Mailing List Archive

On 10/01/2004 12:49 PM, Adrian CAPDEFIER wrote:
> ... or it's discontinued and there's a user-space thingie
> to replace it?
>

There are some alternatives for supermount:

- submount, which is a kernel module. Unlike supermount you don't need
to patch your kernel to make it work. Submount is part of the portage tree.

- dbus+hal+ivman, a userspace solution. Unfortunately the necessary
programs/components are not in the (official) portage tree or are hard
masked.

- dbus+hal+gnome-volume-manager, also a userspace solution, like the one
above, but more gnome-centric.


I am happy with submount, but as always: your mileage may vary...

Jochen

--
gentoo-user@gentoo.org mailing list

Jochen Schalanda wrote:
> On 10/01/2004 12:49 PM, Adrian CAPDEFIER wrote:
>
>> ... or it's discontinued and there's a user-space thingie
>> to replace it?
>>
>
> There are some alternatives for supermount:
>
> - submount, which is a kernel module. Unlike supermount you don't need
> to patch your kernel to make it work. Submount is part of the portage tree.
>
> - dbus+hal+ivman, a userspace solution. Unfortunately the necessary
> programs/components are not in the (official) portage tree or are hard
> masked.
>
> - dbus+hal+gnome-volume-manager, also a userspace solution, like the one
> above, but more gnome-centric.
>
>
> I am happy with submount, but as always: your mileage may vary...
>
> Jochen

I didn't so much like submount, so I re-patched the kernel with the
supermount patch. Then I switched to ck-sources, which still used it up
to the last revision, when I had to manually patch that kernel myself
as well (again).

In any case, it is possible to continue using supermount, if you-- like
me-- want to wait until you feel/hear that dbus+hal+whatever is a bit
more stable than it seems to be at the moment before switching.

Holly


--
gentoo-user@gentoo.org mailing list

Adrian writes:

> Does any one know if the supermount patch will ever be applied again to
> gentoo-dev-sources or it's discontinued and there's a user-space thingie
> to replace it?

The new thing is submount. Supposed to be better than supermount
because it's in user space, but I had the problem that the eject
command does not work any more, it tries to unmount the device. I am
using a version of eject patched by SuSE, downloaded here:
ftp://ftp.suse.com/pub/suse/i386/9.1/suse/

Alex
--
Alex Schuster Wonko@wonkology.org PGP Key available
Wonko@netcologne.de

--
gentoo-user@gentoo.org mailing list

On Fri, 01 Oct 2004 13:49:28 +0300, Adrian CAPDEFIER wrote:

> Does any one know if the supermount patch will ever be applied again
> to gentoo-dev-sources or it's discontinued and there's a user-space
> thingie to replace it?

You can apply the supermount patch to the 2.6.8 gentoo-dev-sources
yourself. I've been doing it for a few weeks now and it works exactly as
before.


--
Neil Bothwick

WORM: (n.) acronym for Write Once, Read Mangled. Used to describe a
normally-functioning computer disk of the very latest design.

Alex Schuster wrote:
> Adrian writes:
>
>
>>Does any one know if the supermount patch will ever be applied again to
>>gentoo-dev-sources or it's discontinued and there's a user-space thingie
>>to replace it?
>
>
> The new thing is submount. Supposed to be better than supermount
> because it's in user space, but I had the problem that the eject
> command does not work any more, it tries to unmount the device.

My issue with supermount was similar, ejecting the media by physically
pushing the button worked at first, then stopped at some point during
any given session (so the drive was locked, or alternatively, the
command buffer would wig out and the drive would open and
close--possibly more than once-- before I could grab the first disk and
insert the second), and the whole setup didn't seem to play nicely with
FAM insofar as getting a correct listing of files on the media was
spotty at best (especially once eject had wigged out), depending on what
state submount had decided the device was in (mounted, unmounted,
ejected, reinserted-- who could tell?)

It really wasn't worth the trouble just to be a "good girl" and work in
userspace, which is why I went back to supermount until these things
work a bit better, or I learn how to make them work better, whichever
comes first.

I am
> using a version of eject patched by SuSE, downloaded here:
> ftp://ftp.suse.com/pub/suse/i386/9.1/suse/
>

This is good to know about if I want to give it another try; thanks for
the tip.

Holly


--
gentoo-user@gentoo.org mailing list

Hi,

Adrian CAPDEFIER wrote:
> Does any one know if the supermount patch will ever be applied again to
> gentoo-dev-sources or it's discontinued and there's a user-space thingie
> to replace it?

It won't be readded to gentoo-dev-sources unless is is majorly revamped and
submitted + accepted into the mainstream kernel.

Daniel

--
gentoo-user@gentoo.org mailing list

Daniel Drake wrote:
> It won't be readded to gentoo-dev-sources unless is is majorly revamped
> and submitted + accepted into the mainstream kernel.

Thanks for the info Daniel. I'll go look into the submount thingie then.


--

Adi

--
gentoo-user@gentoo.org mailing list

Adrian CAPDEFIER wrote:
> Daniel Drake wrote:
>
>> It won't be readded to gentoo-dev-sources unless is is majorly
>> revamped and submitted + accepted into the mainstream kernel.
>
>
> Thanks for the info Daniel. I'll go look into the submount thingie then.
>
>


it fails at emerging.


>>> Unpacking source...
* /usr/src/linux is a symbolic link
* Determining the real directory of the Linux kernel source code
* Building for Linux 2.6.8-gentoo-r3 found in /usr/src/linux
* which outputs to /var/tmp/kernel-output/2.6.8-gentoo-r3
>>> Unpacking submount-0.9.tar.gz to /var/tmp/portage/submount-0.9/work
>>> Source unpacked.
make -C /lib/modules/2.6.8-gentoo-r3/build
SUBDIRS=/var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9 modules
make[1]: Entering directory `/usr/src/linux-2.6.8-gentoo-r3'
CC [M] /var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.o
In file included from /usr/src/linux-2.6.8-gentoo-r3/include/linux/init.h:4,
from
/var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.c:10:
/usr/src/linux-2.6.8-gentoo-r3/include/linux/config.h:4:28:
linux/autoconf.h: No such file or directory


<262K of SNIP>


_vfs_rename_sem' isn't known
/usr/src/linux-2.6.8-gentoo-r3/include/linux/namespace.h:12: error:
storage size of `sem' isn't known
/var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.h:67:
error: storage size of `subfs_dir_inode_operations' isn't known
/var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.h:72:
error: storage size of `subfs_file_ops' isn't known
make[3]: ***
[/var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.o] Error 1
make[2]: ***
[_module_/var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9] Error 2
make[1]: *** [modules] Error 2
make[1]: Leaving directory `/usr/src/linux-2.6.8-gentoo-r3'
make: *** [default] Error 2

!!! ERROR: sys-fs/submount-0.9 failed.
!!! Function kmod_src_compile, Line 278, Exitcode 2
!!! (no error message)


--

Adi

--
gentoo-user@gentoo.org mailing list

Hi,
On пт, 2004-10-01 at 20:30, Adrian CAPDEFIER wrote:
> Adrian CAPDEFIER wrote:
> > Daniel Drake wrote:
> >
> >> It won't be readded to gentoo-dev-sources unless is is majorly
> >> revamped and submitted + accepted into the mainstream kernel.
> >
> >
> > Thanks for the info Daniel. I'll go look into the submount thingie then.
> >
> >
>
>
> it fails at emerging.
>
>
> >>> Unpacking source...
> * /usr/src/linux is a symbolic link
> * Determining the real directory of the Linux kernel source code
> * Building for Linux 2.6.8-gentoo-r3 found in /usr/src/linux
> * which outputs to /var/tmp/kernel-output/2.6.8-gentoo-r3
> >>> Unpacking submount-0.9.tar.gz to /var/tmp/portage/submount-0.9/work
> >>> Source unpacked.
> make -C /lib/modules/2.6.8-gentoo-r3/build
> SUBDIRS=/var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9 modules
> make[1]: Entering directory `/usr/src/linux-2.6.8-gentoo-r3'
> CC [M] /var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.o
> In file included from /usr/src/linux-2.6.8-gentoo-r3/include/linux/init.h:4,
> from
> /var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.c:10:
> /usr/src/linux-2.6.8-gentoo-r3/include/linux/config.h:4:28:
> linux/autoconf.h: No such file or directory
>
>
> <262K of SNIP>
>
>
> _vfs_rename_sem' isn't known
> /usr/src/linux-2.6.8-gentoo-r3/include/linux/namespace.h:12: error:
> storage size of `sem' isn't known
> /var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.h:67:
> error: storage size of `subfs_dir_inode_operations' isn't known
> /var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.h:72:
> error: storage size of `subfs_file_ops' isn't known
> make[3]: ***
> [/var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.o] Error 1
> make[2]: ***
> [_module_/var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9] Error 2
> make[1]: *** [modules] Error 2
> make[1]: Leaving directory `/usr/src/linux-2.6.8-gentoo-r3'
> make: *** [default] Error 2
>
> !!! ERROR: sys-fs/submount-0.9 failed.
> !!! Function kmod_src_compile, Line 278, Exitcode 2
> !!! (no error message)
>
Some days ago tied the same and it too gave me an error.
Don't have the info now but remember that there were messages suggesting
(requiring) you (me) to use koutput to compile kernel modules, as in the
case with submount. Read the docs somewhere on w.g.o - there is one for
koutput.
There were also reports about broken (not always working) koutput. Could
check the ML a day or two ago.
HTH
Rumen

Adrian CAPDEFIER wrote:
> Adrian CAPDEFIER wrote:
>
>> Thanks for the info Daniel. I'll go look into the submount thingie then.
>>
>>
>
>
> it fails at emerging.
>
>
> >>> Unpacking source...
> * /usr/src/linux is a symbolic link
> * Determining the real directory of the Linux kernel source code
> * Building for Linux 2.6.8-gentoo-r3 found in /usr/src/linux
> * which outputs to /var/tmp/kernel-output/2.6.8-gentoo-r3
> >>> Unpacking submount-0.9.tar.gz to /var/tmp/portage/submount-0.9/work
> >>> Source unpacked.
> make -C /lib/modules/2.6.8-gentoo-r3/build
> SUBDIRS=/var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9 modules
> make[1]: Entering directory `/usr/src/linux-2.6.8-gentoo-r3'
> CC [M] /var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.o
> In file included from
> /usr/src/linux-2.6.8-gentoo-r3/include/linux/init.h:4,
> from
> /var/tmp/portage/submount-0.9/work/submount-0.9/subfs-0.9/subfs.c:10:
> /usr/src/linux-2.6.8-gentoo-r3/include/linux/config.h:4:28:
> linux/autoconf.h: No such file or directory
>

<snip>

IIrc, submount specifically was the reason that I had to set up Koutput,
although I suppose you could also use config-kernel to make
/usr/src/linux portage writeable:

See http://www.gentoo.org/doc/en/2.6-koutput-user.xml for more
information on this feature for 2.6 kernels.

Holly

--
gentoo-user@gentoo.org mailing list

Rumen Yotov wrote:
> Hi,
> On пт, 2004-10-01 at 20:30, Adrian CAPDEFIER wrote:
>
>>Adrian CAPDEFIER wrote:
>>
>>>Daniel Drake wrote:
>>>
>>>
>>>>It won't be readded to gentoo-dev-sources unless is is majorly
>>>>revamped and submitted + accepted into the mainstream kernel.
>>>
>>>
>>>Thanks for the info Daniel. I'll go look into the submount thingie then.
>>>
>>>
>>
>>
>>it fails at emerging.
>>
>>
>> >>> Unpacking source...
>> * /usr/src/linux is a symbolic link
>> * Determining the real directory of the Linux kernel source code
>> * Building for Linux 2.6.8-gentoo-r3 found in /usr/src/linux
>> * which outputs to /var/tmp/kernel-output/2.6.8-gentoo-r3
>> >>> Unpacking submount-0.9.tar.gz to /var/tmp/portage/submount-0.9/work
>> >>> Source unpacked.
>><SNIP>
>>!!! ERROR: sys-fs/submount-0.9 failed.
>>!!! Function kmod_src_compile, Line 278, Exitcode 2
>>!!! (no error message)
>>
>
> Some days ago tied the same and it too gave me an error.
> Don't have the info now but remember that there were messages suggesting
> (requiring) you (me) to use koutput to compile kernel modules, as in the
> case with submount. Read the docs somewhere on w.g.o - there is one for
> koutput.
> There were also reports about broken (not always working) koutput. Could
> check the ML a day or two ago.
> HTH
> Rumen
>

Yes, so I remembered correctly, then. I have to confirm that koutput was
definitely a bit problematic for me-- not least because it didn't much
want me to get rid of it once I went back to supermount and didn't need
it to emerge submount anymore (submount being the only outside module
that I emerge that would not compile unless koutput was used; all the
others I had been using fine before the supermount patch was removed and
I switched to submount).

Getting free of koutput was one of the reasons that I switched back to
the supermount kernel patch (as well as the fact that submount didn't
work so fabulously that it was worth the troublesome addon).

I hope that your mileage varies, actually, because for me the whole
endeavor was just a pain in the neck.

Holly


--
gentoo-user@gentoo.org mailing list

Holly Bostick wrote:
> Rumen Yotov wrote:
>
>> Hi,
>> On пт, 2004-10-01 at 20:30, Adrian CAPDEFIER wrote:
>>
>>> Adrian CAPDEFIER wrote:
>>>
>>>> Daniel Drake wrote:
>>>>
>>>>
>>>>> It won't be readded to gentoo-dev-sources unless is is majorly
>>>>> revamped and submitted + accepted into the mainstream kernel.
>>>>
>>>>
>>>>
>>>> Thanks for the info Daniel. I'll go look into the submount thingie
>>>> then.
>>>>
>>>>
>>>
>>>
>>> it fails at emerging.
>>>
>>>
>>> >>> Unpacking source...
>>> * /usr/src/linux is a symbolic link
>>> * Determining the real directory of the Linux kernel source code
>>> * Building for Linux 2.6.8-gentoo-r3 found in /usr/src/linux
>>> * which outputs to /var/tmp/kernel-output/2.6.8-gentoo-r3
>>> >>> Unpacking submount-0.9.tar.gz to /var/tmp/portage/submount-0.9/work
>>> >>> Source unpacked.
>>> <SNIP>
>>> !!! ERROR: sys-fs/submount-0.9 failed.
>>> !!! Function kmod_src_compile, Line 278, Exitcode 2
>>> !!! (no error message)
>>>
>>
>> Some days ago tied the same and it too gave me an error.
>> Don't have the info now but remember that there were messages suggesting
>> (requiring) you (me) to use koutput to compile kernel modules, as in the
>> case with submount. Read the docs somewhere on w.g.o - there is one for
>> koutput.
>> There were also reports about broken (not always working) koutput. Could
>> check the ML a day or two ago.
>> HTH
>> Rumen
>>
>
> Yes, so I remembered correctly, then. I have to confirm that koutput was
> definitely a bit problematic for me-- not least because it didn't much
> want me to get rid of it once I went back to supermount and didn't need
> it to emerge submount anymore (submount being the only outside module
> that I emerge that would not compile unless koutput was used; all the
> others I had been using fine before the supermount patch was removed and
> I switched to submount).
>
> Getting free of koutput was one of the reasons that I switched back to
> the supermount kernel patch (as well as the fact that submount didn't
> work so fabulously that it was worth the troublesome addon).
>
> I hope that your mileage varies, actually, because for me the whole
> endeavor was just a pain in the neck.
>
> Holly
>

I did the configure-kernel thingie (if not submount does not emerge).
I'll just user supermount. Where is the patch from Holly?
supermount-ng.sf.net does not seem to have a 2.6.8 patch.

--

Adi

--
gentoo-user@gentoo.org mailing list

Hi,

Holly Bostick wrote:
> Yes, so I remembered correctly, then. I have to confirm that koutput was
> definitely a bit problematic for me--

The issues are known, it's not just you. koutput/kmod usage is currently being
removed from ebuilds that use it, until its completed.

Daniel

--
gentoo-user@gentoo.org mailing list

Daniel Drake wrote:
> Hi,
>
> Holly Bostick wrote:
>
>> Yes, so I remembered correctly, then. I have to confirm that koutput
>> was definitely a bit problematic for me--
>
>
> The issues are known, it's not just you. koutput/kmod usage is currently
> being removed from ebuilds that use it, until its completed.


so I should use FEATURES="-sandbox" ?
--
Adi

--
gentoo-user@gentoo.org mailing list

Adrian CAPDEFIER wrote:
> Holly Bostick wrote:
>>
>> Getting free of koutput was one of the reasons that I switched back to
>> the supermount kernel patch (as well as the fact that submount didn't
>> work so fabulously that it was worth the troublesome addon).
>>
>> I hope that your mileage varies, actually, because for me the whole
>> endeavor was just a pain in the neck.
>>
>> Holly
>>
>
> I did the configure-kernel thingie (if not submount does not emerge).
> I'll just user supermount. Where is the patch from Holly?
> supermount-ng.sf.net does not seem to have a 2.6.8 patch.

I nicked it from the ck-series of kernel patches:

http://ck.kolivas.org/patches/2.6/2.6.8.1/2.6.8.1-ck3/split-out/supermount-ng204.diff.bz2

Don't let the fact that it's a diff alarm you (it alarmed me, which is
why I mention it); it worked fine without there being an actual orignal
to be diffed. In fact, it went wonky when I tried to first install
supermount and then diff it with the patch (I got 3 identical options in
the kernel, but it worked fine when I just patched a clean copy of the
kernel with the diff alone).

Holly


--
gentoo-user@gentoo.org mailing list

On Fri, 01 Oct 2004 23:00:16 +0200, Holly Bostick wrote:

> I nicked it from the ck-series of kernel patches:
>
> http://ck.kolivas.org/patches/2.6/2.6.8.1/2.6.8.1-ck3/split-out/supermount-ng204.diff.bz2

That's the one I've been using ever since 2.6.8 went ~amd64, it works
fine.


--
Neil Bothwick

Do not meddle in the affairs of dragons, for thou art crunchy.