0
0
mirror of https://gitlab.nic.cz/labs/bird.git synced 2024-09-20 12:25:20 +00:00
Commit Graph

3873 Commits

Author SHA1 Message Date
Maria Matejka
c651cef737 Global interface list renamed to not clash with local lists 2022-02-03 10:30:33 +01:00
Maria Matejka
b6c9263543 Interfaces and neighbor notifications do properly enter protocol loops 2022-02-03 10:30:33 +01:00
Maria Matejka
b5155d5cea Page cleanup routines updated to allow nested birdloops 2022-02-03 10:30:33 +01:00
Maria Matejka
87512ccdff Route table fast prune fixup 2022-02-03 10:30:33 +01:00
Maria Matejka
f22bab9818 Babel: fixed interface pool removal 2022-02-03 10:30:33 +01:00
Maria Matejka
3cf5e0f144 Dropping a nonsensical assert which broke Babel 2022-02-03 10:03:37 +01:00
Maria Matejka
ee84ff14a1 Removing periodic wakeup of mainloop in debug version to search for notification issues 2022-02-03 10:03:37 +01:00
Maria Matejka
935b0b7124 Performance data of sark vs. v2.0.8 with a chart generator 2022-02-03 10:03:17 +01:00
Maria Matejka
ffa6a51c58 Thread documentation: Chapter 4 on memory management 2021-12-22 15:35:49 +01:00
Maria Matejka
a840170e84 Thread documentation: Completely rewritten chapter 3 on loops and locks 2021-12-08 20:31:12 +01:00
Maria Matejka
827c78297e Final version of asynchronous export documentation 2021-12-08 12:39:48 +01:00
Maria Matejka
b6612ec792 Thread documentation: chapter 3, coroutines and locking 2021-12-08 12:39:36 +01:00
Maria Matejka
f459deee9f Thread documentation: chapters 0, 1 and 2 2021-12-08 12:39:28 +01:00
Maria Jan Matejka
493d45d950 Fixed build errors for OpenBSD 2021-12-07 16:59:44 +00:00
Maria Matejka
f9e098c98a Fixed standby memory page counters on shutdown
Bug introduced by commit 38278d94ba.
2021-12-07 14:55:27 +01:00
Maria Matejka
542f24555d GDB: io loop printing support in resource dumps, better iterating over linpool chunks 2021-12-07 13:07:00 +01:00
Maria Matejka
a9efce68b5 Fixed a race condition in channel aux table cleanup 2021-12-07 13:05:20 +01:00
Maria Matejka
4f43d326b4 Event list asserts to prevent running into rotten data 2021-12-07 13:05:20 +01:00
Maria Matejka
e32eafaaa7 Using more Python-ish constructions in BIRD linked-list accessors 2021-12-07 13:05:20 +01:00
Maria Matejka
2117864a87 Main IO loop shouldn't skip sockets when ping is received 2021-12-07 13:05:20 +01:00
Maria Matejka
c5b6b18413 Standby memory pages are accounted for as overhead in show memory command 2021-12-02 11:29:59 +00:00
Maria Matejka
38278d94ba No memory unmapping when shutting down
All the memory is just freed implicitly on exit, no need for
page-by-page unmapping.
2021-12-01 22:02:42 +00:00
Maria Matejka
b2bac7ae91 Faster shutdown and cleanups by freeing route attributes strictly from main loop 2021-12-01 21:52:55 +00:00
Maria Matejka
387b279f60 Faster prune on table deletion 2021-12-01 21:33:38 +00:00
Maria Matejka
5f94d684d0 Table prune routines request export announcements directly 2021-12-01 16:21:35 +00:00
Maria Matejka
7092bdc0c9 fixed duplicate routes propagation 2021-12-01 14:02:23 +01:00
Maria Matejka
b50224a003 Merge branch 'master' into HEAD 2021-12-01 13:04:52 +01:00
Maria Matejka
55ee9961e0 Fix of shutdown: premature log cleanup led to use-after-free 2021-12-01 13:00:54 +01:00
Maria Matejka
bb63e99d78 Page allocator moved from pools to IO loops.
The resource pool system is highly hierarchical and keeping spare pages
in pools leads to unnecessarily complex memory management.

Loops have a flat hiearchy, at least for now, and it is therefore much
easier to keep care of pages, especially in cases of excessive virtual memory
fragmentation.
2021-12-01 13:00:54 +01:00
Maria Matejka
385b3ea395 For safer memory allocations, resources are bound to loops.
Also all loops have their basic resource pool for allocations which are
auto-freed when the loop is stopping.
2021-11-30 21:38:25 +01:00
Maria Matejka
f772afc525 Memory statistics split into Effective and Overhead
This feature is intended mostly for checking that BIRD's allocation
strategies don't consume much memory space. There are some cases where
withdrawing routes in a specific order lead to memory fragmentation and
this output should give the user at least a notion of how much memory is
actually used for data storage and how much memory is "just allocated"
or used for overhead.

Also raising the "system allocator overhead estimation" from 8 to 16
bytes; it is probably even more. I've found 16 as a local minimum in
best scenarios among reachable machines. I couldn't find any reasonable
method to estimate this value when BIRD starts up.

This commit also fixes the inaccurate computation of memory overhead for
slabs where the "system allocater overhead estimation" was improperly
added to the size of mmap-ed memory.
2021-11-27 22:54:15 +01:00
Maria Matejka
ab0994a10c fixup of table reload bug 2021-11-24 23:20:26 +01:00
Maria Matejka
644e9ca94e Directly mapped pages are kept for future use if temporarily not needed 2021-11-24 19:42:52 +00:00
Maria Matejka
57d0ecb9b7 GDB pretty printer for resource pools 2021-11-24 15:48:13 +00:00
Maria Matejka
5395880908 Fixed pipe reload/refeed 2021-11-23 12:06:27 +00:00
Maria Matejka
821344c781 Stored pages release routine 2021-11-23 11:13:11 +00:00
Maria Matejka
4ba991f19c Fixed channel export map confusion 2021-11-23 11:12:32 +00:00
Maria Matejka
44dbedbe3f BGP: End route refresh before another starts 2021-11-22 19:05:44 +01:00
Maria Matejka
3fd1f46184 RPKI has its own loop 2021-11-22 19:05:44 +01:00
Maria Matejka
20ace7f2e6 RPKI: Use the route refresh mechanism also for the first load 2021-11-22 19:05:44 +01:00
Maria Matejka
1e8e3b7c12 More information on RPKI errors 2021-11-22 19:05:44 +01:00
Maria Matejka
aadf690b14 Higher settle times when route refresh in the source table is running 2021-11-22 19:05:44 +01:00
Maria Matejka
df476c2e5d Corking also feed start to keep BIRD running when refeeds would easily cause congestion 2021-11-22 19:05:44 +01:00
Maria Matejka
0fd1c1d091 Route attribute cache is now lockless on read / clone.
Lots of time was spent locking when accessing route attribute cache.
This overhead should be now reduced to a minimum.
2021-11-22 19:05:44 +01:00
Maria Matejka
2a224a9e1e Route sources have their separate global lock 2021-11-22 19:05:44 +01:00
Maria Matejka
1b39473993 Introducing basic RCU primitives for lock-less shared data structures 2021-11-22 19:05:44 +01:00
Maria Matejka
adf37d8eff VRF setting reduced to one argument, using default dummy iface for default vrf 2021-11-22 19:05:44 +01:00
Maria Matejka
dc160e11e1 Route table import-to-export announcement indirection to reduce pipe traffic 2021-11-22 19:05:44 +01:00
Maria Matejka
794a4eefa1 Keeping un-unmmappable pages until they can be reused
On Linux, munmap() may fail with ENOMEM when virtual memory is too
fragmented. Working this around by just keeping such blocks for future
use.
2021-11-22 19:05:44 +01:00
Maria Matejka
4f3fa1623f Pipe runs in parallel. 2021-11-22 19:05:44 +01:00