[OpenWrt-Devel] Regression: Trunk no longer runs on the WRT54G
I have a WRT54G v2 that can run my freshly built Backfire just fine. When I build trunk for it, it never comes up on the network after I flash the .bin. I think this is a regression. It's been known for a while that newer official builds with all the web interface stuff don't run well on the router, due to low memory, but it still would boot in, say, 12.04. Does anyone have a WRT54G and the necessary hardware to debug this over a serial connection? Would anyone want me to mail them one of my routers to work on this? -Adam Novak ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] Regression: Trunk no longer runs on the WRT54G
On 26.09.2013 09:08, Adam Novak wrote: > I have a WRT54G v2 that can run my freshly built Backfire just fine. > > When I build trunk for it, it never comes up on the network after I > flash the .bin. > > I think this is a regression. It's been known for a while that newer > official builds with all the web interface stuff don't run well on the > router, due to low memory, but it still would boot in, say, 12.04. > > Does anyone have a WRT54G and the necessary hardware to debug this > over a serial connection? Would anyone want me to mail them one of my > routers to work on this? > > -Adam Novak > ___ > openwrt-devel mailing list > openwrt-devel@lists.openwrt.org > https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel > i think that's related to the removed switch driver in kernel 3.10, see: https://lists.openwrt.org/pipermail/openwrt-devel/2013-July/020663.html He already received my router with that old switch chip smime.p7s Description: S/MIME Cryptographic Signature ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] Regression: Trunk no longer runs on the WRT54G
* Adam Novak [26.09.2013 09:31]: > Does anyone have a WRT54G and the necessary hardware to debug this > over a serial connection? Would anyone want me to mail them one of my > routers to work on this? will check today with serial console, thanks for reporting. bye, bastian ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] Regression: Trunk no longer runs on the WRT54G
Am 26.09.2013 09:34, schrieb Dirk Neukirchen: > On 26.09.2013 09:08, Adam Novak wrote: >> I have a WRT54G v2 that can run my freshly built Backfire just fine. > i think that's related to the removed switch driver in kernel 3.10, see: hi *, my asus wl-500g also not booting and the watchdog reset the device after 60sec. this model has no vlan switch, i think. christoph ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH 1/2] uboot-imx6: alphabetize wandboard variants
Hi Eric, On Wed, Sep 25, 2013 at 06:04:57PM -0700, Eric Nelson wrote: > Signed-off-by: Eric Nelson > --- > package/boot/uboot-imx6/Makefile | 12 ++-- > 1 file changed, 6 insertions(+), 6 deletions(-) Applied in r38203. Thanks! Luka ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH V2 2/2] uboot-imx6: add support for Nitrogen6x boards
Hi Eric, On Wed, Sep 25, 2013 at 06:08:48PM -0700, Eric Nelson wrote: > Signed-off-by: Eric Nelson > --- > V2 alphabetizes the list and fixes nitrogen6dl/dl2g. > > Depends on patch "uboot-imx6: alphabetize wandboard variants" > > Patch 1/2 was nacked, so no update will be sent. > > package/boot/uboot-imx6/Makefile | 30 ++ > 1 file changed, 30 insertions(+) Applied in r38204. Thanks! Luka ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH 1/2] U-Boot: Use git instead of FTP, consolidate download
Hi Eric, On Wed, Sep 25, 2013 at 05:33:28PM -0700, Eric Nelson wrote: > I wasn't suggesting anything different, and the patch includes the > same tag (v2013.07). > > Decisions about when to update are completely separate from the > protcol, right? Yes they are. But if ftp or http works fine I prefer to have that over git for packaging. Also, by using git we would always have here a red color and not preferable green: https://home.comcast.net/~sdwalker/uscan/index.html Yes, it's all about collors ;) > Is anyone else able to grab this tar-ball from Denx? > Is there/should there be a fallback at openwrt.org? I can only say that this must be a problem on your end... Anyway, Felix has pushed the binary on OpenWrt server. Please try it again and let me know if it still does not work for you. Luka ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
[OpenWrt-Devel] PPP unnumbered
Hello, I'm trying to add support for PPP "unnumbered" connections which are PPP connections which don't use a dedicated IP, instead an IP address of another local interface is "reused" on the PPP interface. Some Internet providers require this, since their PPPoE server pushes the network address when routing multiple IPs, and some customers want to use it to save IP addresses. Reference: http://www.sustworks.com/site/ipr_guide/unnumbered.html I have added an UCI option 'unnumbered' 'INTF' which means "reuse the IP address of INTF" (e.g. unnumbered LAN will put the IP of the LAN interface on the ppp interface as well), and the following script in /etc/ppp/ip-up.d/. It works, but it is not integrated with netifd, e.g. netifd does not know about the IP address we assigned. Any ideas how that could be implemented better and integrated with netifd? Thanks, bruno --- #!/bin/sh #pppoe-WAN eth0 0 192.168.71.5 10.0.0.1 WAN P_LOCAL=$4 P_REMOTE=$5 P_IFACE=$1 P_NET=$6 hook_unnumbered() { local cfg="$1" local proto ifname unnumbered config_get proto "$cfg" proto [ "$proto" = "pppoe" ] || return 0 config_get ifname "$cfg" ifname [ "x$P_IFACE" = "x$ifname" ] || return 0 config_get unnumbered "$cfg" unnumbered [ -n "$unnumbered" ] || return 0 # Get ipaddr and netmask from unnumbered interface. # And that interface must be static one. local proto ipaddr netmask config_get proto "$unnumbered" proto config_get ipaddr "$unnumbered" ipaddr config_get netmask "$unnumbered" netmask [ -n "$ipaddr" -a -n "$netmask" -a "x$proto" = "xstatic" ] || return 0 eval "$(ipcalc.sh $ipaddr $netmask)" # Validate netowrk address if [ "x$P_LOCAL" = "x$NETWORK" ]; then /sbin/ifconfig ${P_IFACE} "$ipaddr" pointopoint ${P_REMOTE} /sbin/route add default gw ${P_REMOTE} fi } . /etc/functions.sh include /lib/network scan_interfaces config_load network hook_unnumbered $P_NET --- ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH] Update wifidog to 20130917 version
On 09/17/2013 06:18 PM, Etienne CHAMPETIER wrote: This update: -adds a ProxyPort option (to use a local transparent proxy) -changes http return code to 302 (was 307, but wifidog speaks http 1.0, and we really need 303) (it was causing some trouble with, for exemple, opera) -switch to github repo -correct a small typo -add drop, log, ulog target for FirewallRule This is only compile tested, but 3 people are asking for it https://github.com/wifidog/wifidog-gateway/issues/8 +1 compiles fine on trunk r36139 i guess it would be a good idea to move this into http://github.com/openwrt-routing/ feed, for easier development? cheers! ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH 1/2] U-Boot: Use git instead of FTP, consolidate download
Thanks Luka, On 09/26/2013 02:05 AM, Luka Perkov wrote: Hi Eric, On Wed, Sep 25, 2013 at 05:33:28PM -0700, Eric Nelson wrote: I wasn't suggesting anything different, and the patch includes the same tag (v2013.07). Decisions about when to update are completely separate from the protcol, right? Yes they are. But if ftp or http works fine I prefer to have that over git for packaging. Also, by using git we would always have here a red color and not preferable green: https://home.comcast.net/~sdwalker/uscan/index.html Yes, it's all about collors ;) Like with my TiVo, you gotta like GREEN... Is anyone else able to grab this tar-ball from Denx? Is there/should there be a fallback at openwrt.org? I can only say that this must be a problem on your end... Anyway, Felix has pushed the binary on OpenWrt server. Please try it again and let me know if it still does not work for you. Hmmm. It must be a problem on my end. I just re-tested (from home) and the download from denx.de worked fine. Regards, Eric ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH 1/2] U-Boot: Use git instead of FTP, consolidate download
On 2013-09-26 05:05, Luka Perkov wrote: But if ftp or http works fine I prefer to have that over git for packaging. Also, by using git we would always have here a red color and not preferable green: https://home.comcast.net/~sdwalker/uscan/index.html Yes, it's all about collors ;) Speaking of "all about colors"...: who manages that page? Is there any chance that the colors can be fixed to use a more accessible pallet? Here's the Color Universal Design pallet: http://jfly.iam.u-tokyo.ac.jp/color/#pallet -- "'tis an ill wind that blows no minds." ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] Target kernel config changes not triggering rebuild?
2013.09.24. 16:56 keltezéssel, Paul Fertser írta: > Hello, > > Gabor Juhos writes: >> Yes, the 'Image/Build/Cameo' macro is broken. >> Don't need to send a patch, I'm working on that already. > > Great cleanup and the fixes on top, thank you! > > Can you please explain what is this "mac" partition about, what does it > hold? I know wireless "eeprom" is in the last "art" partition and that > seems to be all that's needed, at least on my other wr741nd device. The WR741ND has no such partition. To be precise, there is an user_cfg partition before the wifi calibration data. That is used by the factory firmware to store user specific settigns. We are reusign that because the factory firmware can reinitialize that with the factory reset option. On the DIR-600-Ex, the 'mac' partition contains the board specific MAC addresses at least, and it seems that there are other data as well. I guess that the original firmware uses this data to reinialize the nvram variables during factory reset. If that partition is erased, and someone wants to use the original software again, it will not be able to do a factory reset. > I also faced a strange issue which forced me to open the device and > attach the serial (and for no evident reason tftpboot didn't work > either, the device kept sending me arp requests ignoring my replies, so > I had to use Ymodem). > > Apparently, after I did "git pull; make oldconfig; make" and > sysupgraded, the resulting images ended up missing the appropriate > firmware partition splitters so it couldn't mount the rootfs after the > reboot. I have no real clue how that might have happened but "make > target/linux/clean; make" gave me a new sysupgrade image that work fine. If you are rebuilding a kernel after a kernel configuration change, it will use the modified configuration. However if the generic/platform specific kernel patches are modified, or some patches has been added/deleted those will not be used without a target/linux/clean. > Thank you again for your work. Np. -Gabor ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
[OpenWrt-Devel] [PATCH V3 2/2] uboot-imx6: add support for Nitrogen6x boards
Signed-off-by: Eric Nelson --- V2 alphabetizes the list and fixes nitrogen6dl/dl2g. V3 fixes the second part of nitrogen6dl/dl2g (UBOOTS list) Depends on patch "uboot-imx6: alphabetize wandboard variants" Patch 1/2 was nacked, so no update will be sent. package/boot/uboot-imx6/Makefile | 30 ++ 1 file changed, 30 insertions(+) diff --git a/package/boot/uboot-imx6/Makefile b/package/boot/uboot-imx6/Makefile index 21fc9bd..009d579 100644 --- a/package/boot/uboot-imx6/Makefile +++ b/package/boot/uboot-imx6/Makefile @@ -24,6 +24,30 @@ define uboot/Default IMAGE:= endef +define uboot/nitrogen6dl + TITLE:=U-Boot for Nitrogen6x i.MX6Dual-Lite 1GB board +endef + +define uboot/nitrogen6dl2g + TITLE:=U-Boot for Nitrogen6x i.MX6Dual-Lite 2GB board +endef + +define uboot/nitrogen6q + TITLE:=U-Boot for Nitrogen6x/SABRE Lite (MX6Q/1GB) +endef + +define uboot/nitrogen6q2g + TITLE:=U-Boot for Nitrogen6x i.MX6Quad 2GB board +endef + +define uboot/nitrogen6s + TITLE:=U-Boot for Nitrogen6x i.MX6Solo 512MB board +endef + +define uboot/nitrogen6s1g + TITLE:=U-Boot for Nitrogen6x i.MX6Solo 1GB board +endef + define uboot/wandboard_dl TITLE:=U-Boot for the Wandboard Dual Lite endef @@ -37,6 +61,12 @@ define uboot/wandboard_solo endef UBOOTS := \ + nitrogen6dl \ + nitrogen6dl2g \ + nitrogen6q \ + nitrogen6q2g \ + nitrogen6s \ + nitrogen6s1g \ wandboard_dl \ wandboard_quad \ wandboard_solo -- 1.8.1.2 ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] Target kernel config changes not triggering rebuild?
On Thu, Sep 26, 2013 at 08:12:13PM +0200, Gabor Juhos wrote: > On the DIR-600-Ex, the 'mac' partition contains the board specific MAC > addresses > at least, and it seems that there are other data as well. I guess that the > original firmware uses this data to reinialize the nvram variables during > factory reset. If that partition is erased, and someone wants to use the > original software again, it will not be able to do a factory reset. Thank you for the explanation. > If you are rebuilding a kernel after a kernel configuration change, it will > use > the modified configuration. However if the generic/platform specific kernel > patches are modified, or some patches has been added/deleted those will not be > used without a target/linux/clean. This one is sort of nasty, I'm not one of those who gets upset when he needs to attach a serial console, but it would still be nice to avoid, as the majority of users is apparently not always ready for that. Should I file a bug report? -- Be free, use free (http://www.gnu.org/philosophy/free-sw.html) software! mailto:fercer...@gmail.com ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] Regression: Trunk no longer runs on the WRT54G
* Christoph Kottke [26.09.2013 10:46]: > my asus wl-500g also not booting and the watchdog reset the device after today i had a Asus wl500g Premium V1 running with trunk with a serial console and it reboots all the time: - [0.992000] mtd: partition "kernel" must either start or end on erase block boundary or be smaller than an erase block -- forcing read-only [1.008000] 0x00141800-0x007f : "rootfs" [1.012000] mtd: partition "rootfs" must either start or end on erase block boundary or be smaller than an erase block -- forcing read-only [1.032000] mtd: partition "rootfs" set to be root filesystem [1.036000] mtd: partition "rootfs_data" created automatically, ofs=0x35, len=0x4a [1.044000] 0x0035-0x007f : "rootfs_data" [1.056000] 0x007f-0x0080 : "nvram" [1.076000] bcm47xx-wdt bcm47xx-wdt.0: BCM47xx Watchdog Timer enabled (30 seconds, Software Timer) [1.084000] GPIO_WDT: failed to register misc device [1.092000] TCP: cubic registered [1.096000] NET: Registered protocol family 17 [1.10] 8021q: 802.1Q VLAN Support v1.8 [1.116000] VFS: Mounted root (squashfs filesystem) readonly on device 31:4. [1.128000] Freeing unused kernel memory: 172K (802e5000 - 8031) procd: Console is alive procd: - watchdog - [4.648000] pps_core: LinuxPPS API ver. 1 registered [4.652000] pps_core: Software ver. 5.3.6 - Copyright 2005-2007 Rodolfo Giometti [4.664000] PTP clock support registered [4.676000] b44: Broadcom 44xx/47xx 10/100 PCI ethernet driver version 2.0 [4.688000] libphy: b44_eth_mii: probed [4.768000] b53_common: found switch: BCM5325, rev 0 [4.772000] b44 ssb0:0 (unregistered net_device): attached PHY driver [Broadcom B53 (1)] (mii_bus:phy_addr=1:1e) [4.788000] b44 ssb0:0 eth0: Broadcom 44xx/47xx 10/100 PCI ethernet driver 00:1b:fc:57:ad:fe CFE version 1.0.37 for BCM947XX (32bit,SP,LE) Build Date: �| 10�� 12 22:21:19 CST 2006 (root@localhost.localdomain) Copyright (C) 2000,2001,2002,2003 Broadcom Corporation. - interesting: there is no crash, but a really fast reboot. maybe it is the ethernet driver, will do more debugging tomorrow. bye, bastian ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] Regression: Trunk no longer runs on the WRT54G
* Adam Novak [26.09.2013 09:31]: > Does anyone have a WRT54G and the necessary hardware to debug this > over a serial connection? Would anyone want me to mail them one of my > routers to work on this? i have flashed a v2.2 a v3.0 and a GL1.1 today and they run fine. will try trunk on the other models tomorrow and report. (v1.0 and v2.0) bye, bastian ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH V3 2/2] uboot-imx6: add support for Nitrogen6x boards
Hi Eric, On Thu, Sep 26, 2013 at 11:28:44AM -0700, Eric Nelson wrote: > V3 fixes the second part of nitrogen6dl/dl2g (UBOOTS list) You are not in sync with the latest trunk. I've already fixed this in r38205. Luka ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] Target kernel config changes not triggering rebuild?
Le jeudi 26 septembre 2013 22:38:57 Paul Fertser a écrit : > On Thu, Sep 26, 2013 at 08:12:13PM +0200, Gabor Juhos wrote: > > On the DIR-600-Ex, the 'mac' partition contains the board specific MAC > > addresses at least, and it seems that there are other data as well. I > > guess that the original firmware uses this data to reinialize the nvram > > variables during factory reset. If that partition is erased, and someone > > wants to use the original software again, it will not be able to do a > > factory reset. > Thank you for the explanation. > > > If you are rebuilding a kernel after a kernel configuration change, it > > will use the modified configuration. However if the generic/platform > > specific kernel patches are modified, or some patches has been > > added/deleted those will not be used without a target/linux/clean. > > This one is sort of nasty, I'm not one of those who gets upset when he > needs to attach a serial console, but it would still be nice to avoid, > as the majority of users is apparently not always ready for > that. Should I file a bug report? Rebuilding the kernel each time a patch changes would really harm the kernel development workflow with OpenWrt, so I would rather stick with this even though this may be confusing. Rebuilding each and every single time a patch is touched could really make you waste WIP work in build_dir/target* -- Florian ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] [PATCH V3 2/2] uboot-imx6: add support for Nitrogen6x boards
On 09/26/2013 12:11 PM, Luka Perkov wrote: Hi Eric, On Thu, Sep 26, 2013 at 11:28:44AM -0700, Eric Nelson wrote: V3 fixes the second part of nitrogen6dl/dl2g (UBOOTS list) You are not in sync with the latest trunk. I've already fixed this in r38205. Thanks Luka. ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
[OpenWrt-Devel] [PATCH v2] [packages] ser2net: bump to 2.9.1
This version fixes a segmentation fault due to a missing buffer wrap. Signed-off-by: Yegor Yefremov --- Changes: v2: add [packages] to the subject and add more info to the commit msg net/ser2net/Makefile | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/ser2net/Makefile b/net/ser2net/Makefile index 2782972..c67e08f 100644 --- a/net/ser2net/Makefile +++ b/net/ser2net/Makefile @@ -8,12 +8,12 @@ include $(TOPDIR)/rules.mk PKG_NAME:=ser2net -PKG_VERSION:=2.8 +PKG_VERSION:=2.9.1 PKG_RELEASE:=1 PKG_SOURCE:=$(PKG_NAME)-$(PKG_VERSION).tar.gz PKG_SOURCE_URL:=@SF/ser2net -PKG_MD5SUM:=1cffbdaad221f19dcf82681472f7f947 +PKG_MD5SUM:=80011ac0e60bbdcb65f1d7a86251e3f3 PKG_FIXUP:=autoreconf PKG_INSTALL:=1 -- 1.8.1.2 ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel