| Age | Commit message (Collapse) | Author |
|---|
|
* configure.ac: Bump version to 0.99.0
* doc/quagga.info: Version bump (autogenerated)
|
|
if old one is flaged as BGP_INFO_ATTR_CHANGED.
|
|
* zebra/rt_netlink.c: ignore wireless newlink netlink messages.
|
|
use it as well.
* vtysh/vtysh.c: Implement walkup in node tree for vtysh as it already
works in vty.
|
|
"integrated-vtysh-config".
|
|
* log.h: Test for SA_SIGINFO to see whether zlog_signal takes final
two args (siginfo and program_counter).
* log.c: (hex_append) Include this function only if SA_SIGINFO or
HAVE_GLIBC_BACKTRACE is defined.
(zlog_signal) Final two args (siginfo and program_counter) now
depend on whether SA_SIGINFO is defined on this platform.
* sigevent.c: (program_counter) Do not include this function if
SA_SIGINFO is not defined on this platform.
(exit_handler,core_handler) Test for SA_SIGINFO to decide whether
2nd & 3rd arguments are present and to decide how to invoke
zlog_signal.
(trap_default_signals) Test for SA_SIGINFO and invoke sigaction
appropriately.
|
|
functions and some tiny related fixes. No functional changes.
|
|
case of /proc being used. I don't have others to test with at the
moment.
|
|
* log.h: Change prototype for zlog_backtrace_sigsafe to take additional
program_counter argument.
* log.c: (zlog_backtrace_sigsafe) Add additional program_counter
argument. If it is non-NULL, use backtrace_symbols_fd to resolve
the address.
(zlog_signal) Call zlog_backtrace_sigsafe with additional
program_counter argument.
[pullup candidate]
|
|
configured or not. Being not configured is even better reason to
do it.
|
|
[pullup candidate]
|
|
to walk up in the node tree if called from vtysh. Different daemons
might have commands with same syntax in different nodes (for example
"router-id x.x.x.x" commands in zebra/ospfd/ospf6d daemons).
* vtysh.c: Reflect changes in lib. cmd_execute_command() should know
now that it's called from vtysh and must not attempt to walk up in
the node tree.
[pullup candidate]
|
|
* redhat/quagga.spec.in: Fix postun script to avoid misleading error
message saying the postun scriptlet failed when watchquagga
is not running.
(pullup candidate)
|
|
* command.c (print_version): Do not bother even to examine host.name,
since it is always NULL when this function is called from main.
|
|
* command.c (print_version): Don't print host.name if it is NULL.
Fixes segfault on Solaris reported by Goetz von Escher <goetz@open.ch>
(pullup candidate)
|
|
last release..
|
|
* configure.ac: Test for header file <ucontext.h> (for use in
signal processing).
* sigevent.c: (trap_default_signals) Use the SA_SIGINFO flag to
pass additional siginfo_t and ucontext_t arguments to core_handler
and exit_handler.
(core_handler,exit_handler) Now invoked with 3 arguments (using
SA_SIGINFO). Pass additional info to zlog_signal.
(program_counter) New function to find program counter in ucontext_t,
needs to be enhanced to support more platforms (currently works only
on Linux/x86).
* log.h: Change the zlog_signal prototype to add new arguments
siginfo_t * and program_counter.
* log.c: (zlog_signal) Add new arguments siginfo and program_counter.
Include si_addr and program counter (if non-NULL) in message.
And remove #ifdef HAVE_GLIBC_BACKTRACE around hex_append, since
that is now used to render the si_addr and PC pointers.
|
|
* configure.ac: If configure is invoked with --enable-snmp, but
the configure script is unable to find SNMP support on the platform,
then configure should give an error message and exit.
|
|
* redhat/quagga.spec.in: Pass --enable-gcc-rdynamic to configure
to get gcc to link with -rdynamic for better backtraces.
When the rpm is upgraded, the restart logic now works as follows:
1. stop watchquagga; 2. stop all routing daemons; 3. restart zebra
if it was running; 4. start all routing daemons that were running;
and 5. start watchquagga if it was running.
|
|
* watchquagga.c: (run_job) Fix bug: delay.tv_sec was being used
without having been initialized when the "force" argument was true.
(try_restart) When phased restart is stopping the routing daemons,
it should tell run_job to update the restart interval.
(phase_check) In final step of phased restart when the routing
daemons are being started again, tell run_job not to update
the restart interval (this is now done when the daemons are
stopped).
|
|
* zebra.h: If not C99 and no va_copy macro available, fall back to
memcpy (solves a build problem on FreeBSD 4.x).
|
|
* texinfo.tex: Updated to a version which understands @{ and @}
* vtysh.texi: Add node name for @section
* quagga.info: updated build of autogenerated file
|
|
* ospf_packet.h: Remove commented out definition of
OSPF_MAX_PACKET; neither it or the uncommented one are used any more.
* ospf_packet.c (ospf_make_ls_upd): Leave room for authentication
when deciding if an update will fit.
(ospf_packet_authspace): Factor out calculation of size required
for authentication.
(ospf_make_db_desc): Use ospf_max_packet, not OSPF_MAX_PACKET.
Don't confuse readers that there is a macro.
|
|
* configure.ac: Bump version to 0.98.0
* doc/quagga.info: Version bump (autogenerated)
|
|
* configure.ac: Bump version to 0.97.5
* doc/quagga.info: autogenerated, version bump.
|
|
* zserv.c: (zebra_accept) Comment out setting of socket to NONBLOCK
for now, as we dont actually deal with with resending.... See
bugzilla #122, fix from wawa@yandex-team.ru (Vladimir Ivanov).
* kernel_socket.c: (routing_socket) ditto.
|
|
* bgp_packet.c: (bgp_write) set socket to nonblock while writing
this should be generalised. See bugzilla #102. Fix supplied by
wawa@yandex-team.ru (Vladimir Ivanov).
|
|
etc. so configure output shows the choice we made.
(Andrew: please recheck on a buggy system; it configures/compiles fine
on NetBSD.)
|
|
* configure.ac: Added test for broken CMSG_FIRSTHDR macro
(relevant for Solaris 8 and unpatched Solaris 9, don't know
whether other platforms are affected).
* zebra.h: Define ZCMSG_FIRSTHDR appropriately based on whether
config.h indicates HAVE_BROKEN_CMSG_FIRSTHDR (as determined
by the configure test program).
* sockopt.c: (getsockopt_cmsg_data) Use ZCMSG_FIRSTHDR instead
of CMSG_FIRSTHDR.
* rtadv.c: (rtadv_recv_packet,rtadv_send_packet) Use ZCMSG_FIRSTHDR
instead of CMSG_FIRSTHDR.
* ripd.c: (rip_recvmsg) Use ZCMSG_FIRSTHDR instead of CMSG_FIRSTHDR.
* ripngd.c: (ripng_recv_packet) Use ZCMSG_FIRSTHDR instead of
CMSG_FIRSTHDR.
|
|
* NEWS: Note improved logging facilities.
|
|
described in [quagga-dev 2292].
|
|
|
|
|
|
|
|
IOS.
|
|
* ospf_network.c: Improve all setsockopt error messages to give detailed
information on the arguments.
|
|
|
|
* ospf_packet.c: (ospf_db_desc) Reduce severity of "Negotiation done"
messages from LOG_WARNING to LOG_INFO, since this seems to be
normal.
|
|
* ospf_packet.c: (ospf_read) Always look up the interface if
ospf_recv_packet returns NULL ifp, since some platforms such
as Solaris 8 appear to support ifindex retrieval but don't.
|
|
|
|
commit. Replace all if_is_up() calls with if_is_operative().
|
|
* sockopt.c (getsockopt_ipv4_ifindex): Document calling
convention. Beef up comments. Handle the case where the cmsghdr
has a zero controllen, or more specifically when the wanted option
is not present. This is needed for Solaris 8, and in general for
any platform for which configure finds a method and it can fail.
Mark some changes with XXX to be cleaned up post 0.98.
|
|
|
|
* sockopt.c (getsockopt_ipv4_ifindex): Return 0 when passed a NULL
cmsghdr pointer.
I believe this will avoid ospfd crashing on Solaris 8, which seems to
define IP_RECVIF but not actually implement it.
|
|
* configure.ac: Add new option --enable-gcc-rdynamic to link
with -rdynamic.
|
|
* watchquagga.c: In several places, cast pid_t to int for printf to
avoid complaints on Solaris 8.
|
|
* watchquagga.c: Fix headers: get most stuff from zebra.h.
|
|
* sockopt.c: (setsockopt_ipv4_ifindex) Improve error message.
When neither IP_PKTINFO nor IP_RECVIF is defined, make return value
deterministic (-1).
|
|
* thread.c: (funcname_thread_add_timer_msec) Reduce overflow risk.
|
|
|
