First, I wonder if you are a troll.
Test writes:
> I maintain some open source code myself, so I appreciate that code can
> have vulnerabilities, and that this is the work of (semi) volunteers.
>
> I also appreciate that sometimes there's a bit of a DIY element to
> maintaining systems.
>
> However, I really have to question to logic of marking the 4.2.8 release
> as stable and apparently abandoning the 4.2.6 code tree.
Looking forward to seeing how you come to this conclusion...
> Because
> - 4.2.6 is now known to be broken. Normally a developer would give some
> fair notice that 4.2.6 was no longer going to be supported, and give
> people time to move to 4.2.7 or 4.2.8. OK one day old code will be
> broken, but please don't leave it broken on the day you deprecate it.
We've had a *long* history of deprecating the "previous" stable version
when a new one becomes available.
You seem to have missed the part where over 1100 bugfixes and
improvements have been made to 4.2.8 since 4.2.6.
We have long offered to backport patches to EOL'd releases for anybody
who wants to pay for that. No takers yet.
And I fail to see the logic in enabling people to run older, buggier
software when significantly better software is available.
> - 4.2.7 is also still apparently broken for 1 of the 4 major
> vulnerabilities if I read the notice correctly. So there's apparently no
> point in moving to that.
No idea, but 4.2.7 is a *development* release that preceeds 4.2.8. We
don't pre-release security patches, that is an invitation to disaster.
So 4.2.7pLAST + security patches == 4.2.8
> - 4.2.8 doesn't even compile on my systems, so how can it be a stable
> release?
What about 4.2.8p1? I'm not aware of any build problems with that
release. 4.2.8p1 was released over a month ago.
Was your problem a reported bug? Is it fixed now?
If it's not a reported bug, why would you expect us to fix a bug we
don't see and nobody has reported?
> Sorry, it is with regret that I'm going to have to pull my small
> contribution to the pool (via firewall rules) until this is fixed.
See above.
> I run two stratum 1 servers (with DCF 77 and GPS reference clocks for
> diversity) and I thought contributing to the pool was a "good thing",
> but not at the cost of my own operations.
OK, looking at your output below I see 1 problem:
- the package manager probably messed up and provided libraries but not
headers. We don't build or maintain packages for distros. You don't
have sys/capability.h installed. Complain to the debian package
manager.
If they (or anybody else) wants to help build packages for debian or any
other distros, we can offer a way to do those and put this as part of
our build test suite. I don't have time to add this to my pile of work.
Since you didn't provide any information about the rPi, I have no
information for you there. Somebody who also has your problem who is
able to debug this and submit a bug report and better still a fix to
whoever has the problem (likely the packager, less likely NTP) will have
better luck.
And I'd recommend you and everybody else run the latest version of ntp,
which is currently 4.2.8p1.
H
--
>
> ---- debug info below ----
>
> Ubuntu 10.4 -> Server LTS no package available.
>
> 4.2.8 MD5 6972a626be6150db8cfbd0b63d8719e7 ntp-4.2.8.tar.gz
>
> sudo service ntp start
> * Starting NTP server
> ntpd /usr/sbin/ntpd:
> The 'user' option has been disabled. -- built without --enable-clockctl
> or --enable-linuxcaps or --enable-solarisprivs
> ntpd - NTP daemon program - Ver. 4.2.8
> Usage: ntpd [ -<flag> [<val>] | --<name>[{=| }<val>] ]... \
> [ <server1> ... <serverN> ]
> Try 'ntpd --help' for more information
>
> Does not compile with the --enable-linuxcaps option so I have to kludge
> my config file as the user option is no longer supported
>
> ntpd --version
> ntpd [email protected] Sun Dec 21 13:36:58 UTC 2014 (1)
>
>
>
> Here's the compile error after
>
> ./configure --enable-linuxcaps
> make
>
> .......
>
> refclock_acts.c: In function acts_message:
> refclock_acts.c:414: warning: ignoring return value of write, declared
> with attribute warn_unused_result
> refclock_acts.c: In function acts_timeout:
> refclock_acts.c:553: warning: ignoring return value of write, declared
> with attribute warn_unused_result
> CC refclock_arbiter.o
> refclock_arbiter.c: In function arb_start:
> refclock_arbiter.c:203: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> refclock_arbiter.c: In function arb_receive:
> refclock_arbiter.c:288: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> refclock_arbiter.c:295: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> refclock_arbiter.c:297: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> refclock_arbiter.c:302: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> refclock_arbiter.c:308: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> refclock_arbiter.c:314: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> refclock_arbiter.c:325: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> refclock_arbiter.c:354: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> refclock_arbiter.c:404: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> refclock_arbiter.c:410: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> refclock_arbiter.c:428: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> CC refclock_arc.o
> CC refclock_as2201.o
> refclock_as2201.c: In function as2201_receive:
> refclock_as2201.c:343: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> CC refclock_atom.o
> CC refclock_bancomm.o
> CC refclock_chronolog.o
> CC refclock_chu.o
> refclock_chu.c: In function chu_burst:
> refclock_chu.c:1271: warning: assuming signed overflow does not occur
> when simplifying conditional to constant
> CC refclock_conf.o
> CC refclock_datum.o
> CC refclock_dumbclock.o
> CC refclock_fg.o
> CC refclock_gpsdjson.o
> refclock_gpsdjson.c: In function process_tpv:
> refclock_gpsdjson.c:857: warning: declaration of log2 shadows a global
> declaration
> /usr/include/bits/mathcalls.h:145: warning: shadowed declaration is here
> CC refclock_gpsvme.o
> CC refclock_heath.o
> CC refclock_hopfser.o
> CC refclock_hopfpci.o
> CC refclock_hpgps.o
> CC refclock_irig.o
> CC refclock_jjy.o
> CC refclock_jupiter.o
> CC refclock_leitch.o
> CC refclock_local.o
> CC refclock_mx4200.o
> CC refclock_neoclock4x.o
> CC refclock_nmea.o
> refclock_nmea.c: In function eval_gps_time:
> refclock_nmea.c:1782: warning: declaration of recv shadows a global
> declaration
> /usr/include/bits/socket2.h:35: warning: shadowed declaration is here
> CC refclock_oncore.o
> CC refclock_palisade.o
> refclock_palisade.c: In function HW_poll:
> refclock_palisade.c:1157: warning: ignoring return value of write,
> declared with attribute warn_unused_result
> CC refclock_parse.o
> CC refclock_pcf.o
> CC refclock_pst.o
> CC refclock_ripencc.o
> CC refclock_shm.o
> CC refclock_tpro.o
> CC refclock_true.o
> CC refclock_tt560.o
> CC refclock_ulink.o
> CC refclock_wwv.o
> CC refclock_wwvb.o
> CC refclock_zyfer.o
> CC refclock_tsyncpci.o
> AR libntpd.a
> CC ntp_config.o
> CC ntp_io.o
> CC ntp_parser.o
> ntp_parser.y: In function yyparse:
> ../../ntpd/ntp_parser.c:1914: warning: assuming pointer wraparound does
> not occur when comparing P +- C1 with P +- C2
> ../../ntpd/ntp_parser.c:1914: warning: assuming pointer wraparound does
> not occur when comparing P +- C1 with P +- C2
> ../../ntpd/ntp_parser.c:1914: warning: assuming pointer wraparound does
> not occur when comparing P +- C1 with P +- C2
> CC ntp_scanner.o
> CC ntpd.o
> ntpd.c:108:29: error: sys/capability.h: No such file or directory
> ntpd.c: In function ntpdmain:
> ntpd.c:970: error: cap_t undeclared (first use in this function)
> ntpd.c:970: error: (Each undeclared identifier is reported only once
> ntpd.c:970: error: for each function it appears in.)
> ntpd.c:970: error: expected ; before caps
> ntpd.c:976: error: caps undeclared (first use in this function)
> ntpd.c:976: warning: implicit declaration of function cap_from_text
> ntpd.c:983: warning: implicit declaration of function cap_set_proc
> ntpd.c:988: warning: implicit declaration of function cap_free
> make[3]: *** [ntpd.o] Error 1
> make[3]: Leaving directory `/home/xxx/ntp-build-4.2.8/ntp-4.2.8/ntpd'
> make[2]: *** [all] Error 2
> make[2]: Leaving directory `/home/xxx/ntp-build-4.2.8/ntp-4.2.8/ntpd'
> make[1]: *** [all-recursive] Error 1
> make[1]: Leaving directory `/home/xxx/ntp-build-4.2.8/ntp-4.2.8'
> make: *** [all] Error 2
>
>
> $ uname -a
> Linux XXXXXX 2.6.32-70-generic #137-Ubuntu SMP Tue Dec 9 11:49:12 UTC
> 2014 x86_64 GNU/Linux
>
> ~/ntp-build-4.2.8$ gcc --version
> gcc (Ubuntu 4.4.3-4ubuntu5.1) 4.4.3
> Copyright (C) 2009 Free Software Foundation, Inc.
> This is free software; see the source for copying conditions. There is NO
> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
>
>
>
>
> Compiling on a Raspberry Pi
>
> 4.2.8 MD5 6972a626be6150db8cfbd0b63d8719e7 ntp-4.2.8.tar.gz
>
> uname -a
> Linux raspberrypi1 3.12.33+ #722 PREEMPT Sat Nov 22 12:03:51 GMT 2014
> armv6l GNU/Li
>
> Gave up due to too many errors so I've just isolated these systems for now
>
>
>
> _______________________________________________
> pool mailing list
> [email protected]
> http://lists.ntp.org/listinfo/pool
>
_______________________________________________
pool mailing list
[email protected]
http://lists.ntp.org/listinfo/pool
