On 05/06/2016 12:41 PM, Daniel Petre wrote:
Hello, so per jow's last night's info i have rsynced the downloads part
from LEDE to the RCS&RDS ( www.rcs-rds.ro ) mirroring server:
http://mirrors.linux.ro/lede/
Two more questions please:
1. How often can we rsync?
Seems i cannot resync so often
I did a pull, ./scripts/feeds update, make clean, and make V=s -j 1, and it
blows up
as below:
find /home/greearb/git/lede/build_dir/target-arm_cortex-a9+neon_musl-1.1.14_eabi/util-linux-2.28/ipkg-arm_cortex-a9_neon/dmesg -name 'CVS' -o -name '.svn' -o
-name '.#*' -o -name '*~'| xargs -r rm -rf
Hi,
After the last changes to the ncurses/ncursesw library I'm getting warnings
with defconfig and building the current git source repo.
tmp/.config-package.in:28040:error: recursive dependency detected!
tmp/.config-package.in:28040: symbol PACKAGE_libncursesw is selected by
PACKAGE_libncurses
The original iperf package is unmaintained. This switches to the "iperf2"
project on sourceforge, a fork that started where the previous iperf left
off.
Version 2.0.8 fixes the issue that patch 002 handled, so that can be dropped.
Due to a faulty check in configure.ac, this version needs _GNU_SOU
Signed-off-by: Bert Vermeulen
---
package/network/utils/iperf/Makefile | 40 +++---
...et-report-next-time-in-single-thread-mode.patch | 14
2 files changed, 4 insertions(+), 50 deletions(-)
delete mode 100644
package/network/utils/iperf/patches/001-set-re
Here is an initial pass at a script to gather pertinent info about
a LEDE (or similar) system. Suggestions for improvement are welcome.
If someone could add a way to put the git commit ID(s) into
the file system somewhere, I think that would be an excellent
addition.
And once the script is in p
> Bert Vermeulen writes:
> The patch made sure the ncursesw library was not selected to save
> space, but that library doesn't exist in this distribution at all.
This commit message is wrong.
--
Russell Senior, President
russ...@personaltelco.net
_
On Tue, 10 May 2016 12:54:58 +0200
John Crispin wrote:
> On 10/05/2016 11:39, Antony Pavlov wrote:
> > Hi All!
> >
> > I have just seen https://www.lede-project.org/todo.html
> >
> > There is the "Convert ar71xx to devicetree" entry.
> >
> > I'm working on this topic. You can see my patchseri
On 10.5.2016 15:58, Jo-Philipp Wich wrote:
Hi Hannu,
yes, this is caused by one of Felix quickfixes ;)
The buildbot did notice the error, see
http://downloads.lede-project.org/snapshots/faillogs/arm_cortex-a9_neon-vfpv4/packages/htop/compile.txt
Fixes are underway.
Thanks. The fixes made t
On 2016-05-10 6:14 AM, Jo-Philipp Wich wrote:
Hi,
I may be asking a stupid question, but why does the build system have to
build everything every time.
until recently this wasn't possible due to certain technical
limitations, some of which we already addressed [1].
Under OpenWRT a typical bu
On Tue, May 10, 2016 at 6:57 AM, Felix Fietkau wrote:
> On 2016-05-10 15:38, John Crispin wrote:
>>
>>
>> On 10/05/2016 15:35, Bert Vermeulen wrote:
>>> Hi all,
>>>
>>> The iperf package appears to be unmaintained, the maintainers directing
>>> people to iperf3 instead (see https://sourceforge.net
Function mtd_fixtrx was changed during trx improvements.
Signed-off-by: Rafał Miłecki
---
package/system/mtd/src/imagetag.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/package/system/mtd/src/imagetag.c
b/package/system/mtd/src/imagetag.c
index b850837..2ad2076 100644
On 05/10/2016 03:58 PM, Felix Fietkau wrote:
> I'd suggest switching the project to iperf2, keeping the package name
> and dropping the non-multi-threaded variant.
Ok, will do.
--
Bert Vermeulen
b...@biot.com
___
Lede-dev mailing list
Lede-dev@lists.
On 2016-05-10 15:35, Bert Vermeulen wrote:
> Hi all,
>
> The iperf package appears to be unmaintained, the maintainers directing
> people to iperf3 instead (see https://sourceforge.net/projects/iperf/).
>
> There is a fork of the original iperf package that's seeing some
> maintenance, named ip
On 10/05/2016 15:57, Felix Fietkau wrote:
> On 2016-05-10 15:38, John Crispin wrote:
>>
>>
>> On 10/05/2016 15:35, Bert Vermeulen wrote:
>>> Hi all,
>>>
>>> The iperf package appears to be unmaintained, the maintainers directing
>>> people to iperf3 instead (see https://sourceforge.net/projects/i
On 2016-05-10 15:38, John Crispin wrote:
>
>
> On 10/05/2016 15:35, Bert Vermeulen wrote:
>> Hi all,
>>
>> The iperf package appears to be unmaintained, the maintainers directing
>> people to iperf3 instead (see https://sourceforge.net/projects/iperf/).
>>
>> There is a fork of the original ipe
On 10/05/2016 15:35, Bert Vermeulen wrote:
> Hi all,
>
> The iperf package appears to be unmaintained, the maintainers directing
> people to iperf3 instead (see https://sourceforge.net/projects/iperf/).
>
> There is a fork of the original iperf package that's seeing some
> maintenance, named ip
Hi all,
The iperf package appears to be unmaintained, the maintainers directing
people to iperf3 instead (see https://sourceforge.net/projects/iperf/).
There is a fork of the original iperf package that's seeing some
maintenance, named iperf2: https://sourceforge.net/projects/iperf2/
Howeve
Hi Hannu,
yes, this is caused by one of Felix quickfixes ;)
The buildbot did notice the error, see
http://downloads.lede-project.org/snapshots/faillogs/arm_cortex-a9_neon-vfpv4/packages/htop/compile.txt
Fixes are underway.
~ Jo
___
Lede-dev mailing l
From: Josua Mayer
This change adds code to handle a new option on cmdline: overlay=
This is used to find the rootfs_data partition / disk when it has a
non-standard name or location.
It takes either the device node name, or the partition name:
i.e. /dev/mmcblk0p3, mmcblk0p3, rootfs_data
This op
On Tue, 10 May 2016, Jo-Philipp Wich wrote:
Hi,
I may be asking a stupid question, but why does the build system have to
build everything every time.
until recently this wasn't possible due to certain technical
limitations, some of which we already addressed [1].
Under OpenWRT a typical bu
On Tue, 10 May 2016, Bruno Randolf wrote:
On 10/05/16 09:33, Hannu Nyman wrote:
The contents of those files, the related config options and the sources
for each items might maybe be rationalised at the same time.
Good idea.
I think most of this can be beautifully handled by using git tags an
a) LEDE r146 build after "make clean"
b) ar71xx/WNDR3700 build. Only the device and htop selected for minimal config
c) htop does not build, it fails in the configure step due to missing libncurses
d) try building htop after make clean
e) nothing
f) LEDE buildbot has not noticed this failure
I no
Hi everybody,
Please take a look at this draft below. It attempts to solve the issue
of using an overlay device on regular block devices such as sata, emmc
or sdcards WITHOUT any unnecessary block2mtd layer in between.
I extended the find.method of ext4 code to include devname, because
partname i
From: Josua Mayer
This change adds code to handle a new option on cmdline: overlay=
This is used to find the rootfs_data partition / disk when it has a
non-standard name or location.
It takes either the device node name, or the partition name:
i.e. /dev/mmcblk0p3, mmcblk0p3, rootfs_data
This op
Den 2016-05-10 kl. 11:52, skrev Bruno Randolf:
>> On 10/05/16 09:33, Hannu Nyman wrote:
>> * branch builds after a release: Branch codename designation, last
>> > release number known + changes after it, source revision, opkg download
>> > from last release repo
> Would automatically happen as expl
On 10/05/2016 11:39, Antony Pavlov wrote:
> Hi All!
>
> I have just seen https://www.lede-project.org/todo.html
>
> There is the "Convert ar71xx to devicetree" entry.
>
> I'm working on this topic. You can see my patchseries
> in the mips-linux mailing list:
>
> https://www.linux-mips.org/a
Hi,
2016-05-10 12:14 GMT+02:00 Jo-Philipp Wich :
>
> Hi,
>
> > I may be asking a stupid question, but why does the build system have to
> > build everything every time.
>
> until recently this wasn't possible due to certain technical
> limitations, some of which we already addressed [1].
>
> > Und
Hi,
> I may be asking a stupid question, but why does the build system have to
> build everything every time.
until recently this wasn't possible due to certain technical
limitations, some of which we already addressed [1].
> Under OpenWRT a typical build will take 12 - 16+ hours, of which 10+
>
I may be asking a stupid question, but why does the build system have to
build everything every time.
Under OpenWRT a typical build will take 12 - 16+ hours, of which 10+
will be spent building the packages. very few of which will have changed.
Would it be possible to split the base system an
On 10/05/16 09:33, Hannu Nyman wrote:
> The contents of those files, the related config options and the sources
> for each items might maybe be rationalised at the same time.
Good idea.
I think most of this can be beautifully handled by using git tags and
"git describe". Basically the simple idea
Hi All!
I have just seen https://www.lede-project.org/todo.html
There is the "Convert ar71xx to devicetree" entry.
I'm working on this topic. You can see my patchseries
in the mips-linux mailing list:
https://www.linux-mips.org/archives/linux-mips/2016-03/msg00256.html).
The patchseries is a
The patch made sure the ncursesw library was not selected to save space,
but that library doesn't exist in this distribution at all.
Signed-off-by: Bert Vermeulen
---
package/network/utils/iftop/Makefile | 4 ++--
package/network/utils/iftop/patches/0001-force-ncurses.pa
On 8.5.2016 7:35, John Crispin wrote:
On 06/05/2016 02:16, l...@daniel.thecshore.com wrote:
package/base-files/files/etc/openwrt_release | 7 --
package/base-files/files/etc/openwrt_version | 1 -
these 2 files cannot simply be removed as that would break luci.
we need to eit
PR sent: https://github.com/lede-project/staging/pull/22
On Mon, May 9, 2016 at 1:28 PM, John Crispin wrote:
> Hi,
>
> as the files are already on github, fancy sending them as a PR via github ?
>
> John
>
> On 09/05/2016 07:18, Andrew Yong wrote:
>> The following changes since commit 91f
On 10 May 2016 at 09:15, Bjørn Mork wrote:
> Hauke Mehrtens writes:
>> On 05/09/2016 10:19 PM, Bjørn Mork wrote:
>>
>>> How do you ensure that they use the code only like you intended after you
>>> donated it?
>> This copyright notice is useless. The copyright will stay at the
>> individual auth
Hauke Mehrtens writes:
> On 05/09/2016 10:19 PM, Bjørn Mork wrote:
>
>> How do you ensure that they use the code only like you intended after you
>> donated it?
> This copyright notice is useless. The copyright will stay at the
> individual author. At least in Germany it would not be possible for
37 matches
Mail list logo
