Yeah I think that wants a RW filesystem, haha.
My apologies in advance to all other ramips devices, then, I’ll add an
init.d script for
./target/linux/ramips/base-files/etc/init.d/S99bootcount to do this.
On Tue, Apr 26, 2016 at 2:35 PM, John Crispin wrote:
> ok, this might not work, just move i
ok, this might not work, just move it to the init.d folder
On 26/04/2016 08:31, Andrew Yong wrote:
> Alright.
>
> Is it acceptable to call `. /lib/functions/uci-defaults.sh` in a preinit file?
>
> If it does not violate the sanctity of any OpenWrt init conventions,
> I’ll do that and test drive
Alright.
Is it acceptable to call `. /lib/functions/uci-defaults.sh` in a preinit file?
If it does not violate the sanctity of any OpenWrt init conventions,
I’ll do that and test drive it.
On Tue, Apr 26, 2016 at 2:06 PM, John Crispin wrote:
>
>
> On 26/04/2016 03:24, Andrew Yong wrote:
>> "alt
This reverts commit 0fd48fd7dab066db8a46e51775579c9ef507ca25.
We are facing some problem on rolling out the current 15.05-branch to
our devices because the image name changed.
The changed image name breaks compatibility for derived projects and
that's something which should only happen if there i
On 26/04/2016 03:24, Andrew Yong wrote:
> "alternatively call uboot-env directly without the fw_* abstraction layer"
>
> Pardon the obtuse question but how would I go about doing this? It
> sounds like the cleanest way to do things.
>
> I'd like to avoid calling uci-default earlier than it need
Dear Colleagues,
I have a new board with Mikrotik OS loaded in it and I want to load OpenWRT
for my Wireless Project. In fact, I want to use it for my TV White Space
Project in Tanzania due to its cost-benefit as compared to available TVWS
Radio Devices available in the market.
I am following thi
Thanks Jo,
Do you have any idea how it would be possible to reduce upstream opkg
size? Create
a libarchive-mini?
Maybe we can simply revert libarchive adoption (specially if we turn it
into a ./configure option). I guess it is limited to libopkg/pkg_extract.c
file.
Maybe it is better to keep an o
On 25 Apr 2016 1:33 pm, "Jo-Philipp Wich" wrote:
>
> Hi,
>
> this should be handled already, configure uses AC_HEADER_MAJOR and
> mountlist.c does:
>
> #ifdef MAJOR_IN_SYSMACROS
> #include
> #define HAVE_MAJOR
> ...
>
> did you patched findutils proactively or was there an actual build issue?
I
25.04.2016, 19:11, "John Crispin" :
> Hi,
>
> i dont see why this is needed.the patch in the tree does not look broken
> and seems to compile fine.
>
> John
>
Because fixed in changeset #49221
---
serge
___
openwrt-devel mailing list
openwrt-de
"alternatively call uboot-env directly without the fw_* abstraction layer"
Pardon the obtuse question but how would I go about doing this? It
sounds like the cleanest way to do things.
I'd like to avoid calling uci-default earlier than it needs to be, and
potentially messing up other things in th
This gadget driver allow you to debug other devices via EHCI Debug Port
capability.
Signed-off-by: Alexander Couzens
---
package/kernel/linux/modules/usb.mk | 17 +
1 file changed, 17 insertions(+)
diff --git a/package/kernel/linux/modules/usb.mk
b/package/kernel/linux/modules/
Signed-off-by: Alexander Couzens
---
target/linux/omap/Makefile | 2 +-
target/linux/omap/config-4.4 | 525 ++-
2 files changed, 171 insertions(+), 356 deletions(-)
diff --git a/target/linux/omap/Makefile b/target/linux/omap/Makefile
index 20f7517..d70
spi-omap-24xx can be found in TI based SoC like the beaglebone black
Signed-off-by: Alexander Couzens
---
package/kernel/linux/modules/spi.mk | 17 +
1 file changed, 17 insertions(+)
diff --git a/package/kernel/linux/modules/spi.mk
b/package/kernel/linux/modules/spi.mk
index 1c
Signed-off-by: Alexander Couzens
---
target/linux/omap/image/Config.in | 5
target/linux/omap/image/Makefile | 34 +-
target/linux/omap/image/gen_omap_sdcard_img.sh | 33 +
target/linux/omap/image/uEnv.txt
I'm using the beaglebone black as development tool for coreboot.
The bbb can receive debug output via ehci debug and flash
spi chips via spi-omap-24xx.
I'm not quite sure about the config-4.4 config. I copied over the omap
config-4.1
and did make kernel_oldconfig. Is there a better way to create t
Backport of r47911, r47912, r47969 to CC.
All other required patches have already been backported in r48822.
Signed-off-by: Vittorio Gambaletta
---
diff --git a/target/linux/lantiq/base-files/etc/uci-defaults/01_leds
b/target/linux/lantiq/base-files/etc/uci-defaults/01_leds
index 6956d61..5643
Backport of r47914 to CC.
Without this patch the TDW8970 is left without any status LED at all,
even during preinit and failsafe.
Signed-off-by: Vittorio Gambaletta
---
diff --git a/target/linux/lantiq/base-files/etc/diag.sh
b/target/linux/lantiq/base-files/etc/diag.sh
index 26f0a3c..57a7055 1
https://wiki.openwrt.org/toh/omylink/omy-g1
http://www.omylink.com/
Signed-off-by: L. D. Pinney
---
v2 corrects the Device name in Subject
v3 removes changes to mktplinkfw.c
target/linux/ar71xx/base-files/etc/board.d/01_leds| 6
target/linux/ar71xx/base-files/etc/board.d/02_ne
Scanning appears to work ( iw dev wlan1 scan ) ,
As soon as I try to use hostapd to run as an AP I get:
[20475.14] ath10k_pci :00:00.0: firmware crashed! (uuid
b826d45e-e51f-4beb-a489-6b7a4d5c6f76)
[20475.15] ath10k_pci :00:00.0: qca988x hw2.0 (0x4100016d,
0x004000ff sub :
On 25/04/2016 23:10:22 CEST, Martin Blumenstingl wrote:
maybe you could document this in the wiki: [0]
Most people check the wiki when they are looking for configuration
hints like this
Done: https://wiki.openwrt.org/toh/tp-link/td-w8970#leds
Cheers,
Vittorio
__
On 04/24/2016 05:53 PM, Mathias Kresin wrote:
> LZMA compress the kernel without dictionary, otherwise brnboot fails to
> uncompress the kernel.
>
> The filesystem parameter of mkbrncmdline was dropped since the used
> filesystem isn't exported in the kernel build step and at least in my
> tests e
On Mon, Apr 25, 2016 at 5:38 PM, Ben Mulvihill wrote:
>> @Ben: one probably has to enable caldata swapping in the .dts.
>
> I saw your comment on openwrt.ebilan.co.uk that the "ath,eep-swap;"
> flag should be added to the dts in order to use unmodified calibration
> data, but actually this does no
On Mon, Apr 25, 2016 at 8:51 PM, Vittorio G (VittGam)
wrote:
> For the record, if anyone wants the old behaviour they just need to put
> this in /etc/config/system:
>
> config led 'led_system'
> option name 'System'
> option sysfs 'tdw89x0:green:wps'
> option default '1'
>
Hello,
I've seen that Lantiq XRX200 trunk images are only being built for the VG3503J.
Is this intended, or a mistake?
Thank you.
Cheers,
Vittorio
___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailma
Would it be worthwhile keeping these userids in a wiki registry
page maybe? This is a current extract from the packages feed and
the base packages.
22:sshd
22:sshd
51:pulse
52:tor
53:zabbix
54:prosody
56:dmapd
57:bind
59:dovecot
72:ocserv
87:icecast
88:pianod
89:upmpdcli
99:_fdm
105:avahi
121:lld
Hi
On 10/04/2016 14:00, Jens Steinhauser wrote:
> This patch adds support for the TP-Link TL-WR810N.
> https://wiki.openwrt.org/toh/tp-link/tl-wr810n
>
> The device has a slide switch to select its mode of operation when using
> the stock firmware. After looking at how it's implemented for simila
Hello,
On Monday, April 25, 2016 10:53:41 AM David Hutchison wrote:
> So with some modifications to pci.c, hw.h and core.c I was able to get
> the radio initialized! :)
Hey, that's nice! Can you make and post a patch for that?
I'm sure if it's just a matter of adding the new pci and chip
ids to
On 25/04/2016 15:14:31 CEST, John Crispin wrote:
i understand the problem, but the WPS led is for WPS and should be left
free for the WPS feature eventually. so nack on this one i am afraid.
Ok, thank you anyway.
For the record, if anyone wants the old behaviour they just need to put
this in /
Running dnsmasq in a dedicated user/group allows matching its outgoing
traffic more easily using iptables' owner match.
Add UID/GID to the package metadata and append the user/group
parameters to the init script.
Signed-off-by: Daniel Golle
---
package/network/services/dnsmasq/Makefile
Hi John!
On Mon, Apr 25, 2016 at 06:28:47PM +0200, John Crispin wrote:
>
>
> On 22/04/2016 06:26, Daniel Golle wrote:
> > Running dnsmasq in a dedicated user/group allows matching its outgoing
> > traffic more easily using iptables' owner match.
> > Add UID/GID to the package metadata and append
Hi,
this should be handled already, configure uses AC_HEADER_MAJOR and
mountlist.c does:
#ifdef MAJOR_IN_SYSMACROS
#include
#define HAVE_MAJOR
...
did you patched findutils proactively or was there an actual build issue?
~ Jo
___
openwrt-devel mailin
With the update to glibc 2.23, sys/types.h doesn't inherit
sys/sysmacros.h anymore. We need to include it manually.
Signed-off-by: Weedy
--- /dev/null 2016-04-18 05:53:30.273258717 -0400
+++ tools/squashfs4/patches/125-glibc-2.23-sysmacros.patch 2016-04-24
16:22:30.271759670 -0400
@@ -0,0
With the update to glibc 2.23, sys/types.h doesn't inherit
sys/sysmacros.h anymore. We need to include it manually.
Signed-off-by: Weedy
--- /dev/null 2016-04-18 05:53:30.273258717 -0400
+++ tools/squashfs/patches/125-glibc-2.23-sysmacros.patch 2016-04-24
16:16:39.524863687 -0400
@@ -0,0
With the update to glibc 2.23, sys/types.h doesn't inherit
sys/sysmacros.h anymore. We need to include it manually.
Signed-off-by: Weedy
--- /dev/null 2016-04-18 05:53:30.273258717 -0400
+++ tools/mtd-utils/patches/125-glibc-2.23-sysmacros.patch 2016-04-24
15:31:35.050480434 -0400
@@ -0,0
With the update to glibc 2.23, sys/types.h doesn't inherit
sys/sysmacros.h anymore. We need to include it manually.
Signed-off-by: Weedy
--- /dev/null 2016-04-18 05:53:30.273258717 -0400
+++ tools/findutils/patches/125-glibc-2.23-sysmacros.patch 2016-04-24
15:36:39.330172567 -0400
@@ -0,
With the update to glibc 2.23, sys/types.h doesn't inherit
sys/sysmacros.h anymore. We need to include it manually.
Signed-off-by: Weedy
--- /dev/null 2016-04-18 05:53:30.273258717 -0400
+++ tools/e2fsprogs/patches/125-glibc-2.23-sysmacros.patch 2016-04-24
14:58:32.704376584 -0400
@@ -0,
Hello,
So with some modifications to pci.c, hw.h and core.c I was able to get
the radio initialized! :)
pci.c: added QCA9887_DEVICE_ID, modified ath10k_pci_id_table and
ath10k_pci_supp_chips
core.c: Duplicated QCA988X entry in ath10k_hw_params_list and passed
0x4100016d as the ID ( left everythin
Hi,
no, I only cared about make-ext4fs as I maintain this fork.
~ Jo
___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
On 25 Apr 2016 09:09, "Jo-Philipp Wich" wrote:
>
> Hi!
>
> Thanks for the patch, I fixed the include issue directly in Git, so this
> patch can be dropped and make-ext4fs should be updated in tools/ to a
> newer version instead.
>
> ~ Jo
Have you fixed all the packages that need this?
Make-ext4f
>
> please us an id between 0-999. ideally check what debian uses.
>
For most system services, with a very small number exceptions, debian
auto-assigns id < 500(?), and which service gets id depends on order of
package installation (which can be hassle with when trying to do network
filesystems)
On 25/04/2016 18:01, Andrew Yong wrote:
> diff --git a/target/linux/ramips/base-files/lib/preinit/04_handle_bootcount
> b/target/linux/ramips/base-files/lib/preinit/04_handle_bootcount
> new file mode 100644
> index 000..570fe16
> --- /dev/null
> +++ b/target/linux/ramips/base-files/lib/prei
On 22/04/2016 06:26, Daniel Golle wrote:
> Running dnsmasq in a dedicated user/group allows matching its outgoing
> traffic more easily using iptables' owner match.
> Add UID/GID to the package metadata and append the user/group
> parameters to the init script.
>
> Signed-off-by: Daniel Golle
>
The only way to access the u-boot console on the BTHOMEHUBV3A
is by soldering a serial console onto some really tiny pads.
To enable installation without soldering it would be helpful
to be able to modify the uboot environment from within
the ramdisk image used as part of the installation process
Hopefully third time's the charm:
PATCHv1:
This patch adds support for the SamKnows version 8.0 Whitebox, built
around the MT7621 platform. 2.4GHz appears to be working, albeit
poorly; 5GHz not working yet.
PATCHv2:
Fixed LED name in DTS.
PATCHv3:
DTS: Syntax error fixed, LEDs and buttons mappe
The dwc2 USB driver now works on Lantiq ar9, but the
default is still ltq-hcd-ar9. This patch switches to
dwc2 by default.
v2: all ar9 boards (v1 was just for BTHOMEHUBV3A and WBMR)
Signed-off-by: Ben Mulvihill
---
diff -uprN a/target/linux/lantiq/xway/profiles/avm.mk
b/target/linux/lantiq/xway
Working on my email client choice.
Anyway, preinit seems to be before fw_setenv gets its bearings:
Board is SamKnows Whitebox 8, resetting bootcount environment variable...
Cannot parse config file: No such file or directory
Error: environment not initialized
Router still soft bricks after 3 rebo
i'm afraid they are both badly formatted
On 25/04/2016 16:51, Andrew Yong wrote:
> PATCHv1:
> This patch adds support for the SamKnows version 8.0 Whitebox, built
> around the MT7621 platform. 2.4GHz appears to be working, albeit
> poorly; 5GHz not working yet.
>
> PATCHv2:
> Fixed LED name in
On 25/04/2016 17:38, Ben Mulvihill wrote:
> On Mon, 2016-04-25 at 16:29 +0200, Martin Blumenstingl wrote:
>> On Mon, Apr 25, 2016 at 3:00 PM, John Crispin wrote:
>>> is this not a bit risky ? normally we try to keep them read-only so that
>>> users cannot accidentally kill them. is the calib dat
On Mon, 2016-04-25 at 16:29 +0200, Martin Blumenstingl wrote:
> On Mon, Apr 25, 2016 at 3:00 PM, John Crispin wrote:
> > is this not a bit risky ? normally we try to keep them read-only so that
> > users cannot accidentally kill them. is the calib data normally not
> > located in the art partition
please send this as a PR against the luci tree on github
John
On 07/04/2016 18:22, Kevin O'Connor wrote:
> Add support for setting the "anonymous_identity" field on EAP type
> networks.
>
> Signed-off-by: Kevin O'Connor
> ---
>
> I sent a similar request to the openwrt mailing list tha
Christian Lamparter writes:
> On Sunday, April 24, 2016 01:16:11 AM David Hutchison wrote:
>> I spoke too soon:
>> https://pci-ids.ucw.cz/read/PC/168c/0050
>>
>> It is the QCA9887, so it's definitely on the PCIe bus and is being
>> seen.
>
> Does ath10k support the QCA9887? I see no entry for th
Hi,
comments inline
On 18/04/2016 12:05, ad...@yapic.net wrote:
> From: Sergey Sergeev
>
> kernel2minor tool is used for packing openwrt kernel into modified yaffs2
> file system image.
> This is necessary to boot via Routerboot on new Mikrotik NOR flash devices.
> ---
> config/Config-images.
Re-sending, newline issues (oops)
diff --git a/package/boot/uboot-envtools/files/ramips
b/package/boot/uboot-envtools/files/ramips
index a759bcc..9ad5974 100644
--- a/package/boot/uboot-envtools/files/ramips
+++ b/package/boot/uboot-envtools/files/ramips
@@ -24,7 +24,8 @@ linkits7688d | \
wsr-600
PATCHv1:
This patch adds support for the SamKnows version 8.0 Whitebox, built
around the MT7621 platform. 2.4GHz appears to be working, albeit
poorly; 5GHz not working yet.
PATCHv2:
Fixed LED name in DTS.
PATCHv3:
DTS: Syntax error fixed, LEDs and buttons mapped correctly now
- diag.sh updated t
On Mon, Apr 25, 2016 at 3:11 PM, John Crispin wrote:
> i dont see why this is needed.the patch in the tree does not look broken
> and seems to compile fine.
It seems that Hauke already pushed a similar fix:
https://dev.openwrt.org/changeset/49221/
___
op
On Mon, Apr 25, 2016 at 3:00 PM, John Crispin wrote:
> is this not a bit risky ? normally we try to keep them read-only so that
> users cannot accidentally kill them. is the calib data normally not
> located in the art partition or why do you need to copy it after flashing ?
I think the caldata pa
On 25/04/2016 15:32, Andrew Yong wrote:
> Stupid question - what calls these files? I assume I have to call my
> new function from somewhere.
>
> I'm working on it.
preinit gets called automagically during preinit/boot
>
> On Mon, Apr 25, 2016 at 9:22 PM, John Crispin wrote:
>>
>>
>> On 25/0
Stupid question - what calls these files? I assume I have to call my
new function from somewhere.
I'm working on it.
On Mon, Apr 25, 2016 at 9:22 PM, John Crispin wrote:
>
>
> On 25/04/2016 15:18, Andrew Yong wrote:
>> I have just found out that I need to perform this additional step on
>> this
On 21/04/2016 08:39, amine.ahd wrote:
> Hostapd allows more than one RADIUS server to be added in case the main
> server goes down.
> This allows netifd to load a list of RADIUS servers from the wireless confid
> file and add them to hostapd.
> The format of the list in the config file is as fo
yes please resend a new patch with all boards getting switched
John
On 25/04/2016 12:26, Ben Mulvihill wrote:
> The dwc2 USB driver now works on Lantiq ar9, but the
> default is still ltq-hcd-ar9. This patch switches to
> dwc2 by default, on the two boards I can test, namely
> BTHOMEHUBV3
On 25/04/2016 15:18, Andrew Yong wrote:
> I have just found out that I need to perform this additional step on
> this SamKnows Whitebox 8 every boot:
>
> /usr/sbin/fw_setenv bootcount 0
>
> Where would be an appropriate place to put this? I could add (yet
> another) init script in ./target/linu
I have just found out that I need to perform this additional step on this
SamKnows Whitebox 8 every boot:
/usr/sbin/fw_setenv bootcount 0
Where would be an appropriate place to put this? I could add (yet another)
init script in ./target/linux/ramips/base-files/etc/init.d with START=99
(this is ho
please avoid using slang words and add "scripts:" as a subject prefix
On 22/04/2016 11:06, Iblis Lin wrote:
> Signed-off-by: Iblis Lin
> ---
> scripts/brcmImage.pl | 2 +-
> scripts/deptest.sh| 2 +-
> scripts/feeds | 2 +-
> scripts/mkits.sh | 2 +-
> scripts
On 22/04/2016 09:13, Vittorio G (VittGam) wrote:
> On 22/04/2016 08:34:41 CEST, Mathias Kresin wrote:
>> Am 21.04.2016 um 23:43 schrieb Vittorio Gambaletta (VittGam):
>>> After recent patches, the status LED on this router would stay
>>> switched off after boot is completed. A switched off status
Hi,
i dont see why this is needed.the patch in the tree does not look broken
and seems to compile fine.
John
On 23/04/2016 05:55, Сергей Василюгин wrote:
> Sorry, fix bad patch (ticket #22248)
>
> Signed-Off-By: Serge Vasilugin
>
> diff --git
> a/target/linux/ramips/patches-4.4/0069-
Hi!
Thanks for the patch, I fixed the include issue directly in Git, so this
patch can be dropped and make-ext4fs should be updated in tools/ to a
newer version instead.
~ Jo
___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.o
On 25/04/2016 11:41, Weedy wrote:
> With the update to glibc 2.23, sys/types.h doesn't inherit
> sys/sysmacros.h anymore. We need to include it manually.
>
>
> --- /dev/null 2016-04-18 05:53:30.273258717 -0400
> +++ tools/make-ext4fs/patches/125-glibc-2.23-sysmacros.patch 2016-04-24
> 14:50:33
On 24/04/2016 21:29, Ben Mulvihill wrote:
> Remove read-only flag on two partitions on BTHOMEHUBV3A:
> uboot-config - otherwise fw_setenv command cannot be used.
> ath9k-cal- so that ath9k calibration data can be copied
> to the partition on a newly installed board.
>
>
Hi,
few comments inline. apart from those 2 nitpicks the patch looks good.
John
On 25/04/2016 03:05, Andrew Yong wrote:
> This patch adds support for the SamKnows version 8.0 Whitebox, built
> around the MT7621 platform. 2.4GHz appears to be working, albeit
> poorly; 5GHz not working yet
The dwc2 USB driver now works on Lantiq ar9, but the
default is still ltq-hcd-ar9. This patch switches to
dwc2 by default, on the two boards I can test, namely
BTHOMEHUBV3A and WBMR. John, if you are happy with the
principle of switching would you prefer me to send a patch
covering all the ar9 boar
With the update to glibc 2.23, sys/types.h doesn't inherit
sys/sysmacros.h anymore. We need to include it manually.
--- /dev/null 2016-04-18 05:53:30.273258717 -0400
+++ tools/make-ext4fs/patches/125-glibc-2.23-sysmacros.patch2016-04-24
14:50:33.932873523 -0400
@@ -0,0 +1,10 @@
+--- a/ext4_
Change mirror to Github (Gentoo repo) and drop the gzip compression.
Worst case there's about 4kbyte increase in size but most images ends up
beign somewhere between 4-100kbyte smaller due to the lzma compression.
Tested on ar71xx
v2 Now contains the correct patch, sorry!
v3 Turns out that the up
On 25/04/2016 08:54, Bastian Bittorf wrote:
> * John Crispin [25.04.2016 07:39]:
>>> The changed image name breaks compatibility for derived projects and
>>> that's something which should only happen if there is a really good
>>> reason (e.g. security fix).
>>
>> how does it beak compatibility ?
73 matches
Mail list logo
