gns3 / dynamips Goto Github PK

View Code? Open in Web Editor NEW

348.0 44.0 94.0 2.43 MB

Dynamips development

License: GNU General Public License v2.0

C 98.88% Perl 0.01% CMake 1.04% Python 0.07%

emulator cisco

dynamips's Introduction

Dynamips (Cisco Router Emulator)

Overview

Authors of this document: Fabien Devaux, Christophe Fillot, MtvE, Gordon Russell, Jeremy Grossmann and Flávio J. Saraiva.

Converted to markdown format by Daniel Lintott.

This is a continuation of Dynamips, based on the last development version and improved with patches wrote by various people from the community. This fork was named Dynamips-community up to the 0.2.8-community release and renamed to the original Dynamips on the 0.2.9 release.

You can compile two different versions of Dynamips with this code. Edit the Makefile to set the flags to suit your environment. One of the flags, DYNAMIPS_CODE, can be "stable" or "unstable".

Unstable is the code which contains most of the development code, and is in particular suitable for use on a 64 bit Mac. Unfortunately this has proved to be unstable on other platforms.

Stable contains the same code as Unstable, minus some mips64 bit optimisations and tcb code which seems to trigger instability on a number of platforms. You should probably use stable unless you have a very good reason.

For more information on the how to use Dynamips see the README file

License: GNU GPLv2 only

How to compile Dynamips

Dynamips now uses the CMake build system. To compile Dynamips you will need CMake and a working GCC or Clang compiler, as well as the build dependencies.

Build Dependencies

On Debian based systems the following build dependencies are required and can be installed using apt-get:

libelf-dev
libpcap0.8-dev

On Redhat based systems (CentOS, Fedora etc) the following build dependencies are required and can be installed using yum:

elfutils-libelf-devel
libpcap-devel

Similar packages should be available for most distributions, consult your distributions package list to find them.

MacPort & Homebrew:

libelf
cmake

Windows with Cygwin:

Install Winpcap: https://www.winpcap.org/
Install Cygwin 32-bit (setup-x86.exe): https://cygwin.com/install.html
In Cygwin setup, install the make, cmake, gcc-core and git packages
Additionally, install the libelf0 package (important: both bin and src)
Download and unzip Winpcap developer pack: http://www.winpcap.org/devel.htm
Copy the libraries WpdPack\Lib\libpacket.a and WpdPack\Lib\libwpcap.a to cygwin\lib\
Copy all headers from WpdPack\Include to cygwin\usr\include\

Compiling (Linux/Mac)

Either download and extract a source tarball from the releases page or clone the Git repository using:

git clone git://github.com/GNS3/dynamips.git
cd dynamips
mkdir build
cd build
cmake ..

On OSX Yosemite you need to force usage of GCC 4.9:

cmake ..  -DCMAKE_C_COMPILER=/usr/local/bin/gcc-4.9

And for building stable release:

cmake .. -DDYNAMIPS_CODE=stable  -DCMAKE_C_COMPILER=/usr/local/bin/gcc-4.9

This will generate the Makefiles required for compiling Dynamips. To just build Dynamips simple run:

make

or to build and install Dynamips run:

make install

The specify a differant installation location run:

cmake -DCMAKE_INSTALL_PREFIX=/target/path ..

Compiling (Windows)

Open the Cygwin terminal.

First, the libelf has to be manually compiled and installed:

<MIRROR_DOWNLOADS> is the directory used by your Cygwin mirror to download packages. It is possible that the libelf version differs from below.

cp <MIRROR_DOWNLOADS>/x86/release/libelf/libelf0/libelf0-0.8.13-2-src.tar.bz2 .
mkdir libelf && tar xvjf libelf0-0.8.13-2-src.tar.bz2 -C libelf
cd libelf
tar xvzf libelf-0.8.13.tar.gz
cd libelf-0.8.13
./configure
make
make install

Then, Dynamips can be build:

git clone git://github.com/GNS3/dynamips.git
cd dynamips
mkdir build
cd build
cmake ..
make

You will find dynamips.exe in the stable directory. Put cygwin1.dll from the Cygwin bin directory in the same directory as dynamips.exe to be able to start it from outside Cygwin terminal.

Releasing

Update ChangeLog
In common/dynamips.c update sw_version_tag with date
Update RELEASE-NOTE
Update CMakeLists.txt
git tag the release

Useful Information

Forum: https://gns3.com/community

Repository: https://github.com/GNS3/dynamips

Bugtracker: https://github.com/GNS3/dynamips/issues

dynamips's People

Contributors

Stargazers

Watchers

Forkers

nickbeee shiningdrops chiehwen agencia delciotorres jrethmann ststewar sdwyer872 flaviojs kaishuu0123 candlerb jeffball55 legun behnamyazdani asitti loganbr nulng wufengyi smuniz bevaz russlank wolf9s citixenken sanyaade-teachings notyours180 atomicstudiodevelopment oscaritox2000 zhuyue1314 n3wtype istudy78 freeshare2011 spiritspb imyzcn rexsi sujggit llbuaa liyezhong ocochard demohunter jmcabandara ddragic di-ken cybersme gmarciales neotim lastic averroes gitmesam ashang onionknight marciopocebon vishwanathjadhav rbalint wolfking2 fladjdd55 birdluo banyan772 ilayshp wxh0000mm nanamisirone kevanmowery tadryanom redeyecoding luckyfish-tu akheon23 lightiam kyitharko uniqueitmail eldarhuseynov2463 jundong galihuotui xytoki chaos-monkey-island hmidani-abdelilah henahkoo jeroenhuisman soliujing crackercat veryblack dresslerfrank rprabhune rubyw4ves alejandrogom1 katgalraghu mmmrahhal johnyu763 1sumprysfdistku lshm theogenentheos hyperia-cn smezaorellana xyzcommander znley jamestiotio

dynamips's Issues

"Frame is Too Long" error in Wireshark

Description:

~~A packet with size >= 64KiB was dumped to the capture file but wireshark only suports up to 65535 bytes in a packet.~~

Sending and receiving packets is done in separate threads.
libpcap/WinPcap isn't thread safe and two threads were dumping at the same time, resulting in the data of one of the frames being out of order. The next frame header contains garbage data, almost always leading to this error (the chance is over 99.99%).

Analysis:

GNS3 uses the dynamips filter called "capture" to dump packets to a .cap file.
Which means I can probably fix it by splitting big packets into multiple chunks.

Different changes might be needed depending on what is triggering it:

either it's sending a huge (tcp?) packet that is meant to be split by the network card during transmission
or it's a bunch of packets sent together as one
or another scenario I can't think of right now

All the limits I found:

WinPcap NICs (Network Interface Cards) have a packet limit of 32KiB (BPF_MAXBUFSIZE in pcap/bfp.h)
libpcap DLPIs (Data Link Provider Interfaces) have a packet limit of 32KiB or 64KiB
libpcap filters have a packet limit of 50KiB
the other cases aren't limited (analysis stuff ignored)

In short:

libpcap/WinPcap doesn't have a global packet size limit of 64KiB
the limits I found do not affect our particular case
the file format supports packets up to 4GiB in size (some places treat it as signed, so 2GiB is safer)

We are setting low snapshot values (max size of captured packets) and we don't respect it's value when dumping.
I'll change it to ~~be configurable,~~ default to the max value supported by wireshark and respect it while dumping.
NOTE: wireshark ignores the snapshot value

(continued in the next comment)

Links:

http://forum.gns3.net/post31730.html
http://ask.wireshark.org/questions/23620/an-error-while-capturing-with-gns-3

cmake fails

I built a new Linux box (Mint 17.0) and pulled the dynamips repo as per the instructions:

git clone git://github.com/GNS3/dynamips.git
cd dynamips
mkdir build
cd build
cmake ..

but the cmake fails:

user@GNS3v1 ~/GNS3/Source/dynamips/build $ cmake ..
-- CMAKE_MODULE_PATH=/home/user/GNS3/Source/dynamips/cmake
-- The C compiler identification is GNU 4.8.2
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- broken
CMake Error at /usr/share/cmake-2.8/Modules/CMakeTestCCompiler.cmake:61 (message):
  The C compiler "/usr/bin/cc" is not able to compile a simple test program.

  It fails with the following output:

   Change Dir: /home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeTmp

  

  Run Build Command:/usr/bin/make "cmTryCompileExec2932928433/fast"

  /usr/bin/make -f CMakeFiles/cmTryCompileExec2932928433.dir/build.make
  CMakeFiles/cmTryCompileExec2932928433.dir/build

  make[1]: Entering directory
  `/home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeTmp'

  /usr/bin/cmake -E cmake_progress_report
  /home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeTmp/CMakeFiles 1

  Building C object
  CMakeFiles/cmTryCompileExec2932928433.dir/testCCompiler.c.o

  /usr/bin/cc -o CMakeFiles/cmTryCompileExec2932928433.dir/testCCompiler.c.o
  -c
  /home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeTmp/testCCompiler.c

  Linking C executable cmTryCompileExec2932928433

  /usr/bin/cmake -E cmake_link_script
  CMakeFiles/cmTryCompileExec2932928433.dir/link.txt --verbose=1

  /usr/bin/cc CMakeFiles/cmTryCompileExec2932928433.dir/testCCompiler.c.o -o
  cmTryCompileExec2932928433 -rdynamic

  /usr/bin/ld: cannot find crt1.o: No such file or directory

  /usr/bin/ld: cannot find crti.o: No such file or directory

  /usr/bin/ld: cannot find -lc

  /usr/bin/ld: cannot find crtn.o: No such file or directory

  collect2: error: ld returned 1 exit status

  make[1]: Leaving directory
  `/home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeTmp'

  make[1]: *** [cmTryCompileExec2932928433] Error 1

  make: *** [cmTryCompileExec2932928433/fast] Error 2

  

  

  CMake will not be able to correctly generate this project.
Call Stack (most recent call first):
  CMakeLists.txt:27 (project)


-- Configuring incomplete, errors occurred!
See also "/home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeOutput.log".
See also "/home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeError.log".

And the files mentioned above look like this:

user@GNS3v1 ~/GNS3/Source/dynamips/build $ cat /home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeOutput.log
The system is: Linux - 3.13.0-24-generic - i686
Compiling the C compiler identification source file "CMakeCCompilerId.c" succeeded.
Compiler: /usr/bin/cc 
Build flags: 
Id flags: -c

The output was:
0


Compilation of the C compiler identification source "CMakeCCompilerId.c" produced "CMakeCCompilerId.o"

The C compiler identification is GNU, found in "/home/user/GNS3/Source/dynamips/build/CMakeFiles/2.8.12.2/CompilerIdC/CMakeCCompilerId.o"

And

user@GNS3v1 ~/GNS3/Source/dynamips/build $ cat /home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeError.log
Compiling the C compiler identification source file "CMakeCCompilerId.c" failed.
Compiler: /usr/bin/cc 
Build flags: 
Id flags: 

The output was:
1
/usr/bin/ld: cannot find crt1.o: No such file or directory
/usr/bin/ld: cannot find crti.o: No such file or directory
/usr/bin/ld: cannot find -lc
/usr/bin/ld: cannot find crtn.o: No such file or directory
collect2: error: ld returned 1 exit status


Determining if the C compiler works failed with the following output:
Change Dir: /home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeTmp

Run Build Command:/usr/bin/make "cmTryCompileExec2932928433/fast"
/usr/bin/make -f CMakeFiles/cmTryCompileExec2932928433.dir/build.make CMakeFiles/cmTryCompileExec2932928433.dir/build
make[1]: Entering directory `/home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeTmp'
/usr/bin/cmake -E cmake_progress_report /home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeTmp/CMakeFiles 1
Building C object CMakeFiles/cmTryCompileExec2932928433.dir/testCCompiler.c.o
/usr/bin/cc    -o CMakeFiles/cmTryCompileExec2932928433.dir/testCCompiler.c.o   -c /home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeTmp/testCCompiler.c
Linking C executable cmTryCompileExec2932928433
/usr/bin/cmake -E cmake_link_script CMakeFiles/cmTryCompileExec2932928433.dir/link.txt --verbose=1
/usr/bin/cc       CMakeFiles/cmTryCompileExec2932928433.dir/testCCompiler.c.o  -o cmTryCompileExec2932928433 -rdynamic 
/usr/bin/ld: cannot find crt1.o: No such file or directory
/usr/bin/ld: cannot find crti.o: No such file or directory
/usr/bin/ld: cannot find -lc
/usr/bin/ld: cannot find crtn.o: No such file or directory
collect2: error: ld returned 1 exit status
make[1]: Leaving directory `/home/user/GNS3/Source/dynamips/build/CMakeFiles/CMakeTmp'
make[1]: *** [cmTryCompileExec2932928433] Error 1
make: *** [cmTryCompileExec2932928433/fast] Error 2

add hypervisor module for events

GNS3 isn't notified when a router stops, which means it's an inconsistent state when the IOS crashes for some reason.

Events:

notify when a router changes state (stop, start, suspend)
...

C3745 has an emulated NVRAM that is not continuous, so we can generate an invalid checksum when the config is written

Symptoms:

You lose the configuration files when you do write mem or copy run start, stop the router and start again.

Details:

The current code assumes the emulated NVRAM and it's backup are continuous blocks of 128 KiB.

After erasing the nvram, it says 155640 bytes total (155588 bytes free), so it's actually bigger.

After writing the maximum number of files (149) and inspecting the rom file I got these notes:

(block 0xB0000 0x20000)

start 0xB0000

sec 149 0xB0C00
...
sec 40 0xCC000
sec 39 0xCC400
sec 38 0xCC800
sec 37 0xCCC00
sec 36 0xCD000
sec 35 0xCD400
sec 34 0xCD800
sec 33 0xCDC00
sec 32 0xCE000
sec 31 0xCE400
sec 30 0xCE800
sec 29 0xCEC00
sec 28 0xCF000
sec 27 0xCF400
sec 26 0xCF800
sec 25 0xCFC00

(block 0xD0000 0x1C000)

start 0xD0000

sec 149 0xD0C00
...
sec 92 0xDF000
sec 91 0xDF400
sec 90 0xDF800
sec 89 0xDFC00
sec 88 0xE0000
sec 87 0xE0400
sec 86 0xE0800
sec 85 0xE0C00
sec 84 0xE1000
sec 83 0xE1400
sec 82 0xE1800
sec 81 0xE1C00
sec 80 0xE2000
sec 79 0xE2400
sec 78 0xE2800
sec 77 0xE2C00
sec 76 0xE3000
sec 75 0xE3400
sec 74 0xE3800
sec 73 0xE3C00
sec 72 0xE4000
sec 71 0xE4400
sec 70 0xE4800
sec 69 0xE4C00
sec 68 0xE5000
sec 67 0xE5400
sec 66 0xE5800
sec 65 0xE5C00
sec 64 0xE6000
sec 63 0xE6400
sec 62 0xE6800
sec 61 0xE6C00
sec 60 0xE7000
sec 59 0xE7400
sec 58 0xE7800
sec 57 0xE7C00
sec 56 0xE8000
sec 55 0xE8400
sec 54 0xE8800
sec 53 0xE8C00
sec 52 0xE9000
sec 51 0xE9400
sec 50 0xE9800
sec 49 0xE9C00
sec 48 0xEA000
sec 47 0xEA400
sec 46 0xEA800
sec 45 0xEAC00
sec 44 0xEB000
sec 43 0xEB400
sec 42 0xEB800
sec 41 0xEBC00

(block 0xEC000 0x6000)

sec 24 0xEC000
sec 23 0xEC400
sec 22 0xEC800
sec 21 0xECC00
sec 20 0xED000
sec 19 0xED400
sec 18 0xED800
sec 17 0xEDC00
sec 16 0xEE000
sec 15 0xEE400
sec 14 0xEE800
sec 13 0xEEC00
sec 12 0xEF000
sec 11 0xEF400
sec 10 0xEF800
sec 9 0xEFC00
sec 8 0xF0000
sec 7 0xF0400
sec 6 0xF0800
sec 5 0xF0C00
sec 4 0xF1000
sec 3 0xF1400
sec 2 0xF1800
sec 1 0xF1C00

(block 0xF2000 0xA000)

sec 40 0xF2000
sec 39 0xF2400
sec 38 0xF2800
sec 37 0xF2C00
sec 36 0xF3000
sec 35 0xF3400
sec 34 0xF3800
sec 33 0xF3C00
sec 32 0xF4000
sec 31 0xF4400
sec 30 0xF4800
sec 29 0xF4C00
sec 28 0xF5000
sec 27 0xF5400
sec 26 0xF5800
sec 25 0xF5C00
sec 24 0xF6000
sec 23 0xF6400
sec 22 0xF6800
sec 21 0xF6C00
sec 20 0xF7000
sec 19 0xF7400
sec 18 0xF7800
sec 17 0xF7C00
sec 16 0xF8000
sec 15 0xF8400
sec 14 0xF8800
sec 13 0xF8C00
sec 12 0xF9000
sec 11 0xF9400
sec 10 0xF9800
sec 9 0xF9C00
sec 8 0xFA000
sec 7 0xFA400
sec 6 0xFA800
sec 5 0xFAC00
sec 4 0xFB000
sec 3 0xFB400
sec 2 0xFB800
sec 1 0xFBC00

(block 0xFC000 0)

The data is not continuous, so when we write the config file we produce an invalid checksum if there are files stored after the filesystem address 0x1C000.

Unfortunately the sectors start at the end of the filesystem (to easily expand the config files) so, if there's any file other than startup-config and private-config in the filesystem, we'll produce an invalid checksum when writing the config files.

packet loss with multicast traffic

i'm studying for CCIE, in particular to copy multicast topics...

while studying i faced a very weird issues where packets would be dropped "somewhere in the topology" and in that point, ping to the directly connected neighbor would timeout...

after deep investigations i found those two old topics (2008 and 2009)

http://7200emu.hacki.at/viewtopic.php?t=4537
http://ieoc.com/forums/t/3488.aspx

and this more recent in 2012

http://chasingmyccie.wordpress.com/2012/02/02/avoid-ip-multicast-in-your-gns3-labs

in particular what happen is that packets are lost somewhere and is impossible to ping even a directly connected router when the packet loss is isolated on a particular link

this issue is still present with dynamips 0.2.11

there is missing hardware functionality

I tried a c3745 image and got suspicious messages in the console during startup.
In the log file there was evidence that the hardware isn't fully functional.

Symptoms in the router console:

At startup:

BIST FAILED...
Unknown file system detected.

R1#show flash: chips

Chip information NOT available.

R1#format flash:

%Error formatting flash (No device available)

Sample errors from log file:

CPU0: IO_FPGA: read from unknown addr 0x6, ...
CPU0: IO_FPGA: read from unknown addr 0x16, ...
CPU0: IO_FPGA: read from unknown addr 0x2e, ...
CPU0: IO_FPGA: read from unknown addr 0x30, ...
CPU0: IO_FPGA: read from unknown addr 0x4c, ...
CPU0: IO_FPGA: read from unknown addr 0x10000a, ...
CPU0: IO_FPGA: write to unknown addr 0x16, ...
CPU0: IO_FPGA: write to unknown addr 0x2e, ...
CPU0: IO_FPGA: write to unknown addr 0x30, ...
CPU0: IO_FPGA: write to unknown addr 0x4c, ...
CPU0: IO_FPGA: write to unknown addr 0x10000a, ...
CPU0: PCI: write request (data=0xc0000000) for unknown device ...
CPU0: MTS: read access to undefined address 0x3c080022 ......
ROM: unhandled syscall 0x0000003e ...
ROM: unhandled syscall 0x00000047 ...

Related issues

issue #29 - packet loss with multicast traffic

Telnet output speed

If I remember correctly, before the Dynamips community releases the Console/Telnet output was in chunk of text and not line by line like it is now. Most likely a buffer flush was added somewhere in the code. Apparently some users miss this because it looks faster for them?! I think you don't really have this issue on fast computers.

reimplement nvram_export

Reimplement the standalone tool nvram_export:

use the NVRAM filesystem (fs_nvram.c/.h)
support all known formats
perform extra checks when opening the filesystem (addr, ~~checksum~~, backup filesystem)
~~command line options to select the platform or provide custom format settings~~ (meh, too lazy =~~)
extract private-config too (optional extra argument)

Rename issues

If I create a router named "R1" and rename it R2 using vm rename R1 R2 I get the following problems:

vm list returns '101 x (c7200)', the x must be a default value when a vm cannot be found.

vm delete R2 fails with "unable to delete VM 'R2'"

libelf not found error on osx10.8.5

hi,all:
i try to compile the dynamips on my mac os,but it throw errors like below:

make -C stable
Linking rom2c
../common/rom2c.c:15:10: fatal error: 'libelf.h' file not found
#include <libelf.h>
         ^
1 error generated.
make[1]: *** [rom2c] Error 1
make: *** [dynamips.stable] Error 2

Then i use homebrew to install the libelf. And it still doesn't work.I install it on /usr/local/lib/libelf.a and the header file on /usr/local/include/libelf/libelf.h.So i try to modify the files that use it and change the #include <libelf.h> to #include <libelf/libelf.h>.And it finally be compiled.

The compiled result:

make -C stable
Compiling ../common/dynamips.c
..........
Compiling ../common/gen_eth.c
Linking dynamips
ld: warning: directory not found for option '-L/opt/local/lib'
Compiling ../common/nvram_export.c
Linking nvram_export
mv stable/dynamips dynamips.stable
mv stable/nvram_export nvram_export.stable

I wonder if this is a bug on mac os?

max os details:

➜  dynamips-0.2.12  uname -a
Darwin admintekiMacBook-Pro.local 12.5.0 Darwin Kernel Version 12.5.0: Sun Sep 29 13:33:47 PDT 2013; root:xnu-2050.48.12~1/RELEASE_X86_64 x86_64

libelf details:

➜  dynamips-0.2.12  brew info libelf
libelf: stable 0.8.13 (bottled)
http://www.mr511.de/software/
/usr/local/Cellar/libelf/0.8.13 (11 files, 264K) *
  Poured from bottle
From: https://github.com/mxcl/homebrew/commits/master/Library/Formula/libelf.rb

[CRITICAL] Dynamips Router crash with IDLE PC value !!!

Short:
When I run a particular Cisco IOS v15 image on Dynamips it works as long as no idle-pc value attached.
After I attach an IDLE-PC value, IOS would CRASH !!
Without IDLE-PC it boots and works fine. (at the cost of 100% CPU usage per core)
Have you experienced anything similar ? (This never happened to me with older IOS v12.3 series)

date: 25.06.2014.
idlepc = 0x60788d00 (also tried with 0x6077c110 and same result !)
IOS image = c7200-adventerprisek9_sna-mz.150-1.M.bin (also happens with: c7200-spservicesk9-mz.150-1.M.bin and c7200-advipservicesk9-mz.150-1.M.bin)
GNS3 = 0.8.6 on Windows 7 x64
Dynamips v0.2.10-x86 (comes with GNS3 all-in-one installer)
I allocated 512 MB RAM for IOS. (16 GB RAM on host)

Booting IOS v15 on Dynamips with IDLE-PC value attached:

Cisco IOS Software, 7200 Software (C7200-ADVENTERPRISEK9_SNA-M), Version 15.0(1)M, RELEASE SOFTWARE (fc2)
Technical Support: http://www.cisco.com/techsupport
Copyright (c) 1986-2009 by Cisco Systems, Inc.
Compiled Wed 30-Sep-09 07:48 by prod_rel_team
Image text-base: 0x6000F624, data-base: 0x6432CC80

: TLB (store) exception, CPU signal 10, PC = 0x633AE134