Yves
Am 08.03.2015 um 13:06 schrieb Yves Blusseau:
Hi all,
just for your information, i have develop a script to store big binaries in
git repository.
With this we can have a full filesystem with all the versions like this:
4.1
├── 4.1/i386
│ └── 4.1/i386/pkg2.lrp
└── 4.1/x86_64
Hi Folks
I upgraded my WLAN repeater LEAF box today and found a little problem I
would like to discuss.
The previous release was ~5.1.1 and pata-acpi was loaded, probably from
initrd. The current 5.1.3 modules tarball does not appear to include
pata-acpi neither is it used on my ALIX board. I
Yves
Am 11.03.2015 um 12:22 schrieb Yves Blusseau:
Le 11 mars 2015 à 09:39, Erich Titl erich.t...@think.ch a écrit :
Hi Yves
Am 11.03.2015 um 08:56 schrieb Yves Blusseau:
...
I guess with your tool we somehow duplicate a git repository to another
filepath without actually duplicating
Am 11.03.2015 um 16:04 schrieb kp kirchdoerfer:
Am Dienstag, 10. März 2015, 21:28:25 schrieb Erich Titl:
Hi KP
...
I don't know about others, but this is on mine anyway. Rather loose some
weight on the module side.
AFAIK it's not that easy to save space/memory with modules, not shure what
Am 11.03.2015 um 15:54 schrieb kp kirchdoerfer:
Hi Erich;
..
It is part of the release work to put the files into the git repositories
(package/5_1 etc), and it's me who does the work currently.
All archs and packages are rebuild from scratch for a new release, then the
tarballs will
Hi Folks
Is there a reason we don't have a serial tarball for the X86_64 arch? I
want to complete my site with all archs and I am basing on the serial
tarball. It does not make a big difference though.
BTW on our main page there are a number of empty links, I guess they
should be removed.
Hi Folks
I have finished the upgrade function, please consider testing it.
It uses my small web server on leaf.think.ch which has a 5.1.3 for the
following architectures.
i486 i686 geode
You can get a config.lrp with upgrade included at
http://leaf.think.ch/stable/i486/config.lrp
I intend to
Am 13.03.2015 um 14:57 schrieb kp kirchdoerfer:
So it looks like https access works, but the version file used to check
access to the net fails. We need something to check for, version is a
good surrogate.
I've added one.
That's all in the git repo packages, uploading is via git (add,
Hi KP
Am 13.03.2015 um 14:57 schrieb kp kirchdoerfer:
Am Donnerstag, 12. März 2015, 19:15:02 schrieb Erich Titl:
..
I've added one.
That's all in the git repo packages, uploading is via git (add, commit,
push...)
A, that is where it's at, OK clone the packages repo :-(
ET
Hi KP
Am 13.03.2015 um 14:57 schrieb kp kirchdoerfer:
Am Donnerstag, 12. März 2015, 19:15:02 schrieb Erich Titl:
...
I've added one.
That's all in the git repo packages, uploading is via git (add, commit,
push...)
I believe I understand now, weird but if it works.
I would l like
Hi Yves
Am 11.03.2015 um 08:56 schrieb Yves Blusseau:
...
I guess with your tool we somehow duplicate a git repository to another
filepath without actually duplicating the data and we should be able to
point to this partial tree. Is that it?
With this we can manage all the packages in a
Hi KP
Am 10.03.2015 um 19:48 schrieb kp kirchdoerfer:
...
Don't think so, they generate the same links as proposed.
Yes, they _generate_ it. That must be the CMS underneath.
And what's the problem to use http/s access to git repo?
None, it just requires the ssl version of wget.
Hi KP
Am 12.03.2015 um 16:17 schrieb kp kirchdoerfer:
HI Erich;
Am Mittwoch, 11. März 2015, 19:27:00 schrieb Erich Titl:
Am 11.03.2015 um 15:54 schrieb kp kirchdoerfer:
Hi Erich;
..
It is part of the release work to put the files into the git repositories
(package/5_1 etc), and it's me
Hi KP
Am 12.03.2015 um 17:59 schrieb kp kirchdoerfer:
...
Look for
https://sourceforge.net/p/leaf/packages/ci/master/tree/stable/
does that help you?
AP# sh upgrade -v
upgrade: checking web access by retrieving the version file
aborting: wget error code 0
upgrade: abort wget error
Am 12.03.2015 um 19:10 schrieb kp kirchdoerfer:
Hi Erich;
Am Donnerstag, 12. März 2015, 18:55:37 schrieb Erich Titl:
...
OK, be it that way, but if you upload files, do you have any control on
the location, e.g. the URL they can be accessed with?
For the major part no, AFAIK it's a git
Hi KP
Am 12.03.2015 um 17:59 schrieb kp kirchdoerfer:
...
It seems one of us has a misconception of the SF features we can use.
It is my understanding that we only have limited access/space to/on the SF
shell, and it's mainly used to install a CMS, but not to use it as space to
Yves
Am 08.03.2015 um 13:06 schrieb Yves Blusseau:
Hi all,
just for your information, i have develop a script to store big binaries in
git repository.
With this we can have a full filesystem with all the versions like this:
4.1
├── 4.1/i386
│ └── 4.1/i386/pkg2.lrp
What is pkg2.lrp?
Am 10.03.2015 um 16:32 schrieb kp kirchdoerfer:
Hi Erich;
Am Dienstag, 10. März 2015, 11:39:21 schrieb Erich Titl:
Hi Folks
I upgraded my WLAN repeater LEAF box today and found a little problem I
would like to discuss.
The previous release was ~5.1.1 and pata-acpi was loaded, probably
Hi Folks
I am playing a little with the package repository, trying to make it
appear to a browser in a reasonable tree form.
I _believe_ by using the git features branches and tags we could provide
a reasonable tree for our packages.
Just looking at the current settings for the packages
Hi Folks
I adapted the 5.1.4-rc1 branch in the package repository to provide all
necessary data to upgrade.
the main differences to the existing packages are
- modules-$ARCH.tgz tarballs are unpacked into module-$ARCH directories,
this makes no big difference in size, as the modules are
Am 30.03.2015 um 15:48 schrieb Andrew:
30.03.2015 16:33, Erich Titl пишет:
Hi Folks
...
- modules-$ARCH.tgz tarballs are unpacked into module-$ARCH directories,
this makes no big difference in size, as the modules are compressed
anyway. The tarballs can be deleted.
How it'll work
Erich Titl:
Hi Folks
I adapted the 5.1.4-rc1 branch in the package repository to provide all
necessary data to upgrade.
the main differences to the existing packages are
- modules-$ARCH.tgz tarballs are unpacked into module-$ARCH directories,
extra work...
Not really, even less work
Am 30.03.2015 um 17:03 schrieb Andrew:
30.03.2015 17:21, Erich Titl пишет:
...
Impossible, it is too big. Upgrade is not thought as an initial setup tool.
But this will cause a mismatched modules.tgz version. And will cause a
headache for adding new hardware that isn't supported by currently
Am 30.03.2015 um 23:24 schrieb Andrew:
30.03.2015 23:40, Erich Titl пишет:
Am 30.03.2015 um 18:39 schrieb Andrew:
30.03.2015 18:23, Erich Titl пишет:
Am 30.03.2015 um 17:03 schrieb Andrew:
...
...
modules.tgz updating possibility?
No problem on my side, recreating the tarball is a piece
Am 30.03.2015 um 18:39 schrieb Andrew:
30.03.2015 18:23, Erich Titl пишет:
Am 30.03.2015 um 17:03 schrieb Andrew:
...
Correct, you cannot just add hardware without adding modules. As there
is no space on the target machine to hold modules.tgz this is not
relevant. Lots of our systems
Am 30.03.2015 um 18:44 schrieb kp kirchdoerfer:
Hi Erich;
Am Montag, 30. März 2015, 17:59:38 schrieb Erich Titl:
BTW
Can you compile the kernel on master, on my system it crashes miserably
DEPMOD 3.10.69-i686
master branch provides a 3.14 kernel
maint branch provides a 3.10
Am 30.03.2015 um 17:09 schrieb kp kirchdoerfer:
Am Montag, 30. März 2015, 16:40:01 schrieb Erich Titl:
Am 30.03.2015 um 16:10 schrieb kp kirchdoerfer:
Hi Erich;
...
:
are in a ?
Fingers slower than brain :-(
The modules are in a directory tree.
I prefer you load the tarball, which
Hi Folks
Am 24.03.2015 um 12:38 schrieb Erich Titl:
..
I don't know, if we can use this to just give access to the current or
stable files. I did some local experiments
mega@leafbuilder:~/leaf/devel/packages$ du -s .
5961104 .
mega@leafbuilder:~/leaf/devel/packages$ git branch
Hi Yves
Am 28.03.2015 um 18:27 schrieb Yves Blusseau:
Sorry for my intrusion but why don't you use a git clone working dir for the
document root of the http server.
What would that achieve? Isn't is just another repository and thus a
copy of the data?
With that we can change the tree
Am 23.03.2015 um 12:07 schrieb Yves Blusseau:
Le 23 mars 2015 à 09:25, Erich Titl erich.t...@think.ch a écrit :
Hi Yves
Am 23.03.2015 um 08:56 schrieb Yves Blusseau:
...
The feature is to clone a repository without downloading all the history
(several GB)
OK so you basically remove
Am 02.03.2015 um 14:13 schrieb kp kirchdoerfer:
Hi Erich;
...
Sure, and it might be possible to rearrange the links, so they would be
more simple. For example the GIT hash does not help here at all.
It's the download link. No need to rearrange and it's avoided cause it will
break...
Hi KP
Am 02.03.2015 um 16:55 schrieb kp kirchdoerfer:
Am Montag, 2. März 2015, 15:44:01 schrieb Erich Titl:
Am 02.03.2015 um 14:13 schrieb kp kirchdoerfer:
...
The page I got the URL from is generated from a parser.
Will it be any help if we add one or more flatfiles with the architecture
Am 03.03.2015 um 17:09 schrieb kp kirchdoerfer:
Erich;
will something like that in a file help you?
zbxagent.lrp
https:/sourceforge.net/p/leaf/packages/ci/786d2c9d66decc4c2d8409362f1180d9446f0271/tree/5_1/X86_64/zbxagent.lrp?format=raw
zbxutil.lrp
Hi Andrew
Am 03.03.2015 um 23:46 schrieb Andrew:
Hi.
03.03.2015 23:49, Erich Titl пишет:
...
Try
https://sourceforge.net/p/leaf/packages/ci/master/tree/5_1/X86_64/zebra.lrp?format=raw
Yes, such a link could be used for packages. I placed a version file in
the tree to easily find
Hi Yves
I have a question regarding my commit branch which I sync to the repository.
I would like to rebase this branch to tag 5.1.3 to be able to work with
this release.
I read about rebase --onto, but to me this does not make sense, since I
did not branch off a tag, but another branch.
Is
Am 01.03.2015 um 16:52 schrieb kp kirchdoerfer:
Hi Erich;
..
What exactly is different from the sourceforge pages?
The Packages 5.x page is available.
For whatever reason we can't this page to SF, it will not have the expected
content, whereas the page on zetam.org does.
...
I see
Yves
Am 01.03.2015 um 12:10 schrieb Yves Blusseau:
You may have a look into
leaf.zetam.org
What exactly is different from the sourceforge pages?
I worked with Yves recently to provide a new and working Packages page
there,
no further decisions made about mocing webpages permamently -
Hi KP
Am 28.02.2015 um 18:40 schrieb kp kirchdoerfer:
Hi Erich;
Sounds like an interesting and useful task.
Am Mittwoch, 25. Februar 2015, 12:56:15 schrieb Erich Titl:
Hi Folks
I am investigating into an upgrade function for LEAF, I did not grok the
stuff Andrew wrote, so I am
Hi KP
I was bit by an old issue
nslookup: can't resolve '(null)': Name or service not known
which you circumvented for webconf
I believe the issue should be sorted out completely as the proposed
solution requires knowledge of a nameserver (localhost is _not_ always a
nameserver)
Is this a
Hi Folks
I am investigating into an upgrade function for LEAF, I did not grok the
stuff Andrew wrote, so I am looking for easy and easily understandable
ways to get there.
A few questions pop up which I have not been able to find an easy
aproach (probably too lazy)
- How to access the actual
Am 27.03.2015 um 16:43 schrieb kp kirchdoerfer:
Am Freitag, 27. März 2015, 16:22:40 schrieb Erich Titl:
Am 27.03.2015 um 15:53 schrieb kp kirchdoerfer:
Am Freitag, 27. März 2015, 09:12:04 schrieb Erich Titl:
...
Yes it's effcient with branches.But it's not really efficient with
binaries
Hi
Am 27.03.2015 um 15:53 schrieb kp kirchdoerfer:
...
Just play with it and let's know about your findings.
Esp stable is a playground, but I hope we can shrink the other repos as well
later.
I can now access the packages as
http://sourceforge.net/p/leaf/packages/ci/5.1.3/tree/
and
Am 27.03.2015 um 15:53 schrieb kp kirchdoerfer:
Am Freitag, 27. März 2015, 09:12:04 schrieb Erich Titl:
...
Yes it's effcient with branches.But it's not really efficient with binaries,
though as discussed that's a workaround for the missing ftp/whatever-space.
To git a file is a file
Hi Yves
Am 23.03.2015 um 08:56 schrieb Yves Blusseau:
...
The feature is to clone a repository without downloading all the history
(several GB)
OK so you basically remove the history from a git repository.
Now with the git-store command you can retrieve any packages you want using
Am 23.03.2015 um 18:05 schrieb Yves Blusseau:
Le 23 mars 2015 à 16:35, kp kirchdoerfer kap...@users.sourceforge.net a
écrit :
...
How do I use it to get a package from it, and more important, how can I
update
a package in the repo, build new directories??
Go into the directory
Hi Folks
Upgrade has proven to be functional on the i386 platform. I cannot judge
other platforms, as I don't have any hardware to test with.
I would like to know what you judge necessary to enable the packages
repository to serve the necessary files. As KP pointed out, existing
procedures build
Hi Andrew
Am 02.04.2015 um 11:32 schrieb Andrew:
Hi.
02.04.2015 11:22, Erich Titl пишет:
Hi Folks
Upgrade has proven to be functional on the i386 platform. I cannot judge
other platforms, as I don't have any hardware to test with.
I would like to know what you judge necessary to enable
Hi Andrew
Am 02.04.2015 um 22:35 schrieb Andrew:
02.04.2015 22:09, Mega пишет:
...
From other side, we may do other trick: we may store modules in
squashfs instead of tgz, and mount it instead of unpacking. And
even we
may remove moddb.lrp - just mount this squashfs at boot, probe
Am 03.04.2015 um 13:26 schrieb Andrew:
03.04.2015 14:02, Erich Titl пишет:
...
The numbers imply that there is nothing to be gained in size over the
tarball using squashfs. The advantage of using a squashfs is that we do
not need to unpack the tarball for each file involved. This method
Am 13.04.2015 um 21:04 schrieb kp kirchdoerfer:
Hi Erich;
...
Make shure you have the Bering-3.14.36-config -xxx.patches, otherwise
apllying
the patches will fail. This happended to me after your new config, therefor I
had to create the new patch files.
Yes I know, sorry about that.
Hi Andrew
Am 13.04.2015 um 23:43 schrieb Andrew:
Hi all.
I added some stuff to distro (hardware/modules autoprobing from squashfs
at boot, + some squashfs cleanup). It looks like all is OK (at least, my
test VM is booted), but there are some features that are needed to be
realized.
Am 14.04.2015 um 10:46 schrieb Andrew:
14.04.2015 09:36, Erich Titl пишет:
...
Of course, old logic remains here. I just want to add possibility to
remove moddb for systems that have enough storage space for squashfs.
This will also free some memory that is hold by unused drivers from moddb
Original-Nachricht
Betreff:[leaf:bering-uclibc] [93a3cd] - Andrew Denisenko: squashfs:
FRAGMENT_CACHE_SIZE=3 for speedup
Datum: Sat, 02 May 2015 21:25:31 +
Von:Repository LEAF Linux Embedded Appliance Framework
bering-uclibc.git
Hi
Am 16.05.2015 um 00:15 schrieb kp kirchdoerfer:
Am Samstag, 16. Mai 2015, 00:23:08 schrieb Andrew:
Hi all.
I think that it'll be better to generate squashfs by buildtool, like
initrd/initmod, instead of buildimage (this will help also to build
squashfs-based initrd in future). And rename
Hi KP
Am 14.04.2015 um 17:32 schrieb kp kirchdoerfer:
HI Erich;
...
This is all correct; and I may have disturbed it.
I confess I never got used to the idea and when I ran into pb's it was the
short path to adjust the config files.
Once we move to a newer kernel, we, and esp. myself,
Hi KP
I had a quick look in the kernel config files and came up with a few
questions you may be able to answer. All my findings are based on the
latest git master version.
Bering-3.14.20.config-versatile Bering-3.14.36.config-i486.patch
Bering-3.14.36.config
Am 16.04.2015 um 13:06 schrieb Andrew:
Hi.
16.04.2015 11:02, Erich Titl пишет
...
AFAIR we decided to store config files just for cases when patch doesn't
lays on new kernel config, for easier troubleshootong.
So if you don't trust patch, why patch?
So I would like to see the Berning
Am 16.04.2015 um 14:00 schrieb Andrew:
16.04.2015 14:48, Erich Titl пишет:
Am 16.04.2015 um 13:06 schrieb Andrew:
Hi.
16.04.2015 11:02, Erich Titl пишет
...
AFAIR we decided to store config files just for cases when patch doesn't
lays on new kernel config, for easier troubleshootong.
So
Hi KP
Am 16.04.2015 um 19:23 schrieb kp kirchdoerfer:
Hi Erich;
...
I suggest you commit your changes to branch next (I think we need to merge
master into next before?) or a a topic branch. Either way we can test your
work before make it into master.
And even then we need a bit of a
Hi Folks
I have been thinking some more on the way we handle our kernel config
files and looked at the rather complicated way we handle them in
buildtool.mk.
We keep the version number of the kernel in the config files. This
completely defeats the purpose of a version control system, as this is
Hi Yves
Am 17.04.2015 um 09:19 schrieb Yves Blusseau:
Hi all,
i'm working on a project that use openembedded/yocto project. It's really
great to create a full linux distribution and boot images for any type of
processor/architecture.
Just for the informations this is how the configs and
Am 17.04.2015 um 12:12 schrieb Andrew:
Hi.
It sounds good - it'll simplify maintenance.
But IMHO we should leave kernel major/minor versions - just because we
may have different kernel versions for different platforms in future
(for ex., some SoC may require specified kernel version for
Am 08.04.2015 um 21:04 schrieb Andrew:
08.04.2015 17:44, kp kirchdoerfer пишет:
...
We can use parted for automation. fdisk isn't good for this (we can't
handle multiple partitions correctly).
Mhhh I used to automate tasks with fdisk to build multiple
partitions on CF's given. fdisk has
Hi KP
you pointed me to buildimage.pl as the tool you use to prepare the
images. There does not appear to be code there for placing the files in
the packages repository, so you must either do it manually or use
something different.
Could you please enlighten me?
Thanks
ET
smime.p7s
Am 09.04.2015 um 17:03 schrieb Andrew:
09.04.2015 17:47, Erich Titl пишет:
...
They are, but of course they cannot be replaced automatically if they
are not in either the tarball, the squashfs or in the modules directory.
They'll be in modules directory because they'll be in moddb.
Yes
Am 09.04.2015 um 16:36 schrieb Andrew:
...
So let's store both packed modules + unpacked on server side, as we
decided earlier.
Packed modules will be suitable for routers with enough storage space;
unpacked - for tiny systems.
I just voting against excluding packed modules in prefer to
Hi Folks
I have been able to test the upgrade script successfully using a
squashfs file as the source for the modules. This leads us to the
question if we should use this method as a means to distribute the modules.
The use of a squashfs file instead of a compressed tarball for the
distribution
Am 09.04.2015 um 16:39 schrieb kp kirchdoerfer:
HI Erich;
Am Donnerstag, 9. April 2015, 16:30:16 schrieb Erich Titl:
Hi KP
you pointed me to buildimage.pl as the tool you use to prepare the
images. There does not appear to be code there for placing the files in
the packages repository, so
Am 09.04.2015 um 14:25 schrieb Andrew:
09.04.2015 14:35, Erich Titl пишет:
Am 09.04.2015 um 11:12 schrieb Andrew:
Hi.
...
For me, it's much easier to update just one file instead of hundreds of
modules on persist storage. Esp. when one broken (non-updated) module
can break system.
You
Hi Folks
The latest git state requires user interaction during the linux build,
probably while running --oldconfig due to the config file not being of
the same release.
cheers
ET
smime.p7s
Description: S/MIME Cryptographic Signature
Hi KP
Am 11.04.2015 um 10:28 schrieb Repository LEAF Linux Embedded Appliance
Framework bering-uclibc.git:
master: change to minstrel rate control for ath9k
http://sourceforge.net/p/leaf/bering-uclibc/ci/aad8bfb2eb88d1a7d551c38cbecbc0411a1674d9/
Hi KP
It looks like we are shooting at each other right now. My last commit
appears to break the patching chain, so I had to revert it. I don't know
what exactly is going on right now, but it appears my changes either get
reverted or overwritten.
cheers
ET
smime.p7s
Description: S/MIME
Hi KP
I _believe_ I saw an indicetion of 5.1.4 being published. Where did it
go? is this now in 5_1 on packages?
Thanks
ET
smime.p7s
Description: S/MIME Cryptographic Signature
--
BPM Camp - Free Virtual Workshop May
Hi
Am 14.04.2015 um 19:16 schrieb kp kirchdoerfer:
Am Dienstag, 14. April 2015, 19:00:07 schrieb Andrew:
Hi.
For maint it works. For master - it seems like fdisk output changes (new
version?).
Yes, it has been updated to util-linux-ng 2.25; I didn't mention that in
changelog.
Quite
Hi
Am 07.04.2015 um 18:58 schrieb kp kirchdoerfer:
Am Montag, 6. April 2015, 22:34:25 schrieb Andrew:
It seems like script is now enough useful, I included it into maint
branch (it shouldn't break anything).
It fails to work in my test setup (see previous mail) - can we disabe for
maint
Am 07.04.2015 um 19:45 schrieb Andrew:
07.04.2015 20:03, Erich Titl пишет:
..
System may be used for new flash drives creation. And we may just move
away hdsupp.lrp from installed system.
But then install makes even less sense on an installed system.
cheers
ET
smime.p7s
Description: S
Hi Folks
following Andrew's suggestions I made a few tests with squashfs to see
how an agreement on the presentation of data could be reached. I have
not tested with the existing upgrade script, I just wanted to see what
implications a squashfs based distribution of the modules tarball would
Looks like a git clean took care of the problem. I still don't know
where the merge comes from though
Am 08.04.2015 um 11:38 schrieb Erich Titl:
Hi Folks
It looks like something went wild in the master branch, I am getting
tons of deleted and untracked files, as if the master changed a lot
Hi Folks
It looks like something went wild in the master branch, I am getting
tons of deleted and untracked files, as if the master changed a lot.
I am doing a separate clone now to see if by any chance my own
repositore went berserk, else I believe the following inadverted commit
is the culprit
Hi
Part of my tests for upgrade involves testing on a WRAP. I found that
the kernel release 3.10.72 appears not to be capable of rebooting the
system despite the reboot=bios parameter on the command line. Version
3.10.71 does not have this issue.
Bright ideas?
Thanks
ET
smime.p7s
Hi Andrew
Am 02.04.2015 um 12:35 schrieb Andrew:
02.04.2015 13:16, Erich Titl пишет:
Hi Andrew
...
Single tarball is easier for update than hundreds of small files.
No it is not, if you have insufficient space. It is just a bloody pain
in the butt. And then what is IT for if we cannot
Hi
Am 16.05.2015 um 20:38 schrieb Andrew:
Scripts are also fixed/cleaned; currently - kernel modules are in
kmod.sqfs squash file (I added PackageExt option to packages). It seems
like all is OK - at least image is created and boots OK.
Have you also adapted upgrade? Upgrade, the way it used
Hi KP
I am somewhat online again and I would like to get the upgrade thing
running (again). Looking at the packages repo I only see 5_0 and 5_1.
What is the actual content (5.1.x) of 5_1?
Will there be a 5_2 (5.2.x) sometime soon?
Thanks
ET
smime.p7s
Description: S/MIME Cryptographic
Hi KP
Am 06.06.2015 um 01:34 schrieb kp kirchdoerfer:
Hi Erich;
...sorry for being late.
Am Dienstag, 2. Juni 2015, 10:44:39 schrieb Erich Titl:
Hi KP
I am somewhat online again and I would like to get the upgrade thing
running (again). Looking at the packages repo I only see 5_0
An: LEAF Linux Embedded Appliance Framework Git repository
nore...@bering-uclibc.leaf.p.re.sf.net
Branch: master
Merge branch 'master' of ssh://git.code.sf.net/p/leaf/bering-uclibc
By Erich Titl on 07/02/2015 07:59
*View Changes*
http://sourceforge.net/p/leaf/bering-uclibc/ci
Hi Andrew
Am 03.07.2015 um 19:24 schrieb Andrew:
Hi.
Modules that are copied manually, should be saved. Modules that are
loaded from squashfs aren't added to moddb.tgz to reduce it's size.
Maybe it'll be good to run modules probing once before squashfs loading
to load updated modules that
Hi Andrew
Am 04.07.2015 um 09:43 schrieb Andrew:
Hi.
Currently - there is a such order:
1. modules are copied to rootfs from moddb
2. sqfs is mounted
3. h/w autoprobing (with sqfs) is executed.
3. loaded modules are copied to rootfs if they are missing here
4. modules from /etc/modules (+
Hi
Am 04.07.2015 um 09:50 schrieb Andrew:
Hi.
In most cases probing from squashfs is preferred over loading from moddb
- at least, this will not break system after update if it'll require
some modules that aren't in moddb (for ex., some ethernet driver).
We may add some option to use legacy
Hi
Am 04.07.2015 um 15:59 schrieb Andrew:
04.07.2015 16:23, Erich Titl пишет:
Am 04.07.2015 um 15:15 schrieb Andrew:
04.07.2015 15:59, Erich Titl пишет:
Hi
...
Erich
I'm not sure that we should store all modules in moddb by default.
There's no profit for this.
Sure
Hi Andrew
Am 02.07.2015 um 21:59 schrieb Andrew:
Hi.
I think that it isn't a good idea to modify proprietary package test
dataset - at least till we expect to use it as one of speedtest.net
servers, not as stand-alone host. Also, it uses jpeg images as test
dataset - jpeg generation is enough
Hi Andrew
Am 02.07.2015 um 18:47 schrieb kp kirchdoerfer:
HI Andrew;
Am Samstag, 20. Juni 2015, 17:52:38 schrieb Andrew:
Hi all.
I created speedtest mini package for LEAF routers using mini_httpd (it
was needed just to add upload size calculation cgi script); but placing
it into main tree
Hi
Am 04.07.2015 um 13:56 schrieb Andrew:
04.07.2015 11:34, Erich Titl пишет:
Hi
Am 04.07.2015 um 09:50 schrieb Andrew:
Hi.
In most cases probing from squashfs is preferred over loading from moddb
- at least, this will not break system after update if it'll require
some modules that aren't
Hi
Am 04.07.2015 um 14:14 schrieb kp kirchdoerfer:
Hi;
I almost agree with Erich :)
..
I don't think it's necessary to revive tgz; it will just waste disk-space adds
code etc..
I agree
Saving modules in modbd and get rid of modules.sqfs later and until next
update works if I remoce
Hi
Am 04.07.2015 um 15:00 schrieb Andrew:
04.07.2015 15:45, kp kirchdoerfer пишет:
Am Samstag, 4. Juli 2015, 15:31:40 schrieb Andrew:
...
by the user?
Right. On my 5.2 boxes there's no moddb at all.
That is fine for you, on mine there is no modules.sqfs and I doubt there
will ever be.
Am 04.07.2015 um 15:15 schrieb Andrew:
04.07.2015 15:59, Erich Titl пишет:
Hi
...
Erich
I'm not sure that we should store all modules in moddb by default.
There's no profit for this.
Sure there is, detecting dependencies is a fine thing at the first
installation or after a hardware
Hi Andrew
Am 16.08.2015 um 12:25 schrieb Andrew:
Hi.
16.08.2015 13:08, Erich Titl пишет:
Hi Andrew
Am 15.08.2015 um 17:50 schrieb Andrew:
After month of uptime I've got crash again. It seems that crash caused
by shaper removing from ppp interface before it's deletion, which causes
race
Hi Andrew
Am 16.08.2015 um 13:10 schrieb Andrew:
16.08.2015 13:30, Erich Titl пишет:
Or maybe a bug in accel-ppp?
Userland tools shouldn't cause kernel errors - so it's kernel bug.
I don't know how one can accelerate ppp without plugging into the kernel
interface, maybe with ioctl
Hi Andrew
Am 15.08.2015 um 17:50 schrieb Andrew:
After month of uptime I've got crash again. It seems that crash caused
by shaper removing from ppp interface before it's deletion, which causes
race condition.
Why does this cause a race condition? You should be able to remove a
shaper without
Gr.
So we have to update leaf.cfg, what a shame.
It should be automatic.
cheers
ET
Weitergeleitete Nachricht
Betreff:[leaf:bering-uclibc] 2 new commits to Bering-uClibc
Datum: Sun, 16 Aug 2015 15:06:49 +
Von:LEAF Linux Embedded Appliance Framework
Hi Andrew
Am 19.08.2015 um 16:06 schrieb Andrew:
...
How does buildtool know what the end user wants?
Kernel modules that are provided via .lrp packages (in our case -
initmod)
I agree, I would even remove them from /lib/modules when loading moddb
and/or autoprobing. To me stuff in initmod
601 - 700 of 880 matches
Mail list logo