B.A.T.M.A.N

b.a.t.m.a.n@lists.open-mesh.org

15 participants
4308 discussions

[B.A.T.M.A.N.] [PATCH] alfred: Keep order of interfaces from commandline
by Sven Eckelmann 14 Oct '16

14 Oct '16

The first (valid) interface mac address will be used as source mac for an dataset. But alfred added the commandline parameters for a multi-interface master in reverse order. Keep the list in the same order as the commandline to fulfill the user expectations when checking the source entries of datasets. Signed-off-by: Sven Eckelmann <sven.eckelmann(a)open-mesh.com> --- netsock.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/netsock.c b/netsock.c index d72541e..5d37347 100644 --- a/netsock.c +++ b/netsock.c @@ -165,7 +165,7 @@ int netsock_set_interfaces(struct globals *globals, char *interfaces) return -ENOMEM; } - list_add(&interface->list, &globals->interfaces); + list_add_tail(&interface->list, &globals->interfaces); } return 0;

2 1

Re: [B.A.T.M.A.N.] Help in setting up mesh test bed using qemu - BATMAN adv
by Kaushik N P 12 Oct '16

12 Oct '16

Hello, @Linus, I am a bit new to this and haven't much exposure to compiling kernels and all. I will try out and see what I can do with it, but a little more explanation would help a lot. "Then you connect the namespace via veth-pairs." Didn't get the namespace and making veth-pairs part. Anyways, I'll be certain to see what I can do with the code itself. Thanks and regards, Kaushik Thanks and Regards, Kaushik NP LI | Tw | FB SAP Ambassador, General Secretary, Computer Society Team IEEE UVCE, 3rd Year Undergraduate Student, University Visvesvaraya College of Engineering, Bangalore, IN On Wed, Oct 12, 2016 at 10:19 PM, Kaushik N P <kaushik.np.in(a)ieee.org> wrote: > Hello, > > @Linus, I am a bit new to this and haven't much exposure to compiling > kernels and all. I will try out and see what I can do with it, but a little > more explanation would help a lot. > > "Then you connect the namespace via veth-pairs." Didn't get the namespace > and making veth-pairs part. > > Anyways, I'll be certain to see what I can do with the code itself. > > Thanks and regards, > Kaushik > > > On Oct 11, 2016 4:06 PM, "Linus Lüssing" <linus.luessing(a)c0d3.blue> wrote: > > Haven't tried that myself yet, but maybe the new network namespace > support in the git master branch might be an easy alternative to > start with? > > You could first try to compile and load the batman-adv > kernel module. Then you'd create a namespace for each virtual node > you want to have. Then you connect the namespace via veth-pairs. > And finally, within each network namespace you would hook some veth > ends into a batX interface. > > > First you could try wiring two, then three nodes manually. If > that works and looks good you could then write a script to setup > the testbed with an arbitrary amount of virtual nodes. > > Regards, Linus > >

2 2

Re: [B.A.T.M.A.N.] Help in setting up mesh test bed using qemu - BATMAN adv
by Kaushik N P 11 Oct '16

11 Oct '16

Hello, Good to hear it is tough to work on in Linux and that am not the only person finding it difficult working on it. @Ufo : Yes, it might be tough, but we learn from experience on working on such stuff after all. I have two routers but want to setup a bigger mesh bed with more nodes and would prefer not to spend more for now. And mainly I think working on Linux would provide me a better understanding and would allow me to experiment more on it. I have worked with Virtual machines before, though with minimal experimentation at that. And I will definitely check out LEDE. Thank you for the suggestion. @Marek : Yes, I know it was a half baked question I presented, but wasn't sure how to put up my query more clearly. Now I will list out the steps I have followed till now and what the results are. Links I followed primarily: [1] https://wiki.openwrt.org/doc/howto/qemu [2] https://www.open-mesh.org/projects/open-mesh/wiki/Emulation So, the link [1] defines setting up QEMU. Again, I am working on Ubuntu 16.0, 64-bit. As far as I could see, only the Commands in 'Booting from SD card image' was related. Have downloaded openwrt-realview-vmlinux.elf , openwrt-realview-sdcard.img and type in the command below : qemu-system-arm -M realview-pbx-a9 -m 1024M -nographic \ -kernel openwrt-realview-vmlinux.elf \ -sd openwrt-realview-sdcard.img \ -append "console=ttyAMA0 verbose debug root=/dev/mmcblk0p1" This command opened up a BusyBox which looks as follows : BusyBox v1.24.2 () built-in shell (ash) //Says OpenWrt below _______ ________ __ | | .-----.-----.-----.| | | |.----.| |_ | - | | _ | -__| || | | || _|| _| |_______| | __|____|__|__||________||__| |____| |__| W I R E L E S S F R E E D O M ----------------------------------------------------- DESIGNATED DRIVER (Bleeding Edge, 49928) ----------------------------------------------------- * 2 oz. Orange Juice Combine all juices in a * 2 oz. Pineapple Juice tall glass filled with * 2 oz. Grapefruit Juice ice, stir well. * 2 oz. Cranberry Juice ----------------------------------------------------- root@OpenWrt:/# [ 13.897119] smsc911x smsc911x.0 eth0: SMSC911x/921x identified at 0xf0b40000, IRQ: 60 [ 13.934741] device eth0 entered promiscuous mode [ 13.978166] IPv6: ADDRCONF(NETDEV_UP): br-lan: link is not ready [ 15.762764] br-lan: port 1(eth0) entered forwarding state [ 15.764526] br-lan: port 1(eth0) entered forwarding state [ 15.802643] IPv6: ADDRCONF(NETDEV_CHANGE): br-lan: link becomes ready [ 17.762480] br-lan: port 1(eth0) entered forwarding state root@OpenWrt:/# There is no details on what to do once this opens up though. And I still need confirmation as to what this means. Is it that QEMU has loaded OpenWRT in my system like a virtual machine or is it a console to access OpenWRT functionalities, etc. After this I made changes according to "openwrt virtualized using debian, qemu/kvm and a lex twitter system with intel atom d525 and chipset ich8m" section to 'Virtualization proper' too in case it was required, but the Virtualization didn't work. This done, I moved to link [2], where it explains how to set up mesh testbed. It says"OpenWrt, kamikaze trunk version for x86 with minimal modifications" and so I have downloaded 'openwrt-x86-ext2.image' from Kamikaze. Changed the above code used to load QEMU as : qemu-system-arm -M realview-pbx-a9 -m 1024M -nographic \ > -kernel openwrt-realview-vmlinux.elf \ > -sd openwrt- \ > -sd openwrt-x86-ext2.image \ > -append "console=ttyAMA0 verbose debug root=/dev/mmcblk0p1" But loading this hasn't been successful as doing this gives : ...<loading and stuff>... [ 1.304028] VFS: Mounted root (ext2 filesystem) readonly on device 179:1. [ 1.317832] Freeing unused kernel memory: 196K (c0481000 - c04b2000) [ 1.328078] Kernel panic - not syncing: No working init found. Try passing init= option to kernel. See Linux Documentation/init.txt for guidance. [ 1.332338] ---[ end Kernel panic - not syncing: No working init found. Try passing init= option to kernel. See Linux Documentation/init.txt for guidance. Next, running the script written in 'OpenWrt' section doesn't work as it says : /tmp/logserver.sh: 3: /tmp/logserver.sh: cannot open /sys/kernel/debug/batman_adv/bat0/log: No such file /tmp/logserver.sh: 3: /tmp/logserver.sh: cannot open /sys/kernel/debug/batman_adv/bat0/log: No such file Commenting the lines that open this file : while [ 1 ]; do nc -l -p 2050 < /sys/kernel/debug/batman_adv/bat0/log done gives the error : RTNETLINK answers: File exists Cannot find device "br-lan" Cannot find device "br-lan" ./rc.local: 23: ./rc.local: cannot create /proc/net/batman-adv/interfaces: Directory nonexistent ./rc.local: 24: ./rc.local: cannot create /proc/net/batman-adv/interfaces: Directory nonexistent ./rc.local: 25: ./rc.local: cannot create /proc/net/batman-adv/log_level: Directory nonexistent /tmp/logserver.sh: 1: /tmp/logserver.sh: !/bin/sh: not found Cannot find device "bat0" Cannot find device "bat0" Next, I have patched both 'vde2-2.3.2_colour.patch' and 'vde2-2.3.2-wirefilter-ethertype.patch' The final step lists a script for running the scripts to start up all the qemu windows. This throw up a lot many errors. Below is the error it shows: kaushik@kaushik-HP:~/QEMUopenwrt$ ./command.sh Must be run inside a screen session [sudo] password for kaushik: wirefilter(8722): Operation not permitted wirefilter(8724): Operation not permitted wirefilter(8726): Operation not permitted wirefilter(8728): Operation not permitted wirefilter(8730): Operation not permitted wirefilter(8732): Operation not permitted wirefilter(8734): Operation not permitted wirefilter(8736): Operation not permitted wirefilter(8738): Operation not permitted wirefilter(8740): Operation not permitted wirefilter(8742): Operation not permitted wirefilter(8744): Operation not permitted wirefilter(8746): Operation not permitted wirefilter(8748): Operation not permitted wirefilter(8750): Operation not permitted vde_switch(8567): Operation not permitted vde_switch(8575): Operation not permitted vde_switch(8583): Operation not permitted vde_switch(8591): Operation not permitted vde_switch(8599): Operation not permitted vde_switch(8607): Operation not permitted vde_switch(8615): Operation not permitted vde_switch(8623): Operation not permitted vde_switch(8631): Operation not permitted ioctl(TUNSETIFF): Device or resource busy RTNETLINK answers: File exists ioctl(TUNSETIFF): Device or resource busy RTNETLINK answers: File exists ioctl(TUNSETIFF): Device or resource busy RTNETLINK answers: File exists ioctl(TUNSETIFF): Device or resource busy RTNETLINK answers: File exists ioctl(TUNSETIFF): Device or resource busy RTNETLINK answers: File exists ioctl(TUNSETIFF): Device or resource busy . . . This I run in a screen as it is mentioned in the guide. It creates many files and folders named num0 to num9, but doesn't proceed after that. This and some more experimentation I have done with it. So, any help I can get would be most welcome Thanks and Regards, Kaushik NP LI | Tw | FB SAP Ambassador, General Secretary, Computer Society Team IEEE UVCE, 3rd Year Undergraduate Student, University Visvesvaraya College of Engineering, Bangalore, IN On Sat, Oct 8, 2016 at 1:21 AM, Kaushik N P <kaushik.np.in(a)ieee.org> wrote: > Hello, > > Good to hear it is tough to work on in Linux and that am not the only person > finding it difficult working on it. > > @Ufo : Yes, it might be tough, but we learn from experience on working on > such stuff after all. I have two routers but want to setup a bigger mesh bed > with more nodes and would prefer not to spend more for now. And mainly I > think working on Linux would provide me a better understanding and would > allow me to experiment more on it. I have worked with Virtual machines > before, though with minimal experimentation at that. And I will definitely > check out LEDE. Thank you for the suggestion. > > @Marek : Yes, I know it was a half baked question I presented, but wasn't > sure how to put up my query more clearly. Now I will list out the steps I > have followed till now and what the results are. > > Links I followed primarily: > [1] https://wiki.openwrt.org/doc/howto/qemu > [2] https://www.open-mesh.org/projects/open-mesh/wiki/Emulation > > So, the link [1] defines setting up QEMU. Again, I am working on Ubuntu > 16.0, 64-bit. > > As far as I could see, only the Commands in 'Booting from SD card image' > was related. Have downloaded openwrt-realview-vmlinux.elf , > openwrt-realview-sdcard.img and type in the command below : > > qemu-system-arm -M realview-pbx-a9 -m 1024M -nographic \ > -kernel openwrt-realview-vmlinux.elf \ > -sd openwrt-realview-sdcard.img \ > -append "console=ttyAMA0 verbose debug root=/dev/mmcblk0p1" > > > This command opened up a BusyBox which looks as follows : > > BusyBox v1.24.2 () built-in shell (ash) > > //Says OpenWrt below > _______ ________ __ > | | .-----.-----.-----.| | | |.----.| |_ > | - | | _ | -__| || | | || _|| _| > |_______| | __|____|__|__||________||__| |____| > |__| W I R E L E S S F R E E D O M > ----------------------------------------------------- > DESIGNATED DRIVER (Bleeding Edge, 49928) > ----------------------------------------------------- > * 2 oz. Orange Juice Combine all juices in a > * 2 oz. Pineapple Juice tall glass filled with > * 2 oz. Grapefruit Juice ice, stir well. > * 2 oz. Cranberry Juice > ----------------------------------------------------- > root@OpenWrt:/# [ 13.897119] smsc911x smsc911x.0 eth0: SMSC911x/921x > identified at 0xf0b40000, IRQ: 60 > [ 13.934741] device eth0 entered promiscuous mode > [ 13.978166] IPv6: ADDRCONF(NETDEV_UP): br-lan: link is not ready > [ 15.762764] br-lan: port 1(eth0) entered forwarding state > [ 15.764526] br-lan: port 1(eth0) entered forwarding state > [ 15.802643] IPv6: ADDRCONF(NETDEV_CHANGE): br-lan: link becomes ready > [ 17.762480] br-lan: port 1(eth0) entered forwarding state > root@OpenWrt:/# > > There is no details on what to do once this opens up though. And I still > need confirmation as to what this means. Is it that QEMU has loaded OpenWRT > in my system like a virtual machine or is it a console to access OpenWRT > functionalities, etc. > > After this I made changes according to "openwrt virtualized using debian, > qemu/kvm and a lex twitter system with intel atom d525 and chipset ich8m" > section to 'Virtualization proper' too in case it was required, but the > Virtualization didn't work. > > > This done, I moved to link [2], where it explains how to set up mesh > testbed. It says"OpenWrt, kamikaze trunk version for x86 with minimal > modifications" and so I have downloaded 'openwrt-x86-ext2.image' from > Kamikaze. Changed the above code used to load QEMU as : > qemu-system-arm -M realview-pbx-a9 -m 1024M -nographic \ >> -kernel openwrt-realview-vmlinux.elf \ >> -sd openwrt- \ >> -sd openwrt-x86-ext2.image \ >> -append "console=ttyAMA0 verbose debug root=/dev/mmcblk0p1" > > But loading this hasn't been successful as doing this gives : > ...<loading and stuff>... > [ 1.304028] VFS: Mounted root (ext2 filesystem) readonly on device 179:1. > [ 1.317832] Freeing unused kernel memory: 196K (c0481000 - c04b2000) > [ 1.328078] Kernel panic - not syncing: No working init found. Try > passing init= option to kernel. See Linux Documentation/init.txt for > guidance. > [ 1.332338] ---[ end Kernel panic - not syncing: No working init found. > Try passing init= option to kernel. See Linux Documentation/init.txt for > guidance. > > > Next, running the script written in 'OpenWrt' section doesn't work as it > says : > /tmp/logserver.sh: 3: /tmp/logserver.sh: cannot open > /sys/kernel/debug/batman_adv/bat0/log: No such file > /tmp/logserver.sh: 3: /tmp/logserver.sh: cannot open > /sys/kernel/debug/batman_adv/bat0/log: No such file > > Commenting the lines that open this file : > while [ 1 ]; do > nc -l -p 2050 < /sys/kernel/debug/batman_adv/bat0/log > done > > gives the error : > RTNETLINK answers: File exists > Cannot find device "br-lan" > Cannot find device "br-lan" > ./rc.local: 23: ./rc.local: cannot create /proc/net/batman-adv/interfaces: > Directory nonexistent > ./rc.local: 24: ./rc.local: cannot create /proc/net/batman-adv/interfaces: > Directory nonexistent > ./rc.local: 25: ./rc.local: cannot create /proc/net/batman-adv/log_level: > Directory nonexistent > /tmp/logserver.sh: 1: /tmp/logserver.sh: !/bin/sh: not found > Cannot find device "bat0" > Cannot find device "bat0" > > Next, I have patched both 'vde2-2.3.2_colour.patch' and > 'vde2-2.3.2-wirefilter-ethertype.patch' > > The final step lists a script for running the scripts to start up all the > qemu windows. This throw up a lot many errors. > Below is the error it shows: > > kaushik@kaushik-HP:~/QEMUopenwrt$ ./command.sh > Must be run inside a screen session > [sudo] password for kaushik: > wirefilter(8722): Operation not permitted > wirefilter(8724): Operation not permitted > wirefilter(8726): Operation not permitted > wirefilter(8728): Operation not permitted > wirefilter(8730): Operation not permitted > wirefilter(8732): Operation not permitted > wirefilter(8734): Operation not permitted > wirefilter(8736): Operation not permitted > wirefilter(8738): Operation not permitted > wirefilter(8740): Operation not permitted > wirefilter(8742): Operation not permitted > wirefilter(8744): Operation not permitted > wirefilter(8746): Operation not permitted > wirefilter(8748): Operation not permitted > wirefilter(8750): Operation not permitted > vde_switch(8567): Operation not permitted > vde_switch(8575): Operation not permitted > vde_switch(8583): Operation not permitted > vde_switch(8591): Operation not permitted > vde_switch(8599): Operation not permitted > vde_switch(8607): Operation not permitted > vde_switch(8615): Operation not permitted > vde_switch(8623): Operation not permitted > vde_switch(8631): Operation not permitted > ioctl(TUNSETIFF): Device or resource busy > RTNETLINK answers: File exists > ioctl(TUNSETIFF): Device or resource busy > RTNETLINK answers: File exists > ioctl(TUNSETIFF): Device or resource busy > RTNETLINK answers: File exists > ioctl(TUNSETIFF): Device or resource busy > RTNETLINK answers: File exists > ioctl(TUNSETIFF): Device or resource busy > RTNETLINK answers: File exists > ioctl(TUNSETIFF): Device or resource busy > . > . > . > > This I run in a screen as it is mentioned in the guide. It creates many > files and folders named num0 to num9, but doesn't proceed after that. > > > This and some more experimentation I have done with it. So, any help I can > get would be most welcome > > > Thanks and Regards, > > Kaushik NP > LI | Tw | FB > SAP Ambassador, > General Secretary, Computer Society > Team IEEE UVCE, > 3rd Year Undergraduate Student, > University Visvesvaraya College of Engineering, > Bangalore, IN > > On Thu, Oct 6, 2016 at 8:39 PM, Ufo <ufo(a)rund.freifunk.net> wrote: >> >> Testing batman-adv in virtual maschines on laptops is the most complex and >> non-realistic way to start with mesh-networking. So its a nice, very >> demanding student thing. >> >> its much more easier to buy some (more than 2!!) cheap wifi-routers! they >> have proper wifi, more than one ethernet-interface (!) and there are many >> openwrt/lede mesh firmwares with a well-configured batman-adv. after that, >> the next step could be to try to couple your first laptop into that mesh. >> starting with nothing but laptops seems to be too "versatile". >> >> if you are unfamiliar with qemu and virtualization your should first try >> to install ubuntu or windows in a virtual machine. please use >> "virt-manager", a graphical interface for that. >> >> unfortunatly the mentioned "stable" openwrt was forced to use a very old >> batman-adv version from 2014. So you might use trunk firmware, and use LEDE >> instead of Openwrt, to have a proper system. >> >> >> http://ftp.halifax.rwth-aachen.de/lede/snapshots/targets/x86/generic/lede-x… >> >> >> >> >> >> >> On 04.10.2016 12:11, Kaushik N P wrote: >>> >>> Hello, >>> >>> Thank you 'rise.net' for the suggestion. It has given me some ideas, >>> but what I hoped to do was to run the laptop with OpenWRT working with >>> Batman adv like a router. This would enable me in creating a more >>> extensive network without having to buy more routers itself and >>> hopefully provide me with a more versatile network to work on. I will >>> ofcourse be experimenting more along your steps, but I hope to get the >>> QEMU set up and running on my system. >>> >>> If any one has more info on the QEMU setup, pls do contact. >>> >>> >>> Thanks and Regards, >>> >>> Kaushik NP >>> General Secretary, Computer Society >>> Team IEEE UVCE, >>> 3rd Year Undergraduate Student, >>> University Visvesvaraya College of Engineering, >>> Bangalore, IN >>> >>> >>> On Mon, Oct 3, 2016 at 10:38 PM, <leftbydefault(a)riseup.net> wrote: >>>> >>>> Hello Kaushik >>>> >>>> On 2016-10-03 19:44, Kaushik N P wrote: >>>>> >>>>> Hello, >>>>> >>>>> What I hope to find is:- >>>>> >>>>> Another documentation maybe for setting up, or confirmation for >>>>> setting up (because I can't tell sometimes if the instructions have >>>>> been followed correctly). >>>> >>>> >>>> I have no experience in simulating or emulating batman-adv, but I have >>>> experience >>>> in compiling openwrt from source along with batman-adv and batctl. I can >>>> help you with that. batman-adv is built as a kernel module and so it is >>>> included by default in Linux kernels >= 2.x. You have mentioned that you >>>> are >>>> using Linux based operating system, I hope by now, you should be running >>>> Linux kernel 4.x. If you want to play around with batman-adv for >>>> sometime >>>> before emulating, you can use two three Linux based laptops with WiFi >>>> interface cards that support IBSS (adhoc mode). >>>> >>>> You can check that using >>>> >>>> iw list | grep IBSS >>>> >>>> if the command returned something like >>>> Device supports RSN-IBSS. >>>> * IBSS >>>> >>>> then your Wireless NIC can be switched to ad-hoc mode. We have a bash >>>> shell >>>> script that does it automatically for you. You can find the script and >>>> terminal cast in the following links. >>>> >>>> 1. https://github.com/PYMeshnet/scripts >>>> 2. https://asciinema.org/a/38575 >>>> >>>>> Clarification for my doubt regarding the different types of OpenWRT >>>>> images (like Kamikaze, Chaos Calmer, Attitude Adjustment,etc), what do >>>>> they do and which one I should use for working with standard Batman >>>>> adv setup. >>>> >>>> >>>> Chaos Calmer (15.x) is the recent stable release of OpenWRT firmware. >>>> 'Kamikaze, Altitude Adjustment, etc.,' are just different names given to >>>> different versions of OpenWRT releases. OpenWRT are meant to be the >>>> operating system for routers and modems. Depending on the configuration >>>> of >>>> devices, some versions may be supported and some may not be supported on >>>> certain routers and modems. >>>> >>>> Since you are trying to create a virtual environment, I think should go >>>> with >>>> Chaos Calmer. (again I have no experience either with creating a virtual >>>> environment for OpenWRT nor for batman-adv). >>>> >>>> >>>>> Confirmation method to see if qemu having OpenWRT is setup in my >>>>> system - as now I have a setup in terminal which says >>>>> >>>>> BusyBox v1.24.2 () built-in shell (ash) >>>>> >>>>> //Says OpenWrt below >>>>> _______ ________ __ >>>>> | | .-----.-----.-----.| | | |.----.| |_ >>>>> | - | | _ | -__| || | | || _|| _| >>>>> |_______| | __|____|__|__||________||__| |____| >>>>> |__| W I R E L E S S F R E E D O M >>>>> ----------------------------------------------------- >>>>> DESIGNATED DRIVER (Bleeding Edge, 49928) >>>>> ----------------------------------------------------- >>>>> * 2 oz. Orange Juice Combine all juices in a >>>>> * 2 oz. Pineapple Juice tall glass filled with >>>>> * 2 oz. Grapefruit Juice ice, stir well. >>>>> * 2 oz. Cranberry Juice >>>>> ----------------------------------------------------- >>>>> root@OpenWrt:/# [ 13.897119] smsc911x smsc911x.0 eth0: SMSC911x/921x >>>>> identified at 0xf0b40000, IRQ: 60 >>>>> [ 13.934741] device eth0 entered promiscuous mode >>>>> [ 13.978166] IPv6: ADDRCONF(NETDEV_UP): br-lan: link is not ready >>>>> [ 15.762764] br-lan: port 1(eth0) entered forwarding state >>>>> [ 15.764526] br-lan: port 1(eth0) entered forwarding state >>>>> [ 15.802643] IPv6: ADDRCONF(NETDEV_CHANGE): br-lan: link becomes >>>>> ready >>>>> [ 17.762480] br-lan: port 1(eth0) entered forwarding state >>>>> root@OpenWrt:/# >>>>> >>>>> >>>>> I have more doubts, but hopefully, I will get some answers to them >>>>> from the above questions itself. >>>>> Hope to hear soon!!! >>>>> >>>>> Thanks and Regards, >>>>> >>>>> Kaushik NP >>>>> LI | Tw | FB >>>>> SAP Ambassador, >>>>> General Secretary, Computer Society >>>>> Team IEEE UVCE, >>>>> 3rd Year Undergraduate Student, >>>>> University Visvesvaraya College of Engineering, >>>>> Bangalore, IN >>>> >>>> >> >

3 5

[B.A.T.M.A.N.] [PATCH next v2 1/3] batman-adv: Introduce compat-patches support
by Sven Eckelmann 09 Oct '16

09 Oct '16

compat-includes/compat.h can usually be used to solve compatibility problems with older kernels. This works well for functions, defines/enums and sometimes even structures that were introduced. But this can fail when structs changed. Some of these can be solved in crude ways but sometimes it is unavoidable to have a version specific code. Unfortunately, this kind of code is not acceptable in the kernel and thus the compat infrastructure of the external module has to do add it automatically before the source is compiled. This process works by creating a build directory which is prefilled with the source from net/batman-adv/. The patches from compat-patches/ will be applied on top of this copy and then the code is compiled. Signed-off-by: Sven Eckelmann <sven(a)narfation.org> --- v2: - Rebased on top of current next .gitignore | 10 ++++------ Makefile | 31 +++++++++++++++++++++++++++---- compat-patches/README | 23 +++++++++++++++++++++++ compat-sources/Makefile | 6 +++--- 4 files changed, 57 insertions(+), 13 deletions(-) create mode 100644 compat-patches/README diff --git a/.gitignore b/.gitignore index 4c03561..15a99aa 100644 --- a/.gitignore +++ b/.gitignore @@ -1,10 +1,8 @@ +/build/ /compat-autoconf.h /compat-autoconf.h.tmp /compat-sources/**/.* /compat-sources/**/*.o -/net/batman-adv/.* -/net/batman-adv/batman-adv.ko -/net/batman-adv/batman-adv.mod.c -/net/batman-adv/modules.order -/net/batman-adv/Module.symvers -/net/batman-adv/*.o +/modules.order +/Module.symvers +/.tmp_versions diff --git a/Makefile b/Makefile index d42bb56..b105290 100644 --- a/Makefile +++ b/Makefile @@ -33,6 +33,7 @@ export CONFIG_BATMAN_ADV_MCAST=y export CONFIG_BATMAN_ADV_BATMAN_V=n PWD:=$(shell pwd) +BUILD_DIR=$(PWD)/build KERNELPATH ?= /lib/modules/$(shell uname -r)/build # sanity check: does KERNELPATH exist? ifeq ($(shell cd $(KERNELPATH) && pwd),) @@ -41,6 +42,14 @@ endif export KERNELPATH RM ?= rm -f +MKDIR := mkdir -p +PATCH_FLAGS = --batch --fuzz=0 --forward --strip=1 --unified --version-control=never -g0 --remove-empty-files --no-backup-if-mismatch --reject-file=- +PATCH := patch $(PATCH_FLAGS) -i +CP := cp -fpR + +SOURCE = $(wildcard net/batman-adv/*.[ch]) net/batman-adv/Makefile +SOURCE_BUILD = $(wildcard $(BUILD_DIR)/net/batman-adv/*.[ch]) $(BUILD_DIR)/net/batman-adv/Makefile +SOURCE_STAMP = $(BUILD_DIR)/net/batman-adv/.compat-prepared REVISION= $(shell if [ -d "$(PWD)/.git" ]; then \ echo $$(git --git-dir="$(PWD)/.git" describe --always --dirty --match "v*" |sed 's/^v//' 2> /dev/null || echo "[unknown]"); \ @@ -57,7 +66,7 @@ endif include $(PWD)/compat-sources/Makefile -obj-y += net/batman-adv/ +obj-y += build/net/batman-adv/ export batman-adv-y @@ -76,18 +85,32 @@ BUILD_FLAGS := \ CONFIG_BATMAN_ADV_BATMAN_V=$(CONFIG_BATMAN_ADV_BATMAN_V) \ INSTALL_MOD_DIR=updates/ -all: config +all: config $(SOURCE_STAMP) $(MAKE) -C $(KERNELPATH) $(BUILD_FLAGS) modules clean: $(RM) compat-autoconf.h* - $(MAKE) -C $(KERNELPATH) $(BUILD_FLAGS) clean + $(RM) -r $(BUILD_DIR) -install: config +install: config $(SOURCE_STAMP) $(MAKE) -C $(KERNELPATH) $(BUILD_FLAGS) modules_install depmod -a config: $(PWD)/gen-compat-autoconf.sh $(PWD)/compat-autoconf.h +$(SOURCE_STAMP): $(SOURCE) compat-patches/* + $(MKDIR) $(BUILD_DIR)/net/batman-adv/ + @$(RM) $(SOURCE_BUILD) + @$(CP) $(SOURCE) $(BUILD_DIR)/net/batman-adv/ + @set -e; \ + patches="$$(ls -1 compat-patches/|grep '.patch$$'|sort)"; \ + for i in $${patches}; do \ + echo ' COMPAT_PATCH '$${i};\ + cd $(BUILD_DIR); \ + $(PATCH) ../compat-patches/$${i}; \ + cd - > /dev/null; \ + done + touch $(SOURCE_STAMP) + .PHONY: all clean install config diff --git a/compat-patches/README b/compat-patches/README new file mode 100644 index 0000000..3bbddb3 --- /dev/null +++ b/compat-patches/README @@ -0,0 +1,23 @@ +WARNING +======= + +Please avoid using the compat-patches/ to implement support for old kernels. +This should be the last resort. + + * it is nearly always possible to use compat-includes/ to do the same with a + lot less problems + + * maintaining these patches is *censored* + +GENERATING A PATCH +================== + +If it not possible to avoid a patch then please make the patch as small as +possible. Even refactor the code which has to be patched to reduce the +size/number of the changes. + +Please use git-format-patches to generate them and order same inside via the +XXXX- prefix of the patch name. + + git format-patch --abbrev=7 -U3 --diff-algorithm=histogram --no-signature \ + --format=format:'From: %an <%ae>%nDate: %aD%nSubject: [PATCH] %B' -1 diff --git a/compat-sources/Makefile b/compat-sources/Makefile index a45c202..c8bae49 100644 --- a/compat-sources/Makefile +++ b/compat-sources/Makefile @@ -1,3 +1,3 @@ -batman-adv-$(CONFIG_BATMAN_ADV_MCAST) += ../../compat-sources/net/core/skbuff.o -batman-adv-$(CONFIG_BATMAN_ADV_MCAST) += ../../compat-sources/net/ipv4/igmp.o -batman-adv-$(CONFIG_BATMAN_ADV_MCAST) += ../../compat-sources/net/ipv6/mcast_snoop.o +batman-adv-$(CONFIG_BATMAN_ADV_MCAST) += ../../../compat-sources/net/core/skbuff.o +batman-adv-$(CONFIG_BATMAN_ADV_MCAST) += ../../../compat-sources/net/ipv4/igmp.o +batman-adv-$(CONFIG_BATMAN_ADV_MCAST) += ../../../compat-sources/net/ipv6/mcast_snoop.o -- 2.9.3

2 5

[B.A.T.M.A.N.] Broadcast Avoidances, Neighborhood Hash
by Linus Lüssing 06 Oct '16

06 Oct '16

Hi, Here's the second broadcast avoidance patchset (rebased on the former one). Hopefully, those two patchsets together should be an automatic approach and replacement for the (non-upstream) no-rebroadcast flag patch and the cases it is currently used for. Documentation: * https://www.open-mesh.org/projects/batman-adv/wiki/Broadcast-Avoidances Branch: * https://git.open-mesh.org/batman-adv.git/shortlog/refs/heads/linus/neighhash Regards, Linus

3 16

[B.A.T.M.A.N.] Help in setting up mesh test bed using qemu - BATMAN adv
by Kaushik N P 06 Oct '16

06 Oct '16

Hello, I have recently decided to take up the prospect of layer 2 Batman adv protocol as my area of interest, and want to implement a mesh bed to get some statistical readings of my own (am very new to this area, so ask for forgiveness beforehand if I was supposed to provide more info). The plan was to set up OpenWRT in qemu in laptops and work on them with batman adv. But, unfortunately, I have been facing some problems in setting the system up. I have followed (as much as I could) the steps as per documentation (links for them posted below). Links: https://wiki.openwrt.org/doc/howto/qemu (Setting up qemu - problem I face in this is of which commands I should run , and which of the systems are viable for me (like QEMU arm, arm 64, x86-64,etc) to work on a 64-bit laptop having Linux OS ) https://www.open-mesh.org/projects/open-mesh/wiki/Emulation (I am not sure if my qemu is set up properly for there is no confirmation methods that are provided to me to check if its working properly) What I hope to find is:- Another documentation maybe for setting up, or confirmation for setting up (because I can't tell sometimes if the instructions have been followed correctly). Clarification for my doubt regarding the different types of OpenWRT images (like Kamikaze, Chaos Calmer, Attitude Adjustment,etc), what do they do and which one I should use for working with standard Batman adv setup. Confirmation method to see if qemu having OpenWRT is setup in my system - as now I have a setup in terminal which says BusyBox v1.24.2 () built-in shell (ash) //Says OpenWrt below _______ ________ __ | | .-----.-----.-----.| | | |.----.| |_ | - | | _ | -__| || | | || _|| _| |_______| | __|____|__|__||________||__| |____| |__| W I R E L E S S F R E E D O M ----------------------------------------------------- DESIGNATED DRIVER (Bleeding Edge, 49928) ----------------------------------------------------- * 2 oz. Orange Juice Combine all juices in a * 2 oz. Pineapple Juice tall glass filled with * 2 oz. Grapefruit Juice ice, stir well. * 2 oz. Cranberry Juice ----------------------------------------------------- root@OpenWrt:/# [ 13.897119] smsc911x smsc911x.0 eth0: SMSC911x/921x identified at 0xf0b40000, IRQ: 60 [ 13.934741] device eth0 entered promiscuous mode [ 13.978166] IPv6: ADDRCONF(NETDEV_UP): br-lan: link is not ready [ 15.762764] br-lan: port 1(eth0) entered forwarding state [ 15.764526] br-lan: port 1(eth0) entered forwarding state [ 15.802643] IPv6: ADDRCONF(NETDEV_CHANGE): br-lan: link becomes ready [ 17.762480] br-lan: port 1(eth0) entered forwarding state root@OpenWrt:/# I have more doubts, but hopefully, I will get some answers to them from the above questions itself. Hope to hear soon!!! Thanks and Regards, Kaushik NP LI | Tw | FB SAP Ambassador, General Secretary, Computer Society Team IEEE UVCE, 3rd Year Undergraduate Student, University Visvesvaraya College of Engineering, Bangalore, IN

4 4

[B.A.T.M.A.N.] [PATCH 1/3] batman-adv: refactor wifi interface detection
by Marek Lindner 30 Sep '16

30 Sep '16

The ELP protocol requires cfg80211 to auto-detect the WiFi througput to a given neighbor. Use batadv_is_cfg80211_netdev() to determine whether or not an interface is eligible. Signed-off-by: Marek Lindner <mareklindner(a)neomailbox.ch> --- net/batman-adv/bat_v_elp.c | 29 ++++++++++++++--------------- net/batman-adv/hard-interface.c | 26 +++++++++++++++++++++----- net/batman-adv/hard-interface.h | 1 + 3 files changed, 36 insertions(+), 20 deletions(-) diff --git a/net/batman-adv/bat_v_elp.c b/net/batman-adv/bat_v_elp.c index 7d17001..51fc08d 100644 --- a/net/batman-adv/bat_v_elp.c +++ b/net/batman-adv/bat_v_elp.c @@ -90,22 +90,21 @@ static u32 batadv_v_elp_get_throughput(struct batadv_hardif_neigh_node *neigh) * cfg80211 API */ if (batadv_is_wifi_netdev(hard_iface->net_dev)) { - if (hard_iface->net_dev->ieee80211_ptr) { - ret = cfg80211_get_station(hard_iface->net_dev, - neigh->addr, &sinfo); - if (ret == -ENOENT) { - /* Node is not associated anymore! It would be - * possible to delete this neighbor. For now set - * the throughput metric to 0. - */ - return 0; - } - if (!ret) - return sinfo.expected_throughput / 100; + if (!batadv_is_cfg80211_netdev(hard_iface->net_dev)) + /* unsupported WiFi driver version */ + goto default_throughput; + + ret = cfg80211_get_station(hard_iface->net_dev, + neigh->addr, &sinfo); + if (ret == -ENOENT) { + /* Node is not associated anymore! It would be + * possible to delete this neighbor. For now set + * the throughput metric to 0. + */ + return 0; } - - /* unsupported WiFi driver version */ - goto default_throughput; + if (!ret) + return sinfo.expected_throughput / 100; } /* if not a wifi interface, check if this device provides data via diff --git a/net/batman-adv/hard-interface.c b/net/batman-adv/hard-interface.c index 714af8e..478977b 100644 --- a/net/batman-adv/hard-interface.c +++ b/net/batman-adv/hard-interface.c @@ -166,6 +166,26 @@ static bool batadv_is_valid_iface(const struct net_device *net_dev) } /** + * batadv_is_cfg80211_netdev - check if the given net_device struct is a + * cfg80211 wifi interface + * @net_device: the device to check + * + * Return: true if the net device is a cfg80211 wireless device, false + * otherwise. + */ +bool batadv_is_cfg80211_netdev(struct net_device *net_device) +{ + if (!net_device) + return false; + + /* cfg80211 drivers have to set ieee80211_ptr */ + if (net_device->ieee80211_ptr) + return true; + + return false; +} + +/** * batadv_is_wifi_netdev - check if the given net_device struct is a wifi * interface * @net_device: the device to check @@ -185,11 +205,7 @@ bool batadv_is_wifi_netdev(struct net_device *net_device) return true; #endif - /* cfg80211 drivers have to set ieee80211_ptr */ - if (net_device->ieee80211_ptr) - return true; - - return false; + return batadv_is_cfg80211_netdev(net_device); } static struct batadv_hard_iface * diff --git a/net/batman-adv/hard-interface.h b/net/batman-adv/hard-interface.h index a76724d..e0893de 100644 --- a/net/batman-adv/hard-interface.h +++ b/net/batman-adv/hard-interface.h @@ -51,6 +51,7 @@ enum batadv_hard_if_cleanup { extern struct notifier_block batadv_hard_if_notifier; +bool batadv_is_cfg80211_netdev(struct net_device *net_device); bool batadv_is_wifi_netdev(struct net_device *net_device); bool batadv_is_wifi_iface(int ifindex); struct batadv_hard_iface* -- 2.8.0.rc3

3 7

[B.A.T.M.A.N.] Creating alfred code documentation
by Germano Massullo 30 Sep '16

30 Sep '16

Hi, it's Germano from Ninux.org community. For personal purposes, I am commenting vis.c/vis.h alfred code, at the beginning of each function, explaining also most important local variables (keeping in mind documentation guidelines [1]) I would like to know if you are interested in such documentation process. If yes, I could extend this kind of work to the whole alfred code. Have a nice day [1]: https://www.kernel.org/doc/Documentation/kernel-doc-nano-HOWTO.txt

2 3

[B.A.T.M.A.N.] [PATCH v2 0/2] batman-adv: hard interface removal fixes
by Linus Lüssing 29 Sep '16

29 Sep '16

Hi, This patchset contains another, revived patch (1/2). As we had quite some trouble validating the patches back then in Aalborg, I tried to push the overall structure in a hopefully easier to read and comprehend, API like style. While keeping the initial idea (that is, steal and claim forwarding packets from the queue under spinlock, then wait and free without spinlock as a second step) intact. The second patch is another kernel splat I came across while playing with msleep(s)/mdelay()s in the purging routines. Regards, Linus PS: Especially for 1/2 I'm unsure whether to target it for maint/stable. It didn't seem to be a show stopper for anyone within the last three years (had been the only one reporting it at #168, I think?). And the fix is difficult to do as one-liners, unless risking to introduce new race conditions.

2 5

[B.A.T.M.A.N.] Interested in contributing
by leftbydefault＠riseup.net 28 Sep '16

28 Sep '16

Hello folks, I came across the list of improvements and ideas to work on B.A.T.M.A.N-Adv from Gsoc2013 Ideas here https://wiki.freifunk.net/Ideas_GSoC_2013 I am interested to work on implementing a proposed solution to any of the mentioned problem. The wiki has a list of problems but, there is no description about the current status of each problem listed (whether they have been already implemented or still looking for contributions?). If there are any other problems apart from the above list, I would also like to know those to see if I can contribute.

3 3

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

B.A.T.M.A.N