Quentin Young [Wed, 6 Jun 2018 21:53:09 +0000 (21:53 +0000)]
doc: clean up introductory user docs
* Update the list of protocol daemons from 5 to the current 13
* Rewrap and clean up formatting for lots of miscellaneous code blocks
* Improve English in various spots
* Update architecture description and diagram
* Remove duplicate information on reporting bugs
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Donald Sharp [Thu, 7 Jun 2018 23:51:13 +0000 (19:51 -0400)]
bgpd, lib, ospf6d, vtysh: fix possible snprintf possible truncation
With a new version of clang 6.0, the compiler is detecting more
issues where we may be possibly be truncating the output string.
Fix by increasing the size of the output string to make the compiler
happy.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Donald Sharp [Thu, 7 Jun 2018 21:10:38 +0000 (17:10 -0400)]
bgpd: Actually display labeled unicast routes received
The labeled unicast and unicast tables have been combined
into the unicast table. Additionally we have a restriction
where if you configure labeled unicast you cannot configure
unicast. This created a bug with 'show bgp ipv4 labeled-unicast summ'
command where we were displaying NoNeg, because v4 has been intentionally
turned off.
Modify the code so that when we are looking up if we have negotiated
a capapbility we use the correct one, while still using the appropriate
table for prefix count.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Renato Westphal [Thu, 7 Jun 2018 18:45:33 +0000 (15:45 -0300)]
zebra: remove unnecessary preprocessor check
The IFLA_INFO_SLAVE_KIND constant is always defined now that we imported
our own copies of the Linux kernel headers. Remove the preprocessor
checks since they aren't necessary anymore.
Quentin Young [Wed, 6 Jun 2018 19:09:44 +0000 (19:09 +0000)]
lib: fix output mangling with | include
Sometimes output would be mangled when filtering with include as a
result of the following bugs:
* Filters were applied per each call to vty_out() instead of buffering
until a line break and then applying
* Long output would sometimes be cut due to using the wrong buffer
pointer
Also remove the trailing \n as it should no longer be necessary to
ensure the vty prompt ends up on a new line.
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Donald Sharp [Wed, 6 Jun 2018 17:33:19 +0000 (13:33 -0400)]
bgpd: Collapse bgp_aggregate_add into bgp_aggregate_route
These two functions are functionally the same, except
bgp_aggregate_route is meant to handle the addition and
deletion of routes, while aggregate_add is meant for all of them.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Donald Sharp [Wed, 6 Jun 2018 17:13:00 +0000 (13:13 -0400)]
bgpd: Allow bgp to know when to actually add/delete agg route
The aggregated route was being sent in updates to peers every
time a route changed that we were aggregating. Modify
the code such that we only send aggregated route updates
if we actually have something different to tell the peer.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Donald Sharp [Wed, 6 Jun 2018 16:31:17 +0000 (12:31 -0400)]
bgpd: Seperate out install/removal of aggregate from delete function
This is a transitional commit, to get us where we want to go.
Seperate out the install/removal of the aggregate route from
the bgp_aggregate_delete and bgp_aggregate_route functions.
In the future we'll write a bit of code to determine if the
aggregate add has actually changed any information we care
about.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Quentin Young [Tue, 29 May 2018 21:38:18 +0000 (21:38 +0000)]
lib: add vector_compact(), use after str splits
* Add function to move all data to the start of a vector by shifting
over contiguous empty slots
* Use this function to remove empty slots leftover after
frrstr_filter_vec
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Quentin Young [Tue, 29 May 2018 17:13:51 +0000 (17:13 +0000)]
lib: fix static analysis issues, use regfree()
* Fix potential NULL dereference
* Fix use of uninitialized value
* Fix leaking memory by not freeing regex_t
* Fix extra \n when using empty regex filter
* Clean up still-reachable hook memory
* Handle nonexistent pager
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Quentin Young [Fri, 25 May 2018 22:49:53 +0000 (22:49 +0000)]
lib: add vector_remove() to vector.[ch]
An optimized version of this has already been implemented within graph.c
that assumes some specialized constraints for that code. It's generally
useful so this change implements a general purpose version of it.
This fixes cmd_make_strvec() that was broken by some code shuffling in
previous commits.
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Quentin Young [Mon, 14 May 2018 22:13:03 +0000 (18:13 -0400)]
vtysh: add | support
* Rewrite pager implementation
* Replace fprintf() with vty_out()
* Modify vty_out() for better vtysh support
* Remove static global outputfile var
* Remove fp argument from many vtysh functions
* Add some docs for stuff along the way
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Quentin Young [Sun, 13 May 2018 00:14:22 +0000 (20:14 -0400)]
lib: add cli preprocessor for `|` actions
This patch adds a CLI preprocessor function that activates when `|` is
found in the command. This is the start of adding support for some text
processing utilities intended for inline use. The first one implemented
here is `| include`, which provides grep-like filtering of command
output.
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Quentin Young [Sun, 13 May 2018 00:09:08 +0000 (20:09 -0400)]
lib: add hook for preprocessing commands
This patch adds a hook point intended to allow subscribers to modify the
raw text of a CLI command before it is passed to the rest of the CLI
pipeline. To give access to the raw text of the command, a new function
for executing CLI has been defined whose only difference from
`cmd_execute_command` is that it accepts the command to execute as a
string rather than as a string vector.
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Quentin Young [Fri, 11 May 2018 19:32:06 +0000 (15:32 -0400)]
lib: add string utilities
I see lots of the same code being copy-pasted and slightly tweaked for
string processing all over the codebase. Time to start aggregating these
pieces into something consistent and correct.
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Arthur Jones [Wed, 6 Jun 2018 14:47:17 +0000 (07:47 -0700)]
tests/isisd: bypass test_fuzz_isis_tlv when inet_ntop is broken
On Alpine Linux edge, musl does not seem to be RFC 5952 4.2.2
compliant (how to print a single :0: in the IPv6 address). Let's
skip that test, as we get false negatives when running against
that version of musl.
Credit for the idea for the fix and how to fix it is due to
chris@opensourcerouting.org.
Testing done:
make check on alpine linux passes now
Issue: https://github.com/FRRouting/frr/issues/2375 Signed-off-by: Arthur Jones <arthur.jones@riverbed.com>
Donald Sharp [Fri, 1 Jun 2018 18:19:12 +0000 (14:19 -0400)]
bgpd: Trust the bgp_table_top_nolock for aggregates
The safi passed in to short-circuit the aggregate lookup
adds code complexity and little speed improvements for
the case where we actually may have aggregates configured!
Since bgp_table_top_nolock() actually tells us if there
are any aggregates installed and safely returns if there
is nothing to do, trust it. As that we know for those
safi's were we don't want to have, we dissallow the
creation via the cli anyways.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Donald Sharp [Fri, 1 Jun 2018 18:13:58 +0000 (14:13 -0400)]
bgpd: Cleanup bgp_aggregate_set/unset to not worry about safi so much
The bgp_aggregate_set/unset functions are only called from the cli
invocations which control what AFI/SAFI we are looking at. Tests
for safi are unimportant.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Donald Sharp [Fri, 1 Jun 2018 00:51:43 +0000 (20:51 -0400)]
ospfd: Note that we do not care about self origination in this case
When we first get a packet, we need to know if we are self
originated later to make correct decisions. Go ahead and
note that we do not plan to make any decision points
about our self origination here.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Martin Buck [Mon, 4 Jun 2018 13:52:14 +0000 (15:52 +0200)]
pimd: Fix nexthop determination when sending towards RP
When sending a PIM join upwards on the RP-based tree, it may get dropped on
the last hop before the RP if the RP is reachable via a connected route
(i.e. there's no associated nexthop). pimd needs to put the nexthop IP
address into the PIM join payload and fails to do that if that route has a
nexthop of 0.0.0.0. So whenever we look up a route to determine the nexthop
or we receive a nexthop tracking update from Zebra, use the destination
address as the nexthop address for connected routes.
Fixes #2326.
Signed-off-by: Martin Buck <mb-tmp-tvguho.pbz@gromit.dyndns.org>
Donald Sharp [Thu, 31 May 2018 16:56:21 +0000 (12:56 -0400)]
bgpd: Actually release peer from bgp_peer hash
When we are determining the state of a peer, we sometimes
detect that we should update the peer->su. The bgp->peer_hash
keeps a hash of peers based upon the peer->su. This requires
us to release the stored value before we re-insert it.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Renato Westphal [Sun, 3 Jun 2018 01:00:32 +0000 (22:00 -0300)]
build: append -std=gnu99 after SNMP_CFLAGS
Programs that link to libnetsnmp must be compiled using a special set
of flags as specified by the "net-snmp-config --base-cflags" command
(whose output is stored in the SNMP_CFLAGS variable). The problem is
that "net-snmp-config --base-cflags" can output -std=c99 in addition to
other compiler flags in some platforms, and this breaks the build since
FRR souce code makes use of some GNU compiler extensions (e.g. allow
trailing commas in function parameter lists). In order to solve this
problem, append -std=gnu99 after SNMP_CFLAGS in all makefiles where this
variable is used. This way the -std=c99 flag will be overwritten when it's
present. Source files that don't link to libnetsnmp will be compiled using
either -std=gnu99 or -std=gnu11 depending on the compiler availability.
Quentin Young [Fri, 1 Jun 2018 19:25:03 +0000 (19:25 +0000)]
vtysh: fix potential stack buffer overflow
If vtysh was instructed to perform line-by-line processing on the output
of a command executed against a daemon and this output, as received by
vtysh, was not terminated with a newline, vtysh could print contents of
memory to its output device.
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Donald Sharp [Fri, 1 Jun 2018 19:03:57 +0000 (15:03 -0400)]
bgpd: Deprecate and hide the `no bgp multiple-instance` command
This command needs to be deprecated. It partially implements
a refusal to create multiple instances. If you do not need
multiple instances, just don't create them in the cli instead.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Philippe Guibert [Thu, 17 May 2018 13:34:51 +0000 (15:34 +0200)]
zebra: keep one command to handle vty with/without table param
That fix is a workaround from a vtysh limitation.
Because table identifier should be accessible in configuration only for
vrf netns backends, there was a need to differentiate the vty commands.
Unfortunately, vtysh parses the two commands without knowing which
command has really been installed.
Using one single vty command will avoid having this issue in vtysh.
Signed-off-by: Philippe Guibert <philippe.guibert@6wind.com>
Philippe Guibert [Wed, 28 Mar 2018 08:46:14 +0000 (10:46 +0200)]
zebra: table_id election with the vrf backend
As table_id for VRF with netns backend is main table ( RT_TABLE_MAIN or
zebrad.rtm_table_default), this makes possible to return the table id
that wants to be configured for those cases. ( in addition to default
VRF). In other cases ( VRF Lite presumably), then vrf table_id is
returned.
Signed-off-by: Philippe Guibert <philippe.guibert@6wind.com>
Philippe Guibert [Wed, 30 May 2018 09:46:13 +0000 (11:46 +0200)]
doc: replace 'intact' keyword with something more clear
Keeping the config intact might be misunderstood. I say that even if VRF
netns is automatically discovered, it is possible for administrator to
save the netns information in the config file, to bring more clarity (
hence the config commands available).
Signed-off-by: Philippe Guibert <philippe.guibert@6wind.com>
Philippe Guibert [Wed, 30 May 2018 09:38:24 +0000 (11:38 +0200)]
doc: inform the diff between config from zebra and outside
To avoid some confusions, it is precised in the documentation that
the configuration not done from zebra will not be injected in the
configuration context. As consequence, the config file will not be
impacted by underlying network context. But also, this will not be
possible for *Zebra* to attempt to modify outside networking objects.
Signed-off-by: Philippe Guibert <philippe.guibert@6wind.com>
projects / matthieu / frr.git / log