* Re: Minimum kernel version requirement
From: Tom Gundersen @ 2012-01-01 17:28 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <CAG-2HqWXyMz6SSsQSv70DO=AKaPgAhxD+JWD5tDE8ZBKymGM2A@mail.gmail.com>
On Sun, Jan 1, 2012 at 4:47 PM, Kay Sievers <kay.sievers@vrfy.org> wrote:
> The bump has no hard dependency. It's devtmpfs in general which got
> fixes, the 'devname' static module-load stuff, things like the
> in-kernel media-presence polling which udev manages, and some
> architectures which have broken syscall implementations which only got
> fixed later in .34.
>
> Only the broken syscall stuff will prevent udev from brining up these
> old kernels, the rest will only cause some minor details not to work
> as expected, but I guess, it's all that can be worked around.
Thanks for the explanation, now I have a better idea. I was aware of
the devname and media-presence stuff, but didn't think that was
relevant since it arrived post-.34 and can be worked around using
custom rules and static nodes. That custom rules might be needed on
old kernel is already clear from the README so that's fine.
If I understand you correctly, I think the README should be reverted to say
"Requirements:
- Version 2.6.32 of the Linux kernel [...]"
The broken syscalls on pre-.34 kernels on some architectures is
already mentioned explicitly:
"- Some architectures might need a later kernel, that supports accept4(),
or need to backport the accept4() syscall wiring in the kernel."
I agree that it would be best to drop old kernels, but as long as
people would like to try I guess that's up to them.
Cheers,
Tom
^ permalink raw reply
* [PATCH] hwdb: drop useless line freeing
From: Mike Frysinger @ 2012-01-02 21:00 UTC (permalink / raw)
To: linux-hotplug
getline() takes care of extending the buffer as needed, so no point in
constantly freeing it before every call.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
---
udev/udev-builtin-hwdb.c | 5 -----
1 files changed, 0 insertions(+), 5 deletions(-)
diff --git a/udev/udev-builtin-hwdb.c b/udev/udev-builtin-hwdb.c
index e38c6b9..b6af4b6 100644
--- a/udev/udev-builtin-hwdb.c
+++ b/udev/udev-builtin-hwdb.c
@@ -101,11 +101,6 @@ static int lookup_vid_pid(const char *database,
for (;;) {
size_t n;
- if (line) {
- free(line);
- line = NULL;
- }
-
if (getline(&line, &n, f) < 0)
break;
--
1.7.6.1
^ permalink raw reply related
* Re: [PATCH] hwdb: drop useless line freeing
From: Kay Sievers @ 2012-01-02 21:49 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <1325538057-6652-1-git-send-email-vapier@gentoo.org>
On Mon, Jan 2, 2012 at 22:00, Mike Frysinger <vapier@gentoo.org> wrote:
> getline() takes care of extending the buffer as needed, so no point in
> constantly freeing it before every call.
Applied.
Thanks,
Kay
^ permalink raw reply
* [ANNOUNCE] kmod 3
From: Lucas De Marchi @ 2012-01-05 12:27 UTC (permalink / raw)
To: linux-modules, LKML, linux-hotplug
Hi,
I'm glad to announce kmod 3! Signed tarball is available at
http://packages.profusion.mobi/kmod/kmod-3.tar.xz
In this version we added a new implementation of depmod, the only missing tool
to completely replace module-init-tools. We have only a couple of options
missing yet and need more testing, so I'd not say distros could start shipping
kmod instead of module-init-tools yet. This will probably be possible with the
next version.
We removed the several kmod-* tools that mimic the module-init-tools ones. Now
there's only one tool called 'kmod' that does the right thing when symlinks are
used with the previous names.
I'd like to thank again Dave Reisner and Tom Gundersen for the really good work
at testing, reporting bugs and helping on solving them. Right now we received
feedback from them and other people that are testing on the following
architectures: i686, x86_64, sparcv9, sparc64, powerpc64, s390, ARM and sh4.
Any architecture missing here? With this version we also fixed a major bug on
architectures that do not support unaligned accesses.
It's also worth noting that now udev is already using kmod for loading
modules.For more information about new things, please check the NEWS file that
we ship with kmod.
For the next version we plan to review some APIs, extract documentation from
source code and generate the missing man pages. Hopefully we will already be
hosted on kernel.org, too.
Shortlog for kmod 3 is below.
Cristian Rodríguez (2):
Fix leak on error path
index_file_open: fix another fd leak on error path.
Dave Reisner (7):
libkmod-config: remove warning for skipped config files
modprobe: remove --list option
modprobe: use lighter access call instead of stat
kmodprobe: post-remove module deps with 0 refcnt
tools: unify error verbiage and casing
libkmod: Fix casing in error output
libkmod/hash: check for NULL before freeing hash
Gustavo Sverzut Barbieri (33):
Fix unaligned memory access
kmod-modprobe: fix leak in command line option processing.
file: speed up loading non-gzipped modules when zlib is enabled.
elf: minimum size is enough.
elf: add get_symbols()
add test/test-elf
elf: add bind type to kmod_modversion.
elf: implement kmod_module_get_dependency_symbols()
kmod-depmod: initial code (no files generated, untested)
kmod-depmod: dump devname
kmod-depmod: dump modules.symbols
kmod-depmod: dump softdeps.
kmod-depmod: dump aliases
kmod-depmod: dump deps.
kmod-depmod: copy code from module-init-tools/index.c
kmod-depmod: add utility functions to be used by binary dumps.
kmod-depmod: dump modules.builtin.bin
kmod-depmod: dump modules.symbols.bin
kmod-depmod: dump modules.alias.bin
kmod-depmod: dump modules.dep.bin
update TODO
kmod-depmod: implement -F and -E options.
kmod-depmod: implement -A (--quick)
kmod-depmod: refactor extension matching array, support XZ.
utils/read_str_safe(): fix wrong behavior and bugs.
libkmod-file: refactor code to avoid ifdef mess.
elf: zero *array when count is zero.
kmod-depmod: add missing trailing \n to log messages.
kmod-depmod: document --config/-C in help output.
kmod-depmod: fix incorrect math finding out end of dirname.
utils/array: add array_remove_at()
kmod-depmod: fix comparison of module priority.
kmod-depmod: fix replacement of existing modules.
Jan Engelhardt (4):
build: use pkgconfig to detect zlib
build: use AC_ARG_WITH for zlib
Support for loading Xz-compressed modules
libkmod: remove external cflags from .pc file
Kay Sievers (2):
do not (mis-)use the config file generator for non-autobuild stuff
introduce --with-rootlibdir=DIR
Leandro Pereira (3):
kmod tool: don't keep iterating when command has been executed
Do not forget parenthesis around if (streq(A, B)).
libkmod-util: getline_wrapped: return NULL when buffer allocation fails
Lucas De Marchi (63):
gitignore: ignore dist files
Remove useless warning causing trouble in sparc64
Initialize line number with 0
kmod_modprobe: use basename(argv[0]) in help message
build-sys: use MKDIR_P
tools: add skeleton of kmod tool
tools: kmod: Add global options
tools: kmod: check progname for handling compat commands
tools: kmod: Add handling of compat lsmod
tools: kmod: Add handling of compat rmmod
tools: kmod: Add handling of compat insmod
tools: kmod: Add handling of compat modinfo
tools: kmod: Add handling of compat modprobe
tools: kmod: add list command
tools: kmod: print help message when no commands given
tools: kmod: show wrong command in err message
build-sys: use AS_IF
Fix version script with v3 symbols
TODO: add task in modprobe's compatibility
Reverse order of dependency list
Fix kmod_list_remove_n_latest()
Add functions to get ctx blacklists
build-sys: remove unneeded _SOURCES var
Allow to internally get dependencies without copying list
Do not cache softdeps list
Add implementation of modprobe's insertion
Add test for probe insert
kmod-depmod: Fix leak of dependency vector
TODO: format and add task
Log paths if kmod_module_new_from_path() failed
Remove kmod_ prefix from hash implementation
Copy missing hash functions from kmod-depmod to libkmod
build-sys: create libkmod-util.la convenience lib
hash: add iterator
kmod-depmod: use hash implementation from util lib
Move util functions to libkmod-util.c
Fix header guard
tools: kmod: bundle depmod together with the others
Move array implementation from depmode to libkmod-util
Move libkmod-util.c to convenience util lib
Use errno instead of return value of init_module()
Ignore "already loaded" error in module_probe_insert_module()
Move function to the right place
Do not call exported function for mod->name
Use last enum value instead of ARRAY_SIZE
TODO: add list of things that are different on kmod
Fix leak of kmod_module and fix code style
kmod-modprobe: mimic modprobe when removing deps with usecount=0
util: add helper function to compare timestamps
config: save list of config paths with their timestamps
index: save timestamp of each loaded index
Add call to check if resources are valid
Add test to check kmod_validate_resources
TODO: add task for providing man page
Add missing static const
config: take a weakref to ctx
file: take a weakref to ctx
file: use log facilities
build-sys: do not install kmod-* tools
kmod-depmod: demote messages to WRN when non-critical files are missing
Add missing doc for function argument
tools: use basename in help message
kmod 3
--
Lucas De Marchi
^ permalink raw reply
* [ANNOUNCE] udev 176
From: Kay Sievers @ 2012-01-11 21:18 UTC (permalink / raw)
To: linux-hotplug
Here comes a new udev version. Thanks to all who have contributed to
this release.
Until ftp.kernel.org is ready, the signed tarball is here:
http://people.freedesktop.org/~kay/udev/
(This is just a temporary location and the files will be deleted at the moment
ftp.kernel.org is used again)
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 176
====
The 'devtmpfs' filesystem is required now, udev will not create or delete
device nodes anymore, it only adjusts permissions and ownership of device
nodes and maintains additional symlinks.
A writable /run directory (ususally tmpfs) is required now for a fully
functional udev, there is no longer a fallback to /dev/.udev.
The default 'configure' install locations have changed. Packages for systems
with the historic / vs. /usr split need to be adapted, otherwise udev will
be installed in /usr and not work properly. Example configuration options
to install things the traditional way are in INSTALL.
The default install location of the 'udevadm' tool moved from 'sbin'
to /usr/bin. Some tools expect udevadm in 'sbin', a symlink to udevadm
needs to be manually created if needed, or --bindir=/sbin be specified.
The expected value of '--libexecdir=' has changed and must no longer contain
the 'udev' directory.
Kernel modules are now loaded directly by linking udev to 'libkmod'. The
'modprobe' tool is no longer executed by udev.
The 'blkid' tool is no longer executed from udev rules. Udev links
directly to libblkid now.
Firmware is loaded natively by udev now, the external 'firmware' binary
is no longer used.
All built-in tools can be listed and tested with 'udevadm test-builtin'.
The 'udevadm control --reload-rules' option has been renamed to '--reload'.
It now also reloads the kernel module configuration.
The systemd socket files use PassCredentials=yes, which is available in
systemd version 38.
The udev build system only creates a .xz tarball now.
All tabs in the source code used for indentation are replaced by spaces now. :)
^ permalink raw reply
* Re: [ANNOUNCE] udev 176
From: Greg KH @ 2012-01-11 21:25 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <CAPXgP11Pgdmk98VAGkNwQB50t4hkPfZf=_VHojn3zkxFTQBs_A@mail.gmail.com>
On Wed, Jan 11, 2012 at 10:18:33PM +0100, Kay Sievers wrote:
> All tabs in the source code used for indentation are replaced by spaces now. :)
What? Why the change? Are you trying to see if anyone reads these
things? :)
greg k-h
^ permalink raw reply
* Re: [ANNOUNCE] udev 176
From: Kay Sievers @ 2012-01-11 21:47 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <CAPXgP11Pgdmk98VAGkNwQB50t4hkPfZf=_VHojn3zkxFTQBs_A@mail.gmail.com>
On Wed, Jan 11, 2012 at 22:25, Greg KH <greg@kroah.com> wrote:
> On Wed, Jan 11, 2012 at 10:18:33PM +0100, Kay Sievers wrote:
>> All tabs in the source code used for indentation are replaced by spaces now. :)
>
> What? Why the change? Are you trying to see if anyone reads these
> things? :)
Yeah, I know now you read it till the end. :)
Why? Because fixed tabs are so pointless, especially mixing tabs and
trailing spaces like the kernel does all over the place. It just makes
not much sense, we just have been used to it. I personally prefer the
non-tab 8-space model now. It's still kernel style and 8 space indent,
just without the weird tabs and no 80 char limit. :)
And we wanted sync everything up with the systemd tree.
Kay
^ permalink raw reply
* Re: [ANNOUNCE] udev 176
From: Greg KH @ 2012-01-11 22:07 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <CAPXgP11Pgdmk98VAGkNwQB50t4hkPfZf=_VHojn3zkxFTQBs_A@mail.gmail.com>
On Wed, Jan 11, 2012 at 10:47:39PM +0100, Kay Sievers wrote:
> On Wed, Jan 11, 2012 at 22:25, Greg KH <greg@kroah.com> wrote:
> > On Wed, Jan 11, 2012 at 10:18:33PM +0100, Kay Sievers wrote:
> >> All tabs in the source code used for indentation are replaced by spaces now. :)
> >
> > What? Why the change? Are you trying to see if anyone reads these
> > things? :)
>
> Yeah, I know now you read it till the end. :)
>
> Why? Because fixed tabs are so pointless, especially mixing tabs and
> trailing spaces like the kernel does all over the place. It just makes
> not much sense, we just have been used to it. I personally prefer the
> non-tab 8-space model now. It's still kernel style and 8 space indent,
> just without the weird tabs and no 80 char limit. :)
>
> And we wanted sync everything up with the systemd tree.
Ok, that's a good enough reason alone, at least I know that 'git blame'
will never show any lines of code I ever wrote :)
thanks,
greg k-h
^ permalink raw reply
* Re: [ANNOUNCE] udev 176
From: Kay Sievers @ 2012-01-11 22:15 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <CAPXgP11Pgdmk98VAGkNwQB50t4hkPfZf=_VHojn3zkxFTQBs_A@mail.gmail.com>
On Wed, Jan 11, 2012 at 23:07, Greg KH <greg@kroah.com> wrote:
> On Wed, Jan 11, 2012 at 10:47:39PM +0100, Kay Sievers wrote:
>> On Wed, Jan 11, 2012 at 22:25, Greg KH <greg@kroah.com> wrote:
>> > On Wed, Jan 11, 2012 at 10:18:33PM +0100, Kay Sievers wrote:
>> >> All tabs in the source code used for indentation are replaced by spaces now. :)
>> >
>> > What? Why the change? Are you trying to see if anyone reads these
>> > things? :)
>>
>> Yeah, I know now you read it till the end. :)
>>
>> Why? Because fixed tabs are so pointless, especially mixing tabs and
>> trailing spaces like the kernel does all over the place. It just makes
>> not much sense, we just have been used to it. I personally prefer the
>> non-tab 8-space model now. It's still kernel style and 8 space indent,
>> just without the weird tabs and no 80 char limit. :)
>>
>> And we wanted sync everything up with the systemd tree.
>
> Ok, that's a good enough reason alone, at least I know that 'git blame'
> will never show any lines of code I ever wrote :)
Oh, I checked, because it was wondering too, git blame nicely ignores
whitespace by default. It all looks the same as before.
You still some original lines in the .c files from the very first
bitkeeper commit. But yeah, we shuffled a lot around over the years.
Kay
^ permalink raw reply
* [ANNOUNCE] udev 177
From: Kay Sievers @ 2012-01-12 4:45 UTC (permalink / raw)
To: linux-hotplug
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 177
====
Bugfix for rule_generator instalation.
^ permalink raw reply
* udev 177 and kmod3
From: Bruce Dubbs @ 2012-01-12 6:36 UTC (permalink / raw)
To: linux-hotplug
udev-177 appears to require kmod3, yet the kmod3 announcement says:
"I'd not say distros could start shipping
kmod instead of module-init-tools yet."
There appears to be a miscommunication here.
-- Bruce Dubbs
LFS
^ permalink raw reply
* [PATCH] gudev: several minor introspection fixes
From: Evan Nemerson @ 2012-01-12 7:42 UTC (permalink / raw)
To: linux-hotplug
- Include exported package information
- Include C include information
- g_udev_device_get_parent & g_udev_device_get_parent_with_subsystem
transfer ownership of their return values
---
Makefile.am | 2 ++
| 4 ++--
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/Makefile.am b/Makefile.am
index 06500b0..a9dfe3e 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -476,6 +476,8 @@ src/extras/gudev/GUdev-1.0.gir: src/extras/gudev/libgudev-1.0.la $(G_IR_SCANNER)
--output $@ \
--pkg=glib-2.0 \
--pkg=gobject-2.0 \
+ --pkg-export=gudev-1.0 \
+ --c-include=gudev/gudev.h \
-I$(top_srcdir)/src/extras \
-I$(top_builddir)/src/extras \
-D_GUDEV_COMPILATION \
--git a/src/extras/gudev/gudevdevice.c b/src/extras/gudev/gudevdevice.c
index 0c3340f..62a26f9 100644
--- a/src/extras/gudev/gudevdevice.c
+++ b/src/extras/gudev/gudevdevice.c
@@ -373,7 +373,7 @@ g_udev_device_get_device_file_symlinks (GUdevDevice *device)
*
* Gets the immediate parent of @device, if any.
*
- * Returns: A #GUdevDevice or %NULL if @device has no parent. Free with g_object_unref().
+ * Returns: (transfer full): A #GUdevDevice or %NULL if @device has no parent. Free with g_object_unref().
*/
GUdevDevice *
g_udev_device_get_parent (GUdevDevice *device)
@@ -404,7 +404,7 @@ g_udev_device_get_parent (GUdevDevice *device)
* Walks up the chain of parents of @device and returns the first
* device encountered where @subsystem and @devtype matches, if any.
*
- * Returns: A #GUdevDevice or %NULL if @device has no parent with @subsystem and @devtype. Free with g_object_unref().
+ * Returns: (transfer full): A #GUdevDevice or %NULL if @device has no parent with @subsystem and @devtype. Free with g_object_unref().
*/
GUdevDevice *
g_udev_device_get_parent_with_subsystem (GUdevDevice *device,
--
1.7.7.5
^ permalink raw reply related
* Re: [PATCH] gudev: several minor introspection fixes
From: Martin Pitt @ 2012-01-12 8:08 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <1326354142.21249.6.camel@hoplite>
[-- Attachment #1: Type: text/plain, Size: 427 bytes --]
Hello Evan,
Evan Nemerson [2012-01-11 23:42 -0800]:
> - Include exported package information
> - Include C include information
> - g_udev_device_get_parent & g_udev_device_get_parent_with_subsystem
> transfer ownership of their return values
Thanks! Applied.
Martin
--
Martin Pitt | http://www.piware.de
Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org)
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]
^ permalink raw reply
* Re: udev 177 and kmod3
From: Lucas De Marchi @ 2012-01-12 12:56 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <4F0E7F79.9000503@gmail.com>
Hi Bruce,
On Thu, Jan 12, 2012 at 4:36 AM, Bruce Dubbs <bruce.dubbs@gmail.com> wrote:
> udev-177 appears to require kmod3, yet the kmod3 announcement says:
>
> "I'd not say distros could start shipping
> kmod instead of module-init-tools yet."
>
> There appears to be a miscommunication here.
There isn't.
If udev doesn't care about kmod's man pages, then it's good to start
using libkmod. For the udev's use cases, libkmod is pretty complete.
It can be installed in parallel to module-init-tools too if distros
don't seem confident to make a step further and replace it.
So far Archlinux seems to be the first one. kmod is in their testing
repo and they helped us a lot to fix some bugs. Then I'd say kmod 4
will be ready enough for distros.
regards,
Lucas De Marchi
^ permalink raw reply
* Re: udev 177 and kmod3
From: Robby Workman @ 2012-01-12 14:27 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <4F0E7F79.9000503@gmail.com>
On Thu, 12 Jan 2012 10:56:31 -0200
Lucas De Marchi <lucas.demarchi@profusion.mobi> wrote:
> Hi Bruce,
>
> On Thu, Jan 12, 2012 at 4:36 AM, Bruce Dubbs <bruce.dubbs@gmail.com>
> wrote:
> > udev-177 appears to require kmod3, yet the kmod3 announcement says:
> >
> > "I'd not say distros could start shipping
> > kmod instead of module-init-tools yet."
> >
> > There appears to be a miscommunication here.
>
>
> There isn't.
>
> If udev doesn't care about kmod's man pages, then it's good to start
> using libkmod. For the udev's use cases, libkmod is pretty complete.
> It can be installed in parallel to module-init-tools too if distros
> don't seem confident to make a step further and replace it.
>
> So far Archlinux seems to be the first one. kmod is in their testing
> repo and they helped us a lot to fix some bugs. Then I'd say kmod 4
> will be ready enough for distros.
Agreed on that. We don't have it in Slackware yet, but I've been
running it locally since v2 (and fully replaced m-i-t since v3),
and aside from a few minor issues (that have mostly been fixed),
it's fine.
-RW
^ permalink raw reply
* systemd-38 and acls
From: Allin Cottrell @ 2012-01-14 13:56 UTC (permalink / raw)
To: linux-hotplug
I notice that (unlike version 37) systemd-38 has an
unconditional dependence on acls. Even if you pass the
configure flag --disable-acl the new journald drags them in.
Would it be possible to make the acl-dependent component of
journald optional?
--
Allin Cottrell
Department of Economics
Wake Forest University, NC
^ permalink raw reply
* Re: systemd-38 and acls
From: Kay Sievers @ 2012-01-14 14:00 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <alpine.LNX.2.00.1201140841420.10777@waverley.dhcp.wfu.edu>
On Sat, Jan 14, 2012 at 14:56, Allin Cottrell <cottrell@wfu.edu> wrote:
> I notice that (unlike version 37) systemd-38 has an unconditional dependence
> on acls. Even if you pass the configure flag --disable-acl the new journald
> drags them in.
>
> Would it be possible to make the acl-dependent component of journald
> optional?
http://cgit.freedesktop.org/systemd/systemd/commit/?idæ520a0fcebb4d5e07f4c81a51382501b0beb22f
Kay
^ permalink raw reply
* HEADS-UP: udev 176+ exposes some broken netdev driver's firmware loading
From: Kay Sievers @ 2012-01-14 15:33 UTC (permalink / raw)
To: linux-hotplug
Recent udev versions will need fixes for some network drivers blocking
in modprobe and trying to load firmware.
It's caused by:
http://git.kernel.org/?p=linux/hotplug/udev.git;a=commitdiff;hCd5c5f03645c4b842659f9b5bd0ae465e885e92
Depending on the response on netdev, packages or upstream will need to
revert that until the drivers are fixed. In the long run, the 'block
in module_init() until the firmware is loaded' model will no longer
work.
The netdev mail is here:
http://thread.gmane.org/gmane.linux.network/217729
Thanks,
Kay
^ permalink raw reply
* Re: HEADS-UP: udev 176+ exposes some broken netdev driver's firmware loading
From: Allin Cottrell @ 2012-01-15 17:02 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <CAPXgP12YUX2Oxa4VXg4N5THHEwLnybymsSxZm7Yx7tDW5zvAjA@mail.gmail.com>
On Sat, 14 Jan 2012, Kay Sievers wrote:
> Recent udev versions will need fixes for some network drivers blocking
> in modprobe and trying to load firmware.
>
> It's caused by:
> http://git.kernel.org/?p=linux/hotplug/udev.git;a=commitdiff;hCd5c5f03645c4b842659f9b5bd0ae465e885e92
Do you happen to know if this is an issue with the iwlwifi
firmware for Intel (e.g. Centrino Advanced-N 6205 AGN)?
Thanks.
Allin Cottrell
^ permalink raw reply
* Re: HEADS-UP: udev 176+ exposes some broken netdev driver's firmware loading
From: Tom Gundersen @ 2012-01-15 17:06 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <CAPXgP12YUX2Oxa4VXg4N5THHEwLnybymsSxZm7Yx7tDW5zvAjA@mail.gmail.com>
On Sun, Jan 15, 2012 at 6:02 PM, Allin Cottrell <cottrell@wfu.edu> wrote:
> Do you happen to know if this is an issue with the iwlwifi firmware
I'm not able to reproduce this problem with iwlwifi (iwlwifi-3945-2.ucode).
-t
^ permalink raw reply
* Re: HEADS-UP: udev 176+ exposes some broken netdev driver's firmware loading
From: Kay Sievers @ 2012-01-15 18:22 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <CAPXgP12YUX2Oxa4VXg4N5THHEwLnybymsSxZm7Yx7tDW5zvAjA@mail.gmail.com>
On Sun, Jan 15, 2012 at 18:02, Allin Cottrell <cottrell@wfu.edu> wrote:
> On Sat, 14 Jan 2012, Kay Sievers wrote:
>
>> Recent udev versions will need fixes for some network drivers blocking
>> in modprobe and trying to load firmware.
>>
>> It's caused by:
>>
>> Â http://git.kernel.org/?p=linux/hotplug/udev.git;a=commitdiff;hCd5c5f03645c4b842659f9b5bd0ae465e885e92
>
>
> Do you happen to know if this is an issue with the iwlwifi firmware for
> Intel (e.g. Centrino Advanced-N 6205 AGN)? Thanks.
No, the Intel wireless driver is not broken.
Kay
^ permalink raw reply
* update path for undev + systemd
From: Allin Cottrell @ 2012-01-15 20:52 UTC (permalink / raw)
To: linux-hotplug
I wonder if there's a recommended path/method for updating
from udev 175 + systemd 37 to the current versions? Given the
reorganization of files in both packages it's quite easy to
end up with a fscked system (as I did today -- system not
bootable; had to boot from CD, remove the new versions and
reinstall the old).
With udev one can force the use of the old /lib/udev location
by giving --with-rootlibdir=/lib --libexecdir=/lib, but I
didn't find a way of making current git systemd install into
/lib/systemd; even configuring with --with-rootlibdir=/lib,
all files that used to go into /lib/systemd go into
/usr/lib/systemd.
In principle I have no problem with everything going under
/usr/lib, it's just that it seems the chances of a successful
update would be better if the old files were overwritten
rather than the new ones being installed in parallel.
Thanks for any guidance.
--
Allin Cottrell
Department of Economics
Wake Forest University, NC
^ permalink raw reply
* Re: update path for undev + systemd
From: Kay Sievers @ 2012-01-15 21:30 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <alpine.LNX.2.01.1201151539210.14830@waverley.dhcp.wfu.edu>
On Sun, Jan 15, 2012 at 21:52, Allin Cottrell <cottrell@wfu.edu> wrote:
> I wonder if there's a recommended path/method for updating from udev 175 +
> systemd 37 to the current versions? Given the reorganization of files in
> both packages it's quite easy to end up with a fscked system (as I did today
> -- system not bootable; had to boot from CD, remove the new versions and
> reinstall the old).
>
> With udev one can force the use of the old /lib/udev location by giving
> --with-rootlibdir=/lib --libexecdir=/lib, but I didn't find a way of making
> current git systemd install into /lib/systemd; even configuring with
> --with-rootlibdir=/lib, all files that used to go into /lib/systemd go into
> /usr/lib/systemd.
>
> In principle I have no problem with everything going under /usr/lib, it's
> just that it seems the chances of a successful update would be better if the
> old files were overwritten rather than the new ones being installed in
> parallel.
Udev and systemd are in full sync with the configure options now.
You better read NEWS in the udev tree, and not just install new things
in a running system and hope all works. :)
Check out the examples in the INSTALL file in the udev tree, it's
--with-rootprefix= which specifies if stuff installs in /usr or /.
Kay
^ permalink raw reply
* Re: update path for undev + systemd
From: Allin Cottrell @ 2012-01-15 22:45 UTC (permalink / raw)
To: linux-hotplug
In-Reply-To: <alpine.LNX.2.01.1201151539210.14830@waverley.dhcp.wfu.edu>
On Sun, 15 Jan 2012, Kay Sievers wrote:
> On Sun, Jan 15, 2012 at 21:52, Allin Cottrell <cottrell@wfu.edu> wrote:
>> I wonder if there's a recommended path/method for updating from udev 175 +
>> systemd 37 to the current versions? Given the reorganization of files in
>> both packages it's quite easy to end up with a fscked system (as I did today
>> -- system not bootable; had to boot from CD, remove the new versions and
>> reinstall the old). [...]
>
> Udev and systemd are in full sync with the configure options now.
>
> You better read NEWS in the udev tree, and not just install new things
> in a running system and hope all works. :)
I had in fact read udev's NEWS and acted accordingly, i.e. changed
the --libexecdir= configure line, installed libkmod, etc. (I was
already up to date in respect of devtmpfs and /run). I don't think
my booting problem lay with udev, but rather with systemd, whose
package contains no docs pertaining to configuring/installing or
changes from the previous version.
Allin Cottrell
^ permalink raw reply
* udev 177 and echo gina3g sound card.
From: F. Heitkamp @ 2012-01-15 22:56 UTC (permalink / raw)
To: linux-hotplug
I have been trying udev 177.
My system is basically a linux from scratch like system.
Any thoughts?
Fred
bash-4.2$ uname -a
Linux pc1lin 3.2.0-rt10 #2 SMP Fri Jan 13 06:16:51 EST 2012 x86_64 GNU/Linux
Udev seems to mostly work properly however the firmware for the Echo
Gina3g PCI card will not load.
The card firmware loaded properly with older udev versions and also
works with Debian unstable on another drive.
The "timeout" is not a directory. It is a textfile the contents of which
is "60"
I have tried several recent kernels and the results are the same each time.
lspci:
0d:00.0 "Multimedia controller" "Motorola" "DSP56361 Digital Signal
Processor" -r01 "Echo Digital Audio Corporation" "3G"
udev debug output:
path link '/sys/bus/pci/devices/0000:0d:00.0' points to
'../../../devices/pci0000:00/0000:00:1e.0/0000:0d:00.0'
base '/sys/bus/pci/devices/0000:0d:00.0', tail
'devices/pci0000:00/0000:00:1e.0/0000:0d:00.0', back 3
after moving back '/sys'
'UDEV_LOG=7' added
udev_device: 0x8073450 created
'DEVPATH=/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0' added
device 0x8073450 has devpath '/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0'
udev_device: 0x8073450 released
directory not found: /sys/class/firmware/timeout
bash-4.2$ find /sys/class/firmware/ -ls
3874 0 drwxr-xr-x 2 root root 0 Jan 15 16:54
/sys/class/firmware/
3875 0 -rw-r--r-- 1 root root 4096 Jan 15 16:54
/sys/class/firmware/timeout
bash-4.2$
/var/log/messages:
2012-01-15T15:58:23-05:00 pc1lin udevd[819]: error: can not open
'/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/firmware/0000:0d:00.0/loading'
2012-01-15T16:02:44-05:00 pc1lin udevd[804]: error: can not open
'/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/firmware/0000:0d:00.0/loading'
sys file system:
bash-4.2$ find /sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/uevent
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/resource
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/vendor
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/device
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/subsystem_vendor
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/subsystem_device
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/class
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/irq
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/local_cpus
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/local_cpulist
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/modalias
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/dma_mask_bits
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/consistent_dma_mask_bits
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/enable
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/broken_parity_status
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/msi_bus
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/remove
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/rescan
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/subsystem
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/power
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/config
/sys/devices/pci0000:00/0000:00:1e.0/0000:0d:00.0/resource0
bash-4.2$ find /lib/firmware/ea/
/lib/firmware/ea/
/lib/firmware/ea/echo3g_dsp.fw
689 configure:15960: result:
690 udev 177
691 ====
692
693 prefix: /usr
694 rootprefix: /usr
695 sysconfdir: /etc
696 bindir: ${exec_prefix}/bin
697 libdir: ${exec_prefix}/lib
698 rootlibdir: ${exec_prefix}/lib
699 libexecdir: /lib
700 datarootdir: ${prefix}/share
701 mandir: ${datarootdir}/man
702 includedir: ${prefix}/include
703 include_prefix: /usr/include
704 systemdsystemunitdir: /lib/systemd/system
705 firmware path: \"/lib/firmware/\"
706 usb.ids: /usr/share/usb.ids
707 pci.ids: /usr/share/hwdata/pci.ids
708
709 compiler: gcc
710 cflags: -m32 -O2
711 ldflags:
712 xsltproc: /usr//bin/xsltproc
713 gperf: /usr/local//bin/gperf
714
715 logging: yes
716 debug: yes
717 selinux: yes
718
719 gudev: yes
720 gintrospection: yes
721 keymap: yes
722 mtd_probe: yes
723 rule_generator: yes
724 udev_acl: yes
725 floppy: no
726 edd: no
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox