- ipv6calc - lists.deepspace6.net

Re: [ipv6calc] [PATCH] 6rd local prefix calculation
by Peter Bieringer 16 Sep '11

16 Sep '11

Hi Raphael, I did now some adjustments, autodetection and checks and also extend the online help. $ ./ipv6calc --action -h Available action types: auto : Automatic selection of action (default) geneui64 : Converts a MAC address to an EUI-64 address conv6to4 : Converts IPv4 address <-> 6to4 IPv6 address (prefix) genprivacyiid : Generates a privacy interface ID out of a given one and a token prefixmac2ipv6 : Generates an IPv6 address out of a prefix and a MAC address anonymize : Anonymize IPv4/IPv6 address without loosing much information 6rd_local_prefix : Calculate the 6rd prefix from given IPv6 prefix & relay prefix and IPv4 Required options: --6rd_prefix ... --6rd_relay_prefix ... Command line: $ ./ipv6calc/ipv6calc --action 6rd_local_prefix --6rd_prefix 2607:fedc:ff40::/43 --6rd_relay_prefix 6.230.0.0/15 6.231.32.33 No input type specified, try autodetection...found type: ipv4addr No output type specified, try autodetection...found type: ipv6addr 2607:fedc:ff52:210::/60 (autodetect messages can be suppressed using -q) Please try latest version from CVS and confirm proper working - thank you! Regards, Peter Am 15.09.2011 20:07, schrieb Raphael Assenat: > Hello Peter, > > It is true that generally, the IPv6 prefix may change > less often. As long as the local IP stays in the same subnet, > we can reasonably expect the relay prefix and relay IP > (not used here) to stay the same. > > However, with my ISP, I receive this informations from DHCP so I > consider all the parameters to be dyanamic anyway. I will always > call ipv6calc with values according to variables within > a script based on the most recently received information. > Therefore, at least for my use case, having the local IP or > the IPv6 prefix as main parameter makes no difference. > > Best regards, > Raphaël Assénat > > On Thu, Sep 15, 2011 at 07:17:39PM +0200, Peter Bieringer wrote: >> Hi Raphael, >> >> patch applied, working. >> >> But I have a question regarding the main input argument: >> >> Your implementation: >> >> ./ipv6calc --action 6rd_local_prefix --in ipv6addr --out ipv6addr >> --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip 6.231.32.33 >> 2607:fedc:ff40::/43 >> 2607:fedc:ff52:210::/60 >> >> But I assume, the IPv6 prefix is a more static one that the local IP. >> >> Because ipv6calc also supports bulk calculations by taking the variable >> argument from stdin, this would be currently the prefix, while I assume, >> the local IP is better. >> >> echo "2607:fedc:ff40::/43" | ./ipv6calc --action 6rd_local_prefix --in >> ipv6addr --out ipv6addr --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip >> 6.231.32.33 >> >> vs. (implementation must be changed): >> >> echo "6.231.32.33" | ./ipv6calc --action 6rd_local_prefix --in ipv4addr >> --out ipv6addr --6rd_prefix 2607:fedc:ff40::/43 --6rd_relay_prefix >> 6.230.0.0/15 --6rd_local_ip >> >> Opinion? >> >> Peter >> >> At 15.09.2011 05:00, Raphael Assenat wrote: >>> Hello Peter, >>> >>> Here's the patch I omitted in my previous message... >>> >>> Best regards, >>> Raphaël Assénat >>> >>> On Wed, Sep 14, 2011 at 08:39:57PM +0200, Peter Bieringer wrote: >>>> Hi Raphael, >>>> >>>> I can apply your patch but it did not compile: >>>> >>>> librfc5569.c: In function 'librfc5569_calc_6rd_local_prefix': >>>> librfc5569.c:32:26: error: 'struct in6_addr' has no member named 'in6_u' >>>> >>>> Problem is here: >>>> >>>> sixrd_prefix->in6_addr.in6_u.u6_addr8[i>>3] |= 0x80>> (i&0x7); >>>> >>>> Looks like on Fedora 15, the header files are different than on your system. >>>> >>>> Can you rewrite this section using my independend wrapper functions >>>> >>>> ipv6addr_getoctett and ipv6addr_setoctett (looks I have to fix the "tt" >>>> on next release...) instead of deep structure manipulations? >>>> >>>> Please send a patch on top of your previous one. >>>> >>>> Thank you! >>>> Peter >>>> >>>> >>>> At 13.09.2011 19:07, Raphael Assenat wrote: >>>>> Hello, >>>>> >>>>> I recently began playing with the IPv6 service my ISP provides with 6rd. Calculating >>>>> the local prefix manually was tedious so I wrote the attached patch for ipv6calc. >>>>> >>>>> The local prefix is obtained by specifying the 6rd prefix, 6rd relay prefix >>>>> and local IPv4 address to ipv6calc. >>>>> >>>>> The operation is in fact a simple concatenation of the 6rd prefix with part of the >>>>> local IPv4 address (according to the relay prefix length). >>>>> >>>>> Consider the following (fictive) parameters: >>>>> >>>>> 6rd prefix: 2607:fedc:ff40::/43 >>>>> 6rd_relay_prefix: 6.230.0.0/15 >>>>> local ip: 6.231.32.33 >>>>> >>>>> ./ipv6calc --action 6rd_local_prefix --in ipv6addr 2607:fedc:ff40::/43 --out ipv6addr --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip 6.231.32.33 >>>>> 2607:fedc:ff52:210::/60 >>>>> >>>>> Please review and consider applying. In any case, I think equivalent functionality >>>>> would be a very useful feature to have in ipv6calc. >>>>> >>>>> Best regards, >>>>> Raphaël Assénat >>>>> >>>>

1 0

Re: [ipv6calc] [PATCH] 6rd local prefix calculation
by Peter Bieringer 15 Sep '11

15 Sep '11

Hi Raphael, patch applied, working. But I have a question regarding the main input argument: Your implementation: ./ipv6calc --action 6rd_local_prefix --in ipv6addr --out ipv6addr --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip 6.231.32.33 2607:fedc:ff40::/43 2607:fedc:ff52:210::/60 But I assume, the IPv6 prefix is a more static one that the local IP. Because ipv6calc also supports bulk calculations by taking the variable argument from stdin, this would be currently the prefix, while I assume, the local IP is better. echo "2607:fedc:ff40::/43" | ./ipv6calc --action 6rd_local_prefix --in ipv6addr --out ipv6addr --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip 6.231.32.33 vs. (implementation must be changed): echo "6.231.32.33" | ./ipv6calc --action 6rd_local_prefix --in ipv4addr --out ipv6addr --6rd_prefix 2607:fedc:ff40::/43 --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip Opinion? Peter At 15.09.2011 05:00, Raphael Assenat wrote: > Hello Peter, > > Here's the patch I omitted in my previous message... > > Best regards, > Raphaël Assénat > > On Wed, Sep 14, 2011 at 08:39:57PM +0200, Peter Bieringer wrote: >> Hi Raphael, >> >> I can apply your patch but it did not compile: >> >> librfc5569.c: In function 'librfc5569_calc_6rd_local_prefix': >> librfc5569.c:32:26: error: 'struct in6_addr' has no member named 'in6_u' >> >> Problem is here: >> >> sixrd_prefix->in6_addr.in6_u.u6_addr8[i>>3] |= 0x80 >> (i&0x7); >> >> Looks like on Fedora 15, the header files are different than on your system. >> >> Can you rewrite this section using my independend wrapper functions >> >> ipv6addr_getoctett and ipv6addr_setoctett (looks I have to fix the "tt" >> on next release...) instead of deep structure manipulations? >> >> Please send a patch on top of your previous one. >> >> Thank you! >> Peter >> >> >> At 13.09.2011 19:07, Raphael Assenat wrote: >>> Hello, >>> >>> I recently began playing with the IPv6 service my ISP provides with 6rd. Calculating >>> the local prefix manually was tedious so I wrote the attached patch for ipv6calc. >>> >>> The local prefix is obtained by specifying the 6rd prefix, 6rd relay prefix >>> and local IPv4 address to ipv6calc. >>> >>> The operation is in fact a simple concatenation of the 6rd prefix with part of the >>> local IPv4 address (according to the relay prefix length). >>> >>> Consider the following (fictive) parameters: >>> >>> 6rd prefix: 2607:fedc:ff40::/43 >>> 6rd_relay_prefix: 6.230.0.0/15 >>> local ip: 6.231.32.33 >>> >>> ./ipv6calc --action 6rd_local_prefix --in ipv6addr 2607:fedc:ff40::/43 --out ipv6addr --6rd_relay_prefix 6.230.0.0/15 --6rd_local_ip 6.231.32.33 >>> 2607:fedc:ff52:210::/60 >>> >>> Please review and consider applying. In any case, I think equivalent functionality >>> would be a very useful feature to have in ipv6calc. >>> >>> Best regards, >>> Raphaël Assénat >>> >> -- Dr. Peter Bieringer mailto:pb@bieringer.de Heideckstr. phone: +49-89-36109687 D-80637 Muenchen fax: +49-89-36109689 Germany mobile: +49-174-9015046

1 0

Test mesage
by Peter Bieringer 15 Sep '11

15 Sep '11

Hi, sorry, only a test. Regards, Peter

1 0

ipv6calc and ipv6loganon pipe support added/fixed for next release (0.90.0 is on the way)
by Peter Bieringer 15 May '11

15 May '11

Hi, I've commited to CVS latest changes with following major update: - ipv6calc is now able to read input from stdin (got feature request) Example: $ cat /proc/net/if_inet6 | awk '{ print $1 " " $3 }' | ./ipv6calc --ifinet6_to_compressed fe80::38cd:70ff:fe**:****/64 ::1/128 - ipv6loganon is fixed to support pipe also proper, so for on-the-fly anonymization in Apache following can be used: CustomLog "|/usr/local/bin/ipv6loganon -a -f /var/log/httpd/access_log" combined or in conjunction with cronolog: CustomLog "|/usr/bin/ipv6loganon -f |/usr/sbin/cronolog /var/log/httpd/access.log-%Y%m%d" combined Some other features are added like GeoIP IPv6 support and default location for IP2Location database files. Please test current CVS version and report problems, if any. I would probably release final version in a week or so. Regards, Peter

1 0

Re: ipv6calc build system improvements
by Peter Bieringer 06 Jan '09

06 Jan '09

Hi Peter, At 31.12.2008 13:07, Peter Volkov wrote: > Hi. > > В Срд, 31/12/2008 в 11:38 +0100, Peter Bieringer пишет: >>> Second patch is rather big but it modifies rather small number of >>> things: >> I can't apply this patch automagically, patch --dry-run claims a lot. >> Could you please send me a patch against vanilla 0.72.0 - thank you very >> much! > > It was created against vanila ipv6cacl, and I failed to reproduce this > problem. Probably you need to add -p0 option to patch, since by default > it assumes -p1. IOW, Ah, ok, applied and committed to CVS. Please test from CVS, if all is working fine, I would release 0.72.1. BTW: perhaps you have an idea to suppress following message also: config.status: WARNING: 'man/Makefile.in' seems to ignore the --datarootdir setting Regards, Peter

2 2

Re: ipv6calc build system improvements
by Peter Bieringer 31 Dec '08

31 Dec '08

Hi Peter, At 28.12.2008 23:17, Peter Volkov wrote: > Hi Peter. I wrote two patches to improve build system of ipv6calc a bit. > You can find them in attachment. > > First patch optional-geoip makes configure script work as it should in > case user provides --disable-geoip. Ok. > Second patch is rather big but it modifies rather small number of > things: I can't apply this patch automagically, patch --dry-run claims a lot. Could you please send me a patch against vanilla 0.72.0 - thank you very much! Best regards, Peter

1 0

ipv6calcweb and -T
by Peter Bieringer 23 Nov '08

23 Nov '08

Hi, At 19.11.2008 23:32, JC Janos wrote: > make test > > ... > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/ipv6logstats' > make[1]: Entering directory `/usr/local/src/ipv6calc-0.72.0/ipv6calcweb' > # Pseudo random selected > ./test_ipv6calcweb.sh 2001:200:1:ea::1 > Content-type: text/html > > Insecure $ENV{ENV} while running with -T switch at ./ipv6calcweb.cgi line 727. > make[1]: *** [test] Error 2 I can't reproduce this and also I do not understand that Perl claims about it, because I did already the ENV checking and assign only $1 from a regex to the Perl variable. Could you try to figure the reason out and if successful, pls. send me the patch. BTW: I'm using perl-5.10.0-38.fc9.i386 on Fedora 9 Thank you very much! Regards, Peter

1 0

Re: [ipv6calc] Re: ipv6calc 'make update' failure ?
by Peter Bieringer 19 Nov '08

19 Nov '08

JC Janos schrieb: > Hi, > > On Wed, Nov 19, 2008 at 1:43 PM, Peter Bieringer <pb(a)bieringer.de> wrote: >> Pls. try the new 0.72.0 version now, uploaded some minutes before. > > Close, I believe ... > > cd /usr/local/src/ipv6calc-0.72.0/ > > > ./configure \ > --prefix=/usr/local \ > --bindir /usr/local/bin \ > --mandir /usr/local/share/man \ > --enable-geoip > > make update > make > make test > > ... > IPv6/Teredo/RESERVED 0 > IPv6/Teredo/UNKNOWN 0 > > All tests were successfully done! > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/ipv6logstats' > make[1]: Entering directory `/usr/local/src/ipv6calc-0.72.0/ipv6calcweb' > # Pseudo random selected > ./test_ipv6calcweb.sh 2001:200:1:ea::1 > Content-type: text/html > > ?? Use of uninitialized value $bin_ipv6calc in -f at ./ipv6calcweb.cgi line 659. > ?? Error: missing needed program > ?? make[1]: *** [test] Error 1 > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/ipv6calcweb' > make[1]: Entering directory `/usr/local/src/ipv6calc-0.72.0/ipv6calc' > make > make[2]: Entering directory `/usr/local/src/ipv6calc-0.72.0/ipv6calc' > ... That's strange, could you please check: cd /usr/local/src/ipv6calc-0.72.0/ipv6calcweb ll ../ipv6calc/ipv6calc This should already exist > sudo make install > > ... > for f in *.8; do \ > install -m 644 $f /usr/local/share/man/man8; \ > done > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/man' > make[1]: Entering directory `/usr/local/src/ipv6calc-0.72.0/lib' > make[1]: Nothing to be done for `install'. > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/lib' > make[1]: Entering directory `/usr/local/src/ipv6calc-0.72.0/md5' > ?? make[1]: *** No rule to make target `install'. Stop. > make[1]: Leaving directory `/usr/local/src/ipv6calc-0.72.0/md5' Will be fixed. Peter

1 1

Re: ipv6calc 'make update' failure ?
by Peter Bieringer 19 Nov '08

19 Nov '08

JC Janos schrieb: > Hi, > > On Wed, Nov 19, 2008 at 12:56 PM, Peter Bieringer <pb(a)bieringer.de> wrote: >>> I am confused, how am I supposed to run 'make install' as root if it >>> invokes 'make test', obviously as root, and you advise to "avoid" >>> doing that? >> Hmm, "make installonly" does the job for you. > > Though ChangeLog says, > > 0.19: Peter Bieringer <pb(a)bieringer.de> > Extend "TODO" list > Print also info on "-?" > Fixes in spec file, see changelog at the end of that file > Add "COPYING" and "INSTALLING" > ----> Add "installonly" on "Makefile" to support changes in spec file > > There's no such target in the Makefile, > > make installonly > make: *** No rule to make target `installonly'. Stop. > > grep install Makefile > install: > ${MAKE} install; \ Just removed, no longer needed, will update the ChangeLog and create a next 0.72.0... Peter

1 0

Re: ipv6calc 'make update' failure ?
by Peter Bieringer 19 Nov '08

19 Nov '08

JC Janos schrieb: > On Wed, Nov 19, 2008 at 12:37 PM, Peter Bieringer <pb(a)bieringer.de> wrote: >> Yes, that's right... >> >> Could you try now the 0.72.0 release? >> >> ftp://ftp.bieringer.de/pub/linux/IPv6/ipv6calc/ >> >> If it works for you, then I will release it officially. > > Certainly, I'll try it in a moment ... > > I am confused, how am I supposed to run 'make install' as root if it > invokes 'make test', obviously as root, and you advise to "avoid" > doing that? Hmm, "make installonly" does the job for you. > Also, from a cvs a few moments ago, running 'make test' as user "test", > > whoami > test > > Completes, but shows one error, > > ... > make[1]: Leaving directory `/usr/local/src/ipv6calc/ipv6logstats' > make[1]: Entering directory `/usr/local/src/ipv6calc/ipv6calcweb' > # Pseudo random selected > ./test_ipv6calcweb.sh 2001:200:1:ea::1 > Content-type: text/html > > Use of uninitialized value $bin_ipv6calc in -f at ./ipv6calcweb.cgi line 659. > Error: missing needed program > make[1]: *** [test] Error 1 > make[1]: Leaving directory `/usr/local/src/ipv6calc/ipv6calcweb' > make[1]: Entering directory `/usr/local/src/ipv6calc/ipv6calc' > make > make[2]: Entering directory `/usr/local/src/ipv6calc/ipv6calc' > ... That's caused by the non-existend "ipv6calc" program which is required by test_ipv6calcweb.sh And ooops, this points me to a problem in the CGI at all: # Pseudo random selected ./test_ipv6calcweb.sh 2001:200:1:ea::1 Content-type: text/html Support for option '--db-ip2location-ipv4 <IP2Location IPv4 database file>' not compiled in, IP2Location support disabled Support for option '--db-ip2location-ipv6 <IP2Location IPv6 database file>' not compiled in, IP2Location support disabled Support for option '--db-geoip <GeoIP database file>' not compiled in, GeoIP support disabled <html> This should not happen... Peter

1 1