Clarification: ar71xx is fully tested, I've build-tested some others,
but other than ar71xx I have no means to fully verify builds.
Regards,
Daniel
On 16/02/16 02:47 AM, open...@daniel.thecshore.com wrote:
No comment on previous send of this version, tested and working, hence resend.
This pa
From: Daniel Dickinson
Minor fixes so that that kirkwood acts correctly in the
new multi-profile build machinery/menuconfig (allowing multi-profile
selection).
Signed-off-by: Daniel Dickinson
---
target/linux/kirkwood/Makefile| 1 +
target/linux/kirkwood/image/Makefile
From: Daniel Dickinson
Minor fixes so that the 'Default' profile (build all images profile)
not only builds all images but acts correctly in the
new multi-profile build machinery/menuconfig.
Signed-off-by: Daniel Dickinson
---
target/linux/brcm63xx/generic/target.mk | 1 +
target/linux/
From: Daniel Dickinson
Minor fixes so that the 'Generic' profile (build all images profile)
not only builds all images but acts correctly in the
new multi-profile build machinery/menuconfig.
Signed-off-by: Daniel Dickinson
---
target/linux/adm5120/image/Makefile| 4 ++--
t
From: Daniel Dickinson
Minor fixes so that the 'Default' profile (build all images profile)
not only builds all images but acts correctly in the
new multi-profile build machinery/menuconfig
Signed-off-by: Daniel Dickinson
---
target/linux/mvebu/Makefile| 1 +
target/linux/mveb
From: Daniel Dickinson
ar71xx has a large number of images, which makes it especially helpful
to get building a selection of a subset of images working for this target,
therefore this patches enables the multi-profile select capability from
the previous patch for ar71xx.
Signed-off-by: Daniel Di
From: Daniel Dickinson
Certain platforms have large numbers of possible images, and it can be
desirable to build neither all images nor only a single image,
therefore this patch makes selecting target profiles a menu instead of a
single choice, which allows the user to build a specific subset of
From: Daniel Dickinson
Minor fixes so that the former 'Default' profile (now All)
not only builds all profiles but acts correctly in the
new multi-profile build machinery/menuconfig
Signed-off-by: Daniel Dickinson
---
target/linux/ramips/image/Makefile | 7 +--
target/linu
From: Daniel Dickinson
Minor fixes so that the 'Default' profile (build all images profiles)
not only builds all profiles but acts correctly in the
new multi-profile build machinery/menuconfig.
Signed-off-by: Daniel Dickinson
---
target/linux/at91/image/Makefile| 6 +-
tar
No comment on previous send of this version, tested and working, hence resend.
This patch series allows to select to build a subset of the images for a
(sub)target
by allowing to generate a profile that is composed of a selected subset of the
list
of profiles for the (sub)target. This allows to
From: Daniel Dickinson
No comment on previous send of this version, tested and working, hence resend.
Failsafe mode while convenient for development and experimentation
can reasonbly considered a major security loophole (by giving an
easy way to gain passwordless root access to the device), ther
From: Daniel Dickinson
No comment on previous send of this version. Tested and working, hence resend.
Passwordless root login is undesirable by default
on any platform, therefore make requiring a login to
gain root (or any other user) even on hardware console
the default. This is an opt-out op
From: Daniel Dickinson
No comment on previous send of this verison. Tested and working, hence resend.
If packages have already been built by the base system we don't want
a different version in another repo (e.g. built by SDK) because this
could cause conflicts with packaes that depend on it.
From: Daniel Dickinson
Second part to patch series with no comment to previous send.
The patch for preventing SDK from reconfiguring already built packages
uses SRC_PKG_NAME for source package name which conflicts with use of
that variable in polarssl Makefile, so fix polarssl variable be the
mo
From: Daniel Dickinson
Sometimes you wan to enable feeds in the SDK that you left disabled
in the base build. This patch makes menuconfig feed configuration
possible from SDK.
Signed-off-by: Daniel Dickinson
---
target/sdk/files/Config.in | 4
1 file changed, 4 insertions(+)
diff --git
From: Daniel Dickinson
Abstract out the signing-key from base-files (to make duplicating
for use for building different key for SDK easier; the SDK needs to
build a signing-key by default when base-files has a signing key
due to the fact opkg will not accept unsigned package lists when
configured
From: Daniel Dickinson
Because opkg rejects package lists that are not signed
of use of signed lists is enabled, make the default for
signing SDK packages the same as for the base build (that
is is base was signed, also sign SDK, only with a new key).
To facilitate this we duplicate the and rena
I don't remember if I've sent this patch before but I have been using it
succesfully.
Basically split signing-key out of base-files so that it can be copied
into signing-key-sdk in the SDK and make signing packages the default for
SDK built packages as well because opkg rejects unsigned packages
On Tue, Feb 16, 2016 at 8:06 AM, Martin Schiller wrote:
> Hi!
>
> I need to set up the xrx200 switch with one LAN switch part (phy2-phy5) and
> one WAN port (phy1).
>
> The problem is, that I can see the LAN traffic on the WAN port and vice versa.
>
> It doesn't also matter if SW_PORTMAP is defin
From: Daniel Dickinson
No comment on previous send of this patch, hence resent.
We add an 'httpauth' section type that contains the options:
instance: Which uhttpd config section the auth is associated with
prefix: What virtual or real URL is being protected
username: The username for the Basic
Felix didn't like my previous approach which was an lxc-specific tarball
generation.
Do either you have an actual suggestion of what you'd like to see, so
that I can do something that will keep you both happy?
Regards,
Daniel
On 16/02/16 02:10 AM, John Crispin wrote:
i think having lxc spe
Hi,
patch is whitespace broken. please resubmit a fixed version.
John
On 15/02/2016 22:31, Frédéric Basse wrote:
> Enable port 6 of AR8327 switch on DHP-1565 board to allow
> communication between CPU & AR7400 PLC chipset.
>
> Signed-off-by: Frédéric Basse
> ---
> target/linux/ar71xx/
i think having lxc specific hooks sprayed over a pile a scripts is the
wrong approach, so NAK on this one.
John
On 16/02/2016 08:03, open...@daniel.thecshore.com wrote:
> From: Daniel Dickinson
>
> I have a patch that it will be some time before I personally will have some
> time
> to
Hi!
I need to set up the xrx200 switch with one LAN switch part (phy2-phy5) and one
WAN port (phy1).
The problem is, that I can see the LAN traffic on the WAN port and vice versa.
It doesn't also matter if SW_PORTMAP is defined or not.
I think it maybe has something to do with the PCE_PMAPx se
From: Daniel Dickinson
I have a patch that it will be some time before I personally will have some time
to test but have noted that there are others interested in LXC on OpenWrt and
therefore am tossing out this untested patch in case someone is interested in
trying it out.
Basically you can use
Hi,
As I still don't have working patchwork and the patchwork admin hasn't
gotten back to me on the issue, could you drop the patches from
patchwork for removing using of ifconfig/route in favour of ip.
I looked at the actual size of ifconfig/route (<5k) and concluded that
for the amount of
Hi,
Is anyone working on the port of Intel Edison on OpenWRT? And what is
the status?
Cheers,
--
Baptiste
___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Enable port 6 of AR8327 switch on DHP-1565 board to allow
communication between CPU & AR7400 PLC chipset.
Signed-off-by: Frédéric Basse
---
target/linux/ar71xx/files/arch/mips/ath79/mach-dhp-1565-a1.c | 16
1 file changed, 16 insertions(+)
diff --git a/target/linux/ar71xx/files
The sysfixtime script can cause ntpd to stop setting managing the time.
This patch ensures that ntpd is restarted after the time is changed by
sysfixtime.
---
package/base-files/files/etc/init.d/sysfixtime | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/package/base-files
OK, cool. In future then, should I email yourself with patches and CC
openwrt-devel as well?
On Sat, 2016-02-13 at 02:48 +0100, pau wrote:
> Hey I just said this is how we normally work. If you don't want to use
> github, I (or other people from openwrt-routing) can apply your patch. I
> never
Hi!
Am 12.02.2016 um 09:27 schrieb John Crispin:
> Hi,
>
>
>
> On 30/01/2016 23:15, André Valentin wrote:
>> This dependency is needed to be able to create ipsec xfrm states.
>>
>> Signed-off-by: André Valentin
>> ---
>> package/kernel/linux/modules/netsupport.mk | 2 +-
>> 1 file changed, 1 i
Prevent external routes and address being added or deleted when changing
the state of the interface ip settings
Signed-off-by: Hans Dedecker
---
interface-ip.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/interface-ip.c b/interface-ip.c
index eb585f5..26a2865 100644
--- a/interface-
Hi,
On 02/13/2016 03:47 PM, Mauro M. wrote:
> Hello,
>
> I am new to this list and this is my first post.
This was a replay to the " downloads.openwrt.org NOT accessible!!!"
thread and your mail is totally unrelated to that. When you want to
start a new thread next time just send a new mail to t
On 02/15/2016 03:29 PM, Mauro M. wrote:
> Hello,
>
> I would like to get in touch with developers who currently or previously
> worked on the WNDR4500 router.
>
> I would like to collaborate to port OpenWRT to this model.
>
> Thank you.
Hi,
This dpends on the revision of the device.
Version 1
Hello,
I would like to get in touch with developers who currently or previously
worked on the WNDR4500 router.
I would like to collaborate to port OpenWRT to this model.
Thank you.
___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://
On 2016-02-15 12:54, Eyal Birger wrote:
> > > if (offset < sizeof(ub->hdr)) {
> > > - iov[0].iov_base = ((char *) &ub->hdr) + offset;
> > > - iov[0].iov_len = sizeof(ub->hdr) - offset;
> > > + struct ubus_msghdr hdr;
> >
Hi Felix,
Thanks for the quick reply! (sorry for the previous top post...)
On Mon, Feb 15, 2016 at 1:24 PM Felix Fietkau wrote:
> On 2016-02-15 12:19, Eyal Birger wrote:
> > Hi Felix,
> >
> > Thanks for your review!
> >
> > If I understood the code correctly, the header fields are never accesse
Hi Felix,
Thanks for the quick reply!
Interesting... I initialize the iov_base every time to a newly created and
calculated hdr variable before the sendmsg() call, and iov is never used
otherwise - so I wonder how it could be reused in subsequent calls?
Eyal.
On Mon, Feb 15, 2016 at 1:24 PM Fel
On 2016-02-15 12:19, Eyal Birger wrote:
> Hi Felix,
>
> Thanks for your review!
>
> If I understood the code correctly, the header fields are never accessed
> before the header is fully received - at which point the conversion is done.
> My motivation for changing the fields only as they are sent
Hi Felix,
Thanks for your review!
If I understood the code correctly, the header fields are never accessed
before the header is fully received - at which point the conversion is done.
My motivation for changing the fields only as they are sent on the wire is
to avoid renaming the fields or changi
On 2016-02-15 05:09, Eyal Birger wrote:
> Changing the ubus message header fields from 'host' order to 'network' order
> allows passing ubus messages between hosts with different endianity.
>
> Example use (creating a ubus proxy):
>
> on host A (e.g. big endian router already running ubusd), run:
On Mon, 15 Feb 2016 09:27:04 +0100 Rafał Miłecki wrote:
> >> NAK, it is a general pattern of daemons to be installed in disabled
> >> state, to prevent a default config causing security or other issues.
> >
> > I understand it, but, for example, igmpproxy is not disabled by default.
> > Besides u
I discovered the issue is not with vlans but with the way I was routing
traffic from openwrt as a virtual router for virtual machines running on
the same laptop, for which I want routing out the appropriate vlan for
type of host (e.g. home vs. work) when connected to the local network,
and a co
On 14 February 2016 at 15:20, Nerijus Baliunas
wrote:
> On Sun, 14 Feb 2016 08:08:07 +0100 John Crispin wrote:
>
>> NAK, it is a general pattern of daemons to be installed in disabled
>> state, to prevent a default config causing security or other issues.
>
> I understand it, but, for example, ig
44 matches
Mail list logo
