lib: revert datastructure reverts...

[eqvinox]

This restores the full changeset from #3045, by reverting reverts:

• Revert "lib: use DECLARE_SKIPLIST for timers instead of pqueue"
• Revert "lib: remove pqueue_*"

Signed-off-by: David Lamparter equinox@opensourcerouting.org

[louberger]

This is for testing purposes only, right?

[NetDEF-CI]

Continuous Integration Result: SUCCESSFUL

Congratulations, this patch passed basic tests

Tested-by: NetDEF / OpenSourceRouting.org CI System

CI System Testrun URL: https://ci1.netdef.org/browse/FRR-FRRPULLREQ-7412/

This is a comment from an automated CI system.
For questions and feedback in regards to this CI system, please feel free to email
Martin Winter - mwinter (at) opensourcerouting.org.

Warnings Generated during build:

Checkout code: Successful with additional warnings

cp: cannot stat './lib/pqueue.c': No such file or directory
cp: cannot stat './lib/pqueue.h': No such file or directory
Report for thread.c | 2 issues
===============================================
< WARNING: braces {} are not necessary for any arm of this statement
< #1068: FILE: /tmp/f1-24226/thread.c:1068:

---

CLANG Static Analyzer Summary

• Github Pull Request 4229, comparing to Git base SHA 31e944a
• Base image data for Git 31e944a does not exist - compare skipped

12 Static Analyzer issues remaining.

See details at
https://ci1.netdef.org/browse/FRR-FRRPULLREQ-7412/artifact/shared/static_analysis/index.html

[LabN-CI]

💚 Basic BGPD CI results: SUCCESS, 0 tests failed

Results table

_	_
Result	SUCCESS git merge/4229 4cb91dc
Date	04/30/2019
Start	12:15:05
Finish	12:39:30
Run-Time	24:25
Total	1816
Pass	1816
Fail	0
Valgrind-Errors	0
Valgrind-Loss	0
Details	vncregress-2019-04-30-12:15:05.txt
Log	autoscript-2019-04-30-12:15:51.log.bz2
Memory	451 446 374

For details, please contact louberger

[eqvinox]

This is for testing purposes only, right?

No; the reverts that I stacked on #3045 were for testing purposes, but since that got merged with the testing reverts, this PR restores the full changeset.

Either way now we can try and figure out wtf the cause is of those weird timing results you reported... can you please run those tests a few more times so we have timing + deviation info?

[LabN-CI]

💚 Basic BGPD CI results: SUCCESS, 0 tests failed

Results table

_	_
Result	SUCCESS git merge/4229 4cb91dc
Date	05/10/2019
Start	07:54:29
Finish	08:18:51
Run-Time	24:22
Total	1813
Pass	1813
Fail	0
Valgrind-Errors	0
Valgrind-Loss	0
Details	vncregress-2019-05-10-07:54:29.txt
Log	autoscript-2019-05-10-07:55:17.log.bz2
Memory	431 434 359

For details, please contact louberger

[louberger]

I'm still seeing a notable processing time increase on vpn route addition (the same 50000 routes coming from two CPEs, time to all 100K routes seen in VPN instance):
PR:
6.17.4 7.0.0.132 Done Registering, L=260 (8.195 secs) 1 0
6.17.5 0.130_cli All 100000 prefixes rx'ed: 127.0.0.131 1 0
(+1.069)
Master
6.17.4 7.0.0.132 Done Registering, L=260 (7.894 secs) 1 0
6.17.5 0.130_cli All 100000 prefixes rx'ed: 127.0.0.131 1 0
(+0.062)

17.4 is time to add routes
17.5 is time for all routes to be seen in l3vpn safi

It should be possible to modify tests/topotests/bgp_l3vpn_to_bgp_vrf to do this same test - maybe via sharpd.

[eqvinox]

Closing this - doing another rework.

[louberger]

CI:rerun

[NetDEF-CI]

Continuous Integration Result: SUCCESSFUL

Congratulations, this patch passed basic tests

Tested-by: NetDEF / OpenSourceRouting.org CI System

CI System Testrun URL: https://ci1.netdef.org/browse/FRR-FRRPULLREQ-7748/

This is a comment from an automated CI system.
For questions and feedback in regards to this CI system, please feel free to email
Martin Winter - mwinter (at) opensourcerouting.org.

Warnings Generated during build:

Checkout code: Successful with additional warnings

cp: cannot stat './lib/pqueue.c': No such file or directory
cp: cannot stat './lib/pqueue.h': No such file or directory
Report for thread.c | 2 issues
===============================================
< WARNING: braces {} are not necessary for any arm of this statement
< #1068: FILE: /tmp/f1-12278/thread.c:1068:

---

CLANG Static Analyzer Summary

• Github Pull Request 4229, comparing to Git base SHA 31e944a

No Changes in Static Analysis warnings compared to base

12 Static Analyzer issues remaining.

See details at
https://ci1.netdef.org/browse/FRR-FRRPULLREQ-7748/artifact/shared/static_analysis/index.html