I looked into buildbot snapshots and noticed that some platforms have not
successfully compiled for a long time. One example is at91, which seems to be
missing device-related kernel config symbols and the snapshot build crashes
already there.
Please find attached a patch that makes the at91
On 04/15/2012 10:59 AM, Hannu Nyman wrote:
I looked into buildbot snapshots and noticed that some platforms have
not successfully compiled for a long time. One example is at91, which
seems to be missing device-related kernel config symbols and the
snapshot build crashes already there.
During the long pause in buildbot snapshots last month I got interested in
the snapshot builds.
I noticed that some of the platforms are being built daily although they have
not completed the build process successfully for 1-2 years, in some cases
never. It seems like a waste of CPU power and
Hi Hannu
On 04/15/2012 12:26 PM, Hannu Nyman wrote:
During the long pause in buildbot snapshots last month I got interested
in the snapshot builds.
I noticed that some of the platforms are being built daily although they
have not completed the build process successfully for 1-2 years, in
On 15.4.2012 18:27, Hauke Mehrtens wrote:
2 Never:
These have never completed a build succesfully, or at least it has not
been uploaded to the snapshot server:
ps3 kernel config symbols missing
ubicom32 ERROR: Missing site config for target
ubicom32-openwrt-linux-uclibc
These targets are
Ubicom32 and ps3 have been removed from the buildbot list. It would be nice
to see the other platforms fixed, maybe with the buildbot threads that have
been floating around the list as of late will help spur people with these
platforms to submit patches.
Travis
On Sun, Apr 15, 2012 at 11:11 AM,
my wr941nd uses a MARVELL 88e6063 Swtich and I have these lines in
backfire/build_dir/linux-ar71xx/linux-2.6.32.27/.config in r31251
seems it should support dot1q and vlaning as well
CONFIG_NET_DSA=y
# CONFIG_NET_DSA_TAG_DSA is not set
# CONFIG_NET_DSA_TAG_EDSA is not set
Signed-off-by: David Woodhouse david.woodho...@intel.com
---
Updated patch with the final commit ID now it's been merged by DaveM.
.../patches-3.2/130-pppoatm-queue-depth.patch | 193
.../patches-3.3/130-pppoatm-queue-depth.patch | 193
2
Signed-off-by: David Woodhouse david.woodho...@intel.com
---
Too slow; my router is running 3.3.2 now...
include/kernel-version.mk |3 +++
target/linux/x86/geos/target.mk |2 +-
2 files changed, 4 insertions(+), 1 deletions(-)
diff --git a/include/kernel-version.mk
When I try to compile, I get the following error:
--- snip ---
ipnetns.o: In function `netns_exec':
ipnetns.c:(.text+0xfc): undefined reference to `unshare'
ipnetns.o: In function `netns_add':
ipnetns.c:(.text+0x54c): undefined reference to `unshare'
collect2: ld returned 1 exit status
make[5]:
On 2012-04-15 11:53 PM, Jim Henderson wrote:
When I try to compile, I get the following error:
--- snip ---
ipnetns.o: In function `netns_exec':
ipnetns.c:(.text+0xfc): undefined reference to `unshare'
ipnetns.o: In function `netns_add':
ipnetns.c:(.text+0x54c): undefined reference to
On Sun, 15 Apr 2012 23:59:15 +0200, Felix Fietkau wrote:
OpenWrt has been using uClibc 0.9.33 for a while now, whereas you seem
to be using 0.9.32. It looks like 0.9.33 provides that missing function.
Awesome, thanks - seems the last time uClibc was updated I had a similar
issue, and had
On 2012-04-16 12:10 AM, Jim Henderson wrote:
On Sun, 15 Apr 2012 23:59:15 +0200, Felix Fietkau wrote:
OpenWrt has been using uClibc 0.9.33 for a while now, whereas you seem
to be using 0.9.32. It looks like 0.9.33 provides that missing function.
Awesome, thanks - seems the last time uClibc
I am curious as I was not even aware that you could tell the compile process to
use one uClib erosion or another. I thought that was fixed through the
toolchain compile process. How do you get the buildroot process to use a
certain uClib version then?
Kind Regards
Hanno Schupp
Connect
Attached the patch to fix support for the D-Link DVA-G3810BN, generated
by quilt
Signed-Off-By: Jeroen van Bemmel jbem...@zonnet.nl
--- a/arch/mips/bcm63xx/boards/board_bcm963xx.c
+++ b/arch/mips/bcm63xx/boards/board_bcm963xx.c
@@ -2137,6 +2137,71 @@ static struct board_info __initdata boar
Attached patch fixes a typo in the image Makefile, and overrides the
board name for DVA-G3810BN to set the correct board properties
Signed-Off-By: Jeroen van Bemmel jbem...@zonnet.nl
Index: Makefile
===
--- Makefile(revision
On Mon, 16 Apr 2012 11:10:32 +1200, Hanno Schupp wrote:
I am curious as I was not even aware that you could tell the compile
process to use one uClib erosion or another. I thought that was fixed
through the toolchain compile process. How do you get the buildroot
process to use a certain
17 matches
Mail list logo
