Re: [OpenWrt-Devel] problem with uci delete

2016-07-26 Thread nam228

you can not delete an entire config at once by using uci
You can only delete one by one.For example:

uci delete network.lan.ifname

On 24/07/2016 16:56, Stephan Hennig wrote:

[moving to devel, since there has been no response on the users list]

Hi,

uci help has this on the delete command


root@uap:~# uci
Usage: uci []  []

Commands:

[...]

 delete [.[[.][=]]]

Which I interpret as: you can delete an entire config at once (and fill
it again with 'uci set' afterwards).  However, 'uci delete '
returns an error:

   root@uap:~# uci delete network
   uci: Invalid argument
   root@uap:~# uci changes
   root@uap:~#

(And 'uci commit' doesn't seem to do anything.)

Am I missing something?

system:

   OpenWrt 15.05.1 CC
   Nanostation 5 loco

Background: Task is to deploy WDS bridge configuration on devices via
shell script.  Instead of managing separate (network and other)
configuration files for access points and stations that share many
identical lines and are therefore hard to maintain, the idea is to
deploy a parametrized shell script which builds configuration files on
the device in question.  Now, so that existing configuration doesn't get
into the way, the script should be able to remove, i.e., network
configuration, at once and rebuild it using 'uci set' commands afterwards.

Best regards,
Stephan Hennig
___
openwrt-users mailing list
openwrt-us...@lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-users
___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel

___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel


Re: [OpenWrt-Devel] UART1 mt7688

2016-07-26 Thread Lazar Demin
Could you please elaborate on how you accomplished this, ie how to set
the clock in the dts?

On Wed, Apr 27, 2016 at 6:03 AM, Baptiste Clenet  wrote:
> Found the solution ! Need to add uart1 in dts and set the clock
>
> Le 27 avr. 2016 09:24, "Baptiste Clenet"  a écrit :
>>
>> Hi,
>>
>> I want to run an UART communication between the mt7688 and another
>> board, console uses UART0 but I want to configure UART1 to be able to
>> send data on it. Currently, if I send something on /dev/ttyS1, I get:
>> write to fd 4 failed: I/O error
>> or
>> ash: write error: I/O error
>>
>> root@OpenWrt:/# stty -F /dev/ttyS1
>> stty: /dev/ttyS1: I/O error
>>
>> Should I add something in the dts?
>>
>> Thanks
>>
>> --
>> Baptiste
>
>
> ___
> openwrt-devel mailing list
> openwrt-devel@lists.openwrt.org
> https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
>
___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel


[OpenWrt-Devel] Github and selfhosted divergent histories

2016-07-26 Thread Kevin Azzam
Hello folks,

I have some questions about the github move.

The git histories of the selfhosted repository [0] and that of the github
repository [1] are currently divergent (first commits are 753606a [2] and
1427738 [3] respectively -- same timestamp, different authors).

Moving forward, which repository should we take as "canon"? ie, which
repository
is it preferred to code against? The documentation refers to different
repos in
various locations: e.g. this [4] refers the self hosted url [0] but [5]
refers
to the github url [1] (but only for trunk, not for the other branches)

I'm maintaining a fork for personal use and I'd like to know if I should
cherrypick my changes unto the github upstream or not (otherwise rebasing
in the
future may be increasingly problematic -- right now, it would take quite a
while
to rebase and I think it would replay/reapply the whole history, i.e. each
commit would be there twice, so all my commits should be cherrypicked
instead)

Ideally, the history would not diverge, but I'm assuming there was a good
reason
to do so (cleaning up author names/emails + ...?).

Basically, I would like to know the following:

1. Should we indeed move to github and consider that as upstream from now
on?
2. Are there plans on merging the histories, or are you comfortable with the
fact that they diverge and intend on keeping it that way?
If it's a mistake, then I shouldn't "take the plunge" and cherrypick
against
github if that's indeed going to again change in the future.
3. Is there an, or can we have an "official stance" on this? I assume it
is/will
be confusing for many others as well, so some some sort of official
guideline
would be helpful.


Reference: Announcement of the github move:
https://lists.openwrt.org/pipermail/openwrt-devel/2016-June/041857.html
[0]: https://git.openwrt.org/openwrt.git
[1]: git://github.com/openwrt/openwrt.git (or
https://github.com/openwrt/openwrt.git)
[2]: https://git.openwrt.org/openwrt.git?p=openwrt.git;a=commit;h=753606a
[3]: https://github.com/openwrt/openwrt/commit/1427738
[4]: https://wiki.openwrt.org/doc/howto/buildroot.exigence
[5]: https://dev.openwrt.org/wiki/GetSource
___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel


Re: [OpenWrt-Devel] Create Avnet brand in the new data entry page

2016-07-26 Thread Zoltan HERPAI

fengling.qin wrote:

Hi,
I am starting develop with Avnet Zedboard, based Zynq-7000 ARM/FPGA Soc.

I found the code below was added but without documents,  so I wanna 
create a new
data entry page for documents, but without success. 
The page is /https://wiki.openwrt.org/meta/create_new_dataentry_page/


Can some some here add Avnet to the brand list ?

Attach some commits here:
https://github.com/openwrt/openwrt/commit/addbde1976b932f70cf3d2f682e75b55cb1db9e9#diff-a4a7c5b9de6e062ee563f05332a6063d
https://github.com/openwrt/openwrt/commit/24ef2c55861b889503decd6a45d281c0184d0eea#diff-a4a7c5b9de6e062ee563f05332a6063d

Hi Fengling,

Avnet is now added to the list.

Regards,
-w-
___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel