Re: Poudriere failed to build pkg in 13-stable jail under 12-stable

[Baptiste Daroussin]

On Mon, Jan 25, 2021 at 12:46:43PM +0800, Thomas Legg wrote:
> The build of a 13-stable source and kernel were a success under 12-stable
> (though with some issues on freeze-ups and hard reboots that I suspect
> might be related to the bufdaemon issue and my 0x15 gen AMD cpu).
> 
> Created a 13-stable poudriere jail with the knowledge that there may be
> issues with builds under a 12-stable r369076 kernel. I didn't expect this
> failure on packaging pkg.
> 
> > Compressing man pages (compress-man)
> ===>   Installing ldconfig configuration file
> ===
> ===
> ===>  Building package for pkg-1.16.2
> cp: /wrkdirs/usr/ports/ports-mgmt/pkg/work/pkg/pkg-1.16.2.txz: Function not
> implemented
> *** Error code 1
> 

cp on head and 13 is using copy_file_range(2), introduced in 13-CURRENT
recentish, this is what is failing here.

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: pkg thinks kernel is old

[Baptiste Daroussin]

On Fri, Oct 11, 2019 at 03:07:13PM +0200, Bengt Ahlgren wrote:
> Baptiste Daroussin  writes:
> 
> > On Fri, Oct 11, 2019 at 02:50:57PM +0200, Bengt Ahlgren wrote:
> >> Baptiste Daroussin  writes:
> >> 
> >> > On Fri, Oct 11, 2019 at 01:57:06PM +0200, Bengt Ahlgren wrote:
> >> >> Baptiste Daroussin  writes:
> >> >> 
> >> >> > On Fri, Oct 11, 2019 at 11:54:15AM +0200, Bengt Ahlgren wrote:
> >> >> >> Baptiste Daroussin  writes:
> >> >> >> 
> >> >> >> > On Fri, Oct 11, 2019 at 11:38:46AM +0200, Bengt Ahlgren wrote:
> >> >> >> >> Baptiste Daroussin  writes:
> >> >> >> >> 
> >> >> >> >> > On Thu, Oct 10, 2019 at 03:44:03PM +0200, Bengt Ahlgren wrote:
> >> >> >> >> >> Ronald Klop  writes:
> >> >> >> >> >> 
> >> >> >> >> >> > Van: Bengt Ahlgren 
> >> >> >> >> >> > Datum: woensdag, 9 oktober 2019 23:38
> >> >> >> >> >> > Aan: sta...@freebsd.org
> >> >> >> >> >> > Onderwerp: pkg thinks kernel is old
> >> >> >> >> >> >>
> >> >> >> >> >> >> I've ran into this on my 11.3-STABLE system:
> >> >> >> >> >> >>
> >> >> >> >> >> >> # uname -K
> >> >> >> >> >> >> 1103500
> >> >> >> >> >> >> # pkg update -f
> >> >> >> >> >> >> Updating ivyp repository catalogue...
> >> >> >> >> >> >> Fetching meta.txz: 100%560 B   0.6kB/s00:01
> >> >> >> >> >> >> Fetching
> >> >> >> >> >> >> packagesite.txz: 100%  410 KiB 419.9kB/s00:01
> >> >> >> >> >> >> Processing
> >> >> >> >> >> >> entries:   0%
> >> >> >> >> >> >> Newer FreeBSD version for package zziplib:
> >> >> >> >> >> >> To ignore this error set IGNORE_OSVERSION=yes
> >> >> >> >> >> >> - package: 1103000
> >> >> >> >> >> >> - running kernel: 1102509
> >> >> >> >> >> >> Ignore the mismatch and continue? [Y/n]:
> >> >> >> >> >> >>
> >> >> >> >> >> >> I build my own packages with poudriere, and have just 
> >> >> >> >> >> >> switched from an
> >> >> >> >> >> >> 11.2-REL to an 11.3-REL jail.  I didn't force-upgrade 
> >> >> >> >> >> >> everything after
> >> >> >> >> >> >> the switch, perhaps that's needed to make pkg recognise the 
> >> >> >> >> >> >> correct
> >> >> >> >> >> >> kernel version?
> >> >> >> >> >> >>
> >> >> >> >> >> >> Bengt
> >> >> >> >> >> 
> >> >> >> >> >> > This message of pkg is misleading.
> >> >> >> >> >> >
> >> >> >> >> >> > It looks at /bin/sh or something like that for the version. 
> >> >> >> >> >> > And prints
> >> >> >> >> >> > that as the 'kernel' version. So if you did an incremental 
> >> >> >> >> >> > build and
> >> >> >> >> >> > /bin/sh was not changed the version stays the same.
> >> >> >> >> >> > To fix it you can do a clean buildworld/installworld.
> >> >> >> >> >> >
> >> >> >> >> >> > Regards,
> >> >> >> >> >> > Ronald.
> >> >> >> >> >> 
> >> >> >> >> >> That might well be it - I've done some buildworld:s with 
> >> >> >> >> >> -DNO_CLEAN
> >> >> >> >> >> lately.  I'll test - thanks for the tip!
> >> >> >> >> >
> >> >> >> >> > Can you provide the uname -U information (building with 
> >> >> >> >> > -DNO_CLEAN) should be
> >> >> >> >> > perfectly fine.
> >> >> >> >> 
> >> >> >> >> Thanks for looking into this!
> >> >> >> >> 
> >> >> >> >> $ uname -U
> >> >> >> >> 1103500
> >> >> >> >> 
> >> >> >> >> Let me know if there is something else I can check before I do a 
> >> >> >> >> full
> >> >> >> >> buildworld on this system!
> >> >> >> >
> >> >> >> > What are the output of pkg config osversion
> >> >> >> > and pkg -o ABI_FILE=/usr/bin/uname config osversion
> >> >> >> > please
> >> >> >> 
> >> >> >> $ pkg config osversion
> >> >> >> 1102509
> >> >> >> $ pkg -o ABI_FILE=/usr/bin/uname config osversion
> >> >> >> 1102509
> >> >> >
> >> >> > It means you /usr/bin/uname binary has been built with a compiler 
> >> >> > that says it
> >> >> > is building for 1102509.
> >> >> >
> >> >> > Can you show the output of file /usr/bin/uname ?
> >> >> 
> >> >> $ uname -aKU
> >> >> FreeBSD P142.sics.se 11.3-STABLE FreeBSD 11.3-STABLE #9 r353280: Mon
> >> >> Oct 7 22:16:02 CEST 2019
> >> >> r...@p142.sics.se:/storage/obj/usr/src/sys/X250 amd64 1103500
> >> >> 1103500
> >> >> 
> >> >> no signs of 1102509 there...
> >> >> 
> >> >> and "strings -a /usr/bin/uname |grep 110" returns nothing.
> >> >> 
> >> > I am interested in the output of file /usr/bin/uname
> >> 
> >> Ah, right, didn't get it that you meant the file command, sorry...
> >> 
> >> $ file /usr/bin/uname 
> >> /usr/bin/uname: ELF 64-bit LSB executable, x86-64, version 1
> >> (FreeBSD), dynamically linked, interpreter /libexec/ld-elf.so.1, for
> >> FreeBSD 11.2 (1102509), FreeBSD-style, stripped
> >> 
> >
> > That is the problem.
> >
> > So if you don't want to rebuild you can add to your make.conf ABI_FILE= XXX
> > where XXX is the path to any file where when you run file on it says it was
> > built for freebsd 11.3
> 
> Great, thanks for investigating!  It also means that it is safe to
> answer "Y" to the pkg question and just let it carry on!
> 
> Bengt

yes :x
it is.

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: pkg thinks kernel is old

[Baptiste Daroussin]

On Fri, Oct 11, 2019 at 02:50:57PM +0200, Bengt Ahlgren wrote:
> Baptiste Daroussin  writes:
> 
> > On Fri, Oct 11, 2019 at 01:57:06PM +0200, Bengt Ahlgren wrote:
> >> Baptiste Daroussin  writes:
> >> 
> >> > On Fri, Oct 11, 2019 at 11:54:15AM +0200, Bengt Ahlgren wrote:
> >> >> Baptiste Daroussin  writes:
> >> >> 
> >> >> > On Fri, Oct 11, 2019 at 11:38:46AM +0200, Bengt Ahlgren wrote:
> >> >> >> Baptiste Daroussin  writes:
> >> >> >> 
> >> >> >> > On Thu, Oct 10, 2019 at 03:44:03PM +0200, Bengt Ahlgren wrote:
> >> >> >> >> Ronald Klop  writes:
> >> >> >> >> 
> >> >> >> >> > Van: Bengt Ahlgren 
> >> >> >> >> > Datum: woensdag, 9 oktober 2019 23:38
> >> >> >> >> > Aan: sta...@freebsd.org
> >> >> >> >> > Onderwerp: pkg thinks kernel is old
> >> >> >> >> >>
> >> >> >> >> >> I've ran into this on my 11.3-STABLE system:
> >> >> >> >> >>
> >> >> >> >> >> # uname -K
> >> >> >> >> >> 1103500
> >> >> >> >> >> # pkg update -f
> >> >> >> >> >> Updating ivyp repository catalogue...
> >> >> >> >> >> Fetching meta.txz: 100%560 B   0.6kB/s00:01Fetching
> >> >> >> >> >> packagesite.txz: 100%  410 KiB 419.9kB/s00:01Processing
> >> >> >> >> >> entries:   0%
> >> >> >> >> >> Newer FreeBSD version for package zziplib:
> >> >> >> >> >> To ignore this error set IGNORE_OSVERSION=yes
> >> >> >> >> >> - package: 1103000
> >> >> >> >> >> - running kernel: 1102509
> >> >> >> >> >> Ignore the mismatch and continue? [Y/n]:
> >> >> >> >> >>
> >> >> >> >> >> I build my own packages with poudriere, and have just switched 
> >> >> >> >> >> from an
> >> >> >> >> >> 11.2-REL to an 11.3-REL jail.  I didn't force-upgrade 
> >> >> >> >> >> everything after
> >> >> >> >> >> the switch, perhaps that's needed to make pkg recognise the 
> >> >> >> >> >> correct
> >> >> >> >> >> kernel version?
> >> >> >> >> >>
> >> >> >> >> >> Bengt
> >> >> >> >> 
> >> >> >> >> > This message of pkg is misleading.
> >> >> >> >> >
> >> >> >> >> > It looks at /bin/sh or something like that for the version. And 
> >> >> >> >> > prints
> >> >> >> >> > that as the 'kernel' version. So if you did an incremental 
> >> >> >> >> > build and
> >> >> >> >> > /bin/sh was not changed the version stays the same.
> >> >> >> >> > To fix it you can do a clean buildworld/installworld.
> >> >> >> >> >
> >> >> >> >> > Regards,
> >> >> >> >> > Ronald.
> >> >> >> >> 
> >> >> >> >> That might well be it - I've done some buildworld:s with 
> >> >> >> >> -DNO_CLEAN
> >> >> >> >> lately.  I'll test - thanks for the tip!
> >> >> >> >
> >> >> >> > Can you provide the uname -U information (building with 
> >> >> >> > -DNO_CLEAN) should be
> >> >> >> > perfectly fine.
> >> >> >> 
> >> >> >> Thanks for looking into this!
> >> >> >> 
> >> >> >> $ uname -U
> >> >> >> 1103500
> >> >> >> 
> >> >> >> Let me know if there is something else I can check before I do a full
> >> >> >> buildworld on this system!
> >> >> >
> >> >> > What are the output of pkg config osversion
> >> >> > and pkg -o ABI_FILE=/usr/bin/uname config osversion
> >> >> > please
> >> >> 
> >> >> $ pkg config osversion
> >> >> 1102509
> >> >> $ pkg -o ABI_FILE=/usr/bin/uname config osversion
> >> >> 1102509
> >> >
> >> > It means you /usr/bin/uname binary has been built with a compiler that 
> >> > says it
> >> > is building for 1102509.
> >> >
> >> > Can you show the output of file /usr/bin/uname ?
> >> 
> >> $ uname -aKU
> >> FreeBSD P142.sics.se 11.3-STABLE FreeBSD 11.3-STABLE #9 r353280: Mon
> >> Oct 7 22:16:02 CEST 2019
> >> r...@p142.sics.se:/storage/obj/usr/src/sys/X250 amd64 1103500
> >> 1103500
> >> 
> >> no signs of 1102509 there...
> >> 
> >> and "strings -a /usr/bin/uname |grep 110" returns nothing.
> >> 
> > I am interested in the output of file /usr/bin/uname
> 
> Ah, right, didn't get it that you meant the file command, sorry...
> 
> $ file /usr/bin/uname 
> /usr/bin/uname: ELF 64-bit LSB executable, x86-64, version 1 (FreeBSD), 
> dynamically linked, interpreter /libexec/ld-elf.so.1, for FreeBSD 11.2 
> (1102509), FreeBSD-style, stripped
> 

That is the problem.

So if you don't want to rebuild you can add to your make.conf ABI_FILE= XXX
where XXX is the path to any file where when you run file on it says it was
built for freebsd 11.3

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: pkg thinks kernel is old

[Baptiste Daroussin]

On Fri, Oct 11, 2019 at 01:57:06PM +0200, Bengt Ahlgren wrote:
> Baptiste Daroussin  writes:
> 
> > On Fri, Oct 11, 2019 at 11:54:15AM +0200, Bengt Ahlgren wrote:
> >> Baptiste Daroussin  writes:
> >> 
> >> > On Fri, Oct 11, 2019 at 11:38:46AM +0200, Bengt Ahlgren wrote:
> >> >> Baptiste Daroussin  writes:
> >> >> 
> >> >> > On Thu, Oct 10, 2019 at 03:44:03PM +0200, Bengt Ahlgren wrote:
> >> >> >> Ronald Klop  writes:
> >> >> >> 
> >> >> >> > Van: Bengt Ahlgren 
> >> >> >> > Datum: woensdag, 9 oktober 2019 23:38
> >> >> >> > Aan: sta...@freebsd.org
> >> >> >> > Onderwerp: pkg thinks kernel is old
> >> >> >> >>
> >> >> >> >> I've ran into this on my 11.3-STABLE system:
> >> >> >> >>
> >> >> >> >> # uname -K
> >> >> >> >> 1103500
> >> >> >> >> # pkg update -f
> >> >> >> >> Updating ivyp repository catalogue...
> >> >> >> >> Fetching meta.txz: 100%560 B   0.6kB/s00:01Fetching
> >> >> >> >> packagesite.txz: 100%  410 KiB 419.9kB/s00:01Processing
> >> >> >> >> entries:   0%
> >> >> >> >> Newer FreeBSD version for package zziplib:
> >> >> >> >> To ignore this error set IGNORE_OSVERSION=yes
> >> >> >> >> - package: 1103000
> >> >> >> >> - running kernel: 1102509
> >> >> >> >> Ignore the mismatch and continue? [Y/n]:
> >> >> >> >>
> >> >> >> >> I build my own packages with poudriere, and have just switched 
> >> >> >> >> from an
> >> >> >> >> 11.2-REL to an 11.3-REL jail.  I didn't force-upgrade everything 
> >> >> >> >> after
> >> >> >> >> the switch, perhaps that's needed to make pkg recognise the 
> >> >> >> >> correct
> >> >> >> >> kernel version?
> >> >> >> >>
> >> >> >> >> Bengt
> >> >> >> 
> >> >> >> > This message of pkg is misleading.
> >> >> >> >
> >> >> >> > It looks at /bin/sh or something like that for the version. And 
> >> >> >> > prints
> >> >> >> > that as the 'kernel' version. So if you did an incremental build 
> >> >> >> > and
> >> >> >> > /bin/sh was not changed the version stays the same.
> >> >> >> > To fix it you can do a clean buildworld/installworld.
> >> >> >> >
> >> >> >> > Regards,
> >> >> >> > Ronald.
> >> >> >> 
> >> >> >> That might well be it - I've done some buildworld:s with -DNO_CLEAN
> >> >> >> lately.  I'll test - thanks for the tip!
> >> >> >
> >> >> > Can you provide the uname -U information (building with -DNO_CLEAN) 
> >> >> > should be
> >> >> > perfectly fine.
> >> >> 
> >> >> Thanks for looking into this!
> >> >> 
> >> >> $ uname -U
> >> >> 1103500
> >> >> 
> >> >> Let me know if there is something else I can check before I do a full
> >> >> buildworld on this system!
> >> >
> >> > What are the output of pkg config osversion
> >> > and pkg -o ABI_FILE=/usr/bin/uname config osversion
> >> > please
> >> 
> >> $ pkg config osversion
> >> 1102509
> >> $ pkg -o ABI_FILE=/usr/bin/uname config osversion
> >> 1102509
> >
> > It means you /usr/bin/uname binary has been built with a compiler that says 
> > it
> > is building for 1102509.
> >
> > Can you show the output of file /usr/bin/uname ?
> 
> $ uname -aKU
> FreeBSD P142.sics.se 11.3-STABLE FreeBSD 11.3-STABLE #9 r353280: Mon Oct  7 
> 22:16:02 CEST 2019 r...@p142.sics.se:/storage/obj/usr/src/sys/X250  amd64 
> 1103500 1103500
> 
> no signs of 1102509 there...
> 
> and "strings -a /usr/bin/uname |grep 110" returns nothing.
> 
I am interested in the output of file /usr/bin/uname

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: pkg thinks kernel is old

[Baptiste Daroussin]

On Fri, Oct 11, 2019 at 11:54:15AM +0200, Bengt Ahlgren wrote:
> Baptiste Daroussin  writes:
> 
> > On Fri, Oct 11, 2019 at 11:38:46AM +0200, Bengt Ahlgren wrote:
> >> Baptiste Daroussin  writes:
> >> 
> >> > On Thu, Oct 10, 2019 at 03:44:03PM +0200, Bengt Ahlgren wrote:
> >> >> Ronald Klop  writes:
> >> >> 
> >> >> > Van: Bengt Ahlgren 
> >> >> > Datum: woensdag, 9 oktober 2019 23:38
> >> >> > Aan: sta...@freebsd.org
> >> >> > Onderwerp: pkg thinks kernel is old
> >> >> >>
> >> >> >> I've ran into this on my 11.3-STABLE system:
> >> >> >>
> >> >> >> # uname -K
> >> >> >> 1103500
> >> >> >> # pkg update -f
> >> >> >> Updating ivyp repository catalogue...
> >> >> >> Fetching meta.txz: 100%560 B   0.6kB/s00:01Fetching
> >> >> >> packagesite.txz: 100%  410 KiB 419.9kB/s00:01Processing
> >> >> >> entries:   0%
> >> >> >> Newer FreeBSD version for package zziplib:
> >> >> >> To ignore this error set IGNORE_OSVERSION=yes
> >> >> >> - package: 1103000
> >> >> >> - running kernel: 1102509
> >> >> >> Ignore the mismatch and continue? [Y/n]:
> >> >> >>
> >> >> >> I build my own packages with poudriere, and have just switched from 
> >> >> >> an
> >> >> >> 11.2-REL to an 11.3-REL jail.  I didn't force-upgrade everything 
> >> >> >> after
> >> >> >> the switch, perhaps that's needed to make pkg recognise the correct
> >> >> >> kernel version?
> >> >> >>
> >> >> >> Bengt
> >> >> 
> >> >> > This message of pkg is misleading.
> >> >> >
> >> >> > It looks at /bin/sh or something like that for the version. And prints
> >> >> > that as the 'kernel' version. So if you did an incremental build and
> >> >> > /bin/sh was not changed the version stays the same.
> >> >> > To fix it you can do a clean buildworld/installworld.
> >> >> >
> >> >> > Regards,
> >> >> > Ronald.
> >> >> 
> >> >> That might well be it - I've done some buildworld:s with -DNO_CLEAN
> >> >> lately.  I'll test - thanks for the tip!
> >> >
> >> > Can you provide the uname -U information (building with -DNO_CLEAN) 
> >> > should be
> >> > perfectly fine.
> >> 
> >> Thanks for looking into this!
> >> 
> >> $ uname -U
> >> 1103500
> >> 
> >> Let me know if there is something else I can check before I do a full
> >> buildworld on this system!
> >
> > What are the output of pkg config osversion
> > and pkg -o ABI_FILE=/usr/bin/uname config osversion
> > please
> 
> $ pkg config osversion
> 1102509
> $ pkg -o ABI_FILE=/usr/bin/uname config osversion
> 1102509

It means you /usr/bin/uname binary has been built with a compiler that says it
is building for 1102509.

Can you show the output of file /usr/bin/uname ?

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: pkg thinks kernel is old

[Baptiste Daroussin]

On Fri, Oct 11, 2019 at 11:38:46AM +0200, Bengt Ahlgren wrote:
> Baptiste Daroussin  writes:
> 
> > On Thu, Oct 10, 2019 at 03:44:03PM +0200, Bengt Ahlgren wrote:
> >> Ronald Klop  writes:
> >> 
> >> > Van: Bengt Ahlgren 
> >> > Datum: woensdag, 9 oktober 2019 23:38
> >> > Aan: sta...@freebsd.org
> >> > Onderwerp: pkg thinks kernel is old
> >> >>
> >> >> I've ran into this on my 11.3-STABLE system:
> >> >>
> >> >> # uname -K
> >> >> 1103500
> >> >> # pkg update -f
> >> >> Updating ivyp repository catalogue...
> >> >> Fetching meta.txz: 100%560 B   0.6kB/s00:01Fetching
> >> >> packagesite.txz: 100%  410 KiB 419.9kB/s00:01Processing
> >> >> entries:   0%
> >> >> Newer FreeBSD version for package zziplib:
> >> >> To ignore this error set IGNORE_OSVERSION=yes
> >> >> - package: 1103000
> >> >> - running kernel: 1102509
> >> >> Ignore the mismatch and continue? [Y/n]:
> >> >>
> >> >> I build my own packages with poudriere, and have just switched from an
> >> >> 11.2-REL to an 11.3-REL jail.  I didn't force-upgrade everything after
> >> >> the switch, perhaps that's needed to make pkg recognise the correct
> >> >> kernel version?
> >> >>
> >> >> Bengt
> >> 
> >> > This message of pkg is misleading.
> >> >
> >> > It looks at /bin/sh or something like that for the version. And prints
> >> > that as the 'kernel' version. So if you did an incremental build and
> >> > /bin/sh was not changed the version stays the same.
> >> > To fix it you can do a clean buildworld/installworld.
> >> >
> >> > Regards,
> >> > Ronald.
> >> 
> >> That might well be it - I've done some buildworld:s with -DNO_CLEAN
> >> lately.  I'll test - thanks for the tip!
> >
> > Can you provide the uname -U information (building with -DNO_CLEAN) should 
> > be
> > perfectly fine.
> 
> Thanks for looking into this!
> 
> $ uname -U
> 1103500
> 
> Let me know if there is something else I can check before I do a full
> buildworld on this system!

What are the output of pkg config osversion
and pkg -o ABI_FILE=/usr/bin/uname config osversion
please

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: pkg thinks kernel is old

[Baptiste Daroussin]

On Thu, Oct 10, 2019 at 03:44:03PM +0200, Bengt Ahlgren wrote:
> Ronald Klop  writes:
> 
> > Van: Bengt Ahlgren 
> > Datum: woensdag, 9 oktober 2019 23:38
> > Aan: sta...@freebsd.org
> > Onderwerp: pkg thinks kernel is old
> >>
> >> I've ran into this on my 11.3-STABLE system:
> >>
> >> # uname -K
> >> 1103500
> >> # pkg update -f
> >> Updating ivyp repository catalogue...
> >> Fetching meta.txz: 100%560 B   0.6kB/s00:01Fetching
> >> packagesite.txz: 100%  410 KiB 419.9kB/s00:01Processing
> >> entries:   0%
> >> Newer FreeBSD version for package zziplib:
> >> To ignore this error set IGNORE_OSVERSION=yes
> >> - package: 1103000
> >> - running kernel: 1102509
> >> Ignore the mismatch and continue? [Y/n]:
> >>
> >> I build my own packages with poudriere, and have just switched from an
> >> 11.2-REL to an 11.3-REL jail.  I didn't force-upgrade everything after
> >> the switch, perhaps that's needed to make pkg recognise the correct
> >> kernel version?
> >>
> >> Bengt
> 
> > This message of pkg is misleading.
> >
> > It looks at /bin/sh or something like that for the version. And prints
> > that as the 'kernel' version. So if you did an incremental build and
> > /bin/sh was not changed the version stays the same.
> > To fix it you can do a clean buildworld/installworld.
> >
> > Regards,
> > Ronald.
> 
> That might well be it - I've done some buildworld:s with -DNO_CLEAN
> lately.  I'll test - thanks for the tip!

Can you provide the uname -U information (building with -DNO_CLEAN) should be
perfectly fine.

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: pkg thinks kernel is old

[Baptiste Daroussin]

On Thu, Oct 10, 2019 at 01:25:28PM +0200, Ronald Klop wrote:
> This message of pkg is misleading.
> 
> It looks at /bin/sh or something like that for the version. And prints that 
> as the 'kernel' version. So if you did an incremental build and /bin/sh was 
> not changed the version stays the same.
> To fix it you can do a clean buildworld/installworld.
> 
> Regards,
> Ronald.

I do agree the message is misleading, It should say "running world".

It does not look anymore at /bin/sh because of the incremental build problem, it
looks at uname which is always updated for uname -U during incremental builds.
(and falls back on /bin/sh if uname does not exist).

so it says the userland of the running system is not inlined with the version
the packages where built for.

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: pkg issue after FreeBSD 11 upgrade

[Baptiste Daroussin]

On Wed, Aug 30, 2017 at 10:07:51AM -0300, Cassiano Peixoto wrote:
> Ok I know about HANDLE_RC_SCRIPTS, it's a good approach. But how to deal
> with when I need to restart a service without upgrading? Reaper
> functionnality is a trouble for many administrators who made meta ports to
> manage their servers. I really think it could be a option to be
> enabled/disabled. Can you see this possibility?
> 

Yes I could add an option to disable the reaper functionnality (and will
probably to unblock such use case as soon as I have time to do it.)

However I still think this is not the right idea :) and a better one could be
found

Bapt


signature.asc
Description: PGP signature

Re: pkg issue after FreeBSD 11 upgrade

[Baptiste Daroussin]

On Wed, Aug 30, 2017 at 03:01:30PM +0200, Franco Fichtner wrote:
> Hi Cassiano,
> 
> > On 30. Aug 2017, at 2:55 PM, Cassiano Peixoto  
> > wrote:
> > 
> > Why it used to work on FreeBSD 10? It stopped worked on FreeBSD 11 only.
> 
> It was a later 10.x change as far as I know.
> 
> > Is there some flag to disable it? Or some hack that I could do?
> 
> No.  At OPNsense we use a patch to revert the behaviour.
> 
> https://github.com/opnsense/ports/blob/master/ports-mgmt/pkg/files/patch-libpkg_scripts.c

Why and what is your use case, there is a reason why this code has been added,
and I would like to hear more about use cases so we can try to address them.

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: pkg issue after FreeBSD 11 upgrade

[Baptiste Daroussin]

On Wed, Aug 30, 2017 at 09:55:22AM -0300, Cassiano Peixoto wrote:
> Hi Baptiste,
> 
> Why it used to work on FreeBSD 10? It stopped worked on FreeBSD 11 only.

It only worked on FreeBSD 10 prior to 10.2, the reaper functionnality in freebsd
kernel appeared in 10.2
> 
> Cron is just an example, I manage more than 50 FreeBSD servers, and I've
> been using ports for years to update some configs and restart the service
> on all of them. Many times I need to change nginx config, ldap, etc. I just
> need to restart the service.

HANDLE_RC_SCRIPTS=true in your pkg.conf and pkg will automatically restart
anything rc script provide once the package containing it is upgrading.

This is off by default because in many cases it is dangerous (database upgrades,
dovecot like things upgrade etc). But if you know what you are doing it does the
job.

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: pkg issue after FreeBSD 11 upgrade

[Baptiste Daroussin]

On Wed, Aug 30, 2017 at 09:00:55AM -0300, Cassiano Peixoto wrote:
> Hi Matthew,
> 
> Sorry back to this subject. But I really need to restart services with a
> port. I'm quite sure there is a bug with pkg and FreeBSD 11.
> 
> I made a simple port to restart cron service:

It is not a bug, it is by design, pkg becomes the reaper of the scripts it runs
and kills everything once the script is executed.

btw if you install crontab in cron.d you do not need to restart the service,
cron will figure out itself and reload what it needed.

Bapt


signature.asc
Description: PGP signature

Re: 11.1-PRERELEASE, waagent crashes attaching reosurce disc in Azure

[Baptiste Daroussin]

On Fri, May 26, 2017 at 01:26:46PM +0100, Pete French wrote:
> Have upgraded one of my Azure boxes to 11.1-PRERELEASE yesterday, and I
> now find the below backtrace from waagent when it tries to attach
> the resource disc. It looks as if it is finding 'pass1' as the device,
> when it should be finding 'da1'. As the waagents hasnt changed I am
> wondering what might have chnaged in FreeBSD to make this stop working.
> 
> I consider this to be an Azure bug not a FreeBSD bug, and have
> filed an issue, but would be nice to get to the bottom of it myself.
> 
> cheers,
> 
> -pete.
> 

The agent has just been updated in the ports tree a few days ago, can you try
the new version?

Bapt


signature.asc
Description: PGP signature

Re: mandoc 1.14 on FreeBSD 11.1

[Baptiste Daroussin]

On Mon, Mar 20, 2017 at 03:59:35PM -0600, Warner Losh wrote:
> 
> > On Mar 20, 2017, at 3:53 PM, Baptiste Daroussin <b...@freebsd.org> wrote:
> > 
> > Hi all,
> > 
> > mandoc 1.14 has introduced a new database format (removing the requirement 
> > for
> > sqlite). FreeBSD 11.x is having right mandoc 1.13, Given the database 
> > format is
> > a technical detail (aka new format will be generated anyway automatically 
> > and
> > seamlessly replacing the old database). I think it does not fall into POLA. 
> > For
> > such reason I'm planing to upgrade to the 1.14 series (current already have 
> > it).
> > 
> > If that is a matter for you to change that database format please tell me.
> > Otherwise in 2 weeks I will import 1.14 in 11-STABLE in order we have it in
> > 11.1-R
> 
> I tend to agree with our analysis, but have a couple of questions that will 
> help us know what’s the right thing to do.
> 
> First, this is 100% independent of the kernel, so we don’t have any ‘forward 
> / back’ issues with that, correct?

Totally :), there might be an issue with the framework as it is not picking
makewhatis from the cross tools as it should I may need to fix that
> 
> Next, If I reinstall an older version of FreeBSD, it will create the database 
> that’s appropriate for 1.13, and vice versa, so it shouldn’t be visible to 
> the user.

Yes there is a periodic script for that and the database is anyway rebuild when
installing etc files
> 
> What happens if I upgrade and then add/delete packages, including upgrade, 
> add packages, downgraded?

Exactly as now, makewhatis (periodic script) will regenerate the databases at
the proper format
> 
> Is there a dependence on which pkg version you are using?

Nope
> 
> If the answers to these questions are approximately “yes”, “shouldn’t matter 
> / does the right thing”, “won’t be an issue because recreation sets the right 
> format and installworld will update the database with the right things as 
> part of that” and “doesn’t matter”  then I’d agree that this is an 
> uninteresting detail and need not be preserved.

Yup that is the answer to it :)
> 
> Warner




signature.asc
Description: PGP signature

mandoc 1.14 on FreeBSD 11.1

[Baptiste Daroussin]

Hi all,

mandoc 1.14 has introduced a new database format (removing the requirement for
sqlite). FreeBSD 11.x is having right mandoc 1.13, Given the database format is
a technical detail (aka new format will be generated anyway automatically and
seamlessly replacing the old database). I think it does not fall into POLA. For
such reason I'm planing to upgrade to the 1.14 series (current already have it).

If that is a matter for you to change that database format please tell me.
Otherwise in 2 weeks I will import 1.14 in 11-STABLE in order we have it in
11.1-R

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: Uppercase RE matching problems in FreeBSD 11

[Baptiste Daroussin]

On Sun, Nov 06, 2016 at 10:20:54PM +0100, Stefan Bethke wrote:
> 
> > Am 06.11.2016 um 22:06 schrieb Baptiste Daroussin <b...@freebsd.org>:
> > 
> > On Sun, Nov 06, 2016 at 09:57:00PM +0100, Stefan Bethke wrote:
> >> 
> >>> Am 06.11.2016 um 12:07 schrieb Baptiste Daroussin <b...@freebsd.org>:
> >>> 
> >>> On Sat, Nov 05, 2016 at 08:23:25PM -0500, Greg Rivers wrote:
> >>>> I happened to run an old script today that uses sed(1) to extract the 
> >>>> system
> >>>> boot time from the kern.boottime sysctl MIB. On 11.0 this no longer 
> >>>> works as
> >>>> expected:
> >>>> 
> >>>> $ sysctl kern.boottime
> >>>> kern.boottime: { sec = 1478380714, usec = 145351 } Sat Nov  5 16:18:34 
> >>>> 2016
> >>>> $ sysctl kern.boottime | sed -e 's/.*\([A-Z].*\)$/\1/'
> >>>> v  5 16:18:34 2016
> >>>> 
> >>>> sed passes over 'S' and 'N' until it hits 'v', which it considers 
> >>>> uppercase
> >>>> apparently. This is with LANG=en_US.UTF-8. If I set LANG=C, it works as
> >>>> expected:
> >>>> 
> >>>> $ sysctl kern.boottime | LANG=C sed -e 's/.*\([A-Z].*\)$/\1/'
> >>>> Nov  5 16:18:34 2016
> >>>> 
> >>>> Testing every lowercase character separately gives even more inconsistent
> >>>> results:
> >>>> 
> >>>> $ cat < >> 
> >>>> Here sed thinks every lowercase character except for 'a' is uppercase! 
> >>>> This
> >>>> differs from the first test where sed did not think 'o' is uppercase. 
> >>>> Again,
> >>>> the above behaves as expected with LANG=C.
> >>>> 
> >>>> Does anyone have any insight into this? This is likely to break a lot of
> >>>> existing code.
> >>>> 
> >>> 
> >>> Yes A-Z only means uppercase in an ASCII only world in a unicode world it 
> >>> means
> >>> AaBb... Z because there are way more characters that simple A-Z. In 
> >>> FreeBSD 11
> >>> we have a unicode collation instead of falling back in on LC_COLLATE=C 
> >>> which
> >>> means ascii only
> >>> 
> >>> For regrexp for example one should use the classes: :upper: or :lower:.
> >> 
> >> That is rather surprising.  Is there a normative reference for the 
> >> treatment of bracket expressions and character classes when using locales 
> >> other than C and/or encodings like UTF-8?
> > 
> > http://pubs.opengroup.org/onlinepubs/009695399/basedefs/xbd_chap09.html
> > 
> > For example:
> > 
> > "Regular expressions are a context-independent syntax that can represent a 
> > wide
> > variety of character sets and character set orderings, where these character
> > sets are interpreted according to the current locale. While many regular
> > expressions can be interpreted differently depending on the current locale, 
> > many
> > features, such as character class expressions, provide for contextual 
> > invariance
> > across locales.“
> 
> Sorry, maybe I wasn’t clear enough with my question.  When a character class 
> fits the problem, it is clearly advantageous.
> 
> But under what circumstances would [A-Z] mean anything other than a character 
> whose Unicode codepoint is between U+0041 and U+005A, inclusive?  Especially 
> given the locale in the example is en_US.UTF-8.  Or, put another way, why 
> would an implementation interpret [A-Z] as anything other than [ABCDE…XYZ]?

The collation rules for unicode comes from: http://cldr.unicode.org/ and they do
match the one on linux for example and the one on illumos.

On some gnu tool they explicitly decide to be non locale aware to avoid that
kind of "surprises"
> 
> From reading your reference, I can see in 9.3.5.7:
> > In the POSIX locale, a range expression represents the set of collating 
> > elements that fall between two elements in the collation sequence, 
> > inclusive. In other locales, a range expression has unspecified behavior[…]
> 
> So even if the observed behaviour is conforming, I’d think it’s still highly 
> undesirable.
> 
That works for POSIX locale aka C aka ASCII only world

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: Uppercase RE matching problems in FreeBSD 11

[Baptiste Daroussin]

On Sun, Nov 06, 2016 at 09:57:00PM +0100, Stefan Bethke wrote:
> 
> > Am 06.11.2016 um 12:07 schrieb Baptiste Daroussin <b...@freebsd.org>:
> > 
> > On Sat, Nov 05, 2016 at 08:23:25PM -0500, Greg Rivers wrote:
> >> I happened to run an old script today that uses sed(1) to extract the 
> >> system
> >> boot time from the kern.boottime sysctl MIB. On 11.0 this no longer works 
> >> as
> >> expected:
> >> 
> >> $ sysctl kern.boottime
> >> kern.boottime: { sec = 1478380714, usec = 145351 } Sat Nov  5 16:18:34 2016
> >> $ sysctl kern.boottime | sed -e 's/.*\([A-Z].*\)$/\1/'
> >> v  5 16:18:34 2016
> >> 
> >> sed passes over 'S' and 'N' until it hits 'v', which it considers uppercase
> >> apparently. This is with LANG=en_US.UTF-8. If I set LANG=C, it works as
> >> expected:
> >> 
> >> $ sysctl kern.boottime | LANG=C sed -e 's/.*\([A-Z].*\)$/\1/'
> >> Nov  5 16:18:34 2016
> >> 
> >> Testing every lowercase character separately gives even more inconsistent
> >> results:
> >> 
> >> $ cat < 
> >> Here sed thinks every lowercase character except for 'a' is uppercase! This
> >> differs from the first test where sed did not think 'o' is uppercase. 
> >> Again,
> >> the above behaves as expected with LANG=C.
> >> 
> >> Does anyone have any insight into this? This is likely to break a lot of
> >> existing code.
> >> 
> > 
> > Yes A-Z only means uppercase in an ASCII only world in a unicode world it 
> > means
> > AaBb... Z because there are way more characters that simple A-Z. In FreeBSD 
> > 11
> > we have a unicode collation instead of falling back in on LC_COLLATE=C which
> > means ascii only
> > 
> > For regrexp for example one should use the classes: :upper: or :lower:.
> 
> That is rather surprising.  Is there a normative reference for the treatment 
> of bracket expressions and character classes when using locales other than C 
> and/or encodings like UTF-8?

http://pubs.opengroup.org/onlinepubs/009695399/basedefs/xbd_chap09.html

For example:

"Regular expressions are a context-independent syntax that can represent a wide
variety of character sets and character set orderings, where these character
sets are interpreted according to the current locale. While many regular
expressions can be interpreted differently depending on the current locale, many
features, such as character class expressions, provide for contextual invariance
across locales."

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: Uppercase RE matching problems in FreeBSD 11

[Baptiste Daroussin]

On Sun, Nov 06, 2016 at 01:26:51PM +0100, Mark Martinec wrote:
> 2016-11-06 12:07, Baptiste Daroussin wrote:
> > Yes A-Z only means uppercase in an ASCII only world in a unicode world
> > it means
> > AaBb... Z because there are way more characters that simple A-Z. In
> > FreeBSD 11
> > we have a unicode collation instead of falling back in on LC_COLLATE=C
> > which
> > means ascii only
> > 
> > For regrexp for example one should use the classes: :upper: or :lower:.
> 
> It is a good idea to keep LC_COLLATE and LC_NUMERIC (and LC_MONETARY?) at
> "C"
> when LANG or LC_CTYPE is set to something else, otherwise unexpected
> things may happen.
> 

In scripts clearly, the collation rules, numeric rules and monetary rules may
vary depending on the locale.

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: Uppercase RE matching problems in FreeBSD 11

[Baptiste Daroussin]

On Sat, Nov 05, 2016 at 08:23:25PM -0500, Greg Rivers wrote:
> I happened to run an old script today that uses sed(1) to extract the system
> boot time from the kern.boottime sysctl MIB. On 11.0 this no longer works as
> expected:
> 
> $ sysctl kern.boottime
> kern.boottime: { sec = 1478380714, usec = 145351 } Sat Nov  5 16:18:34 2016
> $ sysctl kern.boottime | sed -e 's/.*\([A-Z].*\)$/\1/'
> v  5 16:18:34 2016
> 
> sed passes over 'S' and 'N' until it hits 'v', which it considers uppercase
> apparently. This is with LANG=en_US.UTF-8. If I set LANG=C, it works as
> expected:
> 
> $ sysctl kern.boottime | LANG=C sed -e 's/.*\([A-Z].*\)$/\1/'
> Nov  5 16:18:34 2016
> 
> Testing every lowercase character separately gives even more inconsistent
> results:
> 
> $ cat < > a
> > b
> > c
> > d
> > e
> > f
> > g
> > h
> > i
> > j
> > k
> > l
> > m
> > n
> > o
> > p
> > q
> > r
> > s
> > t
> > u
> > v
> > w
> > x
> > y
> > z
> > !
> b
> c
> d
> e
> f
> g
> h
> i
> j
> k
> l
> m
> n
> o
> p
> q
> r
> s
> t
> u
> v
> w
> x
> y
> z
> 
> Here sed thinks every lowercase character except for 'a' is uppercase! This
> differs from the first test where sed did not think 'o' is uppercase. Again,
> the above behaves as expected with LANG=C.
> 
> Does anyone have any insight into this? This is likely to break a lot of
> existing code.
> 

Yes A-Z only means uppercase in an ASCII only world in a unicode world it means
AaBb... Z because there are way more characters that simple A-Z. In FreeBSD 11
we have a unicode collation instead of falling back in on LC_COLLATE=C which
means ascii only

For regrexp for example one should use the classes: :upper: or :lower:.

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: Dying jail

[Baptiste Daroussin]

On Wed, Oct 26, 2016 at 03:09:31PM +0700, Eugene Grosbein wrote:
> Hi!
> 
> Recently I've upgraded one of my server running 9.3-STABLE with jail 
> containing 4.11-STABLE system.
> The host was source-upgraded upto 10.3-STABLE first and next to 11.0-STABLE
> and jail configuration migrated to /etc/jail.conf. The jail kept intact.
> 
> "service jail start" started the jail successfully
> but "service jail restart" fails due to jail being stuck in "dying" state for 
> long time:
> "jls" shows no running jails and "jls -d" shows the dying jail.
> 
> How do I know why is it stuck and how to forcebly kill it without reboot of 
> the host?
> 

I have the same problem on a FreeBSD 11.0

I have no specific jail.conf

for exec.start is directly calls service cassandra onestart
and stop calls the stop of the very same service.

jail -f /myconf -r nameofthejail

I can see the jail staying in dying mode for multiple minutes
even after sockstat -j has been showing no TCP is left at all.

No processes are left in the jail

I'm mostly clueless on how to debug that and know that the dying jail is waiting
on.

It is painful as it prevents from exporting the zfs pool the jail was sitting
on.

Any one has ideas?

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: stable/11 Problems with Unicode character and ls(1) sorting

[Baptiste Daroussin]

On Fri, Jul 15, 2016 at 11:38:22AM -0400, Jung-uk Kim wrote:
> On 07/15/16 09:55 AM, Thomas Eberhardt wrote:
> > % uname -a
> > FreeBSD clarence.ocp.lan 11.0-BETA1 FreeBSD 11.0-BETA1 #0 r302457: Sat Jul  
> > 9 10:41:40 CEST 2016 
> > tho...@clarence.ocp.lan:/usr/obj/usr/src/sys/GENERIC  amd64
> > 
> > % locale
> > LANG=en_US.UTF-8
> > LC_CTYPE="en_US.UTF-8"
> > LC_COLLATE="en_US.UTF-8"
> > LC_TIME="en_US.UTF-8"
> > LC_NUMERIC="en_US.UTF-8"
> > LC_MONETARY="en_US.UTF-8"
> > LC_MESSAGES="en_US.UTF-8"
> > LC_ALL=
> > 
> > % mkdir x
> > % touch x/Æon1
> > % ls x
> > Æon1
> > % touch x/Æon2
> > % ls -f x 
> > . ..Æon1  Æon2
> > % ls -f x | hd
> >   2e 0a 2e 2e 0a c3 86 6f  6e 31 0a c3 86 6f 6e 32  
> > |...on1...on2|
> > 0010  0a|.|
> > 0011
> > % ls x
> > [hangs and consumes 100% CPU]
> > ^C
> > 
> > % gdb /bin/ls
> > […]
> > (gdb) run x
> > Starting program: /bin/ls x
> > [hangs]
> > ^C
> > Program received signal SIGINT, Interrupt.
> > 0x000800ffa9dd in _collate_lookup (table=, 
> > t=, 
> > len=, pri=, which= > optimized out>, state=)
> > at /usr/src/lib/libc/locale/collate.c:340
> > 340 *pri = table->char_pri_table[*t].pri[which];
> > Current language:  auto; currently minimal
> > (gdb) bt
> > #0  0x000800ffa9dd in _collate_lookup (table=, 
> > t=, 
> > len=, pri=, which= > optimized out>, state=)
> > at /usr/src/lib/libc/locale/collate.c:340
> > #1  0x000800fdc38a in wcscoll_l (ws1=, ws2= > optimized out>, 
> > locale=) at /usr/src/lib/libc/string/wcscoll.c:158
> > #2  0x000800fd9071 in strcoll_l (s=, s2= > optimized out>, locale=0x80124f338)
> > at /usr/src/lib/libc/string/strcoll.c:101
> > #3  0x000800fee253 in qsort (a=, n= > optimized out>, es=, 
> > cmp=0x800f28530 ) at /usr/src/lib/libc/stdlib/qsort.c:130
> > #4  0x000800f27397 in fts_sort (sp=, head= > optimized out>, nitems=)
> > at /usr/src/lib/libc/gen/fts.c:995
> > #5  0x000800f2848e in fts_children (sp=, 
> > instr=Error accessing memory address 0x2: Bad address.
> > ) at /usr/src/lib/libc/gen/fts.c:570
> > #6  0x004030df in traverse (argc=, argv= > optimized out>, 
> > options=) at /usr/src/bin/ls/ls.c:576
> > #7  0x00402eeb in main (argc=, argv= > optimized out>) at /usr/src/bin/ls/ls.c:498
> > #8  0x004020cf in _start ()
> > #9  0x000800629000 in ?? ()
> > #10 0x in ?? ()
> > (gdb) br strcoll.c:101
> > Breakpoint 1 at 0x800fd9063: file /usr/src/lib/libc/string/strcoll.c, line 
> > 101.
> > (gdb) run
> > The program being debugged has been started already.
> > Start it from the beginning? (y or n) y
> > Starting program: /bin/ls x
> > […]
> > 
> > Breakpoint 1, strcoll_l (s=, s2=, 
> > locale=0x80124f338)
> > at /usr/src/lib/libc/string/strcoll.c:101
> > 101 ret = wcscoll_l(w1, w2, locale);
> > (gdb) n
> > [wcscoll_l never returns]
> > 
> > 
> > Does anybody know what’s going on? This is on a ZFS filesystem if that 
> > matters.
> 
> Yes, it is a known problem and bapt is working on it.  If you're in
> hurry, reverting r302324 will fix the regression for now.
> 
> https://svnweb.freebsd.org/changeset/base/302324
> 
> FYI, this problem is tracked as PR211135.
> 
> https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=211135
> 
> Jung-uk Kim
> 


Fixed by: r302916.

I'll merge it in 2 days in stable/11 if re@ agrees

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: FreeBSD_STABLE_10-i386 - Build #661 - Still Failing

[Baptiste Daroussin]

On Tue, Nov 24, 2015 at 01:06:50AM +0800, Li-Wen Hsu wrote:
> On Mon, Nov 23, 2015 at 17:20:36 +0100, Baptiste Daroussin wrote:
> > On Mon, Nov 23, 2015 at 04:10:33PM +, jenkins-ad...@freebsd.org wrote:
> > > FreeBSD_STABLE_10-i386 - Build #661 - Still Failing:
> > > 
> > > Build information: 
> > > https://jenkins.FreeBSD.org/job/FreeBSD_STABLE_10-i386/661/
> > > Full change log: 
> > > https://jenkins.FreeBSD.org/job/FreeBSD_STABLE_10-i386/661/changes
> > > Full build log: 
> > > https://jenkins.FreeBSD.org/job/FreeBSD_STABLE_10-i386/661/console
> > 
> > Someone needs to update the jenkins to the latest head.
> 
> The build happens in a jail built every time the job starts, by fetching
> the latest snapshot from:
> 
> http://ftp.freebsd.org/pub/FreeBSD/snapshots/i386/i386/11.0-CURRENT/base.txz
> 
> It looks there is no /usr/bin/colldef.  I haven't looked into what's
> going on here.
> 
Ok so it will be fixed by the next snapshot

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: FreeBSD_STABLE_10-i386 - Build #661 - Still Failing

[Baptiste Daroussin]

On Mon, Nov 23, 2015 at 04:10:33PM +, jenkins-ad...@freebsd.org wrote:
> FreeBSD_STABLE_10-i386 - Build #661 - Still Failing:
> 
> Build information: https://jenkins.FreeBSD.org/job/FreeBSD_STABLE_10-i386/661/
> Full change log: 
> https://jenkins.FreeBSD.org/job/FreeBSD_STABLE_10-i386/661/changes
> Full build log: 
> https://jenkins.FreeBSD.org/job/FreeBSD_STABLE_10-i386/661/console

Someone needs to update the jenkins to the latest head.

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: [CTF] pkg 1.6.0

[Baptiste Daroussin]

On Tue, Sep 22, 2015 at 06:27:49PM -0400, Adam McDougall wrote:
> On 09/22/2015 03:20, Ranjan1018 . wrote:
> > 2015-09-21 23:27 GMT+02:00 Baptiste Daroussin <b...@freebsd.org>:
> > 
> >> Hi all,
> >>
> >> We are about to release pkg 1.6.0. pkg-devel has been updated to 1.5.99.13
> >> aka
> >> 1.6.0 rc3 that we hope will become the new pkg 1.6.0 btw the end of the
> >> Week
> >> (release planned for Saturday 26th of September or no important issues are
> >> raised)
> 
> > running version 1.5.3 I have this error message:
> > # pkg upgrade
> > Updating FreeBSD repository catalogue...
> > FreeBSD repository is up-to-date.
> > All repositories are up-to-date.
> > Checking for upgrades (40 candidates): 100%
> > Processing candidates (40 candidates): 100%
> > Checking integrity... done (1 conflicting)
> > pkg: Cannot solve problem using SAT solver:
> > dependency rule: package Thunar(l) depends on:
> > xfce4-tumbler(r)xfce4-tumbler(l)
> > upgrade rule: upgrade local xfce4-tumbler-0.1.31_1 to remote
> > xfce4-tumbler-0.1.31_1
> > cannot install package xfce4-tumbler, remove it from request? [Y/n]:
> > pkg: cannot find xfce4-tumbler in the request
> > pkg: cannot solve job using SAT solver
> > Checking integrity... done (0 conflicting)
> > Your packages are up to date.
> > 
> > With this version I have been able to update the packages.
> 
> Same here, I ran into a conflict with Thunar on at least two computers
> with 1.5.6:
> 
> pkg: Cannot solve problem using SAT solver:
> dependency rule: package Thunar(l) depends on:
> xfce4-tumbler(r)xfce4-tumbler(l)
> upgrade rule: upgrade local xfce4-tumbler-0.1.31_1 to remote
> xfce4-tumbler-0.1.31_1
> cannot install package xfce4-tumbler, remove it from request? [Y/n]: ^C
> 
> I upgraded to 1.5.99.13 without any problems and it handles Thunar fine
> without any workarounds.  Just some extra warnings the first time when
> upgrading to 1.5.99 from my own repo:
> 
> # pkg upgrade
> Updating pkg-desktop repository catalogue...
> Fetching meta.txz: 100%260 B   0.3kB/s00:01
> Fetching packagesite.txz: 100%  217 KiB 222.0kB/s00:01
> Processing entries:   0%
> pkg: Skipping unknown key 'messages'
> Processing entries:   2%
> pkg: Skipping unknown key 'messages'
> Processing entries:   4%
> pkg: Skipping unknown key 'messages'
> Processing entries:   5%
> pkg: Skipping unknown key 'messages'
> Processing entries:   6%
> pkg: Skipping unknown key 'messages'
> pkg: Skipping unknown key 'messages'
> ...
> Processing entries: 100%
> pkg-desktop repository update completed. 914 packages processed.
> New version of pkg detected; it needs to be installed first.
> pkg: Skipping unknown key 'messages'
> Checking integrity... done (0 conflicting)
> The following 1 package(s) will be affected (of 0 checked):
> 
> Installed packages to be UPGRADED:
> pkg: 1.5.6 -> 1.5.99.13
> 

Yes those warnings are expected :(
They won't happen in the futur this is a deficiency in "forward" compatibility
design we fixed in pkg 1.6 dev process.

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: [CTF] pkg 1.6.0

[Baptiste Daroussin]

On Tue, Sep 22, 2015 at 09:20:28AM +0200, Ranjan1018 . wrote:
> 2015-09-21 23:27 GMT+02:00 Baptiste Daroussin <b...@freebsd.org>:
> 
> > Hi all,
> >
> > We are about to release pkg 1.6.0. pkg-devel has been updated to 1.5.99.13
> > aka
> > 1.6.0 rc3 that we hope will become the new pkg 1.6.0 btw the end of the
> > Week
> > (release planned for Saturday 26th of September or no important issues are
> > raised)
> >
> > On the list of changes:
> > - Lots of improvements in the solver (in particular fixes the case like the
> >   recent jpeg upgrade)
> > - Lots of fixes in the 3 way merge code
> > - pkg add can now work without a version specified in the dependency line
> > - pkg check -d now also check the required libraries
> > - Improved support for partial upgrades
> > - Improved zsh completion support
> > - Improved linux support (now all regression tests passes on linux)
> > - Messages can now be context aware: (only print a given message during
> >   installation, upgrade - version aware -, removal, or always)
> > - @keywords now accepts new entries to add context aware messages
> > - Add the ability to generate graphiz's dot format representation of the
> >   solver's problem
> > - pkg search now default on showing the comments of of the matched packages
> > - Lots of bug fixes and code cleanup
> > - Plenty of new bugs
> >
> > Please test heavily, I would like to make a release quite soon to limit the
> > number of users suffering from the jpeg->jpeg-turbo update.
> >
> > Best regards,
> > Bapt
> >
> 
> Thank you Batp,
> 
> running version 1.5.3 I have this error message:
> # pkg upgrade
> Updating FreeBSD repository catalogue...
> FreeBSD repository is up-to-date.
> All repositories are up-to-date.
> Checking for upgrades (40 candidates): 100%
> Processing candidates (40 candidates): 100%
> Checking integrity... done (1 conflicting)
> pkg: Cannot solve problem using SAT solver:
> dependency rule: package Thunar(l) depends on:
> xfce4-tumbler(r)xfce4-tumbler(l)
> upgrade rule: upgrade local xfce4-tumbler-0.1.31_1 to remote
> xfce4-tumbler-0.1.31_1
> cannot install package xfce4-tumbler, remove it from request? [Y/n]:
> pkg: cannot find xfce4-tumbler in the request
> pkg: cannot solve job using SAT solver
> Checking integrity... done (0 conflicting)
> Your packages are up to date.
> 
> With this version I have been able to update the packages.
> 
Thanks for the test!

Best regards,
Bapt


signature.asc
Description: PGP signature

[CTF] pkg 1.6.0

[Baptiste Daroussin]

Hi all,

We are about to release pkg 1.6.0. pkg-devel has been updated to 1.5.99.13 aka
1.6.0 rc3 that we hope will become the new pkg 1.6.0 btw the end of the Week
(release planned for Saturday 26th of September or no important issues are
raised)

On the list of changes:
- Lots of improvements in the solver (in particular fixes the case like the
  recent jpeg upgrade)
- Lots of fixes in the 3 way merge code
- pkg add can now work without a version specified in the dependency line
- pkg check -d now also check the required libraries
- Improved support for partial upgrades
- Improved zsh completion support
- Improved linux support (now all regression tests passes on linux)
- Messages can now be context aware: (only print a given message during
  installation, upgrade - version aware -, removal, or always)
- @keywords now accepts new entries to add context aware messages
- Add the ability to generate graphiz's dot format representation of the
  solver's problem
- pkg search now default on showing the comments of of the matched packages
- Lots of bug fixes and code cleanup
- Plenty of new bugs

Please test heavily, I would like to make a release quite soon to limit the
number of users suffering from the jpeg->jpeg-turbo update.

Best regards,
Bapt


signature.asc
Description: PGP signature

Re: 10.2-RELEASE-p2 lost ability to bootstrap pkg with signature_type="pubkey"

[Baptiste Daroussin]

On Wed, Sep 09, 2015 at 09:14:12AM +0200, Marko Cupać wrote:
> On Tue, 8 Sep 2015 23:28:59 +0200
> Baptiste Daroussin <b...@freebsd.org> wrote:
> 
> > On Tue, Sep 08, 2015 at 12:38:38PM +0200, Marko Cupać wrote:
> > > Hi,
> > > 
> > > I just found out that 10.2-RELEASE-p2 lost ability to bootstrap pkg
> > > with signature_type="pubkey".
> > > 
> > > Quick search returns:
> > > https://github.com/freebsd/pkg/issues/1309
> > > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=202622
> > > 
> > > I guess it is not hard to switch repo to fingerprints, however I
> > > would not expect to lose this functionality by updating to
> > > patchlevel.
> > > 
> > Implemented in head: r287579 I will MFC it asap. And see if it cannot
> > be added asap to a next patchlevel update.
> > 
> > Best regards,
> > Bapt
> 
> Thanx!
> 
> Just a few quick not-completely-related questions: poudriere has the
> ability to sign repos with PKG_REPO_SIGNING_KEY, but not with external
> command, right? Is there a plan to support it? Can I build packages in
> poudriere without PKG_REPO_SIGNING_KEY, and sign repo later on with
> external command?
> 

First yes I plan to add the ability to sign the package used to bootstrap via
PKG_REPO_SIGNING_KEY asap in poudriere.

Second you can keep your current configuration of poudriere, the signing with
pubkey works perfectly well. All you need to do is either via a poudriere post
bulk hook or manually go in the directory where your packages lives (in the
Latest directory) and
echo -n "$(sha256 -q pkg.txz)" | openssl dgst -sha256 -sign /thekey \
-binary -out ./pkg.txz.pubkeysig

Last if you want to do all the process manually:

pkg repo /yourrepository /yourkey
cd /yourrepositry/Latest
echo -n "$(sha256 -q pkg.txz)" | openssl dgst -sha256 -sign /yourkey \
-binary -out ./pkg.txz.pubkeysig

I will see if I can avoid the the extra command by merging the signing of the
bootstrap bit directly into pkg repo, that would be more handy

Best regards,
Bapt


pgp0u1kR9laO4.pgp
Description: PGP signature

Re: 10.2-RELEASE-p2 lost ability to bootstrap pkg with signature_type="pubkey"

[Baptiste Daroussin]

On Wed, Sep 09, 2015 at 09:21:24AM -0400, Shawn Webb wrote:
> On Wednesday, 09 September 2015 10:56:20 AM Baptiste Daroussin wrote:
> > On Wed, Sep 09, 2015 at 09:14:12AM +0200, Marko Cupać wrote:
> > > On Tue, 8 Sep 2015 23:28:59 +0200
> > > 
> > > Baptiste Daroussin <b...@freebsd.org> wrote:
> > > > On Tue, Sep 08, 2015 at 12:38:38PM +0200, Marko Cupać wrote:
> > > > > Hi,
> > > > > 
> > > > > I just found out that 10.2-RELEASE-p2 lost ability to bootstrap pkg
> > > > > with signature_type="pubkey".
> > > > > 
> > > > > Quick search returns:
> > > > > https://github.com/freebsd/pkg/issues/1309
> > > > > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=202622
> > > > > 
> > > > > I guess it is not hard to switch repo to fingerprints, however I
> > > > > would not expect to lose this functionality by updating to
> > > > > patchlevel.
> > > > 
> > > > Implemented in head: r287579 I will MFC it asap. And see if it cannot
> > > > be added asap to a next patchlevel update.
> > > > 
> > > > Best regards,
> > > > Bapt
> > > 
> > > Thanx!
> > > 
> > > Just a few quick not-completely-related questions: poudriere has the
> > > ability to sign repos with PKG_REPO_SIGNING_KEY, but not with external
> > > command, right? Is there a plan to support it? Can I build packages in
> > > poudriere without PKG_REPO_SIGNING_KEY, and sign repo later on with
> > > external command?
> > 
> > First yes I plan to add the ability to sign the package used to bootstrap
> > via PKG_REPO_SIGNING_KEY asap in poudriere.
> > 
> > Second you can keep your current configuration of poudriere, the signing
> > with pubkey works perfectly well. All you need to do is either via a
> > poudriere post bulk hook or manually go in the directory where your
> > packages lives (in the Latest directory) and
> > echo -n "$(sha256 -q pkg.txz)" | openssl dgst -sha256 -sign /thekey \
> > -binary -out ./pkg.txz.pubkeysig
> 
> I can't find any documentation in neither Poudriere's manpage nor in 
> poudriere.conf.sample on how toadd a post bulk hook.
> 
> Is the signing_command option to `pkg repo` really only used in generating 
> pkg.txz.sig? Is there any formal documentation about the cryptography design 
> and architecture in relation to pkg's repositories?
> 
> Thanks,

This is the doc we have on hooks: 
https://github.com/freebsd/poudriere/wiki/hooks

Would be nice to get more stuff in there :)

Best regards,
Bapt


pgpYRpTbmc4pz.pgp
Description: PGP signature

Re: 10.2-RELEASE-p2 lost ability to bootstrap pkg with signature_type="pubkey"

[Baptiste Daroussin]

On Tue, Sep 08, 2015 at 12:38:38PM +0200, Marko Cupać wrote:
> Hi,
> 
> I just found out that 10.2-RELEASE-p2 lost ability to bootstrap pkg
> with signature_type="pubkey".
> 
> Quick search returns:
> https://github.com/freebsd/pkg/issues/1309
> https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=202622
> 
> I guess it is not hard to switch repo to fingerprints, however I would
> not expect to lose this functionality by updating to patchlevel.
> 
Implemented in head: r287579 I will MFC it asap. And see if it cannot be added
asap to a next patchlevel update.

Best regards,
Bapt


pgpbbpFbVKTkD.pgp
Description: PGP signature

Re: regression from previous version of pw

[Baptiste Daroussin]

On Mon, Aug 24, 2015 at 03:49:52PM +0200, Willem Jan Withagen wrote:
 Hi,
 
 In the previous version before Bapt upgraded pw(8) I was allowed to do
   pw useradd -V /etc
 
 Even though the manual page instructed otherwise.
 
 The new version sticks to the manual page and generates:
 pw: illegal option -- V
 
 I'm going to change the scripts, but perhaps others will be bitten too.
 Up to Bapt to decide what to do with this.

The code has been changed to respect what the manual and the help output was
showing: -V should be passed before the subcommand

The reason for this is was to make the code more easily maintenable and
understandable.

if really that is too painful I can allow again -V to be also passed after the
subcommand but at the price of complicating the code.

Best regards,
Bapt


pgpsDiMBSUCjr.pgp
Description: PGP signature

Re: regression from previous version of pw

[Baptiste Daroussin]

On Mon, Aug 24, 2015 at 04:07:29PM +0200, Willem Jan Withagen wrote:
 On 24-8-2015 16:01, Baptiste Daroussin wrote:
  On Mon, Aug 24, 2015 at 03:49:52PM +0200, Willem Jan Withagen wrote:
  Hi,
 
  In the previous version before Bapt upgraded pw(8) I was allowed to do
 pw useradd -V /etc
 
  Even though the manual page instructed otherwise.
 
  The new version sticks to the manual page and generates:
  pw: illegal option -- V
 
  I'm going to change the scripts, but perhaps others will be bitten too.
  Up to Bapt to decide what to do with this.
  
  The code has been changed to respect what the manual and the help output was
  showing: -V should be passed before the subcommand
 
 You just changed that now? Or my version is lagging behind?
 Text is much more clear than the 'illegal option' espacially now there
 is also the -R stuff. So it made me wonder.

I changed that when bringing in the -R stuff so long ago. Illegal option is made
by getopt(3) directly but I can make it nicer and show a message that explains.
 
  The reason for this is was to make the code more easily maintenable and
  understandable.
 
 You fixed the code, so it is your prerogative.

Well I also want the feedback from users, as if it is only me, given how things
are documented and the whole organisation of the pw(8) code (even before my
refactoring) it makes more sense as it is now. (this is the only behaviour that
changed on purpose, all others are considered as a bug and should be reported
and fixed).
 
  if really that is too painful I can allow again -V to be also passed after 
  the
  subcommand but at the price of complicating the code.
 
 Nope, I've already changed the code I have.
 Was about 5 lines in a shell script.
 
 My comment was more as general notification, in case somebody else ran
 into the same. Nobody thus far ran into this, so it won't be major anyways.

You are actually the second to report, the first was crochet which has be fixed
straight forward.

Best regards,
Bapt


pgpSzICuwTtq2.pgp
Description: PGP signature

Re: pkg clean question

[Baptiste Daroussin]

On Mon, Aug 03, 2015 at 09:49:33AM -0700, jungle Boogie wrote:
 On 3 August 2015 at 09:13, Paul Mather freebsd-li...@gromit.dlib.vt.edu 
 wrote:
  On Aug 3, 2015, at 11:39 AM, Zoran Kolic zko...@sbb.rs wrote:
 
  Amd64, 9.3. Updated few times, packages also. At first I
  found on laptop that /var directory was at 102%. Using
  pkg clean i removed about 400mb of cache packages from
  /var/cache/pkg. Tried to clean desktop cache and the out-
  put of the command was nothing to do. I see about 400
  mb of files in the cache directory. Is it safe to remove
  them by the hand?
 
 
  Use pkg clean -ay to delete all cached packages, not just the ones that 
  are no longer current or provided by the upstream repository.
 
 
 
 You may want to autoremove, too:
 
 autoremove
  Delete packages which were automatically installed as dependen-
  cies and are not required any more.
 
  Cheers,
 
  Paul.
 
 
Note that since recent updates of pkg there is a new AUTOCLEAN option to add in
pkg.conf which will automatically cleanup the cache after each pkg operation.

Best regards,
Bapt


pgpE1oWcrSDGk.pgp
Description: PGP signature

Re: 10.2-BETA1: pw(8) does not support pw useradd name -u 0

[Baptiste Daroussin]

On Tue, Jul 14, 2015 at 05:08:19PM +1000, Jan Mikkelsen wrote:
 
  On 13 Jul 2015, at 19:10, Baptiste Daroussin b...@freebsd.org wrote:
  
  On Mon, Jul 13, 2015 at 10:36:28AM +0200, Baptiste Daroussin wrote:
  On Mon, Jul 13, 2015 at 04:57:32PM +1000, Jan Mikkelsen wrote:
  Hi,
  
  In our system build scripts we have this command:
  
  /usr/sbin/pw -V $d useradd toor -u 0 -g 0 -d /root -s /bin/sh -c 
  Bourne-again Superuser -g wheel -o
  
  After 10.2-BETA1, the toor account is being added with UID 1001 instead 
  of UID 0. This looks like a problem with line 754 in pw_user.c, which has 
  this test:
  
 /*
  * Check the given uid, if any
  */
 if (id  0) {
 uid = (uid_t) id;
  
 if ((pwd = GETPWUID(uid)) != NULL  conf.checkduplicate)
 errx(EX_DATAERR, uid `%u' has already been 
  allocated, pwd-pw_uid);
 } else {
 struct bitmap   bm;
  
  
  The (id  0) test should probably be (id = 0) to allow “-u 0” to be 
  passed on the command line.
  
  This change is from r285092 by bapt@. Was this change in behaviour 
  intentional?
  
  Nope, I'll fix asap
  
  Thanks for reporting
  
  Best regards,
  Bapt
  
  Fixed in head, will be merged soon in stable, I also added a regression test
  about this.
  
  Please note that you do add -g 0 and -g wheel in your command line, this is
  buggy, only one should be specified.
  
  Best regards,
  bapt
 
 The next problem is that the meaning of the -o option seems to have been 
 reversed. Setting -o sets conf.checkduplicate to true, which is then tested 
 in the code fragment above. Setting -o is meant to prevent duplicate 
 checking, not turn it on.
 
 My guess is that this isn’t intentional either.
 
 Also: The policy for auto-allocating group identifiers seems to have changed. 
 For UIDs  1000 the old pw allocated a GID the same as the UID. This pw 
 allocates the next available above 1000. I can see an argument for both cases 
 and I’ve changed our build scripts to deal with this but I’m curious: Was 
 this intentional also?
 

Both has been fixed head already I do plan to merge them later today so they
will be in BETA2

Best regards,
Bapt


pgp6XSYUz1o2z.pgp
Description: PGP signature

Re: 10.2-BETA1: pw(8) does not support pw useradd name -u 0

[Baptiste Daroussin]

On Tue, Jul 14, 2015 at 05:08:19PM +1000, Jan Mikkelsen wrote:
 
  On 13 Jul 2015, at 19:10, Baptiste Daroussin b...@freebsd.org wrote:
  
  On Mon, Jul 13, 2015 at 10:36:28AM +0200, Baptiste Daroussin wrote:
  On Mon, Jul 13, 2015 at 04:57:32PM +1000, Jan Mikkelsen wrote:
  Hi,
  
  In our system build scripts we have this command:
  
  /usr/sbin/pw -V $d useradd toor -u 0 -g 0 -d /root -s /bin/sh -c 
  Bourne-again Superuser -g wheel -o
  
  After 10.2-BETA1, the toor account is being added with UID 1001 instead 
  of UID 0. This looks like a problem with line 754 in pw_user.c, which has 
  this test:
  
 /*
  * Check the given uid, if any
  */
 if (id  0) {
 uid = (uid_t) id;
  
 if ((pwd = GETPWUID(uid)) != NULL  conf.checkduplicate)
 errx(EX_DATAERR, uid `%u' has already been 
  allocated, pwd-pw_uid);
 } else {
 struct bitmap   bm;
  
  
  The (id  0) test should probably be (id = 0) to allow “-u 0” to be 
  passed on the command line.
  
  This change is from r285092 by bapt@. Was this change in behaviour 
  intentional?
  
  Nope, I'll fix asap
  
  Thanks for reporting
  
  Best regards,
  Bapt
  
  Fixed in head, will be merged soon in stable, I also added a regression test
  about this.
  
  Please note that you do add -g 0 and -g wheel in your command line, this is
  buggy, only one should be specified.
  
  Best regards,
  bapt
 
 The next problem is that the meaning of the -o option seems to have been 
 reversed. Setting -o sets conf.checkduplicate to true, which is then tested 
 in the code fragment above. Setting -o is meant to prevent duplicate 
 checking, not turn it on.
 
 My guess is that this isn’t intentional either.
 
 Also: The policy for auto-allocating group identifiers seems to have changed. 
 For UIDs  1000 the old pw allocated a GID the same as the UID. This pw 
 allocates the next available above 1000. I can see an argument for both cases 
 and I’ve changed our build scripts to deal with this but I’m curious: Was 
 this intentional also?
 
 Regards,
 
All fixes has been merged:
https://svnweb.freebsd.org/base?view=revisionrevision=285536

Do not hesitate to report other regressions if you do find any more.

Thank you very much,
Bapt


pgpKAZHfAKRVC.pgp
Description: PGP signature

Re: 10.2-BETA1: pw(8) does not support pw useradd name -u 0

[Baptiste Daroussin]

On Mon, Jul 13, 2015 at 04:57:32PM +1000, Jan Mikkelsen wrote:
 Hi,
 
 In our system build scripts we have this command:
 
 /usr/sbin/pw -V $d useradd toor -u 0 -g 0 -d /root -s /bin/sh -c 
 Bourne-again Superuser -g wheel -o
 
 After 10.2-BETA1, the toor account is being added with UID 1001 instead of 
 UID 0. This looks like a problem with line 754 in pw_user.c, which has this 
 test:
 
 /*
  * Check the given uid, if any
  */
 if (id  0) {
 uid = (uid_t) id;
 
 if ((pwd = GETPWUID(uid)) != NULL  conf.checkduplicate)
 errx(EX_DATAERR, uid `%u' has already been 
 allocated, pwd-pw_uid);
 } else {
 struct bitmap   bm;
 
 
 The (id  0) test should probably be (id = 0) to allow “-u 0” to be passed 
 on the command line.
 
 This change is from r285092 by bapt@. Was this change in behaviour 
 intentional?

Nope, I'll fix asap

Thanks for reporting

Best regards,
Bapt


pgpDnNIUX4Hk0.pgp
Description: PGP signature

Re: 10.2-BETA1: pw(8) does not support pw useradd name -u 0

[Baptiste Daroussin]

On Mon, Jul 13, 2015 at 10:36:28AM +0200, Baptiste Daroussin wrote:
 On Mon, Jul 13, 2015 at 04:57:32PM +1000, Jan Mikkelsen wrote:
  Hi,
  
  In our system build scripts we have this command:
  
  /usr/sbin/pw -V $d useradd toor -u 0 -g 0 -d /root -s /bin/sh -c 
  Bourne-again Superuser -g wheel -o
  
  After 10.2-BETA1, the toor account is being added with UID 1001 instead of 
  UID 0. This looks like a problem with line 754 in pw_user.c, which has this 
  test:
  
  /*
   * Check the given uid, if any
   */
  if (id  0) {
  uid = (uid_t) id;
  
  if ((pwd = GETPWUID(uid)) != NULL  conf.checkduplicate)
  errx(EX_DATAERR, uid `%u' has already been 
  allocated, pwd-pw_uid);
  } else {
  struct bitmap   bm;
  
  
  The (id  0) test should probably be (id = 0) to allow “-u 0” to be passed 
  on the command line.
  
  This change is from r285092 by bapt@. Was this change in behaviour 
  intentional?
 
 Nope, I'll fix asap
 
 Thanks for reporting
 
 Best regards,
 Bapt

Fixed in head, will be merged soon in stable, I also added a regression test
about this.

Please note that you do add -g 0 and -g wheel in your command line, this is
buggy, only one should be specified.

Best regards,
bapt


pgpXGvYnqU7zZ.pgp
Description: PGP signature

Re: pkg 1.5.0 is out

[Baptiste Daroussin]

On Tue, Jul 07, 2015 at 07:38:18PM +0200, Ulrich Spörlein wrote:
 2015-07-01 13:38 GMT+02:00 Baptiste Daroussin b...@freebsd.org:
 
  On Wed, Jul 01, 2015 at 01:36:14PM +0200, Hans Petter Selasky wrote:
   On 04/21/15 12:34, Slawa Olhovchenkov wrote:
On Tue, Apr 14, 2015 at 10:05:00PM +0200, Baptiste Daroussin wrote:
   
Hi all,
   
Final pkg 1.5.0 has been released.
   
  
   Hi,
  
   Is there a way the external SAT solver functionality can be memory
   optimised? When trying to use this feature having +750 packages
   installed, the memory usage starts growing and growing beyond 4GBytes
   until PKG segfaults, even before the CNF export has started.
  
   env SAT_SOLVER=mysolver pkg upgrade
 
  Probably, but given the little amount of time pkg developers has we will 
  greatly
  appreciate patches :)
 
  AKA this would be greatly appreciated, but very low on the priority list :(
 
  Best regards,
  Bapt
 
 
 Hijacking this, I managed to mess up my local pkg repo somehow.
 
 I build my own set of packages, and typically do pkg upgrade on the
 clients. This time, I tried pkg upgrade -F, which went and downloaded
 everything and that's fine. But now when I run pkg upgrade it claims
 everything is already updated?
 
 root@coyote:~# pkg --version
 1.5.4
 root@coyote:~# pkg upgrade
 Updating acme repository catalogue...
 acme repository is up-to-date.
 All repositories are up-to-date.
 Checking for upgrades (68 candidates): 100%
 Processing candidates (68 candidates): 100%
 Checking integrity... done (0 conflicting)
 Your packages are up to date.
 
 
 So let's try brute forcing this:
 
 root@coyote:~# pkg install `pkg info -aqo`
 Updating acme repository catalogue...
 acme repository is up-to-date.
 All repositories are up-to-date.
 databases/db48 has no direct installation candidates, change it to db5? 
 [Y/n]: Y
 Assertion failed: (0), function pkg_jobs_try_remote_candidate, file
 pkg_jobs.c, line 821.
 Child process pid=60776 terminated abnormally: Abort trap
 Exit 250
 
 
 Using more force:
 
 root@coyote:~# pkg upgrade -f db48
 Updating acme repository catalogue...
 acme repository is up-to-date.
 All repositories are up-to-date.
 db48 has no direct installation candidates, change it to db5? [Y/n]: y
 pkg: sqlite error while executing UPDATE packages SET name=?1  WHERE
 name=?2; in file pkg_jobs.c:1658: UNIQUE constraint failed:
 packages.name
 The following 1 package(s) will be affected (of 0 checked):
 
 Installed packages to be UPGRADED:
 db5: 4.8.30.0_2 - 5.3.28_2
 
 The process will require 37 MiB more space.
 12 MiB to be downloaded.
 
 Proceed with this action? [y/N]: y
 Fetching db5-5.3.28_2.txz: 100%   12 MiB   6.4MB/s00:02
 Checking integrity...Assertion failed: (strcmp(uid, p-uid) != 0),
 function pkg_conflicts_check_local_path, file pkg_jobs_conflicts.c,
 line 368.
 Child process pid=60922 terminated abnormally: Abort trap
 Exit 250
 
 the -debug output has nothing of interest that I can see.
 
 What's up?

Please open a ticket with this and save your /var/db/pkg/* somewhere so we can
analyze (and send them to me or host them in a place when I can fetch them)

in the meantime what you should do is pkg upgrade -f (lower f) and say no to
everything that ask you to switch candidates

Best regards,
Bapt


pgpKCbT7Og8t3.pgp
Description: PGP signature

Re: pkg 1.5.0 is out

[Baptiste Daroussin]

On Wed, Jul 01, 2015 at 01:36:14PM +0200, Hans Petter Selasky wrote:
 On 04/21/15 12:34, Slawa Olhovchenkov wrote:
  On Tue, Apr 14, 2015 at 10:05:00PM +0200, Baptiste Daroussin wrote:
 
  Hi all,
 
  Final pkg 1.5.0 has been released.
 
 
 Hi,
 
 Is there a way the external SAT solver functionality can be memory 
 optimised? When trying to use this feature having +750 packages 
 installed, the memory usage starts growing and growing beyond 4GBytes 
 until PKG segfaults, even before the CNF export has started.
 
 env SAT_SOLVER=mysolver pkg upgrade

Probably, but given the little amount of time pkg developers has we will greatly
appreciate patches :)

AKA this would be greatly appreciated, but very low on the priority list :(

Best regards,
Bapt


pgpBveW065vG4.pgp
Description: PGP signature

Re: protecting some processes from out-of-swap killer

[Baptiste Daroussin]

On Sat, Apr 25, 2015 at 01:31:14PM +0300, Dmitry Morozovsky wrote:
 Hi there colleagues,
 
 I have stable/10 on a rather big machine (2*8*2 e5 Xeon, 64G RAM, SAS+SSD ZFS 
 raid10+ZIL+L2ARC) acting as a PostgreSQL server. 
 
 To use such a big resource pool that is mostly idle, I'd deployed poudriere 
 there (using tmpfs) too.
 
 Most times this combination works like a charm: LA could be 60+ and no visual 
 latency increase on SQL queries.
 
 However, sometimes postgres processes got killed by 'out of swap space'.  
 I suppose the source of problem could be that VSZ size of postgres processes 
 (8-9 G) is bigger than swap congigured (4G).
 
 Is there any way to prevent this, besides reallocating space for swap?
 
 Quick googling does not help, at least I could not find answers relevant 
 enough.
 
 Thanks!

protect(1) ?

Best regards,
Bapt


pgpmab3b4cR5_.pgp
Description: PGP signature

Re: pkg 1.5.0 is out

[Baptiste Daroussin]

On Sun, Apr 19, 2015 at 02:22:15PM +0300, Dmitry Sivachenko wrote:
 
  On 14 апр. 2015 г., at 23:05, Baptiste Daroussin b...@freebsd.org wrote:
  
  Final pkg 1.5.0 has been released.
  
 
 Thank a lot for working on pkg!
 
  
  For pkg 1.6.0 among other things and depending on the time, here is what we 
  do
  plan to work on:
  - please add here you personal ideas
  
 
 
 What I really miss a lot is support for package profiles: an ability to 
 build the same port with different OPTIONs combination.
 For example:
 minimal nginx version;
 nginx version with passenger module (for puppet server)
 nginx version with some other rare options turned on for custom application.
 
 Right now I achieve this with manually renaming /var/db/ports/*/options files 
 and some manipulations in /usr/ports/packages/All.
 But a framework to automatically handle this would be very useful.
 
 Thanks.
 
This is already supported on pkg side, the work needs to be done in ports side,
what you call profiles, I usually call it flavours but that is the same.

Dependencies made on provides/requires allows to handle dependencies in this:

For example
A can depend on nginx (whatever profile):
so it requires nginx
B can only depend on nginx-passenger:
so it requires nginx-passenger

nginx itself provides: nginx
nginx-passenger provides: nginx and nginx-passenger

that should do what you want unfortunatly that mean both nginx nginx-passenger
both have one origin: www/nginx and this will break
portupgrade/portmaster/tinderbox/poudriere (I can easily make poudriere friendly
for that but I can't for the others).

So most of the work is now on ports land (and maybe a couple of bug fixes in
pkg)

Best regards,
Bapt


pgpcHHlwZ_Rny.pgp
Description: PGP signature

Re: pkg 1.5.0 is out

[Baptiste Daroussin]

On Wed, Apr 15, 2015 at 10:57:28AM +0100, Big Lebowski wrote:
 On Tue, Apr 14, 2015 at 9:05 PM, Baptiste Daroussin b...@freebsd.org wrote:
  Hi all,
 
  Final pkg 1.5.0 has been released.
 
 
  Best regards,
  Bapt
 
 Congratulations to entire pkg team and thanks for your hard work!
 
 I've one question - how does the OSX support work? What are plans for
 it? What packages are working? And so on...

That should been asked to the people that made the support for OSX I'm just the
messenger here, I do not have access to any OSX to be able to test/code, I'm
just making sure I do not break their work :)

Best regards,
Bapt


pgpt9iQjclSkf.pgp
Description: PGP signature

pkg 1.5.0 is out

[Baptiste Daroussin]

Hi all,

Final pkg 1.5.0 has been released.

What happened since pkg 1.4.0:
- Initial provides/requires support
- Lots of new regression tests have been added
- Initial support for OS X
- Initial support for NetBSD/EdgeBSD
- Update most of the bundled third party software has been updated to their
  latest version
- Improve the messages reported by pkg
- Properly support file flags
- Implement argument support for custom keywords
- Extend setting credential via plist to allow to set file flags
- Make credential syntax via plist more flexible allow to only defines the first
- pkg updating now supports case insensitive matching
- pkg create now support a verbose mode
- Add an option to change the default on question, until now the default answer
  was No with that option set it would be Yes
- Lots of fixes to pkg audit -r
- Global memory usage reduction and speed up
- Improvements and cleanup on pkg alias
- pkg annotate --show --all has been fixed
- Make pkg.h C++ friendly
- Lots of improvements in the solver
- Lots of fixes on 32 bits platforms
- Add support for: pkg create -M ./plop.ucl -p ./plop.plist
- New pkg -r rootdir that will install in the given rootdir without chrooting
- Export PKG_ROOTDIR to scripts allow to make them as portable as possible
- Stop trying to remove all installed package with the argument of pkg delete is
  a local file
- Be more explicit about why the solver it going to reinstall, remove or upgrade
  (when possible)
- Plenty of bug fixes
- Plenty of new bugs
- pkg shlibs now support -q
- pkg lock gained a new --has-locked-packages option
- pkg now resumes fetch if possible
- CONSERVATIVE_UPGRADE is now on by default
- pkg alias now have a -l argument to list aliases
- A sample pkg.conf is now installed with a bunch of aliases set by default
- Fix the backup script to properly export an sql which will be importable via
  pkg shell and/or sqlite out of box

I would like to thank anyone that has been contributing to pkg to make this
release happen (via code, bug report, feature request, testing and 
documentation)

For pkg 1.6.0 among other things and depending on the time, here is what we do
plan to work on:
- Safe cherry-picking of upgrades (aka: pkg upgrade something)
- New context dependant messages:
  * messages that only appears during upgrades
  * messages that only appears on deinstall
  * messages that only appears on install
- Extend provides/requires to support flexible dependencies
- Linux package backend (?)
- Allow multiple versions of a given package in a repo
- Add more regression tests
- Improve documentation
- please add here you personal ideas

Best regards,
Bapt


pgpMh21EM6XGC.pgp
Description: PGP signature

Re: boostrapping pkg and ipv6

[Baptiste Daroussin]

On Sat, Apr 04, 2015 at 10:02:04AM +0200, Marten wrote:
 Hi all,
 
 fyi:
 pkg does not honor -4 or such config /usr/local/etc/pkg.conf
 http://vuxml.freebsd.org/freebsd/vuln.xml.bz2 
 http://vuxml.freebsd.org/freebsd/vuln.xml.bz2 is not reachable over ipv6
 
 adding ipv4 addresses in /etc/hosts are a workaround
 
 data below,
 
 kind regards,
 Marten
 
 on this box with ipv4 and 6 configured:
 
 /usr/local/etc/pkg.conf   = IP_VERSION: 4
 
 pkg -4 audit -F
 pkg: http://vuxml.freebsd.org/freebsd/vuln.xml.bz2: Operation timed out
 pkg: cannot fetch vulnxml file
 
 
 fetch http://vuxml.freebsd.org/freebsd/vuln.xml.bz2
 vuln.xml.bz2
 fetch: transfer timed out
 
 tcpdump shows no ipv4 traffic though is sees traffic to 
 2001:1900:2254:206a::50:0
 
 
 host  2001:1900:2254:206a::50:0
 0.0.0.0.0.5.0.0.0.0.0.0.0.0.0.0.a.6.0.2.4.5.2.2.0.0.9.1.1.0.0.2.ip6.arpa 
 domain name pointer wfe0.ysv.freebsd.org
 
 host vuxml.freebsd.org
 vuxml.freebsd.org is an alias for wfe0.ysv.freebsd.org.
 wfe0.ysv.freebsd.org has address 8.8.178.110
 wfe0.ysv.freebsd.org has IPv6 address 2001:1900:2254:206a::50:0
 wfe0.ysv.freebsd.org mail is handled by 0 .
 2001:1900:2254:206a::50:0
 
 
 while on another box without ipv6 configured:
 fetch http://vuxml.freebsd.org/freebsd/vuln.xml.bz2
 vuln.xml.bz2  100% of  474 kB  434 kBps
 
 
 adding the next line to /etc/hosts
 
 8.8.178.110 vuxml.freebsd.org wfe0.ysv.freebsd.org 
 http://wfe0.ysv.freebsd.org/
 
 
 pkg -4 audit -F
 pkg: vulnxml file up-to-date
 0 problem(s) in the installed packages found.
 
Can you open a ticket on http://github.com/freebsd/pkg ?

Best regards,
Bapt


pgpFnZljdwxek.pgp
Description: PGP signature

Re: [jos...@joshualokken.com: Re: [CFT] Call for testing pkg 1.5.0]

[Baptiste Daroussin]

On Fri, Apr 03, 2015 at 08:50:01AM -0700, Joshua Lokken wrote:
 From: Joshua Lokken jos...@joshualokken.com
 To: freebsd-questi...@freebsd.org
 Date: Fri, 3 Apr 2015 08:36:56 -0700
 Subject: Re: [CFT] Call for testing pkg 1.5.0
 
 On Mar/31 09:03PM, Baptiste Daroussin wrote:
  Hi all,
  
  We just released pkg 1.5.0 beta1 (in ports-mgmt/pkg-devel),
  
  Here is what happened since pkg 1.4.0:
 8 -- [snip] -- * 
  Please test and report as much bugs as you can!
  We could be very grateful if regressions tests could be provided along with 
  the
  bug reports :)
  Bapt
 
 
 
 jrlokken@satellite:~ % uname -a
 FreeBSD satellite.workgroup.local 10.1-STABLE FreeBSD 10.1-STABLE 
 #1 r280923: Tue Mar 31 15:32:16 PDT 2015 
 root@satellite.workgroup.local:/usr/obj/usr/src/sys/SATELLITE  amd64
 
 jrlokken@satellite:~ % pkg version -x kde
 kcm-polkit-kde-0.0.20121008_3  =
 kde-4.14.2 
 kde-base-artwork-4.14.2
 kde-baseapps-4.14.2_1  
 kde-runtime-4.14.2_3   
 kde-wallpapers-4.14.2  
 kde-workspace-4.11.13_2
 ...
 
 jrlokken@satellite:~ % pkg version kde
 Child process pid=9630 terminated abnormally: Segmentation fault
 
 Seems to work for day-to-day operation, but if I attempt to pass a bad
 argument to pkg version, it segfaults.
 
 I know, I know...  Then don't pass bad arguments to pkg version :)
 
Fixed will be in next beta/rc

Best regards,
Bapt


pgpNOp83Hvp0O.pgp
Description: PGP signature

Re: [CFT] Call for testing pkg 1.5.0

[Baptiste Daroussin]

On Fri, Apr 03, 2015 at 02:35:43PM +0100, Big Lebowski wrote:
 
  Please test and report as much bugs as you can!
  We could be very grateful if regressions tests could be provided along with 
  the
  bug reports :)
 
 
 Mine just did something like that:
 
 foobar# uname -a
 FreeBSD foobar.org 11.0-CURRENT FreeBSD 11.0-CURRENT #0 r280994M: Thu
 Apr  2 20:16:53 CEST 2015
 r...@pd.valinor.palantiri.org:/usr/obj/usr/src/sys/FOOBAR  amd64
 foobar [uname -a] ~/
   15-04-03
  3:33PM
 foobar# pkg --version
 1.4.99.15
 foobar [pkg --version] ~/
   15-04-03
  3:33PM
 foobar# pkg version
 Child process pid=41119 terminated abnormally: Segmentation fault
 foobar [pkg version] ~/
   15-04-03
  3:33PM
 foobar# pkg version -t 0.2.o.git20150311 0.2.o.20150402
 
 foobar [pkg version -t 0.2.o.git20150311 0.2.o.20150402] ~/
   15-04-03
  3:33PM
 pd#

Let me fix that :)

Best regards,
Bapt


pgptyF_fLXsK2.pgp
Description: PGP signature

Re: [CFT] Call for testing pkg 1.5.0

[Baptiste Daroussin]

On Thu, Apr 02, 2015 at 07:21:19AM -0700, Chris H wrote:
 On Tue, 31 Mar 2015 21:03:23 +0200 Baptiste Daroussin b...@freebsd.org wrote
 
  Hi all,
  
  We just released pkg 1.5.0 beta1 (in ports-mgmt/pkg-devel),
  
 ..
  Please test and report as much bugs as you can!
  We could be very grateful if regressions tests could be provided along with
  the bug reports :)
  
  Plan is to release 1.5.0 as soon as possible
  
  Best regards,
  Bapt
 Hello, Baptiste.
 I just wanted to take the time to thank you for all the
 work you've put into this.
 
 Thanks!

Thanks much appreciated, I want to share that with vsevolod@ and az@ who also
spent a lot of time working on it!

Best regards,
Bapt


pgpXzG081o0eW.pgp
Description: PGP signature

Re: pkgng cannot fetch from PACKAGESITE with nanobsd.sh in FreeBSD 10.1

[Baptiste Daroussin]

On Wed, Apr 01, 2015 at 12:48:21PM +0200, Zenny wrote:
 Hi,
 
 1. In order to comply with the pkgng in FreeBSD 10.1, the following changes
 were appended to my customized nanobsd.conf:
 
 customize_cmd cust_NANOBSD_setup
 
  cust_NANOBSD_packages() {
  chroot ${NANO_WORLDDIR} sh -c 'cd packages; pkg; pkg install nano; pkg
  clean;'
  }
 
  customize_cmd cust_NANOBSD_packages
 
 
 Certainly I had tuned chrooted /etc/resolv.conf with:
 
 #tune resolv.conf
  echo '
  nameserver 8.8.8.8
  nameserver 8.8.4.4
  '  ${NANO_WORLDDIR}/etc/resolv.conf
 

This tuning is not needed you can specify the nameserver on pkg command line:
pkg -o NAMESERVER=8.8.8.8 -c ${NANO_WORLDDIR} update

You can also simply add NAMESERVER in your environement variable
 
 Still the package cannot be downloaded from PACKAGESITE and got built into
 nanobsd image.
 
 2. Also tried with:
 
 cust_pkg() {
  pkg -c ${NANO_WORLDDIR} update
  pkg -c ${NANO_WORLDDIR} install nano
  pkg -c ${NANO_WORLDDIR} clean
  }
 
  customize_cmd cust_pkg
 
 
 Both ways, I get the same error that reads as of below:
 
 cd: packages: No such file or directory
  pkg: Error fetching
  http://pkg.FreeBSD.org/freebsd:10:x86:64/latest/Latest/pkg.txz: No
  address record
  A pre-built version of pkg could not be found for your system.
  Consider changing PACKAGESITE or installing it from ports:
  'ports-mgmt/pkg'.
  The package management tool is not yet installed on your system.
  Do you want to fetch and install it now? [y/N]: Bootstrapping pkg from pkg+
  http://pkg.FreeBSD.org/freebsd:10:x86:64/latest, please wait...
  pkg: Error fetching
  http://pkg.FreeBSD.org/freebsd:10:x86:64/latest/Latest/pkg.txz: No
  address record
  A pre-built version of pkg could not be found for your system.
  Consider changing PACKAGESITE or installing it from ports:
  'ports-mgmt/pkg'.
  The package management tool is not yet installed on your system.
  Do you want to fetch and install it now? [y/N]: Bootstrapping pkg from pkg+
  http://pkg.FreeBSD.org/freebsd:10:x86:64/latest, please wait...
  pkg: Error fetching
  http://pkg.FreeBSD.org/freebsd:10:x86:64/latest/Latest/pkg.txz: No
  address record
  A pre-built version of pkg could not be found for your system.
  Consider changing PACKAGESITE or installing it from ports:
  'ports-mgmt/pkg'.
  The package management tool is not yet installed on your system.
  Do you want to fetch and install it now? [y/N]: Bootstrapping pkg from pkg+
  http://pkg.FreeBSD.org/freebsd:10:x86:64/latest, please wait...
 
 
Is pkg installed on your host? it seems not?

Best regards,
Bapt


pgpBmjsn_s9a8.pgp
Description: PGP signature

Re: [CFT] Call for testing pkg 1.5.0

[Baptiste Daroussin]

On Wed, Apr 01, 2015 at 11:48:27AM +, Thomas Mueller wrote:
 Excerpt from Baptiste Daroussin:
 
 - Initial support for OS X
 - Initial support for NetBSD/EdjeBSD
 
 How would pkg-1.5.0 integrate with NetBSD pkgsrc?
 
 I didn't think there were any plans to port FreeBSD ports to NetBSD.  Or is 
 such a plan in the works?
 
There are people looking at integrating pkg with pkgsrc yes, don't know the
status, (pkg is already in pkgsrc-wip but no further integration for what I do
know

And yes you are right it is EdgeBSD not EdjeBSD sorry.

Best regards,
Bapt


pgpwv8cOvqD9U.pgp
Description: PGP signature

Re: [CFT] Call for testing pkg 1.5.0

[Baptiste Daroussin]

On Wed, Apr 01, 2015 at 01:56:38PM +0100, David Chisnall wrote:
 On 1 Apr 2015, at 05:03, Rui Paulo rpa...@me.com wrote:
  
  That is expected.  WITH_PKG=devel is a make(1) option that only affects 
  ports (non-binary pkgs).
 
 Are you sure?  I have it in make.conf on one of my systems where I never 
 build ports manually (and don't even have a ports tree installed) and there I 
 get this:
 
 $ pkg -v
 1.4.99.13
 
 In a jail on the same machine without the make.conf entry, I get the stable 
 version.  This is how I've been testing pkg-devel for a while.  Is there a 
 different recommended way?
 
That is because you enforced installing pkg-devel one day via:
pkg install pkg-devel
probably

then it will stay on pkg-devel :) pkg itself is not aware of make.conf

Best regards
Bapt


pgp6UBoFxwzrv.pgp
Description: PGP signature

[CFT] Call for testing pkg 1.5.0

[Baptiste Daroussin]

Hi all,

We just released pkg 1.5.0 beta1 (in ports-mgmt/pkg-devel),

Here is what happened since pkg 1.4.0:
- pkg has grown with an initial support for provides/requires: this is a naive
  version but good enough to at least make major upgrade of php safer as well as
  making pear/pecl maintenance saner (note that this will need modifications in
  the ports tree)
- Lots of new regression tests has been added, which allows us not less break
  your systems an unexpected way (do not worry there are still rooms for
  breakage)
- Initial support for OS X
- Initial support for NetBSD/EdjeBSD
- Update most of the bundled third party software has been updated to their
  latest version
- Improve the messages reported by pkg (and probably make some other worse)
- Properly support file flags
- Implement argument support for custom keywords
- Extend setting credential via plist to allow to set file flags
- Make credential syntax via plist more flexible allow to only defines the first
  fields and not latest for example @(user,,) can now be written just @(user)
- pkg updating now supports case insensitive matching
- pkg create now support a verbose mode
- Add an option to change the default on question, until now the default answer
  was No with that option set it would be Yes
- lots of fixes to pkg audit -r
- Global memory usage reduction and speed up
- Improvements and cleanup on pkg alias
- pkg annotate --show --all has been fixed
- Make pkg.h C++ friendly
- Lots of improvements in the solver
- Lots of fixes on 32 bits platforms
- Add support for: pkg create -M ./plop.ucl -p ./plop.plist
- New pkg -r rootdir that will install in the given rootdir without chrooting
- Export PKG_ROOTDIR to scripts allow to make them as portable as possible
- Stop trying to remove all installed package with the argument of pkg delete is
  a local file
- Be more explicit about why the solver it going to reinstall, remove or upgrade
  (when possible)
- Plenty of bug fixes
- Plenty of new bugs

Please test and report as much bugs as you can!
We could be very grateful if regressions tests could be provided along with the
bug reports :)

Plan is to release 1.5.0 as soon as possible

Best regards,
Bapt


pgp9RSzgT2h8U.pgp
Description: PGP signature

Re: [CFT] Call for testing pkg 1.5.0

[Baptiste Daroussin]

On Tue, Mar 31, 2015 at 03:19:15PM -0400, Shawn Webb wrote:
 On Tue, 2015-03-31 at 21:03 +0200, Baptiste Daroussin wrote:
  Hi all,
  
  We just released pkg 1.5.0 beta1 (in ports-mgmt/pkg-devel),
 
 Hey Baptiste,
 
 Great work to you and all those involved in this project! I'm grateful
 to have such an awesome tool.
 
 For those of us who run our own package repos via Poudriere, what kinds
 of changes should we expect to make once pkg 1.5.0 is released? Do we
 need to do a full rebuild of our package repo? I'm also assuming that
 the upgrade path from 1.4.x to 1.5.0 will simply be as easy as running
 `pkg upgrade`, right?
 
 Thanks again for your hard work.
 
Add WITH_PKG=devel in your build make.conf

then pkg upgrade will want you to upgrade to 1.4.99.16 (which is pkg 1.5.0
beta1)

Best regards,
Bapt


pgpmXuAYEDP8s.pgp
Description: PGP signature

Re: Troubles with dependencies (file -L is broken?)

[Baptiste Daroussin]

On Sun, Sep 08, 2013 at 12:09:32AM -0700, Maxim Sobolev wrote:
 Hi,
 
 I am trying to portupgrade my subversion and it keeps trying to install
 databases/db42 over already installed version (same version), missing the
 shared library dependency.
 
 Adding some debug into bsd.port.mk, I see:
 
 ===   subversion-1.8.3 depends on package: libtool=2.4 - found
 set -x; set -e ;  for i in libdb-4.2.so:/usr/ports/databases/db42; do
 lib=${i%%:*} ;  dir=${i#*:}  ;  target=install;  depends_args=;  echo
 -n ===   subversion-1.8.3 depends on shared library: ${lib} ;  found=0
 ;  dirs=/lib /usr/lib /usr/local/lib `/bin/cat
 /usr/local/libdata/ldconfig/* 2/dev/null || : ` ;  for libdir in $dirs;
 do  test -f ${libdir}/${lib} || continue;  if [ -x /usr/bin/file ]; then  [
 `file -b -L --mime-type ${libdir}/${lib}` = application/x-sharedlib ] ||
 continue ;  fi ;  found=1 ;  echo  - found;  done ;  if [ ${found} -eq 0
 ]; then  echo  - not found;  echo ===Verifying for $lib in $dir;
 if [ ! -d $dir ] ; then  echo = No directory for $lib.
 Skipping..;  else  if [ -n  -o -n  ]; then  subpkgfile=`(cd $dir; make
 $depends_args -V PKGFILE)`;  subpkgname=${subpkgfile%-*} ;
 subpkgname=${subpkgname##*/} ;  if [ -r ${subpkgfile} -a $target =
 install ]; then  echo ===   Installing existing package
 ${subpkgfile};  if [ -n  -a ${subpkgname} = pkg ]; then  [ -d
 /usr/ports/devel/subversion/work ] || /bin/mkdir -p
 /usr/ports/devel/subversion/work ;  /usr/bin/tar xf ${subpkgfile} -C
 /usr/ports/devel/subversion/work -s ,/.*/,,g */pkg-static ;
 /usr/ports/devel/subversion/work/pkg-static add ${subpkgfile};  /bin/rm -f
 /usr/ports/devel/subversion/work/pkg-static;  else  /usr/sbin/pkg_add
 ${subpkgfile};  fi;  elif [ -n  -a ${target} = install ]; then  echo
 ===   subversion-1.8.3 depends on package: ${subpkgfile} - not found;
 echo ===   USE_PACKAGE_DEPENDS_ONLY set - will not build from source;
 exit 1;  else  (cd $dir; make -DINSTALLS_DEPENDS $target $depends_args) ;
 fi;  else  (cd $dir; make -DINSTALLS_DEPENDS $target $depends_args) ;  fi;
 echo ===   Returning to build of subversion-1.8.3;  fi ;  fi ;  done
 + set -e
 + lib=libdb-4.2.so
 + dir=/usr/ports/databases/db42
 + target=install
 + depends_args=''
 + echo -n '===   subversion-1.8.3 depends on shared library: libdb-4.2.so'
 ===   subversion-1.8.3 depends on shared library: libdb-4.2.so+ found=0
 + /bin/cat /usr/local/libdata/ldconfig/compat7x
 /usr/local/libdata/ldconfig/mysql /usr/local/libdata/ldconfig/portupgrade
 /usr/local/libdata/ldconfig/pth
 + dirs='/lib /usr/lib /usr/local/lib /usr/local/lib/compat
 /usr/local/lib/mysql
 /usr/local/lib/compat/pkg
 /usr/local/lib/pth'
 + test -f /lib/libdb-4.2.so
 + continue
 + test -f /usr/lib/libdb-4.2.so
 + continue
 + test -f /usr/local/lib/libdb-4.2.so
 + [ -x /usr/bin/file ]
 + file -b -L --mime-type /usr/local/lib/libdb-4.2.so
 + [ inode/symlink = application/x-sharedlib ]
 + continue
 + test -f /usr/local/lib/compat/libdb-4.2.so
 + continue
 + test -f /usr/local/lib/mysql/libdb-4.2.so
 + continue
 + test -f /usr/local/lib/compat/pkg/libdb-4.2.so
 + continue
 + test -f /usr/local/lib/pth/libdb-4.2.so
 + continue
 + [ 0 -eq 0 ]
 + echo ' - not found'
  - not found
 + echo '===Verifying for libdb-4.2.so in /usr/ports/databases/db42'
 ===Verifying for libdb-4.2.so in /usr/ports/databases/db42
 
 So, file(1) call is the culprit here. Trying to reproduce in the console:
 
 [sobomax@pioneer ~]$ file -b -L --mime-type /usr/local/lib/libdb-4.2.so
 inode/symlink
 
 However:
 
 [sobomax@pioneer ~]$ hexdump -C /usr/local/lib/libdb-4.2.so | head -n 20
   7f 45 4c 46 02 01 01 09  00 00 00 00 00 00 00 00
 |.ELF|
 0010  03 00 3e 00 01 00 00 00  d0 40 02 00 00 00 00 00
 |...п@..|
 0020  40 00 00 00 00 00 00 00  b8 a6 0e 00 00 00 00 00
 |@...╦і..|
 0030  00 00 00 00 40 00 38 00  05 00 40 00 1d 00 1a 00  |@.8...@
 .|
 0040  01 00 00 00 05 00 00 00  00 00 00 00 00 00 00 00
 ||
 0050  00 00 00 00 00 00 00 00  00 00 00 00 00 00 00 00
 ||
 0060  dc 45 0e 00 00 00 00 00  dc 45 0e 00 00 00 00 00
 |эE..эE..|
 0070  00 00 20 00 00 00 00 00  01 00 00 00 06 00 00 00  |..
 .|
 0080  00 50 0e 00 00 00 00 00  00 50 2e 00 00 00 00 00
 |.P...P..|
 0090  00 50 2e 00 00 00 00 00  fc 36 00 00 00 00 00 00
 |.P..Э6..|
 00a0  d8 3a 00 00 00 00 00 00  00 00 20 00 00 00 00 00  |ь:
 .|
 00b0  02 00 00 00 06 00 00 00  e8 54 0e 00 00 00 00 00
 |ХT..|
 00c0  e8 54 2e 00 00 00 00 00  e8 54 2e 00 00 00 00 00
 |ХT..ХT..|
 00d0  a0 01 00 00 00 00 00 00  a0 01 00 00 00 00 00 00
 |═...═...|
 00e0  08 00 00 00 00 00 00 00  50 e5 74 64 04 00 00 00
 |PЕtd|
 00f0  5c 39 0d 00 00 00 00 00  5c 39 0d 00 00 00 00 00
 |\9..\9..|
 0100  5c 39 0d 00 00 00 00 00  b4 34 00 00 00 00 00 00
 |\9..Є4..|
 0110  b4 34 00 00 00 00 

Re: java (openjdk6) segfaults when built with 9-stable clang

[Baptiste Daroussin]

On Fri, Aug 02, 2013 at 02:15:25PM +0200, Patrick Lamaiziere wrote:
 Le Fri, 26 Jul 2013 00:05:50 +0200,
 Baptiste Daroussin b...@freebsd.org a écrit :
 
 Hi all, Baptiste,
 
   Hi
   
   I recently upgraded my home NAS from 9.1-RELEASE to 9-stable
   (r253470 (9.2-BETA1))
   
   I also upgraded my poudriere building jail. Since then,
   multimedia/xbmc port fails to build in configure stage : java
   segfaults (sig11). I use WITH_CLANG_IS_CC=YES, for world and
   build-jails.
   
   I found following workarounds:
 - use previously (with 9.1-RELEASE world and clang) build
   openjdk6 pkg (same version).
 - use USE_GCC=YES for java port.
   
   It's the only one place I use java (openjdk6-b27_4). So I cannot
   say if java works otherwise.
   
   Is this a java or clang bug ?
   
  
  Here is the bug
  
  http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6636110
  
  Fixed in b27_6
 
 Hmm, Isn't Openjdk7 affected too ?
 
 http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=2219304
 
 Someone complains on the FreeBSD forums for openjdk7, I think it needs
 to be patched too.  http://forums.freebsd.org/showthread.php?t=41181
 

To be honnest I do not have time to investigate neither fix before a while, I
have CC to the java list, in the hope someone will have a look at it.

regards,
Bapt


pgps0Yykd1UYt.pgp
Description: PGP signature

Re: java (openjdk6) segfaults when built with 9-stable clang

[Baptiste Daroussin]

On Wed, Jul 24, 2013 at 10:51:00AM +0200, Arnaud Houdelette wrote:
 Hi
 
 I recently upgraded my home NAS from 9.1-RELEASE to 9-stable (r253470  
 (9.2-BETA1))
 
 I also upgraded my poudriere building jail. Since then, multimedia/xbmc 
 port fails to build in configure stage : java segfaults (sig11).
 I use WITH_CLANG_IS_CC=YES, for world and build-jails.
 
 I found following workarounds:
   - use previously (with 9.1-RELEASE world and clang) build openjdk6 pkg 
 (same version).
   - use USE_GCC=YES for java port.
 
 It's the only one place I use java (openjdk6-b27_4). So I cannot say if 
 java works otherwise.
 
 Is this a java or clang bug ?
 

Here is the bug

http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6636110

Fixed in b27_6

regards,
Bapt


pgpj_xoM64r9k.pgp
Description: PGP signature

Re: [HEADSUP] New pkg-devel 1.1.0 beta1

[Baptiste Daroussin]

On Mon, Jun 03, 2013 at 07:17:24PM +0400, Slawa Olhovchenkov wrote:
 On Thu, May 30, 2013 at 05:20:54PM +0200, Baptiste Daroussin wrote:
 
  The pkg developement team is proud to announce the new 1.1.0 beta1 release 
  of
  pkg.
 
  - new experimental pkg convert (can convert from and to legacy pkg database)
pkg2ng now uses pkg convert (still recommanded to use pkg2ng)
 
 Converting packages from /var/db/pkg
 Converting pkg-1.1.0.b3_1...
 pkg: unknown keyword display, ignoring @display
 Installing pkg-1.1.0.b3_1...Segmentation fault (core dumped)
 
 ___
 freebsd-curr...@freebsd.org mailing list
 http://lists.freebsd.org/mailman/listinfo/freebsd-current
 To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org

Have you run pkg2ng?

regards,
Bapt


pgpKioBuIVSWS.pgp
Description: PGP signature

Re: [HEADSUP] New pkg-devel 1.1.0 beta1

[Baptiste Daroussin]

On Mon, Jun 03, 2013 at 07:39:03PM +0400, Slawa Olhovchenkov wrote:
 On Mon, Jun 03, 2013 at 05:34:19PM +0200, Baptiste Daroussin wrote:
 
  On Mon, Jun 03, 2013 at 07:17:24PM +0400, Slawa Olhovchenkov wrote:
   On Thu, May 30, 2013 at 05:20:54PM +0200, Baptiste Daroussin wrote:
   
The pkg developement team is proud to announce the new 1.1.0 beta1 
release of
pkg.
   
- new experimental pkg convert (can convert from and to legacy pkg 
database)
  pkg2ng now uses pkg convert (still recommanded to use pkg2ng)
   
   Converting packages from /var/db/pkg
   Converting pkg-1.1.0.b3_1...
   pkg: unknown keyword display, ignoring @display
   Installing pkg-1.1.0.b3_1...Segmentation fault (core dumped)
   
   ___
   freebsd-curr...@freebsd.org mailing list
   http://lists.freebsd.org/mailman/listinfo/freebsd-current
   To unsubscribe, send any mail to freebsd-current-unsubscr...@freebsd.org
  
  Have you run pkg2ng?
 
 Yes, this is run pkg2ng.

Ok I'll have a look and fix asap.

regards,
Bapt


pgpqoZlMsKLc4.pgp
Description: PGP signature

[HEADSUP] New pkg-devel 1.1.0 beta1

[Baptiste Daroussin]

Hi,

The pkg developement team is proud to announce the new 1.1.0 beta1 release of
pkg.

Here is the list of new features that happened in pkg 1.1:
- new simpler and more reliable solver
- shared libraries are now always tracked
- ssh:// is supported as a protocol to distribute packages (needs pkg 1.1+ on
  the server hosting the packages)
- multirepository is no longer considered experimental and works by default.
- incremental update of the catalog (only if the repository was created by pkg
  1.1+)
- simplification of the public API
- stabilisation of the public API (we will now try to keep it stable and if
  change are needed there will be deprecation time before removal of some old
  functions)
- new experimental pkg convert (can convert from and to legacy pkg database)
  pkg2ng now uses pkg convert (still recommanded to use pkg2ng)
- new pkg lock/unlock to prevent any manipulation of a given package (no
  upgrade,delete,etc)
- improved UI (now you can see the progress of an upgrade what is left to be
  done)
- new pkg annotation to allow one to add annotations (free form key/value) to a
  package)
- pkg audit is now able to directly parse the vuxml native format and not only
  the compact version
- pkg -vv now shows all available options and their current settings
- pkg -vvv now shows a description of all the available options
- pkg info now automatically considers the query as globbing if * is in the
  requested pattern
- new hook plugin interface (allows users to create hooks that get called at
  anytime during and upgrade/installation/deletion of a package)
- new cmd plugin interface (allows users to create new sub command available for
  pkg)
- pkg register can now register a port installation in the legacy database
  format
- repository can be defined in simple yaml files

Internal:
- massive usage of hash tables (uthash), which simplifies a lot of the code,
  and improves performances
- lots of optimisation in plist and manifest parsing
- lots of optimisation in loading packages (mmap used when possible)
- lots of cleanup in memory usage
- regression test framework is now ready (using atf) regression test are slowly
  being added and populated.

To use this new version:
  Ports users (or in building factories: poudriere/tinderbox):
Add WITH_PKGNG=devel to your make.conf
pkg set -o ports-mgmt/pkg:ports-mgmt/pkg-devel

  Binary package users, if the remote repository is providing pkg 1.1:
pkg set -o ports-mgmt/pkg:ports-mgmt/pkg-devel
pkg upgrade


Note that pkg 1.1 can use a repository created for pkg 1.0 and vis versa.

Huge thanks to all the people that have contributed to the pkg developement: 
- may that be by code
- documentation
- bug report
- feedback
- ideas

List of people who contributed code:
Baptiste Daroussin, Matthew Seaman, Bryan Drewery, Vsevolod Stakhov,
Marin Atanasov Nikolov, Alexandre Perrin, Romain Tartière, Julien Laffaye,
Glen Barber, John Marino, Alex Kozlov, Roman Naumann, Sofian Brabez,
Alberto Villa, Will Andrews, Eitan Adler, Dan McGregor, namor, niamtokik,
Arthur Gautier, Garrett Cooper, Andrew Turner, Jeremy Chadwick,
Hajimu UMEMOTO, Mark Lokowich, Eygene Ryabinkin, Pietro Cerutti,
Rolf Grossmann, Ed Schouten, Dimitry Andric, David Forsythe, Stefan Grundmann,
Craig Rodrigues, Antoine Brodin, Andrey Zonov, Joel Dahl

Stats between 1.0 and 1.1:
287 files changed, 63418 insertions(+), 18763 deletions(-)

1198 commits

regards,
Bapt


pgp5ZK50yeDyl.pgp
Description: PGP signature

Re: Empty @cwd in some ports

[Baptiste Daroussin]

On Wed, Mar 20, 2013 at 09:13:08PM +1030, Daniel O'Connor wrote:
 I noticed I have a lot of ports which have an empty @cwd line in the 
 +CONTENTS file (even after reinstalling).
 
 For example a2ps..
 @comment PKG_FORMAT_REVISION:1.1
 @name a2ps-a4-4.13b_4
 @comment ORIGIN:print/a2ps-a4
 @cwd /usr/local
 @pkgdep xineramaproto-1.2.1
 @comment DEPORIGIN:x11/xineramaproto
 snip
 @exec /sbin/ldconfig -m /usr/local/lib
 @unexec /sbin/ldconfig -R
 @comment OPTIONS:+NLS -I18N +EMACS
 @cwd 
 @dirrm share/licenses/a2ps-a4-4.13b_4
 
 And curl..
 @comment PKG_FORMAT_REVISION:1.1
 @name curl-7.24.0_2
 @comment ORIGIN:ftp/curl
 @cwd /usr/local
 @pkgdep ca_root_nss-3.14.3
 @comment DEPORIGIN:security/ca_root_nss
 snip
 @comment OPTIONS:-CARES -CURL_DEBUG -GNUTLS +IPV6 -KERBEROS4 -LDAP -LDAPS 
 -LIBIDN -LIBSSH2 -NTLM +OPENSSL +CA_BUNDLE +PROXY -RTMP
 @comment OPTIONS:-TRACKMEMORY
 @cwd 
 @dirrm share/licenses/curl-7.24.0_2
 @unexec rmdir %D/share/licenses 2/dev/null || true
 
 This seems pretty broken but I am not sure what causes it :(
 [midget 21:09] /var/db/pkg egrep -l '@cwd $' */+CONTENTS| wc -l
  300
 
 --
 Daniel O'Connor software and network engineer
 for Genesis Software - http://www.gsoft.com.au
 The nice thing about standards is that there
 are so many of them to choose from.
   -- Andrew Tanenbaum
 GPG Fingerprint - 5596 B766 97C0 0E94 4347 295E E593 DC20 7B3F CE8C
 
 
 
 
 
 
 ___
 freebsd-stable@freebsd.org mailing list
 http://lists.freebsd.org/mailman/listinfo/freebsd-stable
 To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org

Empty cwd is normal, it is equivalent to @cwd %%PREFIX%%

regards,
Bapt


pgpBq5HRWAoxm.pgp
Description: PGP signature

Re: Why does poudriere always rebuild nginx and GraphicsMagick13?

[Baptiste Daroussin]

On Tue, Feb 12, 2013 at 10:59:28PM +0100, Rainer Duffner wrote:
 Hi,
 
 poudriere 2.2 here, running on 9.1-amd64
 
 Of the 730-ish ports, whenever I run a build, it always rebuilds the above 
 two ports.
 Even if nothing changed.
 
 Is there are specific reason for this?
 
 I don't really mind nginx, because it builds so quickly - but GraphicsMagick 
 takes a bit too long.

2.2 is so old :)
please upgrade to at least 2.3.1, lots of things as been fixed since, and soon
2.3.2 with lots of other bug fixes.

I'm sure 2.3.1 fixes your problems, or at least will explain you why something
is rebuilt (it is now explained during the sanity check)

regards,
Bapt


pgpjbBfVNxMOW.pgp
Description: PGP signature

Re: pkgng and updated packages

[Baptiste Daroussin]

On Mon, Jan 28, 2013 at 11:51:20AM +0100, Rainer Duffner wrote:
 Hi,
 
 I run my own pkgng repo (via poudriere) and it seems to work great.
 However, I've got more questions...

Glad to hear it works well :)

 
  - when PHP got updated to 5.3.21 (from 5.3.20), all the other stuff
that depends on it (PECL-*, pear-*) got rebuilt, too.
(In my ports-tree, I just imported the update of PHP and nothing
else)
  - but on the client, when I ran pkg upgrade, only the php-packages got
upgraded. Is that correct? Shouldn't it also update the PECL and pear
packages?

What happen is that poudriere always want to make sure that you have sane
packages, so what it does is rebuilding everything that depends on something
that has been upgraded, so that it always have working packages and do not fall
into the pcre/png/$pickyourfavorite problem when shlib are bumped and not all
the need revision bump has be done.

pkg itself does not care about that and trust the revision number to only what
as changed is really updated. So force updating everything that depends on php a
simple pkg install -fR php should do it correctly for you.

  - how do I re-install all packages? I will soon upgrade some systems
from 8.3 to 9.1 and I want to reduce the number of systems I build
packages for. Before pkgng, I would just pkg_delete everything then
re-install and do a diff on the output of pkg_info |cut -f 1 -d  .
I though there might be an easier way...

pkg upgrade -f will upgrade everything. to avoid you to shoot yourself when
upgrading from 8.3 to 9.1 we also provide a pkg-static binary so that even if
pkg itself is linked against libraries that has disappeared in 9.1 you can still
reisntall everything.

  - is there a make.conf switch to completely disable pkg_* or do I have
to remove the binaries myself? Will they re-appear after a
freebsd-update -r or even after a simple freebsd-update? If I install
postfix, there is a switch in rc.conf do disable it completely and I
think there should be something similar.

WITHOUT_PKGTOOLS will remove all the pkg_* tools if you are building world, but
nothing like this in binary upgrade like freebsd-update what you can do is
manually rm /usr/sbin/pkg_*

regards,
Bapt


pgpklMaQ9icmN.pgp
Description: PGP signature

Re: Will we get a RELEASE-9.1 for Christmas?

[Baptiste Daroussin]

On Fri, Dec 21, 2012 at 09:55:00AM -0600, ajtiM wrote:
 On Friday 21 December 2012 08:12:49 Walter Hurry wrote:
  On Thu, 20 Dec 2012 23:40:19 +, Matthew Seaman wrote:
   I've just finished committing fixes for this problem to the master and
   release-1.0 branches in Github.  If anyone is feeling enthusiastic and
   would like to try pulling those sources and confirming that their
   favourite packages no longer generate scads of nasty errors when
   re-analysing shlib dependencies, I'd be grateful.
  
  Before I had a chance to try the git version, pkg-1.0.4 turned up in the
  ports. Good work!
  
  Hmm...
  
  $ pkg info pkg
  pkg-1.0.4  New generation package manager
  $ pkg -v
  1.0.3
  $
  
  Forgotten (by bapt I assume) again?
  
  A definite improvement in 'pkg ckeck -Ba' though; many fewer reported
  exceptions than before (openjdk7 and firefox are OK now):
  
  $ sudo pkg check -Ba
  pkg: (libreoffice-3.5.7_1) /usr/local/lib/libreoffice/program/
  configmgr.uno.so - shared library libxmlreader.so not found
  pkg: (virtualbox-ose-4.1.22) /usr/local/bin/VBoxBalloonCtrl - shared
  library VBoxRT.so not found
  other virtualbox-ose exceptions snipped for brevity
  
  $ locate libxmlreader.so
  /usr/local/lib/libreoffice/ure/lib/libxmlreader.so
  $ locate VBoxRT.so
  /usr/local/lib/virtualbox/VBoxRT.so
  $ file /usr/local/lib/libreoffice/ure/lib/libxmlreader.so
  /usr/local/lib/libreoffice/ure/lib/libxmlreader.so: ELF 64-bit LSB shared
  object, x86-64, version 1 (FreeBSD), dynamically linked, stripped
  $ file /usr/local/lib/virtualbox/VBoxRT.so
  /usr/local/lib/virtualbox/VBoxRT.so: ELF 64-bit LSB shared object,
  x86-64, version 1 (FreeBSD), dynamically linked, not stripped
  $
  
  ___
  freebsd-stable@freebsd.org mailing list
  http://lists.freebsd.org/mailman/listinfo/freebsd-stable
  To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
 
 I try to update but I am not sure how. I am usin FreeBSD 9.1 RC3 and I have 
 in 
 my make.conf WITH_PKGNG=yes and it works (I did pkg2ng long time ago). I use 
 portmaster. If I put as I got after error of update:
 WITHOUT_PKGNG=yes
 
 - switch to pkgng:
 1) Add WITHOUT_PKGNG to /etc/make.conf
 2) Install ports-mgmt/pkg
 3) Convert your package database by running pkg2ng
 4) Remove WITHOUT_PKGNG from /etc/make.conf
 
 *** [pre-everything] Error code 1
 
 Stop in /usr/ports/ports-mgmt/pkg.

That is the seat belt that means that you have WITH_PKGNG but you also have an
incomplete conversion to pkgng

It tests if there are any old package installed, and should leave you alone if
you have completly and cleanly switch to pkgng.

That should have been added long ago.

regards,
Bapt


pgpYH9EbNPtke.pgp
Description: PGP signature

Re: Will we get a RELEASE-9.1 for Christmas?

[Baptiste Daroussin]

On Fri, Dec 21, 2012 at 11:53:16AM -0600, ajtiM wrote:
 On Friday 21 December 2012 11:26:19 Baptiste Daroussin wrote:
  On Fri, Dec 21, 2012 at 09:55:00AM -0600, ajtiM wrote:
   On Friday 21 December 2012 08:12:49 Walter Hurry wrote:
On Thu, 20 Dec 2012 23:40:19 +, Matthew Seaman wrote:
 I've just finished committing fixes for this problem to the master
 and release-1.0 branches in Github.  If anyone is feeling
 enthusiastic and would like to try pulling those sources and
 confirming that their favourite packages no longer generate scads of
 nasty errors when re-analysing shlib dependencies, I'd be grateful.

Before I had a chance to try the git version, pkg-1.0.4 turned up in
the ports. Good work!

Hmm...

$ pkg info pkg
pkg-1.0.4  New generation package manager
$ pkg -v
1.0.3
$

Forgotten (by bapt I assume) again?

A definite improvement in 'pkg ckeck -Ba' though; many fewer reported
exceptions than before (openjdk7 and firefox are OK now):

$ sudo pkg check -Ba
pkg: (libreoffice-3.5.7_1) /usr/local/lib/libreoffice/program/
configmgr.uno.so - shared library libxmlreader.so not found
pkg: (virtualbox-ose-4.1.22) /usr/local/bin/VBoxBalloonCtrl - shared
library VBoxRT.so not found
other virtualbox-ose exceptions snipped for brevity

$ locate libxmlreader.so
/usr/local/lib/libreoffice/ure/lib/libxmlreader.so
$ locate VBoxRT.so
/usr/local/lib/virtualbox/VBoxRT.so
$ file /usr/local/lib/libreoffice/ure/lib/libxmlreader.so
/usr/local/lib/libreoffice/ure/lib/libxmlreader.so: ELF 64-bit LSB
shared object, x86-64, version 1 (FreeBSD), dynamically linked,
stripped $ file /usr/local/lib/virtualbox/VBoxRT.so
/usr/local/lib/virtualbox/VBoxRT.so: ELF 64-bit LSB shared object,
x86-64, version 1 (FreeBSD), dynamically linked, not stripped
$

___
freebsd-stable@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-stable
To unsubscribe, send any mail to
freebsd-stable-unsubscr...@freebsd.org
   
   I try to update but I am not sure how. I am usin FreeBSD 9.1 RC3 and I
   have in my make.conf WITH_PKGNG=yes and it works (I did pkg2ng long time
   ago). I use portmaster. If I put as I got after error of update:
   WITHOUT_PKGNG=yes
   
   - switch to pkgng:
   1) Add WITHOUT_PKGNG to /etc/make.conf
   2) Install ports-mgmt/pkg
   3) Convert your package database by running pkg2ng
   4) Remove WITHOUT_PKGNG from /etc/make.conf
   
   *** [pre-everything] Error code 1
   
   Stop in /usr/ports/ports-mgmt/pkg.
  
  That is the seat belt that means that you have WITH_PKGNG but you also have
  an incomplete conversion to pkgng
  
  It tests if there are any old package installed, and should leave you alone
  if you have completly and cleanly switch to pkgng.
  
  That should have been added long ago.
  
  regards,
  Bapt
 
 And how can I find it if was not okay? When I ran pkg2ng was everythin okay 
 (looks like, no errors).
 I will run pkg2ng again. 
 Do I need to keep WITH_PKGNG=yes in make.conf as was instructed, please?

What says pkg_info ?

regards,
Bapt


pgpYBnWj1u4Vd.pgp
Description: PGP signature

Re: Will we get a RELEASE-9.1 for Christmas?

[Baptiste Daroussin]

On Fri, Dec 21, 2012 at 11:53:16AM -0600, ajtiM wrote:
 On Friday 21 December 2012 11:26:19 Baptiste Daroussin wrote:
  On Fri, Dec 21, 2012 at 09:55:00AM -0600, ajtiM wrote:
   On Friday 21 December 2012 08:12:49 Walter Hurry wrote:
On Thu, 20 Dec 2012 23:40:19 +, Matthew Seaman wrote:
 I've just finished committing fixes for this problem to the master
 and release-1.0 branches in Github.  If anyone is feeling
 enthusiastic and would like to try pulling those sources and
 confirming that their favourite packages no longer generate scads of
 nasty errors when re-analysing shlib dependencies, I'd be grateful.

Before I had a chance to try the git version, pkg-1.0.4 turned up in
the ports. Good work!

Hmm...

$ pkg info pkg
pkg-1.0.4  New generation package manager
$ pkg -v
1.0.3
$

Forgotten (by bapt I assume) again?

A definite improvement in 'pkg ckeck -Ba' though; many fewer reported
exceptions than before (openjdk7 and firefox are OK now):

$ sudo pkg check -Ba
pkg: (libreoffice-3.5.7_1) /usr/local/lib/libreoffice/program/
configmgr.uno.so - shared library libxmlreader.so not found
pkg: (virtualbox-ose-4.1.22) /usr/local/bin/VBoxBalloonCtrl - shared
library VBoxRT.so not found
other virtualbox-ose exceptions snipped for brevity

$ locate libxmlreader.so
/usr/local/lib/libreoffice/ure/lib/libxmlreader.so
$ locate VBoxRT.so
/usr/local/lib/virtualbox/VBoxRT.so
$ file /usr/local/lib/libreoffice/ure/lib/libxmlreader.so
/usr/local/lib/libreoffice/ure/lib/libxmlreader.so: ELF 64-bit LSB
shared object, x86-64, version 1 (FreeBSD), dynamically linked,
stripped $ file /usr/local/lib/virtualbox/VBoxRT.so
/usr/local/lib/virtualbox/VBoxRT.so: ELF 64-bit LSB shared object,
x86-64, version 1 (FreeBSD), dynamically linked, not stripped
$

___
freebsd-stable@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-stable
To unsubscribe, send any mail to
freebsd-stable-unsubscr...@freebsd.org
   
   I try to update but I am not sure how. I am usin FreeBSD 9.1 RC3 and I
   have in my make.conf WITH_PKGNG=yes and it works (I did pkg2ng long time
   ago). I use portmaster. If I put as I got after error of update:
   WITHOUT_PKGNG=yes
   
   - switch to pkgng:
   1) Add WITHOUT_PKGNG to /etc/make.conf
   2) Install ports-mgmt/pkg
   3) Convert your package database by running pkg2ng
   4) Remove WITHOUT_PKGNG from /etc/make.conf
   
   *** [pre-everything] Error code 1
   
   Stop in /usr/ports/ports-mgmt/pkg.
  
  That is the seat belt that means that you have WITH_PKGNG but you also have
  an incomplete conversion to pkgng
  
  It tests if there are any old package installed, and should leave you alone
  if you have completly and cleanly switch to pkgng.
  
  That should have been added long ago.
  
  regards,
  Bapt
 
 And how can I find it if was not okay? When I ran pkg2ng was everythin okay 
 (looks like, no errors).
 I will run pkg2ng again. 
 Do I need to keep WITH_PKGNG=yes in make.conf as was instructed, please?
 
 Mitja
 
 http://www.redbubble.com/people/lumiwa
 ___
 freebsd-stable@freebsd.org mailing list
 http://lists.freebsd.org/mailman/listinfo/freebsd-stable
 To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org

Ok this is a problem with the seat belt and portmaster not with your setup.

I disabled the seat belt for now until we find a better way.

regards,
Bapt


pgpr3oEmX0uST.pgp
Description: PGP signature

Re: Will we get a RELEASE-9.1 for Christmas?

[Baptiste Daroussin]

On Fri, Dec 21, 2012 at 12:48:45PM -0600, ajtiM wrote:
 On Friday 21 December 2012 11:58:00 Baptiste Daroussin wrote:
  On Fri, Dec 21, 2012 at 11:53:16AM -0600, ajtiM wrote:
   On Friday 21 December 2012 11:26:19 Baptiste Daroussin wrote:
On Fri, Dec 21, 2012 at 09:55:00AM -0600, ajtiM wrote:
 On Friday 21 December 2012 08:12:49 Walter Hurry wrote:
  On Thu, 20 Dec 2012 23:40:19 +, Matthew Seaman wrote:
   I've just finished committing fixes for this problem to the
   master and release-1.0 branches in Github.  If anyone is feeling
   enthusiastic and would like to try pulling those sources and
   confirming that their favourite packages no longer generate
   scads of nasty errors when re-analysing shlib dependencies, I'd
   be grateful.
  
  Before I had a chance to try the git version, pkg-1.0.4 turned up
  in the ports. Good work!
  
  Hmm...
  
  $ pkg info pkg
  pkg-1.0.4  New generation package manager
  $ pkg -v
  1.0.3
  $
  
  Forgotten (by bapt I assume) again?
  
  A definite improvement in 'pkg ckeck -Ba' though; many fewer
  reported exceptions than before (openjdk7 and firefox are OK now):
  
  $ sudo pkg check -Ba
  pkg: (libreoffice-3.5.7_1) /usr/local/lib/libreoffice/program/
  configmgr.uno.so - shared library libxmlreader.so not found
  pkg: (virtualbox-ose-4.1.22) /usr/local/bin/VBoxBalloonCtrl -
  shared library VBoxRT.so not found
  other virtualbox-ose exceptions snipped for brevity
  
  $ locate libxmlreader.so
  /usr/local/lib/libreoffice/ure/lib/libxmlreader.so
  $ locate VBoxRT.so
  /usr/local/lib/virtualbox/VBoxRT.so
  $ file /usr/local/lib/libreoffice/ure/lib/libxmlreader.so
  /usr/local/lib/libreoffice/ure/lib/libxmlreader.so: ELF 64-bit LSB
  shared object, x86-64, version 1 (FreeBSD), dynamically linked,
  stripped $ file /usr/local/lib/virtualbox/VBoxRT.so
  /usr/local/lib/virtualbox/VBoxRT.so: ELF 64-bit LSB shared object,
  x86-64, version 1 (FreeBSD), dynamically linked, not stripped
  $
  
  ___
  freebsd-stable@freebsd.org mailing list
  http://lists.freebsd.org/mailman/listinfo/freebsd-stable
  To unsubscribe, send any mail to
  freebsd-stable-unsubscr...@freebsd.org
 
 I try to update but I am not sure how. I am usin FreeBSD 9.1 RC3 and
 I have in my make.conf WITH_PKGNG=yes and it works (I did pkg2ng
 long time ago). I use portmaster. If I put as I got after error of
 update: WITHOUT_PKGNG=yes
 
 - switch to pkgng:
 1) Add WITHOUT_PKGNG to /etc/make.conf
 2) Install ports-mgmt/pkg
 3) Convert your package database by running pkg2ng
 4) Remove WITHOUT_PKGNG from /etc/make.conf
 
 *** [pre-everything] Error code 1
 
 Stop in /usr/ports/ports-mgmt/pkg.

That is the seat belt that means that you have WITH_PKGNG but you also
have an incomplete conversion to pkgng

It tests if there are any old package installed, and should leave you
alone if you have completly and cleanly switch to pkgng.

That should have been added long ago.

regards,
Bapt
   
   And how can I find it if was not okay? When I ran pkg2ng was everythin
   okay (looks like, no errors).
   I will run pkg2ng again.
   Do I need to keep WITH_PKGNG=yes in make.conf as was instructed, please?
  
  What says pkg_info ?
  
  regards,
  Bapt
 pkg_info shows many corrupted packages. For example:
 fontsproto-2.1.1Fonts extension headers
 foomatic-filters-4.0.7_1 Foomatic wrapper scripts
 fotoxx-11.03_1  Application to organize and edit image collections
 freeglut-2.8.0  An alternative to the OpenGL Utility Toolkit (GLUT) 
 library
 freetype-1.3.1_4A free and portable TrueType font rendering engine
 freetype-tools-1.3.1_9 Tools for FreeType 1
 freetype2-2.4.9_1   A free and portable TrueType font rendering engine
 frei0r-1.3_1Minimalistic plugin API for video effects
 fribidi-0.19.2_1A Free Implementation of the Unicode Bidirectional 
 Algorith
 pkg_info: the package info for package 'gamin-0.1.10_4' is corrupt
 gawk-4.0.1  The GNU version of Awk
 gcc-4.6.3   GNU Compiler Collection 4.6
 gccmakedep-1.0.2Create dependencies in makefiles using 'gcc -M'
 pkg_info: the package info for package 'gconf2-2.32.0_3' is corrupt
 gd-2.0.35_8,1   A graphics library for fast creation of images
 gdbm-1.9.1  The GNU database manager
 pkg_info: the package info for package 'gdk-pixbuf-2.23.5_3' is corrupt
 geeqie-1.1  Lightweight image viewer forked from GQview
 pkg_info: the package info for package 'gegl-0.1.8_4' is corrupt
 ...
 ...

So your conversion wasn't totally complete, for each of them (aka non corrupted)
test if they are in your pkgng database

pkg info gcc gawk etc

Re: nullfs changes MFC

[Baptiste Daroussin]

On Sat, Dec 08, 2012 at 12:35:42PM +0200, Andriy Gapon wrote:
 on 08/12/2012 05:06 Konstantin Belousov said the following:
  On Sat, Dec 08, 2012 at 03:58:16AM +0100, Baptiste Daroussin wrote:
  On Sat, Dec 08, 2012 at 03:01:09AM +0200, Konstantin Belousov wrote:
  ...
  The merge is available at 
  http://people.freebsd.org/~kib/misc/nullfs_9.1.patch
 
 
  Sorry I haven't checked the latest zfs related MFC, but for some time
  there was an issue with nullfs improvements and zfs, will this mfc be
  synchronize with the mfc of the related zfs fixes?
  
  Corresponding zfs fixes were already merged to stable/9, as I was told.
  Cc:ed Andrey to confirm it once more.
 
 Yes, a while ago.
 

Perfect thank you both! And sorry for noise :)

regards,
Bapt


pgpbYLSRsMLFP.pgp
Description: PGP signature

Re: nullfs changes MFC

[Baptiste Daroussin]

On Sat, Dec 08, 2012 at 03:01:09AM +0200, Konstantin Belousov wrote:
 Hi,
 I am going to merge latest batch of the nullfs improvements into
 stable/9. This will bring up significant performance enchancements due
 to use of the shared locks for lookups if the lower layer supports it,
 much better caching on the nullfs layer, and proper handling of the text
 segments on the nullfs. Also, it should improve the error recovery and
 some corner cases with locking.
 
 Unfortunately, the merge would break KBI for VFS, since it needs 5 new
 VOP slots, and only three spares are left. We already are very liberal
 with the VFS KBI, so I do not feel that the merge is not acceptable, due
 to the benefits it brings to the nullfs.
 
 The merge is available at 
 http://people.freebsd.org/~kib/misc/nullfs_9.1.patch


Sorry I haven't checked the latest zfs related MFC, but for some time there was
an issue with nullfs improvements and zfs, will this mfc be synchronize with the
mfc of the related zfs fixes?

regards,
Bapt


pgpZ7YTsbpo9F.pgp
Description: PGP signature

Re: pkgng (Was: Why Are You NOT Using FreeBSD ?)

[Baptiste Daroussin]

On Fri, Jun 01, 2012 at 04:09:53PM -0400, Chris Nehren wrote:
 On Fri, Jun 01, 2012 at 14:50:22 -0500 , Bryan Drewery wrote:
  FWIW, there is freebsd-update(8) now for binary updating of base, and
  pkgng[1] will allow binary upgrading of packages/ports similar to apt-get.
  
  [1] http://wiki.freebsd.org/pkgng
 
 The thing that really has me attracted to pkgng is that it's based on a
 C library with a public API that developers can use / abuse. It's not
 (AFAIK) officially released yet, but some early work I've been doing
 with it has shown it to be useful enough.
 
 -- 
 Thanks and best regards,
 Chris Nehren
 ___
 freebsd-stable@freebsd.org mailing list
 http://lists.freebsd.org/mailman/listinfo/freebsd-stable
 To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org

Do not hesitate to propose patches or bring feedback on how we can improve the
situation with library.

The API is public but will only be consider as stable before 2.0, because it can
be greatly improved and we hope to have reviews and feedback and cleanup before
marking it stable.

regards,
Bapt


pgpaZhs06qJxn.pgp
Description: PGP signature

Re: problem with chpass after recent libutil change

[Baptiste Daroussin]

On Wed, Jun 13, 2012 at 06:57:38PM +0200, Claude Buisson wrote:
 Hi,
 
 I have some 8-STABLE and 9-STABLE systems which have been upgraded from 
 sources
 csuped on June 3, and I found that I could no more change the shell for toor:
 
 #chpass -s /bin/csh toor
 chpass: entry inconsistent
 chpass: pw_copy: Invalid argument
 
 and seemingly ONLY for toor.
 
 reverting lib/lib/libutil/pw_util.c from r236451 (on 9.X) or r236452 (on 8.X)
 to the previous version seems to be workaround.
 
 Claude Buisson

Crap I'll work on a fix asap.

regards,
Bapt


pgp7TKGicWM0u.pgp
Description: PGP signature

Re: problem with chpass after recent libutil change

[Baptiste Daroussin]

On Wed, Jun 13, 2012 at 06:57:38PM +0200, Claude Buisson wrote:
 Hi,
 
 I have some 8-STABLE and 9-STABLE systems which have been upgraded from 
 sources
 csuped on June 3, and I found that I could no more change the shell for toor:
 
 #chpass -s /bin/csh toor
 chpass: entry inconsistent
 chpass: pw_copy: Invalid argument
 
 and seemingly ONLY for toor.
 
 reverting lib/lib/libutil/pw_util.c from r236451 (on 9.X) or r236452 (on 8.X)
 to the previous version seems to be workaround.
 
 Claude Buisson

Can you try with the following patch?

http://people.freebsd.org/~bapt/add-special-case-for-toor.diff

Regards,
Bapt


pgpzug7i7qFFB.pgp
Description: PGP signature

Re: Why Are You NOT Using FreeBSD?

[Baptiste Daroussin]

 Look at the comment of the maintainer of LibreOffice ...

btw I tested and libreoffice is still working as expected.

Most of the time the libreoffice failures is from people tuning their system
without knowing the impact/risk of doing such, for example building some c++
libraries with g++47 let's imagine cppunit or anyother library depended on by
libreoffice, and building libreoffice with clang (which is default) and using
the libstdc++ from base (which also is default) and the mix of libstdc++ is
producing tons of problems.

problem i can't fix.

regards,
Bapt


pgp9XqKXJF7QQ.pgp
Description: PGP signature

Re: dhclient script adjustments

[Baptiste Daroussin]

On Tue, Feb 14, 2012 at 02:47:00AM -0500, Jason Hellenthal wrote:
 
 Anyone ?
 

Sorry for mess, I'm working on this to figure out why it does that.

Thanks for reporting,

regards,
Bapt


pgp2hdjzuz7zb.pgp
Description: PGP signature

shouldn't libarchive be MFCed before release?

[Baptiste Daroussin]

Hi all,

I'm working on suport xz compressed packages for pkg_install, and
libarchive on current supports natively xz compression but seems not
having been MFCed (it should have beed two weeks after the 10th of may
according to the commit message) I still have 2.7.0 on stable (from
yesterday)

I send this mail just to know if libarchive will be mfced on not
before the release.

regards,
Bapt
___
freebsd-stable@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-stable
To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org