Hello.
I've developed a daemon for monitoring network traffic. It is called yantd (Yet
Another Network Traffic Daemon) and is available at
http://github.com/joeroback/yantd
I would like to see yantd offered in the feeds/packages repository for the
OpenWrt community to use. The git repo contain
robust and allow the database to be stored in an arbitrary directory. If
no one has any complaints or other thoughts, I'll do just that and post some
patches in the near future.
Joe
On Jul 10, 2010, at 12:09 PM, Joseph Roback wrote:
> Hello.
>
> I've developed a daemon for
I am experiencing the same errors. It happens to me even with `make -j1'. The
miniupnpd package is being built before the iptables package, which miniupnpd
depends on. It uses header files from the iptables package.
The miniupnpd Makefile tries to set CFLAGS to
CFLAGS="... -I$(firstword $(wildc
/libtool compile
make[2] -C package/iptables compile
make[2] -C feeds/packages/net/miniupnpd compile
The problem, repeatable every build, only seems to occur after `make clean &&
make' of the entire trunk
Joe
On Jul 25, 2010, at 9:18 PM, Joseph Roback wrote:
> I am experiencing
, at 10:01 PM, Jim Henderson wrote:
> On Sun, 25 Jul 2010 21:44:52 -0600, Joseph Roback wrote:
>
>> Also interesting, if invoked to build miniupnpd manually, the iptables
>> dependency seems to be honored.
>>
>> $ make package/feeds/packages/miniupnpd-compile Collecting
t I am not sure.
Joe
On Jul 25, 2010, at 11:26 PM, Jim Henderson wrote:
> On Sun, 25 Jul 2010 22:05:12 -0600, Joseph Roback wrote:
>
>> I using r22388. After building miniupnpd manually and continuing the
>> build, uhttpd failed because its dependency openssl was there. I
>
The init.d script for netatalk is called afpd, the same as the binary daemon.
The stop() function of the init.d actually kills itself, along with the daemon
and the killall of cnid_metad never completes. The afpd daemon writes a pid
file and can be terminated using the pid in the pidfile instead
Seems people are filing tickets about it now...
https://dev.openwrt.org/ticket/7674
I had the same thing happen to me with uhttpd, (minus the lua stuff), where the
build system didn't build openssl first. It seems that the build goes straight
to 'install' targets, skipping the 'compile' targets
add a new cachesize option to /etc/config/dhcp which allows to set the
cachesize for dnsmasq
https://dev.openwrt.org/ticket/7625
Thanks!
___
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/mailman/listinfo/openwrt-
vnstat's makefile uses the '-D' option which is only part of GNU install, thus
compiling vnstat for OpenWrt on FreeBSD or Mac OS X fails because of the
missing option.
Attached is a patch to have the install target separately create the
destination directory before installing the vnstat.conf fi
Hi,
Currently the menuconfig option transmission-cli includes both the
transmission-cli binary and transmission-remote. The transmission-cli on the
order of 100k in size. This patch separates transmission-cli and
transmission-remote into two menuconfig options to have transmission shell
script
Hi,
I've modified 25 Makefiles to enable parallel builds using the
PKG_BUILD_PARALLEL setting. I was told the best way to submit these are to the
mailing list, so here are 25 Makefile patches. Most are one liners, some just
require the additional $(PKG_JOBS) added to the make call, other were a
Enable PKG_BUILD_PARALLEL in busybox. I'm seeing over a 50% reduction in
busybox build times on 8-core and 12-core build machines.
Signed-off-by: Joe Roback
---
Index: package/busybox/Makefile
===
--- package/busybox/Makefile(
Oh, sorry. Should I re-submit? Additionally, I submitted 25 other patches for
enabling parallel builds in an earlier mailing, should I re-submit them as well?
Joe
On Sep 6, 2010, at 2:31 PM, Felix Fietkau wrote:
> On 2010-09-06 9:47 PM, Joseph Roback wrote:
>> Enable PKG_BUILD_PA
n Sep 6, 2010, at 3:31 PM, Michael Büsch wrote:
> On Mon, 2010-09-06 at 15:24 -0600, Joseph Roback wrote:
>> Oh, sorry. Should I re-submit? Additionally, I submitted 25 other patches
>> for enabling parallel builds in an earlier mailing, should I re-submit them
>> as well?
>
&
26 Makefile patches for enabling parallel builds. Tested on Linux (8-core) and
Darwin (12-core), individually and from clean build_root. The first set is from
trunk, the second is from feeds/packages.
Signed-off-by: Joe Roback
---
++
| trunk/packages |
++
Ind
Subject says it all. Compile tested on Linux and Darwin hosts.
Signed-off-by: Joe Roback
---
Index: feeds/packages/libs/uclibc++/Makefile
===
--- feeds/packages/libs/uclibc++/Makefile (revision 22968)
+++ feeds/packages/libs/
I cannot this to apply.. Is it because of the spaces instead of tabs?
$ patch -p0 -i dir825-led-fix.diff
patching file target/linux/ar71xx/files/include/linux/ath9k_platform.h
patch: malformed patch at line 10: };
On Sep 7, 2010, at 11:38 PM, Leonid Protasov wrote:
> Signed-off-by: Leonid
iproute2 tc/Makefile does not define proper dependencies for the Bison/Flex
files used. Therefore parallel builds are broken.
It also contains a 3 year old patch for broken flex implementation. Is this
still broken? Removing this patch worked for me on Ubuntu 10.04 and Darwin
10.6.4.
The follo
Thanks Mike. Both of your patches worked well for me. Run .27 since you
mailing, running .28 right now.
Joe
On Oct 22, 2010, at 6:31 AM, Mike Brady wrote:
> This updates the avahi package from 0.6.25 to 0.6.28, which, according
> to the avahi developers, incorporates bugfixes and fixes minor se
For parallel builds, make targets that build need to be separated from targets
that install ;-)
Signed-off-by: Joe Roback
---
Index: feeds/packages/libs/jpeg/Makefile
===
--- feeds/packages/libs/jpeg/Makefile (revision 24262)
+
21 matches
Mail list logo
