Randy McMurchy wrote:
> On Fri, 2006-04-07 at 16:05 -0700, Dan Nicholson wrote:
>
>
> But the "you don't need HAL thing" with others saying the userspace
> app is dependent on HAL, just has me totally confused at this point.
>
>
A lot of the packages don't *require* build and function, Building
On Fri, 2006-04-07 at 16:05 -0700, Dan Nicholson wrote:
> Andy seems to have left the building, but I recall that he was using
> the header sanitization script when that topic was hot on lfs-dev.
> I'd bet that he's using newer kernel headers.
Seeing how that about 4-5 people have given replies,
On 4/7/06, Joe Ciccone <[EMAIL PROTECTED]> wrote:
> Randy McMurchy wrote:
> > But you never answered my question, what did you do to get inotify to
> > work on a stock LFS system. It is becoming clear to me now, but how
> > you're getting inotify to work is still a mystery to me.
> >
> I figured ou
Randy McMurchy wrote:
> But you never answered my question, what did you do to get inotify to
> work on a stock LFS system. It is becoming clear to me now, but how
> you're getting inotify to work is still a mystery to me.
>
I figured out why I have it, I have /usr/include/sys/inotify.h provided
On Fri, 2006-04-07 at 15:39 -0700, Dan Nicholson wrote:
> I believe inotify got into the released version of the kernel in
> 2.6.13. So, you could get the header from there.
FWIW:
[EMAIL PROTECTED]: find /usr/src/linux* -name inotify.h
/usr/src/linux-2.6.14.3/include/linux/inotify.h
/usr/src
On Sat, 2006-04-08 at 00:35 +0200, Jürg Billeter wrote:
> I thought he just didn't compile gnome-vfs with HAL support, that's
> indeed not necessary for automounting. As previously written,
> gnome-volume-manager is the gnome automounter, not gnome-vfs, and
> gnome-volume-manager has an unconditio
On 4/7/06, Randy McMurchy <[EMAIL PROTECTED]> wrote:
> On Fri, 2006-04-07 at 16:23 -0600, Archaic wrote:
>
> > Add inotify.h and a newer kernel. ;) There are patches for inotify
> > support for the llh used in trunk, but I don't recall where.
>
> Sorry to be so lame in my knowledge about this stuff
On Fre, 2006-04-07 at 17:22 -0500, Randy McMurchy wrote:
> On Sat, 2006-04-08 at 00:14 +0200, Jürg Billeter wrote:
>
> > It's the other way round. udev notifies HAL of new devices, older HAL
> > versions call fstab-sync to create fstab entries on demand and
> > the /etc/fstab change gets noticed b
On Fri, 2006-04-07 at 16:23 -0600, Archaic wrote:
> Add inotify.h and a newer kernel. ;) There are patches for inotify
> support for the llh used in trunk, but I don't recall where.
Sorry to be so lame in my knowledge about this stuff, but where does
one get inotify.h? And exactly what version of
On 4/7/06, Archaic <[EMAIL PROTECTED]> wrote:
> On Fri, Apr 07, 2006 at 05:18:50PM -0500, Randy McMurchy wrote:
> >
> > But you never answered my question, what did you do to get inotify to
> > work on a stock LFS system. It is becoming clear to me now, but how
> > you're getting inotify to work is
On Sat, 2006-04-08 at 00:14 +0200, Jürg Billeter wrote:
> It's the other way round. udev notifies HAL of new devices, older HAL
> versions call fstab-sync to create fstab entries on demand and
> the /etc/fstab change gets noticed by FAM which notifies gnome-vfs.
I understand everything you're say
On Fri, Apr 07, 2006 at 05:18:50PM -0500, Randy McMurchy wrote:
>
> But you never answered my question, what did you do to get inotify to
> work on a stock LFS system. It is becoming clear to me now, but how
> you're getting inotify to work is still a mystery to me.
Add inotify.h and a newer kern
On 4/7/06, Andrew Benton <[EMAIL PROTECTED]> wrote:
>
> Exactly right. It's worked for more than a year, it worked with FAM and
> it works with Gamin. If Gnome VFS can now use inotify directly it makes
> me wonder if Gamin is still needed.
No, they're trying to get rid of it. But, alas, not every
On 4/7/06, Andrew Benton <[EMAIL PROTECTED]> wrote:
> Randy McMurchy wrote:
> > When you plug in a device, there is no filesystem for that device yet,
> > so how does FAM notify anything?
>
> I don't know
Possibly it isn't FAM at all. Randy might be right.
> > And then what does it notify so tha
On Fri, 2006-04-07 at 23:09 +0100, Andrew Benton wrote:
> Exactly right. It's worked for more than a year, it worked with FAM and
> it works with Gamin. If Gnome VFS can now use inotify directly it makes
> me wonder if Gamin is still needed.
Cool.
But you never answered my question, what did y
On Fre, 2006-04-07 at 23:09 +0100, Andrew Benton wrote:
> Dan Nicholson wrote:
> > Probably having newer kernel headers. From gnome-vfs configure, it
> > checks for linux/inotify.h or sys/inotify.h (from glibc-2.3.90+) to
> > enable inotify. This is what I got with l-l-h-2.6.12.0 and
> > glibc-2.
On Fre, 2006-04-07 at 16:48 -0500, Randy McMurchy wrote:
> On Fri, 2006-04-07 at 14:40 -0700, Dan Nicholson wrote:
>
> > In that case, I'm guessing fam (gamin) is providing the notification.
>
> This is a good discussion, I am going to learn something here. It was
> my understanding that the File
Dan Nicholson wrote:
Probably having newer kernel headers. From gnome-vfs configure, it
checks for linux/inotify.h or sys/inotify.h (from glibc-2.3.90+) to
enable inotify. This is what I got with l-l-h-2.6.12.0 and
glibc-2.3.6:
FS monitor backends: fam
In that case, I'm guessing f
Randy McMurchy wrote:
When you plug in a device, there is no filesystem for that device yet,
so how does FAM notify anything?
I don't know
And then what does it notify so that an appropriate fstab entry is
created. And what tells the operating how to create that fstab entry?
The fstab entri
On Fri, 2006-04-07 at 14:40 -0700, Dan Nicholson wrote:
> In that case, I'm guessing fam (gamin) is providing the notification.
This is a good discussion, I am going to learn something here. It was
my understanding that the File Alteration Monitor is used to tell the
operating system when there a
I thought this was interesting. Similar to what's in
http://www.linuxfromscratch.org/lfs/view/udev_update/chapter07/udev.html
This is from Kay Sievers' (current udev maintainer) blog:
http://vrfy.org/log/
--
Dan
--
http://linuxfromscratch.org/mailman/listinfo/lfs-dev
FAQ: http://www.linuxfroms
On 4/7/06, Randy McMurchy <[EMAIL PROTECTED]> wrote:
> On Fri, 2006-04-07 at 22:20 +0100, Andrew Benton wrote:
>
> > FS monitor backends: inotify fam
>
> I didn't know inotify worked with the current LFS setup. My
> installations don't show an inotify enabled FAM. What do you do
> dif
On Fri, 2006-04-07 at 22:20 +0100, Andrew Benton wrote:
> Probably inotify. From the gnome-vfs buildlog
> Gnome VFS configuration summary:
>
> IPv6 support:yes
> SSL support: yes
> Avahi support: no
> Howl supp
Randy McMurchy wrote:
And what mechanism, if not D-Bus, tells the system that plug-in hardware
has been introduced to the system? This isn't a test, I just simply
don't know and am curious.
Probably inotify. From the gnome-vfs buildlog
Gnome VFS configuration summary:
IPv6 support:
On Fri, 2006-04-07 at 21:26 +0100, Andrew Benton wrote:
> It's only an optional dependency. Gnome-VFS and Gnome-2.14 works fine
> without HAL or D-Bus. I plug in some flash memory and it automatically
> mounts it as /dev/sda1 on /mnt/mp3. Some problem with HAL or D-Bus is no
> reason not to us
On 4/7/06, M.Canales.es <[EMAIL PROTECTED]> wrote:
> El Viernes, 7 de Abril de 2006 22:27, Dan Nicholson escribió:
>
> > Yes, I get the failures without /etc/mtab. That's the issue.
>
> Due that in udev_update no mounts are done inside the chroot, /etc/mtab isn't
> created.
>
> I think that we sho
El Viernes, 7 de Abril de 2006 22:27, Dan Nicholson escribió:
> Yes, I get the failures without /etc/mtab. That's the issue.
Due that in udev_update no mounts are done inside the chroot, /etc/mtab isn't
created.
I think that we should to add the "touch /etc/mtab" in chapter06/e2fsprogs.xml
ju
On 4/7/06, M.Canales.es <[EMAIL PROTECTED]> wrote:
> El Viernes, 7 de Abril de 2006 22:08, M.Canales.es escribió:
>
> > After the removal of /etc/mtab, the test failures are here againg.
>
> More news:
>
> A plain "touch $LFS/etc/mtab" allow to pass successfully all e2fsprogs test.
Yes, I get the
Randy McMurchy wrote:
Not sure what do about this HAL/D-Bus thing. Best I can tell GNOME
2.14.0 wants the bleeding edge HAL/D-Bus. Gnome-VFS wants HAL-0.5.7.
It's only an optional dependency. Gnome-VFS and Gnome-2.14 works fine
without HAL or D-Bus. I plug in some flash memory and it automatic
El Viernes, 7 de Abril de 2006 22:08, M.Canales.es escribió:
> After the removal of /etc/mtab, the test failures are here againg.
More news:
A plain "touch $LFS/etc/mtab" allow to pass successfully all e2fsprogs test.
--
Manuel Canales Esparcia
Usuario de LFS nº2886: http://www.linuxfro
On 4/7/06, M.Canales.es <[EMAIL PROTECTED]> wrote:
> El Viernes, 7 de Abril de 2006 22:07, Dan Nicholson escribió:
>
> > I just built e2fsprogs in chroot using mount --bind and no other
> > modifications except that it's building on top of a full system. No
> > test failures:
>
> Do you have an $L
El Viernes, 7 de Abril de 2006 22:07, Dan Nicholson escribió:
> I just built e2fsprogs in chroot using mount --bind and no other
> modifications except that it's building on top of a full system. No
> test failures:
Do you have an $LFS/etc/mtab file?
--
Manuel Canales Esparcia
Usuario de LFS n
El Viernes, 7 de Abril de 2006 21:43, Archaic escribió:
> However, for size recording, my script does one thing in chroot to make
> / appear in /etc/mtab:
>
> mount -f -t $fs_type /dev/$partition /
>
> Maybe that's why it works for me?
Bingo!
When I made the umount of /dev and remount all ker
On 4/7/06, M.Canales.es <[EMAIL PROTECTED]> wrote:
>
> Confirmed :-/
>
> Using mount -bind:
>
> 2 tests succeeded 79 tests failed
>
> Using the old method to populate $LFS/dev:
>
> 81 tests succeeded 0 tests failed
I just built e2fsprogs in chroot using mount --bind and no other
modifications exce
On 4/7/06, Archaic <[EMAIL PROTECTED]> wrote:
>
> I see now what you are saying and agree. However, this sort of
> information seems most useful to developers and the more highly advanced
> readers. Perhaps a note should be placed in chap5's intro linking to
> this advanced information with a cavea
Archaic wrote:
> On Fri, Apr 07, 2006 at 01:36:39PM -0400, Chris Staub wrote:
>> Not really. The package installation page just lists what is needed to
>> build that package. The "buildorder" page lists exactly what needs to be
>> built *in a particular order*. The point is to specify that util-l
Uli Fahrenberg wrote:
Archaic, Apr 7, 13:30 -0600:
However, this sort of information seems most useful to developers and
the more highly advanced readers.
it is sort of like an index of knowledge gained and applicable to
development, but not really applicable to following the book to
produc
On Fri, Apr 07, 2006 at 10:26:10AM -0700, Dan Nicholson wrote:
>
> This may have to do with mount --bind. I can't think of any other
> reasons for it. Definitely needs investigation. Could you post
> anything that sticks out about these failures?
For some reason, I cannot duplicate that proble
Archaic, Apr 7, 13:30 -0600:
However, this sort of information seems most useful to developers and
the more highly advanced readers.
it is sort of like an index of knowledge gained and applicable to
development, but not really applicable to following the book to produce
a working system.
C
M.Canales.es wrote:
> Well, all that is beyond my capabilities. Real developers should to
> try to solve this issue.
Not that I'm necessarily a "real developer", but I do understand C, so
I'll see if I can replicate the failing environment here and do some
tests. I have e2fsprogs, but the rest (t
On Fri, Apr 07, 2006 at 01:36:39PM -0400, Chris Staub wrote:
>
> Not really. The package installation page just lists what is needed to
> build that package. The "buildorder" page lists exactly what needs to be
> built *in a particular order*. The point is to specify that util-linux
> in Chapte
El Miércoles, 5 de Abril de 2006 14:57, William Zhou escribió:
> BTW, in the section Important, the last sentence got a word misspelled.
> It is "to aid" not "to aide".
Fixed, thanks.
--
Manuel Canales Esparcia
Usuario de LFS nº2886: http://www.linuxfromscratch.org
LFS en castellano: htt
On Fri, 2006-04-07 at 14:04 -0500, Bruce Dubbs wrote:
> Overall, I am not that eager to do a BLFS 6.1.1 any more. I think LFS
> 6.2 will be in the testing phase relatively soon and running it against
> BLFS is one of the big tests. Releasing a new BLFS release, as you
> know, is a huge amount of
El Viernes, 7 de Abril de 2006 21:05, Bryan Kadzban escribió:
> The rest of the function is a bit hairy though. Probably the best way
> to figure out what exactly it's complaining about is to set the DEBUG
> preprocessor define to something other than zero; this should be doable
> if you cd into
On Fri, Apr 07, 2006 at 08:39:18PM +0200, Uli Fahrenberg wrote:
>
> ISTR some problems with symlinking man pages, the most obvious being that
BLFS has a compressdoc switch. It is mentioned in the LFS book. Seeing
as how a proper method is given, a workaround for other methods that may
be broken i
On Fri, 2006-04-07 at 12:00 -0700, Dan Nicholson wrote:
> That's a good question that needs to be pursued. I don't follow KDE
> at all, so I wouldn't know where to look for this info. Do you know
> which KDE applications use HAL?
Well, there are only two packages to get KDE up and running (plus
M.Canales.es wrote:
> + ext2fs_check_if_mount: No such file or directory while determining
> whether ./test.img is mounted.
Hmm. There are a few different places where that message appears in the
e2fsprogs source. Most of the tests seem to run e2fsck on an image,
though, so it's probably coming
Randy McMurchy wrote:
> On Thu, 2006-04-06 at 20:53 -0600, Archaic wrote:
>> 1) What base system is being targetted?
>>a) If 6.1.1, then what about the gcc4-specific stuff?
>>b) If 6.1.1, how much testing has gone on with that base version?
>
> I would say it would be for 6.1.1 and beyond
El Viernes, 7 de Abril de 2006 20:36, Bryan Kadzban escribió:
> Any idea which tests succeeded / failed?
e_icount_normal: inode counting abstraction optimized for storing inode
counts: ok
e_icount_opt: inode counting abstraction optimized for counting: ok
> What happens if you build with the n
On 4/7/06, Randy McMurchy <[EMAIL PROTECTED]> wrote:
> On Fri, 2006-04-07 at 13:20 -0500, Bruce Dubbs wrote:
>
> > Thanks Dan. It is interesting, but I'm not sure where to go with this
> > right now.
>
> Can of worms!
I'm not proposing adding pam_console to the book. Just put it out
there for in
El Viernes, 7 de Abril de 2006 20:20, M.Canales.es escribió:
> I have keeped both build trees, if you need some info from them.
Diffing the build trees all dfferences are in the build/test/* files.
All files on that drectory for the "mount -bind" build have ths additional
line:
+ ext2fs_check
Archaic, Apr 5, 12:00 -0600:
If symlinking vim to vi (which tends to suggest vi isn't installed), the
vim manpage should also be symlinked. If there are no objections, I'll
either do this or ticket it for later.
No need for ticketing, it's already been. Twice, at least:
http://wiki.l
M.Canales.es wrote:
> Confirmed :-/
>
> Using mount -bind:
>
> 2 tests succeeded 79 tests failed
>
> Using the old method to populate $LFS/dev:
>
> 81 tests succeeded 0 tests failed
>
> The build logs don't show differences beyond "ok" or "failed" for
> each test.
>
> I have keeped both build
On Fri, 2006-04-07 at 20:31 +0200, Christoph Berg wrote:
> It's abit off topic: I have a working LFS and part of BLFS on my laptop,
> which
> is an AMD64. I had to use patches, which aren't mentioned in the BLFS book or
> wiki. Can I just register for Trac and add these notes to the packages?
On Fri, 2006-04-07 at 13:20 -0500, Bruce Dubbs wrote:
> Thanks Dan. It is interesting, but I'm not sure where to go with this
> right now.
Can of worms!
Not sure what do about this HAL/D-Bus thing. Best I can tell GNOME
2.14.0 wants the bleeding edge HAL/D-Bus. Gnome-VFS wants HAL-0.5.7.
This i
No problem. I hope, I can help BLFS again in the future. :)
It's abit off topic: I have a working LFS and part of BLFS on my laptop, which
is an AMD64. I had to use patches, which aren't mentioned in the BLFS book or
wiki. Can I just register for Trac and add these notes to the packages?
Am Fre
El Viernes, 7 de Abril de 2006 20:06, Dan Nicholson escribió:
> On 4/7/06, M.Canales.es <[EMAIL PROTECTED]> wrote:
> > I have yet all the kernel filesystems mounted. I will do now a new
> > E2fsprogs build ith mount -bind. After that i will to umount /dev and
> > mount it again like is done in tru
Dan Nicholson wrote:
> On 4/5/06, Bruce Dubbs <[EMAIL PROTECTED]> wrote:
>> As far a PAM goes, we could point to
>> http://www.kernel.org/pub/linux/libs/pam/modules.html for external PAM
>> modules.
>>
>> pam_console is from Red Hat. If we need it, I think we would have to
>> extract it from the s
On Thu, 2006-04-06 at 20:53 -0600, Archaic wrote:
> I agree with your sentiments wholeheartedly, but have a few questions:
My apologies for not answering them sooner.
> 1) What base system is being targetted?
>a) If 6.1.1, then what about the gcc4-specific stuff?
>b) If 6.1.1, how much
On 4/7/06, Christoph Berg <[EMAIL PROTECTED]> wrote:
>
> I just want to ask, if the BLFS developers are aware of the current problem,
> the xorg-server-1.0.2 package has with most GTK+ programs as mentioned here:
> http://bugs.gentoo.org/show_bug.cgi?id=127333
I forgot. Thanks, Christoph for trac
On 4/7/06, M.Canales.es <[EMAIL PROTECTED]> wrote:
> El Viernes, 7 de Abril de 2006 19:51, Dan Nicholson escribió:
>
> >
> > Unfortunately, there's not a lot of info there. Do you still have the
> > source directory? How about, now that the base system is installed,
> > try to rebuild e2fsprogs a
El Viernes, 7 de Abril de 2006 19:51, Dan Nicholson escribió:
>
> Unfortunately, there's not a lot of info there. Do you still have the
> source directory? How about, now that the base system is installed,
> try to rebuild e2fsprogs and see if the tests still fail.
I have yet all the kernel fil
On 4/7/06, M.Canales.es <[EMAIL PROTECTED]> wrote:
>
> > This may have to do with mount --bind. I can't think of any other
> > reasons for it. Definitely needs investigation. Could you post
> > anything that sticks out about these failures?
>
> Attached the full E2fsprogs build log.
Unfortunate
El Viernes, 7 de Abril de 2006 19:26, Dan Nicholson escribió:
> Out of curiosity, what kind of hardware do you have? I've been
> getting three error here using an Athlon-XP for a while now. I like
> to think these are processor specific, but I haven't really
> investigated.
An Intel(R) Pentium(
Archaic wrote:
On Fri, Apr 07, 2006 at 12:48:29PM -0400, Chris Staub wrote:
I disagree - all such dependencies should be listed regardless of
whether they are accounted for by alphabetical order.
It's starting to sound like you want to duplicate the info in the
individual package pages. I don'
On 4/5/06, Bruce Dubbs <[EMAIL PROTECTED]> wrote:
>
> As far a PAM goes, we could point to
> http://www.kernel.org/pub/linux/libs/pam/modules.html for external PAM
> modules.
>
> pam_console is from Red Hat. If we need it, I think we would have to
> extract it from the source RPM as I can't find a
On Fri, Apr 07, 2006 at 12:48:29PM -0400, Chris Staub wrote:
>
> I disagree - all such dependencies should be listed regardless of
> whether they are accounted for by alphabetical order.
It's starting to sound like you want to duplicate the info in the
individual package pages. I don't agree wit
Chris Staub wrote:
> Chris Staub wrote:
>> Bruce Dubbs wrote:
>>>
>>> The lists seem a bit too prominent.
>>
>> I don't mind just having a single sentence for listing a couple of
>> dependencies like this, but in a couple instances the list of deps.
>> for a certain package is fairly long. In the c
On 4/7/06, M.Canales.es <[EMAIL PROTECTED]> wrote:
>
> Chapter 06 GCC test suite summary:
>
> === gcc Summary ===
>
> # of expected passes 35544
> # of unexpected successes 3
> # of expected failures 92
> # of untested testcases 28
> # of unsupported tests 326
> /sources/gcc-build/gcc/xgcc v
El Jueves, 6 de Abril de 2006 22:37, Archaic escribió:
> On Thu, Apr 06, 2006 at 10:27:23PM +0200, M.Canales.es wrote:
> > The build will take some hours on my system, then the commit will be made
> > tomorrow if there is no build issues.
>
> Sounds good, Manuel. Thanks for all the help! :)
Build
El Viernes, 7 de Abril de 2006 17:38, Chris Staub escribió:
> I've started working on a page describing the reasoning behind the
> package build order for LFS. Take a look at it here -
> http://linuxfromscratch.org/~chris/lfs-book/chapter05/buildorder.html.
> Any comments - page layout, location, e
Replying on -dev.
On 4/7/06, Ag Hatzim <[EMAIL PROTECTED]> wrote:
> Randy McMurchy([EMAIL PROTECTED])@Fri, Apr 07, 2006 at 11:09:14AM -0500:
> > So, that said, how many modules are you speaking of Ag?
> >
> Lets see,and if we go with the right order and without to count the
> Mesa/libdrm must be
Chris Staub wrote:
Bruce Dubbs wrote:
The lists seem a bit too prominent.
I don't mind just having a single sentence for listing a couple of
dependencies like this, but in a couple instances the list of deps. for
a certain package is fairly long. In the case of Coreutils, there are
actuall
Bruce Dubbs wrote:
The lists seem a bit too prominent.
I would reword the paragraphs like Coreutils to something like:
Coreutils must be installed before Bash and Diffutils because they
hard-code Coreutils binary locations.
I don't mind just having a single sentence for listing a couple of
Archaic wrote:
Some preliminary thoughts:
I would probably not go into such detail as if this book is a teaching
aide, the above paragraph would bog down and confuse the reader. Some
lighter, general reading along the lines of:
"Some pkgs have circular deps [insert very light blurb as to
Chris Staub wrote:
> I've started working on a page describing the reasoning behind the
> package build order for LFS. Take a look at it here -
> http://linuxfromscratch.org/~chris/lfs-book/chapter05/buildorder.html.
> Any comments - page layout, location, etc., are welcome. I don't have
> much act
On Fri, Apr 07, 2006 at 12:25:56PM -0400, Chris Staub wrote:
>
> I'm also updating the dependencies for each package. I think it would be
> a good idea to separate build deps. from testsuite deps. Anyone agree?
Definitely.
--
Archaic
Want control, education, and security from your operating s
Chris Staub wrote:
I've started working on a page describing the reasoning behind the
package build order for LFS. Take a look at it here -
http://linuxfromscratch.org/~chris/lfs-book/chapter05/buildorder.html.
I'm also updating the dependencies for each package. I think it would be
a good i
On Fri, Apr 07, 2006 at 11:38:46AM -0400, Chris Staub wrote:
> I've started working on a page describing the reasoning behind the
> package build order for LFS. Take a look at it here -
> http://linuxfromscratch.org/~chris/lfs-book/chapter05/buildorder.html.
Some preliminary thoughts:
"Au
I've started working on a page describing the reasoning behind the
package build order for LFS. Take a look at it here -
http://linuxfromscratch.org/~chris/lfs-book/chapter05/buildorder.html.
Any comments - page layout, location, etc., are welcome. I don't have
much actual information there yet
80 matches
Mail list logo
