Avoid show() methods from being a NULL function when it is called.
Meanwhile, I did const'ify the handlers so only its debug field can be changed.
Signed-off-by: Vincent Jardin <vincent.jardin@6wind.com>
During signterm (shutdown) ospf6_clean calls route_remove
for brouter_table, during route_remove brouter_table remove
hook function is invoked which in turns calls router_remove
is any of the LSA has max age or cost is infinity,
which leads to stack corruption trying to delete same
route node.
Similar to 'no router ospf6' or 'no area ..' command
where lsdb is cleanedup then brouter route table
clean up is called.
Clean some of route trace to have route related fields.
Ticket:CM-17932
Signed-off-by: Chirag Shah <chirag@cumulusnetworks.com>
Incomplete commands like "debug ospf6 route mem" were being ignored. The
changes in ripd and ripngd are intended to make the code easier to read,
no bugs were fixed in these two daemons.
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
Convert the list_delete(struct list *) function to use
struct list **. This is to allow the list pointer to be nulled.
I keep running into uses of this list_delete function where we
forget to set the returned pointer to NULL and attempt to use
it and then experience a crash, usually after the developer
has long since left the building.
Let's make the api explicit in it setting the list pointer
to null.
Cynical Prediction: This code will expose a attempt
to use the NULL'ed list pointer in some obscure bit
of code.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
During the loop we save a pointer to the next route in the table in case
brouter is deleted during the course of the loop iteration. However when
we call ospf6_route_remove this can trigger ospf6_route_remove on other
routes in the table, one of which could be pointed at by said pointer.
Since ospf6_route_next locks the route that it returns, it won't
actually be deleted, instead the refcount will go to 1. In the next loop
iteration, nbrouter becomes brouter, and calling ospf6_route_next on
this one will finally decrement the refcount to 0, resulting in a free,
which causes subsequent reads on brouter to be UAF. Since the route will
have OSPF6_ROUTE_WAS_REMOVED set, provided the memory was not
overwritten before we got there, we'll continue on to the next one so it
is unlikely this will cause a crash in production.
Solution implemented is to check if we've deleted the route and continue
if so.
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
This is a fallout from PR #1022 (zapi consolidation). In the early days,
the client daemons would allocate enough memory to send all nexthops
to zebra. Then zebra would add all nexthops to the RIB and respect
MULTIPATH_NUM only when installing the routes in the kernel. Now things
are different and the client daemons can send at most MULTIPATH_NUM
nexthops to zebra, and failure to respect that will result in a buffer
overflow. The MULTIPATH_NUM limit in the new zebra API is a small price
we pay to avoid allocating memory for each route sent to zebra.
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
Commit 427f8e61bb introduced prefix list callbacks
to handle when a prefix list is changed. Unfortunately
if you have ospf6 running but not configured it crashes.
Modify ospf6d to not crash when we are not properly configured
yet for prefix-lists handling.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Oops, I tested this with prefix lists configured and failed to notice it
blows up when nothing is set...
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Using the previously-added vty_frame() support, this gets rid of all the
pointless empty "interface XYZ" blocks that get added for any interface
that shows up in the system (e.g. dummys, tunnels, etc.)
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
FLAG_BLACKHOLE is used for different things in different places. remove
it from the zclient API, instead indicate blackholes as proper nexthops
inside the message.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Add the RMAP_COMPILE_SUCCESS and switch over to using it.
Refactoring allows a removal of a if statement to just
use the switch statement already in place. Additionally
the reworking cleans up memory freeing in a couple of spots.
In one spot we no longer will leak memory too.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
With the old API, ospf6d always needed to send a nexthop address and a
nexthop interface when advertising a route to zebra. In the case where
the nexthop address didn't exist (e.g. connected route), zebra would
take care of ignore it in the zread_ipv6_add() function.
Now, if we have a nexthop interface but not a nexthop address, we not
only can but we should send a nexthop of type NEXTHOP_TYPE_IFINDEX. zebra
won't fix bad nexthops anymore because the clients have a proper API to
send correct messages.
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
Some differences compared to the old API:
* Now the redistributed routes are sent using address-family
independent messages (ZEBRA_REDISTRIBUTE_ROUTE_ADD and
ZEBRA_REDISTRIBUTE_ROUTE_DEL). This allows us to unify the ipv4/ipv6
zclient callbacks in the client daemons and thus remove a lot of
duplicate code;
* Now zebra sends all nexthops of the redistributed routes to the client
daemons, not only the first one. This shouldn't have any noticeable
performance implications and will allow us to remove an ugly exception
we had for ldpd (which needs to know all nexthops of the redistributed
routes). The other client daemons can simply ignore the nexthops if
they want or consult just the first one (e.g. ospfd/ospf6d/ripd/ripngd).
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
Register add/delete hooks with the prefix list code to properly change
ospf6_area's prefix list in/out pointers.
There are 2 other uncached uses of prefix lists in the ASBR route-map
code and the interface code; these should probably be cached too. (To
be fixed another day...)
Fixes: #453
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
This version string has not been updated in over 11 years.
I cannot see any viable reason that we should use or update
or anything with this value, remove.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Start creating a counterpart to frr_init and frr_late_init.
Unfortunately, some daemons don't do any exit handling, this doesn't
change that just yet.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
called list_delete instead of list_free
Moved MTYPE_STATIC in ospfd/zebra.c
Revert changes in ospf6_zebra.c where malloc is called for
multiple nexthops.
Signed-off-by: Chirag Shah <chirag@cumulusnetworks.com>
GCC 7.1.1 returned warnings about buffer sizes
not being big enough to handle the full string
that could be generated.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Free route node upon asbr redistribute route cleanup from
external_id_table route tale.
Free route node when route_remove is called and
node->info is set to null.
Decrement route node lock in route_lookup api as it
is incremented as part of node_lookup api.
use local variable for nexthop vs. malloc in zebra parse
routine.
two of the memory leaks related to nexthops per route were not freed.
two of the memory leak detected per frr service restart
Signed-off-by: Chirag Shah <chirag@cumulusnetworks.com>
This reverts commit c14777c6bf.
clang 5 is not widely available enough for people to indent with. This
is particularly problematic when rebasing/adjusting branches.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
The label initializer & nhrpd variable are just to shut up GCC 7,
the other two are actual bugs.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Signed-off-by: Daniel Walton <dwalton@cumulusnetworks.com>
This allows frr-reload.py (or anything else that scripts via vtysh)
to know if the vtysh command worked or hit an error.
Instead of having an ?: expression embedded in every single caller of
vty_out, just expand \n to \r\n in the vty code if neccessary.
(Deprecation warnings will be enabled in the next commits which will do
the search-and-replace over the codebase.)
[This reverts commit 4d5f445 "lib: add vty_outln()"]
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
rip out this pile of open-coded goo and replace it with uses of the API
that table.h provides.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
route_node->parent and route_node->link shouldn't be touched by user
code since that is a recipe for trouble once we have a hash table in
parallel.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Previous fix was missing the possibility of having to modify the io
buffer size if the kernel reports an new mtu value. This fix adds
that check.
Signed-off-by: Don Slice <dslice@cumulusnetworks.com>
Problem reported by customer that if an mtu value was set in the kernel,
quagga/frr would get very confused about what had been configured and
what had been learned. This caused peers to not be successfully established.
Resolved by keeping a configuration value separate than the operational value
and set the operational accordingly. If configured, it wins unless the config
defines a value that is higher than the kernel supports.
Ticket: CM-16876
Signed-off-by: Don Slice <dslice@cumulusnetworks.com>
Reviewed By: CCR-6399
Testing Done: Manual testing successful, submitter tested, ospf-smoke completed
with no new failures.
log.c provides functionality for associating a constant (typically a
protocol constant) with a string and finding the string given the
constant. However this is highly delicate code that is extremely prone
to stack overflows and off-by-one's due to requiring the developer to
always remember to update the array size constant and to do so correctly
which, as shown by example, is never a good idea.b
The original goal of this code was to try to implement lookups in O(1)
time without a linear search through the message array. Since this code
is used 99% of the time for debugs, it's worth the 5-6 additional cmp's
worst case if it means we avoid explitable bugs due to oversights...
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
ospf6_route_remove may free the ospf6_route passed to it if the refcount
reaches zero, in which case zeroing the ->flag field constitutes a uaf
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
If the user enters a decimal, display a decimal.
If the user enters a dotted quad, display a dotted quad.
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
The 'show zebra' command really shouldn't be owned
by ospf6. This command is a specialized command
to show some basic information about ospf6 and zebra
so limit it to ospf6.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
The command 'debug ospf6 lsa unknown' was
always showing up, upon starting of the ospf6 daemon.
Remove this from happening. Also fix some help strings
while we are in there.
Ticket: CM-7913
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
When we are receiving the 'no debug ospf6 lsa unknown ...'
Allow the removal of the originate examine and flooding
keywords.
Ticket: CM-12805
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
The command 'debug ospf6 lsa unknown' was
always showing up, upon starting of the ospf6 daemon.
Remove this from happening. Also fix some help strings
while we are in there.
Ticket: CM-7913
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
We only needed to add/change the vrf callbacks when we initialize
the vrf subsystem. As such it is not necessary to handle the callbacks
in any other way than through the init function.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
The FSF's address changed, and we had a mixture of comment styles for
the GPL file header. (The style with * at the beginning won out with
580 to 141 in existing files.)
Note: I've intentionally left intact other "variations" of the copyright
header, e.g. whether it says "Zebra", "Quagga", "FRR", or nothing.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Pass pointer to pointer instead of assigning by return value. See
previous commit message.
To ensure that the behavior stays functionally correct, any assignments
with the result of a thread_add* function have been transformed to set
the pointer to null before passing it. These can be removed wherever the
pointer is known to already be null.
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
The way thread.c is written, a caller who wishes to be able to cancel a
thread or avoid scheduling it twice must keep a reference to the thread.
Typically this is done with a long lived pointer whose value is checked
for null in order to know if the thread is currently scheduled. The
check-and-schedule idiom is so common that several wrapper macros in
thread.h existed solely to provide it.
This patch removes those macros and adds a new parameter to all
thread_add_* functions which is a pointer to the struct thread * to
store the result of a scheduling call. If the value passed is non-null,
the thread will only be scheduled if the value is null. This helps with
consistency.
A Coccinelle spatch has been used to transform code of the form:
if (t == NULL)
t = thread_add_* (...)
to the form
thread_add_* (..., &t)
The THREAD_ON macros have also been transformed to the underlying
thread.c calls.
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
The generic callbacks need to be enabled in order
for some route-map commands to properly work.
I've tried to match up to what is programmed to work
within ospfv3.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
The function parse_type_spec was always looking in argv[0] of
for figuring out the type of the command. Since the type location
could change, use the passed in idx_lsa.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
1) linkstate-id was made optional in one case
2) The ipv4 address was being looked at in the wrong spot
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
The SNMP modules include <net-snmp/net-snmp-config.h>, which won't be
found in off-searchpath directories without SNMP_CFLAGS. Unfortunately
in my tests the files were on the search path even without the flags.
(SNMP_LIBS is not needed because only libfrrsnmp calls into net-snmp
functions.)
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
OSPF distance commands were broken in a variety of ways. Fix 'em.
* `distance ospf` and `distance ospf6` were accepted commands
* Inconsistent doc strings
* Make use of {keyword|tokens}
* Add ability to reset specific distance without specifying a value
Ex: ~# no distance ospf6 intra
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
These have copies in vtysh that do the node-switch locally and are
listed in extract.pl's ignore list. The ignore list however is
redundant since DEFUN_NOSH does the same thing...
ldpd is a bit hacky, but Renato is reworking this anyway.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
This means there are no ties into the SNMP code anymore other than the
init call at startup.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
The protocols enum serves no purpose other than adding potential for
bugs and making it complicated to add a new protocol... nuke.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Contains the fetch-and-run-thread logic, and vty startup (which is the
last thing happening before entering the main loop).
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Centralise read_config/daemonize/dryrun/pidfile/vty_serv into libfrr.
This also makes multi-instance pid/config handling available as part of
the library. It's only wired up in ospfd, but the code is in lib/.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Start centralising startup & option parsing into the library.
FRR_DAEMON_INFO is a bit weird, but it will become useful later (e.g.
for killing the ZLOG_* enum, and having the daemon name available)
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Converting a 'struct prefix6' to a 'struct prefix'
and then doing a memcpy of the contents writes
beyond the end of the data structure.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
The library libzebra that is installed with FRR will
conflict with Quagga. So let's rename it to libfrr.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Since zebra might be sending srcdest routes down to the various daemons,
they need to understand the presence of the field at the very least.
Sadly, that's also the best we can do at this point since none of the
protocols has support for handling srcdest routes. The only consistent
thing to do is to ignore them throughout.
If an administrator wants to have the srcdest route as non-srcdest in a
protocol, setting a non-srcdest static route (possibly blackhole) is
probably the best way to go.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
This introduces ZAPI_MESSAGE_SRCPFX, and if set adds a source prefix
field to ZAPI IPv6 route messages sent from daemons to zebra. The
function calls all have a new prefix_ipv6 * argument specifying the
source, or NULL. All daemons currently supply NULL.
Zebra support for processing the field was added in the previous patch,
however, zebra does not do anything useful with the value yet.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Somewhere in the past we switched from
using the auto-generated redistribute statements
to a non-generated version. This caused us to
loose new protocols to redistribute as they are
added. Put it back.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
The next patch will bulk-replace these, and monotime() has time_t as
return value, not an error indicator.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Conflicts (CLI vs. atol()):
- bgpd/bgp_vty.c
- ospfd/ospf_vty.c
- zebra/zebra_vty.c
NB: pull req #65 (LabNConsulting/working/2.0/afi-safi-vty/c) was
excluded from this merge.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
The problem reported was that routers downstream from the ABR would have
prefixes in the IPv6 rib that should have been summarized. Testing showed
that the prefixes were absent in the lsdb but not removed from the rib.
The problem (and others) stemmed from determining the link-state id to use
for the area-range statement only from other entries in the range-table. Since
the area range statement creates inter-prefix LSAs (summaries) which are in all
aspects identical to other inter-prefix LSAs, the number space for the lsid
needs to be unique across all inter-prefix LSAs, not just unique between area
range statements. This fix removes the use of range-table specific lsids and
acquires the lsid just as any other inter-prefix LSA.
Ticket: CM-13626
Signed-off-by: Don Slice
Reviewed By: CCR-5489
Testing Done: Manual testing, repeated runs of the test that previously failed,
ospf-smoke
All of the autogenerated macros in lib/route_types.pl are now called
FRR_* instead of QUAGGA_*.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
This replaces Quagga -> FRR in most configure.ac settings as well as
a handful of preprocessor macros in the source code.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
VTY_DECLVAR_CONTEXT already contains a NULL check, vty warning message
and return statement. These are not needed.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
If we fail to set any socket's buffer size, try again with a smaller value
and keep going until it succeeds. This is better than just giving up or,
even worse, abort the creation of a socket (ospf6d and ripd).
Fix broken ospf6d on FreeBSD.
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
These now generate warnings which will break the build with -Werror.
Note this may have enabled commands that should be disabled, or the
other way around...
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
This moves all install_element calls into the file where the DEFUNs are
located. This fixes several small related bugs:
- ospf6d wasn't installing a "no interface FOO" command
- zebra had a useless copy of "interface FOO"
- pimd's copy of "interface FOO" was not setting qobj_index, which means
"description LINE" commands would fail with an error
The next commit will do the actual act of making "foo_cmd" static.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Additionally:
* Add [ip] to a couple bgp show commands
* Quick refactor of a couple ISIS commands
* Quick refactor of a couple OSPF6 commands
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Headers using struct cmd_token should directly include command.h so the
struct is actually defined.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
vtysh's extract.pl script doesn't cope with install_element(var, ) where
"var" is not one of the FOO_NODE constants. Also, the future defun_lex
tool doesn't deal well with preprocessor expansions in the same regard.
This tries simplifying out some of these. lib/distribute.c needs
further thinking.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Since we have autoconf results from a wide swath of target platforms, we
can go remove checks that have the same result on all systems.
This also removes several "fallback" implementations of functions that,
at some point in the history, weren't available on all target platforms.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
The loop should only iterate to array_size - 1.
Signed-off-by: Christian Franke <chris@opensourcerouting.org>
Signed-off-by: Christian Franke <chris@opensourcerouting.org>
Acked-by: Donald Sharp <sharpd@cumulusnetworks.com>
Until today the admin distance cannot be configured for any IPv6
routing protocol. This patch implements it for ospf6.
Signed-off-by: Maitane Zotes <maz@open.ch>
Signed-off-by: Roman Hoog Antink <rha@open.ch>
Consolidate the routemap initialization into one
function.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Daniel Walton <dwalton@cumulusnetworks.com>
If a command is put into the VIEW_NODE, it is going into the
ENABLE_NODE as well. This is especially true for show commands.
As such if a command is in both consolidate it down to VIEW_NODE.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
* zclient.c: prefix length on router-id and interface address add
messages not sanity checked. fix.
* */*_zebra.c: Prefix length on zebra route read was not checked, and
clients use it to write to storage. An evil zebra could overflow
client structures by sending overly long prefixlen.
Prompted by discussions with:
Donald Sharp <sharpd@cumulusnetworks.com>
Only one of these variables is in fact an interface index...
(this fixes -Werror build.)
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Since recently zebra uses only the ZEBRA_REDISTRIBUTE_* messages
to advertise redistributed routes to its clientes. Now the old
ZEBRA_IPV*_ROUTE_* messages are only used for client->zebra communication.
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
This is a rather large mechanical commit that splits up the memory types
defined in lib/memtypes.c and distributes them into *_memory.[ch] files
in the individual daemons.
The zebra change is slightly annoying because there is no nice place to
put the #include "zebra_memory.h" statement.
bgpd, ospf6d, isisd and some tests were reusing MTYPEs defined in the
library for its own use. This is bad practice and would break when the
memtype are made static.
Acked-by: Vincent JARDIN <vincent.jardin@6wind.com>
Acked-by: Donald Sharp <sharpd@cumulusnetworks.com>
[CF: rebased for cmaster-next]
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Signed-off-by: Christian Franke <chris@opensourcerouting.org>
Move over to the new allocation counting added in the previous commit.
(This commit is mostly mechanical.)
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Acked-by: Vincent JARDIN <vincent.jardin@6wind.com>
Signed-off-by: Daniel Walton <dwalton@cumulusnetworks.com>
Reviewed-by: Donald Sharp <sharpd@cumulusnetworks.com>
Ticket: CM-12773
- change "as-ext" to "as-external"
- drop "grp-mbr" option, it does not have a handler
- drop "type7" option, it does not have a handler
When sending ancillary data on OpenBSD, we need to tell the compiler
that the ancillary data buffer should be aligned on the stack to the
minimum alignment of the first ancillary data object that will be sent.
Additionally, HAVE_SIN6_SCOPE_ID is not defined anywhere, check for
HAVE_STRUCT_SOCKADDR_IN6_SIN6_SCOPE_ID instead.
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
Make it easier to see which bits in *_snmp.c are actually referenced
from non-SNMP parts of the code.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
- HAVE_POLL is overloaded by net-snmp
- missing includes
- ospf6_snmp converted to vrf_iflist()
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Added a default log file named /var/log/quagga/Quagga.log to every daemon
to capture log entries if no log file is defined. This also allows the
capture of logged information prior to reading each daemon's config file.
If a log file is defined manually, it will override this default file name.
Ticket: CM-10987
Signed-off-by: Don Slice
Reviewed By: Donald Sharp
Testing Done: Manual testing
Resolved several memory leaks caused by ifdown/ifup the vrf device or
a swp port. For bgp/zebra/ospf/ospf6, bouncing the vrf device would cause
a linked list, Interface, and route-table to get leaked. For ospf6,
bouncing the swp device also caused leaks of Connected and Prefix entries.
Ticket: CM-10841
Signed-off-by: Don Slice
Reviewed-By: Donald Sharp
Testing Done: Manual testing, bgp and ospf mins passed, smokes had fewer failures than base
Ticket: CM-11256
Signed-off-by: Radhika Mahankali <radhika@cumulusnetworks.com>
Reviewed-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Kanna Rajagopal <kanna@cumulusnetworks.com>
Testing: Unit, PTM smoke, OSPF smoke, BGP Smoke
Issue:
BFD client registrations are not being sent to PTM from BGP/OSPF clients when the quagga clients have no BFD configuration. This can create stale BFD sessions in PTM when BFD is removed from quagga configuration before quagga is restarted.
BFD client de-registrations from BGP/OSPF also go missing sometimes when quagga is restarted. This also will cause stale BFD sessions in PTM.
Root Cause:
BFD client registrations were being sent at the time of BGP/OSPF daemon initialization. But, they were being sent to zebra before the socket connection between zebra and BGP/OSPF was established. This causes the missing BFD client registrations.
BFD client de-registrations are sent from zebra when zebra detects socket close for BGP/OSPF daemons. Based on the timing, the de-registrations may happen after socket between PTM and zebra is closed. This will result in missing de-registrations.
Fix:
Moved sending of BFD client registration messages to zebra connected callback to make sure that they are sent after the BGP/OSPF daemons connect with zebra.
Added BFD client de-registrations for BGP/OSPF to be also sent when zebra daemon gets restart signal. They are sent from the signal handler only if it was not already handled in zebra client socket close callback.
Prior to this change, interface bandwidth could not be defined above 10G. With
the use of higher speed interfaces, the ability to effectively define the path
links was highly impacted. Additionally, the default auto-cost reference-bandwidth
for ospf and ospfv3 was set to 100M, which relects a much earlier time. Changed both
the range of interface bandwidth definitions and reference bandwidths to be up to
100G. Set the default interface bandwidth (if not defined) to 10G to make the ratio
continue to cause a cost of 10 as before. Manual testing as well as ospf-min and
ospf-smoke passed successfully.
Ticket: CM-10756
Signed-of-by: Don Slice
Reviewed-by: Donald Sharp
As the comments in if.h, it is better to call ifname2ifindex()
instead of if_nametoindex().
And ifname2ifindex() can work for VRF by appending a parameter
while if_nametoindex() can not.
Signed-off-by: Feng Lu <lu.feng@6wind.com>
Reviewed-by: Alain Ritoux <alain.ritoux@6wind.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Acked-by: Vincent JARDIN <vincent.jardin@6wind.com>
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit 395828eea809e8b2b8c5824d3639cefedd7aa9f0)
There are preexisting fields u.lp.id and u.lp.adv_router in struct
prefix that do the same thing as these type-punning pointer derefs.
Use these and shut up the strict-aliasing warnings.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit 0de0138a9146074f268245193e45c9376d998722)
Some places, particularly headers, were spewing warnings since they
don't include neccessary other headers to get struct/enum definitions.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit 388f8857eb81ef75014060976776523a58a99389)
On higher warning levels, compilers expect %p printf arguments to be
void *. Since format string / argument warnings can be useful
otherwise, let's get rid of this noise by sprinkling casts to void *
over printf calls.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Since we can't assume time_t to be long, int, or even long long, this
consistently uses %lld/long long (or %llu/unsigned long long in a few
cases) to print time_t/susecond_t values. This should fix a bunch of
warnings, on NetBSD in particular.
(Unfortunately, there seems to be no "PRId64" style printing macro for
time_t...)
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit ef008d2f8dc8f7160d8a3d24a15f2fad79ef3242)
Quagga was using a mix of srand/rand and srandom/random.
Consolidate to use srandom/random which are the POSIX
versions of random number generators
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
INCLUDES in configure.ac was not used at all, and INCLUDES in
Makefile.am is supposed to be AM_CPPFLAGS these days.
Reduces warnings spewed during bootstrap/autoreconf.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Acked-by: Greg Troxel <gdt@ir.bbn.com>
Acked-by: Feng Lu <lu.feng@6wind.com>
Acked-by: Paul Jakma <paul@jakma.org>
(cherry picked from commit 237aac56960575f6ad2451ba2796d94bd5ae4b33)
Fix lots of warnings. Some const and type-pun breaks strict-aliasing
warnings left but much reduced.
* bgp_advertise.h: (struct bgp_advertise_fifo) is functionally identical to
(struct fifo), so just use that. Makes it clearer the beginning of
(struct bgp_advertise) is compatible with with (struct fifo), which seems
to be enough for gcc.
Add a BGP_ADV_FIFO_HEAD macro to contain the right cast to try shut up
type-punning breaks strict aliasing warnings.
* bgp_packet.c: Use BGP_ADV_FIFO_HEAD.
(bgp_route_refresh_receive) fix an interesting logic error in
(!ok || (ret != BLAH)) where ret is only well-defined if ok.
* bgp_vty.c: Peer commands should use bgp_vty_return to set their return.
* jhash.{c,h}: Can take const on * args without adding issues & fix warnings.
* libospf.h: LSA sequence numbers use the unsigned range of values, and
constants need to be set to unsigned, or it causes warnings in ospf6d.
* md5.h: signedness of caddr_t is implementation specific, change to an
explicit (uint_8 *), fix sign/unsigned comparison warnings.
* vty.c: (vty_log_fixed) const on level is well-intentioned, but not going
to fly given iov_base.
* workqueue.c: ALL_LIST_ELEMENTS_RO tests for null pointer, which is always
true for address of static variable. Correct but pointless warning in
this case, but use a 2nd pointer to shut it up.
* ospf6_route.h: Add a comment about the use of (struct prefix) to stuff 2
different 32 bit IDs into in (struct ospf6_route), and the resulting
type-pun strict-alias breakage warnings this causes. Need to use 2
different fields to fix that warning?
general:
* remove unused variables, other than a few cases where they serve a
sufficiently useful documentary purpose (e.g. for code that needs
fixing), or they're required dummies. In those cases, try mark them as
unused.
* Remove dead code that can't be reached.
* Quite a few 'no ...' forms of vty commands take arguments, but do not
check the argument matches the command being negated. E.g., should
'distance X <prefix>' succeed if previously 'distance Y <prefix>' was set?
Or should it be required that the distance match the previously configured
distance for the prefix?
Ultimately, probably better to be strict about this. However, changing
from slack to strict might expose problems in command aliases and tools.
* Fix uninitialised use of variables.
* Fix sign/unsigned comparison warnings by making signedness of types consistent.
* Mark functions as static where their use is restricted to the same compilation
unit.
* Add required headers
* Move constants defined in headers into code.
* remove dead, unused functions that have no debug purpose.
(cherry picked from commit 7aa9dcef80b2ce50ecaa77653d87c8b84e009c49)
Conflicts:
bgpd/bgp_advertise.h
bgpd/bgp_mplsvpn.c
bgpd/bgp_nexthop.c
bgpd/bgp_packet.c
bgpd/bgp_route.c
bgpd/bgp_routemap.c
bgpd/bgp_vty.c
lib/command.c
lib/if.c
lib/jhash.c
lib/workqueue.c
ospf6d/ospf6_lsa.c
ospf6d/ospf6_neighbor.h
ospf6d/ospf6_spf.c
ospf6d/ospf6_top.c
ospfd/ospf_api.c
zebra/router-id.c
zebra/rt_netlink.c
zebra/rt_netlink.h
batch-fix all warnings that come up when enabling AgentX SNMP support.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
(cherry picked from commit dfee58f1d41a2e36c7f5f38a3ef5712224131824)
This fix is probably correct on 32bit systems,
but i think it will not work on 64bit systems.
sizeof(signed long) would be 8 and therefore the
cast from u_int32_t will map all the values to
non-negative part of long int.
You would like to use int (like in ospfd) and
change the type of seqnuma, seqnumb to that.
The type int32_t would be even more proper, but
sizeof(int) is 4 on relevant platforms.
Signed-off: Ondrej Zajicek <santiago@crfreenet.org>
Acked-by: Feng Lu <lu.feng@6wind.com>
Acked-by: Yasuhiro Ohara <yasu@jaist.ac.jp>
(cherry picked from commit bdd8cd70a042473477f9144c9cedb8dde11ba2c1)
* Remove the old change from '08 to add in PIE arguments at automake level.
Versions of libtool since then know how to deal with -fpie and do the right
thing according to whether its building shared or executable objects.
So just pass '-fpie' as CFLAG and let libtool do its thing.
Use correct % code in printf statement.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Don Slice <dslice@cumulusnetworks.com>
Reviewed-by: Daniel Walton <dwalton@cumulusnetworks.com>
CM-10680
Issue: When BGP daemon is stopped, all the BGP BFD sessions are not getting deleted from PTM.
Root cause: BGP daemon stop causes BFD de-register message to be sent for every peer on which BFD is enabled. But, all the de-register messages from bgpd to zebra are not processed before the socket close. This results in some stale BGP BFD sessions.
Fix: Support for client de-register message has been added in PTM/BFD. Changes in Quagga to support BFD client de-registrations:
− The BFD clients de-registration is sent directly from zebra daemon when zebra client (bgpd, ospfd and ospf6d) socket close is detected.
− Introduced a BFD flag for the zebra clients to prevent BFD de-registration messages from being sent to zebra daemon when the client is shutting down. This reduces the BFD messaging.
CM-10540
Issue: Invalid ptm status “fail” instead of “n/a” being displayed for VRF interfaces.
Root cause: ptm status is not being initialized to “unknown” status when VRF interface is added or changed. The uninitialized value is ‘0’ which is the value for “fail”
Fix: Initialized the ptm status to the correct value.
Signed-off-by: Radhika Mahankali <radhika@cumulusnetworks.com>
Reviewed-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Kanna Rajagopal <kanna@cumulusnetworks.com>
Ticket: CM-10680, CM-10540
Reviewed By: CCR-4653
Testing Done: PTM smoke, BGP smoke and ptmd_test.py:TestMultipleAddrsIntfOspfBgp
We want the ability to start up quagga in a varied set of
environments. This needs to be done in SysV and systemd
startups. As such refactor the code to allow us to
allow end users to easily switch between the two
sysV:
edit the /etc/quagga/daemons file
service quagga [start|stop|reload|restart]
Systemd:
edit the /etc/quagga/daemons file
systemctl [start|stop|reload|restart] quagga
Ticket: CM-10634
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Daniel Walton <dwalton@cumulusnetworks.com>
The file if.c has a iflist that had the list of interfaces
in the default vrf. Remove this variable and replace
with a vrf_iflist lookup on the default vrf where it
was used.
Additionally, modify ptm code to iterate over all vrf's
when enabling ptm.
Ticket: CM-10338
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Don Slice <dslice@cumulusnetworks.com>
Reviewed-by: Radhika Mahankali <radhika@cumulusnetworks.com>
bfd was receiving a callback with an interface name string
but was ignoring the passed in vrf to find the ifp pointer.
This commit fixes that code path in bfd.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Radhika Mahankali <radhika@cumulusnetworks.com>
All daemons changed to flag an interface that has been moved to a vrf as DELETED instead of INTERNAL.
When they were flagged as IFINDEX_INTERNAL, ospf, rip, and isis would re-install them in the default
assuming that they were being "pre-defined" before the kernel definitions.
Ticket: CM-9265
Signed-off-by: Don Slice
Reviewed-by: Donald Sharp
Currently if you have this setup:
router ospf6
interface lo area 0.0.0.0
Ospf is scheduling and sending hello's out the loopback interface:
2016/03/18 15:26:12.463248 OSPF6: Could not send entire message
2016/03/18 15:26:22.463475 OSPF6: sendmsg failed: ifindex: 1: Network is unreachhable (101)
Adding a check to not schedule hello's for a loopback interface resolves this issue.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Don Slice <dslice@cumulusnetworks.com>
Ticket: CM-7615, CM-7773
Reviewed By: CCR-3610, CCR-3708
Testing Done: Unit, BGP Smoke and OSPF Smoke
Changes (70790261926b17200c8c9377c4576cd3b486fcef) ported from 2.5
Issue (related to CM-7615): 1. CM-7615: There is mismatch in the client name between ptm display of client BFD sessions and the zebra logs. For example, if bgpd added BFD session, zebra logs will show the client as “bgp” but the ptm display will show it as “quagga”
2. Bigger problem is when 2 clients (for example OSPF and BGP) from Quagga register for same BFD session and only one client de-registers the BFD session. This results in BFD session deletion from PTM even though other client still has the BFD registration.
Root Cause: Even though BGP, OSPF and OSPF6 are 3 different clients from Quagga that are trying to register/deregister BFD sessions with PTM, all 3 are represented as one client “quagga” from zebra. This makes it hard for PTM/BFD to distinguish between all three when BFD peer registration/deregistration happens from the clients.
Fix: Send the actual client name bgp, ospf or ospf6 from zebra with BFD reg/dereg messages instead of one unified client name “quagga”
CM-7773: BFD sessions are not getting cleaned from PTM even though no BGP peering exists in Quagga.
Root Cause: PTM cleans up stale BFD sessions from a client when it finds a change in seq id advertised by the client. But, if PTM never detects a change in the seq id then the stale BFD sessions never get cleaned up. The test restarts the quagga without saving the configuration, which results in no BGP peering. No BGP peers are registered with PTM after restart and PTM does not detect a client seq id change resulting in stale BFD sessions.
Fix: New client registration message was added in PTM. Every client that is interested in BFD monitoring will register with PTM with the client seq id. Client will register with a different seq id (typically pid) every time it restarts. This will help in detecting the change in seq id and cleanup of stale BFD sessions for a client.
Code Changes: To support the new client registration message following changes have been made
- Added support for client registration messaging in zebra for sending messages to PTM.
- Added support for client registration messaging between zebra and clients (BGP, OSPF and OSPF6) in BFD library.
- Expanded the reg/de reg peer messaging between zebra and clients to support client specific seq id to distinguish between multiple clients registering for BFD peer rather than one “quagga” client.
- Changes in bgpd, ospfd and ospf6d to send client registrations at the time of daemon initialization and on receiving BFD peer replay message.
Check on area that interface points to before scheduling SPF.
Signed-off-by: Vivek Venkatraman <vivek@cumulusnetworks.com>
Ticket: CM-7911, CM-8934
Reviewed By: CCR-4256
Testing Done: Verification by Anitha
When a user specifies a non area 0 network in an ospf6d
config the daemon could crash.
Ticket: CM-9134
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
If a default route is passed into ospf6d then we need to make
sure that ospf6d is actually running before we accept and process it.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Daniel Walton <dwalton@cumulusnetworks.com>
Clients (BGP, OSPF etc.) register with Zebra for information about
a VRF such as Router ID, interfaces and redistribution. Add API to
support unregister also which is required for the non-default VRF.
Signed-off-by: Vivek Venkatraman <vivek@cumulusnetworks.com>
Reviewed-by: Donald Sharp <sharpd@cumulusnetworks.com>
Ticket: CM-9128
Reviewed By: CCR-4098
Testing Done: Manual testing
Modify the daemons to integrate with systemd, if it is enabled via configure,
and to notify systemd that they are running/stopping and to send watch
notifications.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
When OSPF6 is creating the header for the ROUTER LSA type
if the packet being sent has interface information to add
to the data, the Sequence Number is at least double incremented.
This change moves the header creation to outside the loop over
all interfaces in the area. Additionally the header is created
at the bottom of the function now.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
During CR for nexthop upstream it was noticed that usage
of prefix2str was not consistent. This fixes this problem
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Several routing protocols use the zapi_ipv[4|6] api to talk to
zebra. There are some instances where the api.vrf_id was not
being set. Since the practice is to declare the api structure
on the stack, the data inside is not being set to 0. As
such random vrf_id values were being passed to zebrad
causing rage and confusion.
Ticket: CM-8287
Reviewed-by: CCR-3841
Testing: Test suites no longer crashing and burning
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
The API messages are used by zebra to exchange the interfaces, addresses,
routes and router-id information with its clients. To distinguish which
VRF the information belongs to, a new field "VRF ID" is added in the
message header. And hence the message version is increased to 3.
* The new field "VRF ID" in the message header:
Length (2 bytes)
Marker (1 byte)
Version (1 byte)
VRF ID (2 bytes, newly added)
Command (2 bytes)
- Client side:
- zclient_create_header() adds the VRF ID in the message header.
- zclient_read() extracts and validates the VRF ID from the header,
and passes the VRF ID to the callback functions registered to
the API messages.
- All relative functions are appended with a new parameter "vrf_id",
including all the callback functions.
- "vrf_id" is also added to "struct zapi_ipv4" and "struct zapi_ipv6".
Clients need to correctly set the VRF ID when using the API
functions zapi_ipv4_route() and zapi_ipv6_route().
- Till now all messages sent from a client have the default VRF ID
"0" in the header.
- The HELLO message is special, which is used as the heart-beat of
a client, and has no relation with VRF. The VRF ID in the HELLO
message header will always be 0 and ignored by zebra.
- Zebra side:
- zserv_create_header() adds the VRF ID in the message header.
- zebra_client_read() extracts and validates the VRF ID from the
header, and passes the VRF ID to the functions which process
the received messages.
- All relative functions are appended with a new parameter "vrf_id".
* Suppress the messages in a VRF which a client does not care:
Some clients may not care about the information in the VRF X, and
zebra should not send the messages in the VRF X to those clients.
Extra flags are used to indicate which VRF is registered by a client,
and a new message ZEBRA_VRF_UNREGISTER is introduced to let a client
can unregister a VRF when it does not need any information in that
VRF.
A client sends any message other than ZEBRA_VRF_UNREGISTER in a VRF
will automatically register to that VRF.
- lib/vrf:
A new utility "VRF bit-map" is provided to manage the flags for
VRFs, one bit per VRF ID.
- Use vrf_bitmap_init()/vrf_bitmap_free() to initialize/free a
bit-map;
- Use vrf_bitmap_set()/vrf_bitmap_unset() to set/unset a flag
in the given bit-map, corresponding to the given VRF ID;
- Use vrf_bitmap_check() to test whether the flag, in the given
bit-map and for the given VRF ID, is set.
- Client side:
- In "struct zclient", the following flags are changed from
"u_char" to "vrf_bitmap_t":
redist[ZEBRA_ROUTE_MAX]
default_information
These flags are extended for each VRF, and controlled by the
clients themselves (or with the help of zclient_redistribute()
and zclient_redistribute_default()).
- Zebra side:
- In "struct zserv", the following flags are changed from
"u_char" to "vrf_bitmap_t":
redist[ZEBRA_ROUTE_MAX]
redist_default
ifinfo
ridinfo
These flags are extended for each VRF, as the VRF registration
flags. They are maintained on receiving a ZEBRA_XXX_ADD or
ZEBRA_XXX_DELETE message.
When sending an interface/address/route/router-id message in
a VRF to a client, if the corresponding VRF registration flag
is not set, this message will not be dropped by zebra.
- A new function zread_vrf_unregister() is introduced to process
the new command ZEBRA_VRF_UNREGISTER. All the VRF registration
flags are cleared for the requested VRF.
Those clients, who support only the default VRF, will never receive
a message in a non-default VRF, thanks to the filter in zebra.
* New callback for the event of successful connection to zebra:
- zclient_start() is splitted, keeping only the code of connecting
to zebra.
- Now zclient_init()=>zclient_connect()=>zclient_start() operations
are purely dealing with the connection to zbera.
- Once zebra is successfully connected, at the end of zclient_start(),
a new callback is used to inform the client about connection.
- Till now, in the callback of connect-to-zebra event, all clients
send messages to zebra to request the router-id/interface/routes
information in the default VRF.
Of corse in future the client can do anything it wants in this
callback. For example, it may send requests for both default VRF
and some non-default VRFs.
Signed-off-by: Feng Lu <lu.feng@6wind.com>
Reviewed-by: Alain Ritoux <alain.ritoux@6wind.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Acked-by: Donald Sharp <sharpd@cumulusnetworks.com>
Conflicts:
lib/zclient.h
lib/zebra.h
zebra/zserv.c
zebra/zserv.h
Conflicts:
bgpd/bgp_nexthop.c
bgpd/bgp_nht.c
bgpd/bgp_zebra.c
isisd/isis_zebra.c
lib/zclient.c
lib/zclient.h
lib/zebra.h
nhrpd/nhrp_interface.c
nhrpd/nhrp_route.c
nhrpd/nhrpd.h
ospf6d/ospf6_zebra.c
ospf6d/ospf6_zebra.h
ospfd/ospf_vty.c
ospfd/ospf_zebra.c
pimd/pim_zebra.c
pimd/pim_zlookup.c
ripd/rip_zebra.c
ripngd/ripng_zebra.c
zebra/redistribute.c
zebra/rt_netlink.c
zebra/zebra_rnh.c
zebra/zebra_rnh.h
zebra/zserv.c
zebra/zserv.h
Later, an interface will belong to a specific VRF, and the interface
initialization will be a part of the VRF initialization. So now call
if_init() from vrf_init(), and if_terminate() from vrf_terminate().
Daemons have the according changes:
- if if_init() was called or "iflist" was initialized, now call
vrf_init() instead;
- if if_terminate() was called or "iflist" was destroyed, now call
vrf_terminate() instead.
Signed-off-by: Feng Lu <lu.feng@6wind.com>
Reviewed-by: Alain Ritoux <alain.ritoux@6wind.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Acked-by: Vincent JARDIN <vincent.jardin@6wind.com>
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Conflicts:
bgpd/bgp_main.c
pimd/pim_iface.c
pimd/pim_iface.h
pimd/pim_main.c
pimd/pimd.c
Ticket: CM-6768
Reviewed By: CCR-3207
Testing Done: bgpsmoke, smoke, topo to create failure
Redistributing routes goes through a del/add cycle whenever a redistributed
is updated. This del/add cycle causes disruption by causing traffic loss
for brief/long periods of time(6-8 s in case of OSPF). The modifications in
this patch remove the del/add cycle to ensure that this disruption doesn't
happen.
Also fixed sending no forwarding address when announcing IPv4 routes with IPv6
nexthops, and sending nexthop only when there is a single path.
Signed-off-by: Dinesh G Dutt <ddutt@cumulusnetworks.com>
Reviewed-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Vivek Venkatraman <vivek@cumulusnetworks.com>
Ticket: CM-7649
Reviewed By: Donald
Testing Done:
This is porting of the patch, ospf6d-bfd-fix-dereg-miss.patch from br2.5.
Issue: The IPv6 OSPF BFD sessions stay down after ifdown/ifup on logical interfaces. This problem doesn’t exist for BFD sessions created by BGP and IPv4 OSPF.
Root cause: When the interface is brought down the IPv6 neighbors discovered on that interface are deleted. This deletion happens without first bringing down the neighbor and the BFD deregistration happens only when the neighbor state changes. This leaves an orphaned BFD session in PTM. Also, the BFD session socket that is bound to the interface that was brought down loses connection. The socket has to be rebound to the interface when it comes up. This problem will not happen if the client deleted the sessions and re-adds it when interface goes down and come up. IPv4 OSPF and BGP work exactly like that.
Fix: Added the BFD deregistration code to IPv6 OSPF neighbor delete.
zclient.c depended upon link time inclusion of a
extern struct thread_master *master. This is a violation of the
namespace of the calling daemon. If a library needs the pointer
pass it in and save it for future use.
This code change also makes the zclient code consistent with
the other lib functions that need to schedule work on your behalf
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Ticket:CM-6802, CM-6952
Reviewed By: Donald, Kanna
Testing Done:
Double commit of b76943235e09472ec174edcf7204fc82d27fe966 from br2.5. But, manually resolved all the compilation errors. Also, modified the shows to support the json format which was not supported in br2.5.
CM-6802 – Currently, BFD session status can be monitored only through ptmctl. There is no way to check the BFD status of a peer/neighbor through Quagga. Debugging becomes easier if BFD status is shown in Quagga too. BFD status is relevant when it is shown against the BGP peer/OSPF neighbor. For, this following code changes have been done:
- Only down messages from PTM were being propagated from Zebra daemon to clients (bgpd, ospfd and ospf6d). Now, both up and down messages are redistributed to the clients from zebra. BFD status field has been added to the messaging. Handling of BFD session up messages has been added to the client code. BGP/OSPF neighbor is brought down only if the old BFD session status is ‘Up’ to handle extra/initial down messages.
- BFD status and last update timestamp fields have been added to the common BFD info structure. Also, common show functions for showing BFD information have been added to BFD lib.
- Modified the BGP neighbor show functions to call common BFD lib functions.
- For ospf and ospf6, BFD information was maintained only at interface level. To show BFD status per neighbor, BFD information has been added at neighbor level too. “show ip ospf interface”, “show ip ospf neighbor detail”, “show ipv6 ospf6 interface” and “show ipv6 ospf6 neighbor detail” output have been modified to show BFD information.
CM-6952 - IBGP peers were always assumed to be multi-hop since there was no easy way to determine whether an IBGP peer was single hop or multihop unlike EBGP. But, this is causing problem with IBGP link local peers since BFD doesn't allow multihop BFD session with link local IP addresses. Link local peers were discovered when the interface peering was enabled. Interface peering is always singlehop. So, added checks to treat all interface based peers as single hop irrespective of whether the peer is IBGP or EBGP.
Ticket: Trivial
Reviewed by: Trivial
Testing: arm/powerpc/amd64 targets now build
Apparently we have a arm build. Who knew? This fixes the compile warnings/errors
produced when you compile a arm build with -Werror.
OSPF BFD command enhancement to configure BFD parameters (detect multiplier, min rx and min tx).
interface <if-name>
ip ospf bfd <detect mult> <min rx> <min tx>
This patch also adds BFD support for IPv6 OSPF. ospf6d will dynamically register/deregister IPv6 neighbors with BFD for monitoring the connectivity of the neighbor. Neighbor is registered with BFD when 2-way adjacency is established and deregistered when adjacency goes down if the BFD is enabled on the interface through which the neighbor was discovered.
OSPF6 BFD command added to configure BFD and parameters (detect multiplier, min rx and min tx).
interface <if-name>
ipv6 ospf6 bfd <detect mult> <min rx> <min tx>
Signed-off-by: Radhika Mahankali <radhika@cumulusnetworks.com>
Reviewed-by: Donald Sharp <sharpd@cumulusnetworks.com>
Reviewed-by: Kanna Rajagopal <kanna@cumulusnetworks.com>
This is to avoid a crash triggered by process termination
when ospf6d daemon is running and 'router ospf6' config has not been done yet.
Signed-off-by: Vipin Kumar <vipin@cumulusnetworks.com>
Reviewed-by: Daniel Walton <dwalton@cumulusnetworks.com>
When a route_node has multiple ospf6_routes under it (common subnet case),
then the current implementation has an issue in adjusting the route_node->info
on a ospf6_route_remove() call.
The main reason is that it ends up using exact match to determine if the next
ospf6_route belongs to the same route_node or not. Fixing that part to use
rnode (the existing back-pointer to the route_node) from the ospf6_route to
determine that.
Also fixing some of the walks to turn them safe so that the route deletion is
fine.
Signed-off-by: Vipin Kumar <vipin@cumulusnetworks.com>
Reviewed-by: Vivek Venkatraman <vivek@cumulusnetworks.com>
As per the RFC, when the NU bit is set, prefix should be ignored.
However, the code is currently ignoring prefix with LA bit too.
Fixing that part.
In future, we should also set LA bit for the loopback addresses. Not doing this
part right away, as quagga wont be backward compatible with its own previous
releases. Maybe after a release or so, we should start setting LA bit too.
Signed-off-by: Vipin Kumar <vipin@cumulusnetworks.com>
Reviewed-by: Daniel Walton <dwalton@cumulusnetworks.com>
The problem is that zclient->redist[ZEBRA_ROUTE_MAX] used for storing a
client’s redist state, has no address-family qualification. This means
a client can only store its interest in a protocol (connected, static etc.),
but cant choose IPv4 or ipv6 with that. This hindered implementation on
client sides to manage redistribution of ipv4 and ipv6 both.
BGP's redistribution of protocols like connected/static is one such place.
One fix could be to overload this and flap the redist connection each time
any new afi is added for redist, but that may have side-effects on the
existing afi redist.
The cleaner way is to modify redist data-structure to also take AFI, and adjust
routines that deal with it, so that a client can register for a protocol
redistribution based on the AFI. BGP already maintains redistribution state
based on afi and protocol (bgp->redist[AFI_MAX][ZEBRA_ROUTE_MAX]). This patch
takes care of filling up the gap in zclient/zserv redistribution state to
also use AFI qualification.
Signed-off-by: Vipin Kumar <vipin@cumulusnetworks.com>
Reviewed-by: Daniel Walton <dwalton@cumulusnetworks.com>
Reviewed-by: Dinesh G Dutt <ddutt@cumulusnetworks.com>
——————————————-------------
- etc/init.d/quagga is modified to support creating separate ospf daemon
process for each instance. Each individual instance is monitored by
watchquagga just like any protocol daemons.(requires initd-mi.patch).
- Vtysh is modified to able to connect to multiple daemons of the same
protocol (supported for OSPF only for now).
- ospfd is modified to remember the Instance-ID that its invoked with. For
the entire life of the process it caters to any command request that
matches that instance-ID (unless its a non instance specific command).
Routes/messages to zebra are tagged with instance-ID.
- zebra route/redistribute mechanisms are modified to work with
[protocol type + instance-id]
- bgpd now has ability to have multiple instance specific redistribution
for a protocol (OSPF only supported/tested for now).
- zlog ability to display instance-id besides the protocol/daemon name.
- Changes in other daemons are to because of the needed integration with
some of the modified APIs/routines. (Didn’t prefer replicating too many
separate instance specific APIs.)
- config/show/debug commands are modified to take instance-id argument
as appropriate.
Guidelines to start using multi-instance ospf
---------------------------------------------
The patch is backward compatible, i.e for any previous way of single ospf
deamon(router ospf <cr>) will continue to work as is, including all the
show commands etc.
To enable multiple instances, do the following:
1. service quagga stop
2. Modify /etc/quagga/daemons to add instance-ids of each desired
instance in the following format:
ospfd=“yes"
ospfd_instances="1,2,3"
assuming you want to enable 3 instances with those instance ids.
3. Create corresponding ospfd config files as ospfd-1.conf, ospfd-2.conf
and ospfd-3.conf.
4. service quagga start/restart
5. Verify that the deamons are started as expected. You should see
ospfd started with -n <instance-id> option.
ps –ef | grep quagga
With that /var/run/quagga/ should have ospfd-<instance-id>.pid and
ospfd-<instance-id>/vty to each instance.
6. vtysh to work with instances as you would with any other deamons.
7. Overall most quagga semantics are the same working with the instance
deamon, like it is for any other daemon.
NOTE:
To safeguard against errors leading to too many processes getting invoked,
a hard limit on number of instance-ids is in place, currently its 5.
Allowed instance-id range is <1-65535>
Once daemons are up, show running from vtysh should show the instance-id
of each daemon as 'router ospf <instance-id>’ (without needing explicit
configuration)
Instance-id can not be changed via vtysh, other router ospf configuration
is allowed as before.
Signed-off-by: Vipin Kumar <vipin@cumulusnetworks.com>
Reviewed-by: Daniel Walton <dwalton@cumulusnetworks.com>
Reviewed-by: Dinesh G Dutt <ddutt@cumulusnetworks.com>
SYMPTOM:
If some of the ospfv3 commands like 'show ipv6 ospf6 route' are executed
with ospf6d daemon running but before having any ospfv3 configuration, then
ospf6d crash is seen.
ISSUE:
There are a few show commands, which are (unlike others) not checking if
ospf6 instance is initialized already.
FIX:
Add the missing checks, by using OSPF6_CMD_CHECK_RUNNING() in the commands
where its needed and not yet used.
Signed-off-by: Vipin Kumar <vipin@cumulusnetworks.com>
Reviewed-by: Pradosh Mohapatra <pmohapat@cumulusnetworks.com>
Dinesh Dutt <ddutt@cumulusnetworks.com>
SYMPTOM:
With quagga running on Linux, 'ifdown <if-name>' followed by 'ifup <ifname>
can cause OSPFv3 to not receive Hello packets on the interface.
ISSUE:
Operating System's interface IPv6 readiness may not be guaranteed at the
time of interface-up event. Thats because the ipv6 components in an OS may
also be listening to the same interface-up event that (in this case) is
relayed to OSPFv3.
In this failure case, setsockopt with option IPV6_JOIN_GROUP on the interface
returned EINVAL.
Error logs -
OSPF6: Zebra Interface state change: swp1 index 3 flags 11043 metric 1 mtu 1500
OSPF6: Interface Event swp1: [InterfaceUp]
OSPF6: Network: setsockopt (20) on ifindex 3 failed: Invalid argument
FIX:
To take care of this possible race condition, any address-family related
setting should be retried. Given it's a rare condition and window of this
race should be short, the patch adds a limited retry mechanism for the
IPV6 membership setting on the socket.
Signed-off-by: Vipin Kumar <vipin@cumulusnetworks.com>
Reviewed-by: Dinesh Dutt <ddutt@cumulusnetworks.com>
Satish Ashok <sashok@cumulusnetworks.com>
Support stubby and totally stubby areas in OSPFv3
Signed-off-by: Dinesh G Dutt <ddutt at cumulusnetworks.com>
Reviewed-by: Pradosh Mohapatra <pmohapat at cumulusnetworks.com>
OSPFv3: Add ABR support and make ECMP > 4.
Signed-off-by: Dinesh G Dutt <ddutt at cumulusnetworks.com>
Signed-off-by: Pradosh Mohapatra <pmohapat at cumulusnetworks.com>
Quagga sources have inherited a slew of Page Feed (^L, \xC) characters
from ancient history. Among other things, these break patchwork's
XML-RPC API because \xC is not a valid character in XML documents.
Nuke them from high orbit.
Patches can be adapted simply by:
sed -e 's%^L%%' -i filename.patch
(you can type page feeds in some environments with Ctrl-V Ctrl-L)
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Add support for keyword commands.
Includes new documentation for DEFUN() in lib/command.h, for preexisting
features as well as new keyword specification.
Signed-off-by: Christian Franke <chris@opensourcerouting.org>
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
When iterating over a list, also the last node should be unlocked again.
Signed-off-by: Christian Franke <chris@opensourcerouting.org>
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
This command allows the user to change to default reference bandwidth
for cost calculations. The default value is 100 Mbps. With a default
bandwidth of 10 MBps, the default cost becomes 10. Those values are
consistent with OSPFv2.
[DL: resolved conflicts in vty command additions & docs]
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Previously, the interface cost was a fixed default value that a user
was allowed to change with "ipv6 ospf6 cost XX". As what is done with
OSPFv2, we change this behaviour to compute the default interface cost
from the interface bandwidth.
The user can still force a cost with "ipv6 ospf6 cost XX". He can get
the default value with "no ipv6 ospf6 cost". Moreover, the default
cost value was 1. The cost is now computed from the bandwidth and a
default reference bandwidth of 100 MBps (as for OSPFv2). Since the
default bandwidth for an interface is 10 MBps, the "default" cost
becomes 10 instead of 1.
[DL: resolved conflict in ospf6d/ospf6_interface.c]
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
Do not send ospf6d hellos on fresh created interfaces without
configuration (ie. no vlan configured). Ospf6d use ip6_mtu, if it's not
initalised, Ospf6d tries to alloc indefinite size of memory.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
interface_down() - which also handles some nonobvious cases like the
last linklocal address disappearing - was previously not cancelling the
hello timer. This had the effect of multiple such threads ending up
scheduled after a quick down-up cycle.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
This fixes a SEGV when we receive a higher-SeqNum copy of a LSA that we
originated ourselves, before a reboot of ospf6d. We create a new
copy of the LSA to resync the SeqNum, but then half an hour later
the old refresh thread ends up trying to refresh the free()'d old LSA.
The SEGV is triggered by this chain:
ospf6_lsdb_maxage_remover
-> thread_execute(ospf6_lsa_refresh)
-> old->refresh = NULL
Which assumes that old->refresh is no longer scheduled to run, as it is
being run right there. But the thread_execute() doesn't know about
old->refresh and therefore didn't remove it.
(Found by ANVL OSPFV3-16.17)
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
