8 One of the most frequent decisions to make while writing code for FRR is what
9 to log, what level to log it at, and when to log it. Here is a list of
10 recommendations for these decisions.
16 ``printfrr()`` is FRR's modified version of ``printf()``, designed to make
17 life easier when printing nontrivial datastructures. The following variants
20 .. c:function:: ssize_t snprintfrr(char *buf, size_t len, const char *fmt, ...)
21 .. c:function:: ssize_t vsnprintfrr(char *buf, size_t len, const char *fmt, va_list)
23 These correspond to ``snprintf``/``vsnprintf``. If you pass NULL for buf
24 or 0 for len, no output is written but the return value is still calculated.
26 The return value is always the full length of the output, unconstrained by
27 `len`. It does **not** include the terminating ``\0`` character. A
28 malformed format string can result in a ``-1`` return value.
30 .. c:function:: ssize_t csnprintfrr(char *buf, size_t len, const char *fmt, ...)
31 .. c:function:: ssize_t vcsnprintfrr(char *buf, size_t len, const char *fmt, va_list)
33 Same as above, but the ``c`` stands for "continue" or "concatenate". The
34 output is appended to the string instead of overwriting it.
36 .. c:function:: char *asprintfrr(struct memtype *mt, const char *fmt, ...)
37 .. c:function:: char *vasprintfrr(struct memtype *mt, const char *fmt, va_list)
39 These functions allocate a dynamic buffer (using MTYPE `mt`) and print to
40 that. If the format string is malformed, they return a copy of the format
41 string, so the return value is always non-NULL and always dynamically
44 .. c:function:: char *asnprintfrr(struct memtype *mt, char *buf, size_t len, const char *fmt, ...)
45 .. c:function:: char *vasnprintfrr(struct memtype *mt, char *buf, size_t len, const char *fmt, va_list)
47 This variant tries to use the static buffer provided, but falls back to
48 dynamic allocation if it is insufficient.
50 The return value can be either `buf` or a newly allocated string using
51 `mt`. You MUST free it like this::
53 char *ret = asnprintfrr(MTYPE_FOO, buf, sizeof(buf), ...);
55 XFREE(MTYPE_FOO, ret);
57 .. c:function:: ssize_t bprintfrr(struct fbuf *fb, const char *fmt, ...)
58 .. c:function:: ssize_t vbprintfrr(struct fbuf *fb, const char *fmt, va_list)
60 These are the "lowest level" functions, which the other variants listed
61 above use to implement their functionality on top. Mainly useful for
62 implementing printfrr extensions since those get a ``struct fbuf *`` to
63 write their output to.
65 .. c:macro:: FMT_NSTD(expr)
67 This macro turns off/on format warnings as needed when non-ISO-C
68 compatible printfrr extensions are used (e.g. ``%.*p`` or ``%Ld``.)::
70 vty_out(vty, "standard compatible %pI4\n", &addr);
71 FMT_NSTD(vty_out(vty, "non-standard %-47.*pHX\n", (int)len, buf));
73 When the frr-format plugin is in use, this macro is a no-op since the
74 frr-format plugin supports all printfrr extensions. Since the FRR CI
75 includes a system with the plugin enabled, this means format errors will
76 not slip by undetected even with FMT_NSTD.
80 ``printfrr()`` does not support the ``%n`` format.
85 ``printfrr()`` are AS-Safe under the following conditions:
87 * the ``[v]as[n]printfrr`` variants are not AS-Safe (allocating memory)
88 * floating point specifiers are not AS-Safe (system printf is used for these)
89 * the positional ``%1$d`` syntax should not be used (8 arguments are supported
91 * extensions are only AS-Safe if their printer is AS-Safe
96 ``printfrr()`` format strings can be extended with suffixes after `%p` or `%d`.
97 Printf features like field lengths can be used normally with these extensions,
98 e.g. ``%-15pI4`` works correctly, **except if the extension consumes the
99 width or precision**. Extensions that do so are listed below as ``%*pXX``
100 rather than ``%pXX``.
102 The extension specifier after ``%p`` or ``%d`` is always an uppercase letter;
103 by means of established pattern uppercase letters and numbers form the type
104 identifier which may be followed by lowercase flags.
106 You can grep the FRR source for ``printfrr_ext_autoreg`` to see all extended
107 printers and what exactly they do. More printers are likely to be added as
108 needed/useful, so the list here may be outdated.
112 The ``zlog_*``/``flog_*`` and ``vty_out`` functions all use printfrr
113 internally, so these extensions are available there. However, they are
114 **not** available when calling ``snprintf`` directly. You need to call
115 ``snprintfrr`` instead.
117 Networking data types
118 ^^^^^^^^^^^^^^^^^^^^^
120 .. role:: frrfmtout(code)
122 .. frrfmt:: %pI4 (struct in_addr *, in_addr_t *)
126 ``%pI4s``: :frrfmtout:`*` — print star instead of ``0.0.0.0`` (for multicast)
128 .. frrfmt:: %pI6 (struct in6_addr *)
130 :frrfmtout:`fe80::1234`
132 ``%pI6s``: :frrfmtout:`*` — print star instead of ``::`` (for multicast)
134 .. frrfmt:: %pEA (struct ethaddr *)
136 :frrfmtout:`01:23:45:67:89:ab`
138 .. frrfmt:: %pIA (struct ipaddr *)
140 :frrfmtout:`1.2.3.4` / :frrfmtout:`fe80::1234`
142 ``%pIAs``: — print star instead of zero address (for multicast)
144 .. frrfmt:: %pFX (struct prefix *)
146 :frrfmtout:`1.2.3.0/24` / :frrfmtout:`fe80::1234/64`
148 This accepts the following types:
151 - :c:struct:`prefix_ipv4`
152 - :c:struct:`prefix_ipv6`
153 - :c:struct:`prefix_eth`
154 - :c:struct:`prefix_evpn`
155 - :c:struct:`prefix_fs`
157 It does **not** accept the following types:
159 - :c:struct:`prefix_ls`
160 - :c:struct:`prefix_rd`
161 - :c:struct:`prefix_ptr`
162 - :c:struct:`prefix_sg` (use :frrfmt:`%pPSG4`)
163 - :c:union:`prefixptr` (dereference to get :c:struct:`prefix`)
164 - :c:union:`prefixconstptr` (dereference to get :c:struct:`prefix`)
168 ``%pFXh``: (address only) :frrfmtout:`1.2.3.0` / :frrfmtout:`fe80::1234`
170 .. frrfmt:: %pPSG4 (struct prefix_sg *)
172 :frrfmtout:`(*,1.2.3.4)`
174 This is *(S,G)* output for use in pimd. (Note prefix_sg is not a prefix
175 "subclass" like the other prefix_* structs.)
177 .. frrfmt:: %pSU (union sockunion *)
179 ``%pSU``: :frrfmtout:`1.2.3.4` / :frrfmtout:`fe80::1234`
181 ``%pSUs``: :frrfmtout:`1.2.3.4` / :frrfmtout:`fe80::1234%89`
182 (adds IPv6 scope ID as integer)
184 ``%pSUp``: :frrfmtout:`1.2.3.4:567` / :frrfmtout:`[fe80::1234]:567`
187 ``%pSUps``: :frrfmtout:`1.2.3.4:567` / :frrfmtout:`[fe80::1234%89]:567`
188 (adds port and scope ID)
190 .. frrfmt:: %pRN (struct route_node *, struct bgp_node *, struct agg_node *)
192 :frrfmtout:`192.168.1.0/24` (dst-only node)
194 :frrfmtout:`2001:db8::/32 from fe80::/64` (SADR node)
196 .. frrfmt:: %pNH (struct nexthop *)
198 ``%pNHvv``: :frrfmtout:`via 1.2.3.4, eth0` — verbose zebra format
200 ``%pNHv``: :frrfmtout:`1.2.3.4, via eth0` — slightly less verbose zebra format
202 ``%pNHs``: :frrfmtout:`1.2.3.4 if 15` — same as :c:func:`nexthop2str()`
204 ``%pNHcg``: :frrfmtout:`1.2.3.4` — compact gateway only
206 ``%pNHci``: :frrfmtout:`eth0` — compact interface only
208 .. frrfmt:: %pBD (struct bgp_dest *)
210 :frrfmtout:`fe80::1234/64`
212 (only available in bgpd.)
214 .. frrfmt:: %dPF (int)
218 Prints an `AF_*` / `PF_*` constant. ``PF`` is used here to avoid confusion
219 with `AFI` constants, even though the FRR codebase prefers `AF_INET` over
222 .. frrfmt:: %dSO (int)
224 :frrfmtout:`SOCK_STREAM`
226 Time/interval formats
227 ^^^^^^^^^^^^^^^^^^^^^
229 .. frrfmt:: %pTS (struct timespec *)
231 .. frrfmt:: %pTV (struct timeval *)
233 .. frrfmt:: %pTT (time_t *)
235 Above 3 options internally result in the same code being called, support
236 the same flags and produce equal output with one exception: ``%pTT``
237 has no sub-second precision and the formatter will never print a
238 (nonsensical) ``.000``.
240 Exactly one of ``I``, ``M`` or ``R`` must immediately follow after
241 ``TS``/``TV``/``TT`` to specify whether the input is an interval, monotonic
242 timestamp or realtime timestamp:
244 ``%pTVI``: input is an interval, not a timestamp. Print interval.
246 ``%pTVIs``: input is an interval, convert to wallclock by subtracting it
247 from current time (i.e. interval has passed **s**\ ince.)
249 ``%pTVIu``: input is an interval, convert to wallclock by adding it to
250 current time (i.e. **u**\ ntil interval has passed.)
252 ``%pTVM`` - input is a timestamp on CLOCK_MONOTONIC, convert to wallclock
253 time (by grabbing current CLOCK_MONOTONIC and CLOCK_REALTIME and doing the
254 math) and print calendaric date.
256 ``%pTVMs`` - input is a timestamp on CLOCK_MONOTONIC, print interval
257 **s**\ ince that timestamp (elapsed.)
259 ``%pTVMu`` - input is a timestamp on CLOCK_MONOTONIC, print interval
260 **u**\ ntil that timestamp (deadline.)
262 ``%pTVR`` - input is a timestamp on CLOCK_REALTIME, print calendaric date.
264 ``%pTVRs`` - input is a timestamp on CLOCK_REALTIME, print interval
265 **s**\ ince that timestamp.
267 ``%pTVRu`` - input is a timestamp on CLOCK_REALTIME, print interval
268 **u**\ ntil that timestamp.
270 ``%pTVA`` - reserved for CLOCK_TAI in case a PTP implementation is
271 interfaced to FRR. Not currently implemented.
275 If ``%pTVRs`` or ``%pTVRu`` are used, this is generally an indication
276 that a CLOCK_MONOTONIC timestamp should be used instead (or added in
277 parallel.) CLOCK_REALTIME might be adjusted by NTP, PTP or similar
278 procedures, causing bogus intervals to be printed.
280 ``%pTVM`` on first look might be assumed to have the same problem, but
281 on closer thought the assumption is always that current system time is
282 correct. And since a CLOCK_MONOTONIC interval is also quite safe to
283 assume to be correct, the (past) absolute timestamp to be printed from
284 this can likely be correct even if it doesn't match what CLOCK_REALTIME
285 would have indicated at that point in the past. This logic does,
286 however, not quite work for *future* times.
288 Generally speaking, almost all use cases in FRR should (and do) use
289 CLOCK_MONOTONIC (through :c:func:`monotime()`.)
291 Flags common to printing calendar times and intervals:
293 ``p``: include spaces in appropriate places (depends on selected format.)
295 ``%p.3TV...``: specify sub-second resolution (use with ``FMT_NSTD`` to
296 suppress gcc warning.) As noted above, ``%pTT`` will never print sub-second
297 digits since there are none. Only some formats support printing sub-second
298 digits and the default may vary.
300 The following flags are available for printing calendar times/dates:
302 (no flag): :frrfmtout:`Sat Jan 1 00:00:00 2022` - print output from
303 ``ctime()``, in local time zone. Since FRR does not currently use/enable
304 locale support, this is always the C locale. (Locale support getting added
305 is unlikely for the time being and would likely break other things worse
308 ``i``: :frrfmtout:`2022-01-01T00:00:00.123` - ISO8601 timestamp in local
309 time zone (note there is no ``Z`` or ``+00:00`` suffix.) Defaults to
310 millisecond precision.
312 ``ip``: :frrfmtout:`2022-01-01 00:00:00.123` - use readable form of ISO8601
313 with space instead of ``T`` separator.
315 The following flags are available for printing intervals:
317 (no flag): :frrfmtout:`9w9d09:09:09.123` - does not match any
318 preexisting format; added because it does not lose precision (like ``t``)
319 for longer intervals without printing huge numbers (like ``h``/``m``).
320 Defaults to millisecond precision. The week/day fields are left off if
321 they're zero, ``p`` adds a space after the respective letter.
323 ``t``: :frrfmtout:`9w9d09h`, :frrfmtout:`9d09h09m`, :frrfmtout:`09:09:09` -
324 this replaces :c:func:`frrtime_to_interval()`. ``p`` adds spaces after
325 week/day/hour letters.
327 ``d``: print decimal number of seconds. Defaults to millisecond precision.
329 ``x`` / ``tx`` / ``dx``: Like no flag / ``t`` / ``d``, but print
330 :frrfmtout:`-` for zero or negative intervals (for use with unset timers.)
332 ``h``: :frrfmtout:`09:09:09`
334 ``hx``: :frrfmtout:`09:09:09`, :frrfmtout:`--:--:--` - this replaces
335 :c:func:`pim_time_timer_to_hhmmss()`.
337 ``m``: :frrfmtout:`09:09`
339 ``mx``: :frrfmtout:`09:09`, :frrfmtout:`--:--` - this replaces
340 :c:func:`pim_time_timer_to_mmss()`.
342 FRR library helper formats
343 ^^^^^^^^^^^^^^^^^^^^^^^^^^
345 .. frrfmt:: %pTH (struct thread *)
347 Print remaining time on timer thread. Interval-printing flag characters
348 listed above for ``%pTV`` can be added, e.g. ``%pTHtx``.
350 ``NULL`` pointers are printed as ``-``.
352 .. frrfmt:: %pTHD (struct thread *)
354 Print debugging information for given thread. Sample output:
359 {(thread *)0x55a3b5818910 arg=0x55a3b5827c50 timer r=7.824 mld_t_query() &mld_ifp->t_query from pimd/pim6_mld.c:1369}
360 {(thread *)0x55a3b5827230 arg=0x55a3b5827c50 read fd=16 mld_t_recv() &mld_ifp->t_recv from pimd/pim6_mld.c:1186}
362 (The output is aligned to some degree.)
364 General utility formats
365 ^^^^^^^^^^^^^^^^^^^^^^^
367 .. frrfmt:: %m (no argument)
369 :frrfmtout:`Permission denied`
371 Prints ``strerror(errno)``. Does **not** consume any input argument, don't
374 (This is a GNU extension not specific to FRR. FRR guarantees it is
375 available on all systems in printfrr, though BSDs support it in printf too.)
377 .. frrfmt:: %pSQ (char *)
379 ([S]tring [Q]uote.) Like ``%s``, but produce a quoted string. Options:
381 ``n`` - treat ``NULL`` as empty string instead.
383 ``q`` - include ``""`` quotation marks. Note: ``NULL`` is printed as
384 ``(null)``, not ``"(null)"`` unless ``n`` is used too. This is
387 ``s`` - use escaping suitable for RFC5424 syslog. This means ``]`` is
390 If a length is specified (``%*pSQ`` or ``%.*pSQ``), null bytes in the input
391 string do not end the string and are just printed as ``\x00``.
393 .. frrfmt:: %pSE (char *)
395 ([S]tring [E]scape.) Like ``%s``, but escape special characters.
398 ``n`` - treat ``NULL`` as empty string instead.
400 Unlike :frrfmt:`%pSQ`, this escapes many more characters that are fine for
401 a quoted string but not on their own.
403 If a length is specified (``%*pSE`` or ``%.*pSE``), null bytes in the input
404 string do not end the string and are just printed as ``\x00``.
406 .. frrfmt:: %pVA (struct va_format *)
408 Recursively invoke printfrr, with arguments passed in through:
410 .. c:struct:: va_format
412 .. c:member:: const char *fmt
414 Format string to use for the recursive printfrr call.
416 .. c:member:: va_list *va
418 Formatting arguments. Note this is passed as a pointer, not - as in
419 most other places - a direct struct reference. Internally uses
420 ``va_copy()`` so repeated calls can be made (e.g. for determining
423 .. frrfmt:: %pFB (struct fbuf *)
425 Insert text from a ``struct fbuf *``, i.e. the output of a call to
426 :c:func:`bprintfrr()`.
428 .. frrfmt:: %*pHX (void *, char *, unsigned char *)
430 ``%pHX``: :frrfmtout:`12 34 56 78`
432 ``%pHXc``: :frrfmtout:`12:34:56:78` (separate with [c]olon)
434 ``%pHXn``: :frrfmtout:`12345678` (separate with [n]othing)
436 Insert hexdump. This specifier requires a precision or width to be
437 specified. A precision (``%.*pHX``) takes precedence, but generates a
438 compiler warning since precisions are undefined for ``%p`` in ISO C. If
439 no precision is given, the width is used instead (and normal handling of
440 the width is suppressed).
442 Note that width and precision are ``int`` arguments, not ``size_t``. Use
448 snprintfrr(out, sizeof(out), "... %*pHX ...", (int)len, buf);
450 /* with padding to width - would generate a warning due to %.*p */
451 FMT_NSTD(snprintfrr(out, sizeof(out), "... %-47.*pHX ...", (int)len, buf));
453 .. frrfmt:: %*pHS (void *, char *, unsigned char *)
455 ``%pHS``: :frrfmtout:`hex.dump`
457 This is a complementary format for :frrfmt:`%*pHX` to print the text
458 representation for a hexdump. Non-printable characters are replaced with
466 These formats currently only exist for advanced type checking with the
467 ``frr-format`` GCC plugin. They should not be used directly since they will
468 cause compiler warnings when used without the plugin. Use with
469 :c:macro:`FMT_NSTD` if necessary.
471 It is possible ISO C23 may introduce another format for these, possibly
472 ``%w64d`` discussed in `JTC 1/SC 22/WG 14/N2680 <http://www.open-std.org/jtc1/sc22/wg14/www/docs/n2680.pdf>`_.
474 .. frrfmt:: %Lu (uint64_t)
478 .. frrfmt:: %Ld (int64_t)
488 If it is something that the user will want to look at and maybe do
489 something, it is either an **error** or a **warning**.
491 We're expecting that warnings and errors are in some way visible to the
492 user (in the worst case by looking at the log after the network broke, but
493 maybe by a syslog collector from all routers.) Therefore, anything that
494 needs to get the user in the loop—and only these things—are warnings or
497 Note that this doesn't necessarily mean the user needs to fix something in
498 the FRR instance. It also includes when we detect something else needs
499 fixing, for example another router, the system we're running on, or the
500 configuration. The common point is that the user should probably do
503 Deciding between a warning and an error is slightly less obvious; the rule
504 of thumb here is that an error will cause considerable fallout beyond its
505 direct effect. Closing a BGP session due to a malformed update is an error
506 since all routes from the peer are dropped; discarding one route because
507 its attributes don't make sense is a warning.
509 This also loosely corresponds to the kind of reaction we're expecting from
510 the user. An error is likely to need immediate response while a warning
511 might be snoozed for a bit and addressed as part of general maintenance.
512 If a problem will self-repair (e.g. by retransmits), it should be a
513 warning—unless the impact until that self-repair is very harsh.
515 Examples for warnings:
517 * a BGP update, LSA or LSP could not be processed, but operation is
518 proceeding and the broken pieces are likely to self-fix later
519 * some kind of controller cannot be reached, but we can work without it
520 * another router is using some unknown or unsupported capability
524 * dropping a BGP session due to malformed data
525 * a socket for routing protocol operation cannot be opened
526 * desynchronization from network state because something went wrong
527 * *everything that we as developers would really like to be notified about,
528 i.e. some assumption in the code isn't holding up*
531 Informational messages
532 ^^^^^^^^^^^^^^^^^^^^^^
534 Anything that provides introspection to the user during normal operation
535 is an **info** message.
537 This includes all kinds of operational state transitions and events,
538 especially if they might be interesting to the user during the course of
539 figuring out a warning or an error.
541 By itself, these messages should mostly be statements of fact. They might
542 indicate the order and relationship in which things happened. Also covered
543 are conditions that might be "operational issues" like a link failure due
544 to an unplugged cable. If it's pretty much the point of running a routing
545 daemon for, it's not a warning or an error, just business as usual.
547 The user should be able to see the state of these bits from operational
548 state output, i.e. `show interface` or `show foobar neighbors`. The log
549 message indicating the change may have been printed weeks ago, but the
550 state can always be viewed. (If some state change has an info message but
551 no "show" command, maybe that command needs to be added.)
555 * all kinds of up/down state changes
557 * interface coming up or going down
558 * addresses being added or deleted
559 * peers and neighbors coming up or going down
561 * rejection of some routes due to user-configured route maps
562 * backwards compatibility handling because another system on the network
563 has a different or smaller feature set
566 The previously used **notify** priority is replaced with *info* in all
567 cases. We don't currently have a well-defined use case for it.
570 Debug messages and asserts
571 ^^^^^^^^^^^^^^^^^^^^^^^^^^
573 Everything that is only interesting on-demand, or only while developing,
574 is a **debug** message. It might be interesting to the user for a
575 particularly evasive issue, but in general these are details that an
576 average user might not even be able to make sense of.
578 Most (or all?) debug messages should be behind a `debug foobar` category
579 switch that controls which subset of these messages is currently
580 interesting and thus printed. If a debug message doesn't have such a
581 guard, there should be a good explanation as to why.
583 Conversely, debug messages are the only thing that should be guarded by
584 these switches. Neither info nor warning or error messages should be
587 **Asserts** should only be used as pretty crashes. We are expecting that
588 asserts remain enabled in production builds, but please try to not use
589 asserts in a way that would cause a security problem if the assert wasn't
590 there (i.e. don't use them for length checks.)
592 The purpose of asserts is mainly to help development and bug hunting. If
593 the daemon crashes, then having some more information is nice, and the
594 assert can provide crucial hints that cut down on the time needed to track
595 an issue. That said, if the issue can be reasonably handled and/or isn't
596 going to crash the daemon, it shouldn't be an assert.
598 For anything else where internal constraints are violated but we're not
599 breaking due to it, it's an error instead (not a debug.) These require
600 "user action" of notifying the developers.
604 * mismatched :code:`prev`/:code:`next` pointers in lists
605 * some field that is absolutely needed is :code:`NULL`
606 * any other kind of data structure corruption that will cause the daemon
607 to crash sooner or later, one way or another
609 Thread-local buffering
610 ----------------------
612 The core logging code in :file:`lib/zlog.c` allows setting up per-thread log
613 message buffers in order to improve logging performance. The following rules
614 apply for this buffering:
616 * Only messages of priority *DEBUG* or *INFO* are buffered.
617 * Any higher-priority message causes the thread's entire buffer to be flushed,
618 thus message ordering is preserved on a per-thread level.
619 * There is no guarantee on ordering between different threads; in most cases
620 this is arbitrary to begin with since the threads essentially race each
621 other in printing log messages. If an order is established with some
622 synchronization primitive, add calls to :c:func:`zlog_tls_buffer_flush()`.
623 * The buffers are only ever accessed by the thread they are created by. This
624 means no locking is necessary.
626 Both the main/default thread and additional threads created by
627 :c:func:`frr_pthread_new()` with the default :c:func:`frr_run()` handler will
628 initialize thread-local buffering and call :c:func:`zlog_tls_buffer_flush()`
631 If some piece of code runs for an extended period, it may be useful to insert
632 calls to :c:func:`zlog_tls_buffer_flush()` in appropriate places:
634 .. c:function:: void zlog_tls_buffer_flush(void)
636 Write out any pending log messages that the calling thread may have in its
637 buffer. This function is safe to call regardless of the per-thread log
638 buffer being set up / in use or not.
640 When working with threads that do not use the :c:struct:`thread_master`
641 event loop, per-thread buffers can be managed with:
643 .. c:function:: void zlog_tls_buffer_init(void)
645 Set up thread-local buffering for log messages. This function may be
646 called repeatedly without adverse effects, but remember to call
647 :c:func:`zlog_tls_buffer_fini()` at thread exit.
651 If this function is called, but :c:func:`zlog_tls_buffer_flush()` is
652 not used, log message output will lag behind since messages will only be
653 written out when the buffer is full.
655 Exiting the thread without calling :c:func:`zlog_tls_buffer_fini()`
656 will cause buffered log messages to be lost.
658 .. c:function:: void zlog_tls_buffer_fini(void)
660 Flush pending messages and tear down thread-local log message buffering.
661 This function may be called repeatedly regardless of whether
662 :c:func:`zlog_tls_buffer_init()` was ever called.
667 The actual logging subsystem (in :file:`lib/zlog.c`) is heavily separated
668 from the actual log writers. It uses an atomic linked-list (`zlog_targets`)
669 with RCU to maintain the log targets to be called. This list is intended to
670 function as "backend" only, it **is not used for configuration**.
672 Logging targets provide their configuration layer on top of this and maintain
673 their own capability to enumerate and store their configuration. Some targets
674 (e.g. syslog) are inherently single instance and just stuff their config in
675 global variables. Others (e.g. file/fd output) are multi-instance capable.
676 There is another layer boundary here between these and the VTY configuration
682 .. c:struct:: zlog_target
684 This struct needs to be filled in by any log target and then passed to
685 :c:func:`zlog_target_replace()`. After it has been registered,
686 **RCU semantics apply**. Most changes to associated data should make a
687 copy, change that, and then replace the entire struct.
689 Additional per-target data should be "appended" by embedding this struct
690 into a larger one, for use with `containerof()`, and
691 :c:func:`zlog_target_clone()` and :c:func:`zlog_target_free()` should be
692 used to allocate/free the entire container struct.
694 Do not use this structure to maintain configuration. It should only
695 contain (a copy of) the data needed to perform the actual logging. For
696 example, the syslog target uses this:
701 struct zlog_target zt;
705 static void zlog_syslog(struct zlog_target *zt, struct zlog_msg *msgs[], size_t nmsgs)
707 struct zlt_syslog *zte = container_of(zt, struct zlt_syslog, zt);
710 for (i = 0; i < nmsgs; i++)
711 if (zlog_msg_prio(msgs[i]) <= zt->prio_min)
712 syslog(zlog_msg_prio(msgs[i]) | zte->syslog_facility, "%s",
713 zlog_msg_text(msgs[i], NULL));
717 .. c:function:: struct zlog_target *zlog_target_clone(struct memtype *mt, struct zlog_target *oldzt, size_t size)
719 Allocates a logging target struct. Note that the ``oldzt`` argument may be
720 ``NULL`` to allocate a "from scratch". If ``oldzt`` is not ``NULL``, the
721 generic bits in :c:struct:`zlog_target` are copied. **Target specific
722 bits are not copied.**
724 .. c:function:: struct zlog_target *zlog_target_replace(struct zlog_target *oldzt, struct zlog_target *newzt)
726 Adds, replaces or deletes a logging target (either ``oldzt`` or ``newzt`` may be ``NULL``.)
728 Returns ``oldzt`` for freeing. The target remains possibly in use by
729 other threads until the RCU cycle ends. This implies you cannot release
730 resources (e.g. memory, file descriptors) immediately.
732 The replace operation is not atomic; for a brief period it is possible that
733 messages are delivered on both ``oldzt`` and ``newzt``.
737 ``oldzt`` must remain **functional** until the RCU cycle ends.
739 .. c:function:: void zlog_target_free(struct memtype *mt, struct zlog_target *zt)
741 Counterpart to :c:func:`zlog_target_clone()`, frees a target (using RCU.)
743 .. c:member:: void (*zlog_target.logfn)(struct zlog_target *zt, struct zlog_msg *msgs[], size_t nmsg)
745 Called on a target to deliver "normal" logging messages. ``msgs`` is an
746 array of opaque structs containing the actual message. Use ``zlog_msg_*``
747 functions to access message data (this is done to allow some optimizations,
748 e.g. lazy formatting the message text and timestamp as needed.)
752 ``logfn()`` must check each individual message's priority value against
753 the configured ``prio_min``. While the ``prio_min`` field is common to
754 all targets and used by the core logging code to early-drop unneeded log
755 messages, the array is **not** filtered for each ``logfn()`` call.
757 .. c:member:: void (*zlog_target.logfn_sigsafe)(struct zlog_target *zt, const char *text, size_t len)
759 Called to deliver "exception" logging messages (i.e. SEGV messages.)
760 Must be Async-Signal-Safe (may not allocate memory or call "complicated"
761 libc functions.) May be ``NULL`` if the log target cannot handle this.
766 :file:`lib/zlog_targets.c` provides the standard file / fd / syslog targets.
767 The syslog target is single-instance while file / fd targets can be
768 instantiated as needed. There are 3 built-in targets that are fully
769 autonomous without any config:
771 - startup logging to `stderr`, until either :c:func:`zlog_startup_end()` or
772 :c:func:`zlog_aux_init()` is called.
773 - stdout logging for non-daemon programs using :c:func:`zlog_aux_init()`
774 - crashlogs written to :file:`/var/tmp/frr.daemon.crashlog`
776 The regular CLI/command-line logging setup is handled by :file:`lib/log_vty.c`
777 which makes the appropriate instantiations of syslog / file / fd targets.
781 :c:func:`zlog_startup_end()` should do an explicit switchover from
782 startup stderr logging to configured logging. Currently, configured logging
783 starts in parallel as soon as the respective setup is executed. This results
784 in some duplicate logging.