for people like me who need to support real upgrade paths between distribution releases.

udev 129 says it supports linux 2.6.19 but prints a warning

Mikael Magnusson — 2008-10-03T20:33:11

I just updated udev to 129 and got this, udev: deprecated sysfs layout (CONFIG_SYSFS_DEPRECATED) is unsupported However, I have no such option in my kernel config (2.6.19). I assume this means 2.6.19 will soon not be supported in udev. It would perhaps be nice if the message said which kernel version would be required after the deprecation, and how far in the future this will happen?

Re: ide dvd burner detected as ide-tape

Bartlomiej Zolnierkiewicz — 2008-10-03T17:33:41

ide-tape driving DVD burner is a "CANNOT happen" case... Thomas, which kernel version is it? Please post a full dmesg output. Thanks, Bart -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo< at >vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html

Re: udev doesn't detect ddf-format raid array

Peter Leipold — 2008-10-03T16:53:13

Great, it works, thanks! :) # ./vol_id /dev/sda ID_FS_USAGE=raid ID_FS_TYPE=ddf_raid_member ID_FS_VERSION=02.00.0 ID_FS_UUID=_VsJ__B__VsJ0bsJ_nsJ____ ID_FS_UUID_ENC=\x08VsJ\x95\x10B\x02\x08VsJ0bsJ\x18nsJ\xff\xff\xff\xff ID_FS_LABEL= ID_FS_LABEL_ENC= # ./vol_id /dev/sdb ID_FS_USAGE=raid ID_FS_TYPE=ddf_raid_member ID_FS_VERSION=02.00.0 ID_FS_UUID=_VsJ__B__VsJ0bsJ_nsJ____ ID_FS_UUID_ENC=\x08VsJ\x95\x10B\x02\x08VsJ0bsJ\x18nsJ\xff\xff\xff\xff ID_FS_LABEL= ID_FS_LABEL_ENC= Ok, it's not crucial, I have a workaround to my original problem, so I can wait until debian-unstable starts using this upgraded libvolume_id. Peter -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo< at >vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html

Re: ide dvd burner detected as ide-tape

Kay Sievers — 2008-10-03T13:36:14

What udev rule? It's the kernel which need to know that this device isn't a tape but an optical drive. :) Kay -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo< at >vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html

Re: udev doesn't detect ddf-format raid array

Kay Sievers — 2008-10-03T13:33:29

Here is what I get now from the sector you sent: $ dd if=~/Desktop/sdb_ddf1.dat bs=1 seek=200049641472 of=ddf-header.img $ losetup /dev/loop0 ddf-header.img $ ./vol_id /dev/loop0 ID_FS_USAGE=raid ID_FS_TYPE=ddf_raid_member ID_FS_VERSION=02.00.0 ID_FS_UUID=8Z_I__B_8Z_I_e_I_q_I____ ID_FS_UUID_ENC=8Z\xb0I\x95\x10B\x028Z\xb0I\xc8e\xb0I\xb8q\xb0I\xff\xff\xff\xff ID_FS_LABEL= ID_FS_LABEL_ENC= If you want to try, let me know what it says: $ git clone git://git.kernel.org/pub/scm/linux/hotplug/udev.git $ cd udev $ ./autogen.sh $ cd extras/volume_id/ $ make $ ./vol_id /dev/sda $ ./vol_id /dev/sdb Note, that ./vol_id is just a autotools wrapper script, which runs the program with libvolume_id from the source tree, so don't copy the script to your system. :) Kay -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo< at >vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html

Re: ide dvd burner detected as ide-tape

Thomas Drillich — 2008-10-03T13:27:18

I did some more research on the LG GH22LP20 DVD burner http://www.lge.com/products/model/detail/gh22lp20.jhtml looking at device '/devices/pci0000:00/0000:00:1f.0/ide0/0.0': KERNEL=="0.0" SUBSYSTEM=="ide" DRIVER=="" ATTR{media}=="tape" ATTR{drivename}=="hda" ATTR{modalias}=="ide:m-tape" ATTR{model}=="HL)DP-ST< at >V< at >-RAI CH22HP20" ATTR{firmware}=="1.00" ATTR{serial}==""

Re: udev doesn't detect ddf-format raid array

Peter Leipold — 2008-10-03T08:59:06

Thank you Kay! As I fell into the mistake of buying such a card and believing it's true raid, I really appreciate your work! Peter -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo< at >vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html

Re: udev doesn't detect ddf-format raid array

Kay Sievers — 2008-10-03T08:17:37

That's fine, there have been no changes to ddf after udev 111 I guess it's because configure uses the default --prefix=/usr anf therefore /usr/local/dev/ to look for device nodes. But 125 sounds fine anyway. Ah, they don't conform to the ddf standard and put their signatures not exactly to the end of the volume, I'll add that to vol_id. Looks good: hexdump -C sda_ddf1.dat | head -1 00000000 11 de 11 de e2 e6 4c ae 38 5a b0 49 95 10 42 02 |......L.8Z.I..B.| 0xDE11DE11 is the signature we are looking for. I'll see if I can make that work. Kay -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo< at >vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html

Re: udev doesn't detect ddf-format raid array

Peter Leipold — 2008-10-03T07:59:07

0.125-7 But I downloaded 0.129 yesterday and compiled it (I cannot install it system-wide, as I'm not that expert to know what stuff needs to be copied where, and I don't want to break this system). I tried to run udevadm and vol_id from the place of the compilation, the former didn't work, the latter gave the same result as in the 0.125 version: # ./udev-129/udev/udevadm info --query=all --name=sda device node not found # ./udev-129/extras/volume_id/vol_id /dev/sda2 ID_FS_USAGE=filesystem ID_FS_TYPE=ext3 ID_FS_VERSION=1.0 ID_FS_UUID=db631065-7ca8-45fc-bf60-a9259a732dec ID_FS_UUID_ENC=db631065-7ca8-45fc-bf60-a9259a732dec ID_FS_LABEL= ID_FS_LABEL_ENC= ID_FS_LABEL_SAFE= Yes, it's their simplest 2-port card, model name 1220SA I think. It has a Silicon Image chip on it, the kernel use the sata_sil24 driver to reach the drives. Attached. I hope I did the right thing. I got the offset and size with dmraid's help (dumping raid metadata). So these dd commands resulted the same files as dmraid -rD: d

Re: udev doesn't detect ddf-format raid array

Kay Sievers — 2008-10-02T22:47:23

Likely, that volume_id needs a fix. I never really tested with a real ddf volume. What udev version are you running? Is this an Adaptec controller? Can you send me a copy of the sectors containing the ddf header? Along with the exact dd commandline you used to extract it, so can copy it to a drive and test it. Thanks, Kay -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo< at >vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html

udev doesn't detect ddf-format raid array

Peter — 2008-10-02T21:09:49

Hi! I have an adaptec hostraid controller (fakeraid) with 2 drives connected in a RAID1 array (metadata is in DDF format) which is nicely detected and enabled by dmraid, there is no problem with it after it's enabled. Recently debian's dmraid-booter has switched to use udev. But udev does not detect my drives as raid type, ID_FS_USAGE is missing from the drive info (it should be "raid"), so the result is that dmraid does not called at boot. Also vol_id sees the partitions as normal partitions. I've recognised that one of my disks has a HPA, the other hasn't. I'm not sure if this has any effect, but because of this, the DDF1-headers are not at the same position on the 2 drives. Below are some outputs. The last one is a 'dmraid -n' output, and what is strange for me at first sight is that both drives has 2 DDF1-headers. Is this normal? Or do you think I should try to rebuild the raid array with the bios utility (maybe after some bios upgrade)? Or is this rather an udev problem? Thanks, Peter $

Re: udev-129 compile error with --disable-logging

Matthias Schwarzott — 2008-10-02T17:00:26

wrote: Works now, thanks. Matthias -- To unsubscribe from this list: send the line "unsubscribe linux-hotplug" in the body of a message to majordomo< at >vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html