NetBSD source update | 1 May 2012 05:09
Picon

daily CVS update output


Updating src tree:
P src/distrib/sets/lists/tests/mi
P src/distrib/sets/lists/tests/module.mi
P src/distrib/utils/sysinst/configmenu.c
P src/distrib/utils/sysinst/msg.mi.de
P src/share/man/man5/login.conf.5
P src/sys/compat/netbsd32/netbsd32_execve.c
P src/sys/dev/usb/ums.c
P src/sys/fs/union/union_vfsops.c
P src/sys/kern/kern_exec.c
P src/sys/kern/kern_malloc.c
P src/sys/kern/vfs_syscalls.c
P src/sys/miscfs/nullfs/null_vfsops.c
P src/sys/miscfs/overlay/overlay_vfsops.c
P src/sys/miscfs/procfs/procfs_vfsops.c
P src/sys/miscfs/umapfs/umap_vfsops.c
P src/sys/netinet/ip_mroute.c
P src/sys/netinet/ip_output.c
P src/sys/sys/cdefs.h
P src/sys/sys/exec.h
P src/sys/sys/malloc.h
P src/sys/sys/spawn.h
P src/sys/ufs/chfs/chfs_vfsops.c
P src/sys/ufs/ext2fs/ext2fs_vfsops.c
P src/sys/ufs/lfs/lfs_vfsops.c

Updating xsrc tree:

Killing core files:
(Continue reading)

Toru Nishimura | 1 May 2012 06:05

dhcpcd(8) and automatic hostname

Hi,

I've switched dhclient(8) to dhcpcd(8).  The transition was flawless
except of the issue.   My /etc/rc.conf does not set hostname variable to 
let it determined by dhclient(8).  It worked well for me so far, but
dhcpcd(8) does not any more.  I would like to have the feature to be
added in furture release.

Toru Nishimura /  ALKYL Technology

Paul Goyette | 1 May 2012 15:02

Strange test-bed panic - subr_pool (amd64)

Not sure if this is repeatable, but my home amd64 test-bed just had the 
following panic while running the ATF tests.  This is with sources that 
were checked out at 2012-05-01 01:20:04 UTC

lib/libc/gen/posix_spawn/t_fileactions (211/515): 6 test cases
     t_spawn_empty_fileactions: [0.905798s] Passed.
     t_spawn_fileactions: [0.060030s] Passed.
     t_spawn_open_nonexistent: [0.043759s] Passed.
     t_spawn_open_nonexistent_diag: [0.043292s] Passed.
     t_spawn_openmode: panic: kernel diagnostic assertion "object != 
NULL" failed: file "/test-bed/src/sys/kern/subr_pool.c", line 2611
cpu0: Begin traceback...
kern_assert() at netbsd:kern_assert+0x48
pool_cache_put_paddr() at netbsd:pool_cache_put_paddr+0xb5
pathbuf_destroy() at netbsd:pathbuf_destroy+0x30
do_open() at netbsd:do_open+0x12a
sys_open() at netbsd:sys_open+0x59
syscall() at netbsd:syscall+0xac
cpu0: End traceback...

-------------------------------------------------------------------------
| Paul Goyette     | PGP Key fingerprint:     | E-mail addresses:       |
| Customer Service | FA29 0E3B 35AF E8AE 6651 | paul at whooppee.com    |
| Network Engineer | 0786 F758 55DE 53BA 7731 | pgoyette at juniper.net |
| Kernel Developer |                          | pgoyette at netbsd.org  |
-------------------------------------------------------------------------

Paul Goyette | 1 May 2012 15:59

Re: Strange test-bed panic - subr_pool (amd64)

This is now repeatable - a second pass on the test bed has failed with 
the exact same back-trace.

On Tue, 1 May 2012, Paul Goyette wrote:

> Not sure if this is repeatable, but my home amd64 test-bed just had the 
> following panic while running the ATF tests.  This is with sources that were 
> checked out at 2012-05-01 01:20:04 UTC
>
>
>
> lib/libc/gen/posix_spawn/t_fileactions (211/515): 6 test cases
>    t_spawn_empty_fileactions: [0.905798s] Passed.
>    t_spawn_fileactions: [0.060030s] Passed.
>    t_spawn_open_nonexistent: [0.043759s] Passed.
>    t_spawn_open_nonexistent_diag: [0.043292s] Passed.
>    t_spawn_openmode: panic: kernel diagnostic assertion "object != NULL" 
> failed: file "/test-bed/src/sys/kern/subr_pool.c", line 2611
> cpu0: Begin traceback...
> kern_assert() at netbsd:kern_assert+0x48
> pool_cache_put_paddr() at netbsd:pool_cache_put_paddr+0xb5
> pathbuf_destroy() at netbsd:pathbuf_destroy+0x30
> do_open() at netbsd:do_open+0x12a
> sys_open() at netbsd:sys_open+0x59
> syscall() at netbsd:syscall+0xac
> cpu0: End traceback...
>
> -------------------------------------------------------------------------
> | Paul Goyette     | PGP Key fingerprint:     | E-mail addresses:       |
> | Customer Service | FA29 0E3B 35AF E8AE 6651 | paul at whooppee.com    |
(Continue reading)

gson | 1 May 2012 16:03

Re: Strange test-bed panic - subr_pool (amd64)

Paul Goyette wrote:
> Not sure if this is repeatable, but my home amd64 test-bed just had the 
> following panic while running the ATF tests. [...]
> 
>      t_spawn_openmode: panic: kernel diagnostic assertion "object != 
> NULL" failed: file "/test-bed/src/sys/kern/subr_pool.c", line 2611

The TNF automated test server has the cause narrowed down to the
following set of commits:

  commit 2012.04.30.21.19.58 rmind src/sys/compat/netbsd32/netbsd32_execve.c 1.35
  commit 2012.04.30.21.19.58 rmind src/sys/kern/kern_exec.c 1.351
  commit 2012.04.30.21.19.58 rmind src/sys/sys/exec.h 1.136
  commit 2012.04.30.21.19.58 rmind src/sys/sys/spawn.h 1.3
  commit 2012.04.30.22.51.27 rmind src/sys/fs/union/union_vfsops.c 1.68
  commit 2012.04.30.22.51.27 rmind src/sys/kern/kern_malloc.c 1.142
  commit 2012.04.30.22.51.27 rmind src/sys/miscfs/nullfs/null_vfsops.c 1.84
  commit 2012.04.30.22.51.27 rmind src/sys/miscfs/overlay/overlay_vfsops.c 1.57
  commit 2012.04.30.22.51.27 rmind src/sys/sys/malloc.h 1.114
  commit 2012.04.30.22.51.28 rmind src/sys/miscfs/procfs/procfs_vfsops.c 1.87
  commit 2012.04.30.22.51.28 rmind src/sys/miscfs/umapfs/umap_vfsops.c 1.88
  commit 2012.04.30.22.51.28 rmind src/sys/netinet/ip_mroute.c 1.124
  commit 2012.04.30.22.51.28 rmind src/sys/netinet/ip_output.c 1.215
  commit 2012.04.30.22.51.28 rmind src/sys/ufs/chfs/chfs_vfsops.c 1.4
  commit 2012.04.30.22.51.28 rmind src/sys/ufs/ext2fs/ext2fs_vfsops.c 1.164
  commit 2012.04.30.22.51.28 rmind src/sys/ufs/lfs/lfs_vfsops.c 1.296

--

-- 
Andreas Gustafsson, gson <at> gson.org

(Continue reading)

Martin Husemann | 1 May 2012 17:04
Picon

Re: Strange test-bed panic - subr_pool (amd64)

On Tue, May 01, 2012 at 05:03:31PM +0300, gson <at> gson.org wrote:
> The TNF automated test server has the cause narrowed down to the
> following set of commits:

The interesting question now is, wether it was this batch:

>   commit 2012.04.30.21.19.58 rmind src/sys/compat/netbsd32/netbsd32_execve.c 1.35
>   commit 2012.04.30.21.19.58 rmind src/sys/kern/kern_exec.c 1.351
>   commit 2012.04.30.21.19.58 rmind src/sys/sys/exec.h 1.136
>   commit 2012.04.30.21.19.58 rmind src/sys/sys/spawn.h 1.3

Or the supposedly unrelated rest:

>   commit 2012.04.30.22.51.27 rmind src/sys/fs/union/union_vfsops.c 1.68
>   commit 2012.04.30.22.51.27 rmind src/sys/kern/kern_malloc.c 1.142
>   commit 2012.04.30.22.51.27 rmind src/sys/miscfs/nullfs/null_vfsops.c 1.84
>   commit 2012.04.30.22.51.27 rmind src/sys/miscfs/overlay/overlay_vfsops.c 1.57
>   commit 2012.04.30.22.51.27 rmind src/sys/sys/malloc.h 1.114
>   commit 2012.04.30.22.51.28 rmind src/sys/miscfs/procfs/procfs_vfsops.c 1.87
>   commit 2012.04.30.22.51.28 rmind src/sys/miscfs/umapfs/umap_vfsops.c 1.88
>   commit 2012.04.30.22.51.28 rmind src/sys/netinet/ip_mroute.c 1.124
>   commit 2012.04.30.22.51.28 rmind src/sys/netinet/ip_output.c 1.215
>   commit 2012.04.30.22.51.28 rmind src/sys/ufs/chfs/chfs_vfsops.c 1.4
>   commit 2012.04.30.22.51.28 rmind src/sys/ufs/ext2fs/ext2fs_vfsops.c 1.164
>   commit 2012.04.30.22.51.28 rmind src/sys/ufs/lfs/lfs_vfsops.c 1.296

Martin

Andreas Gustafsson | 1 May 2012 18:06

Re: Strange test-bed panic - subr_pool (amd64)

Martin Husemann wrote:
> The interesting question now is, wether it was this batch:
> 
> >   commit 2012.04.30.21.19.58 rmind src/sys/compat/netbsd32/netbsd32_execve.c 1.35
> >   commit 2012.04.30.21.19.58 rmind src/sys/kern/kern_exec.c 1.351
> >   commit 2012.04.30.21.19.58 rmind src/sys/sys/exec.h 1.136
> >   commit 2012.04.30.21.19.58 rmind src/sys/sys/spawn.h 1.3

Yes - the regression has now been narrowed down to that first batch:

  http://releng.netbsd.org/b5reports/i386/commits-2012.04.html#2012.04.30.21.19.58

--

-- 
Andreas Gustafsson, gson <at> gson.org

Roy Marples | 1 May 2012 20:12
Favicon
Gravatar

Re: dhcpcd(8) and automatic hostname

Hi

On 01/05/2012 5:05, Toru Nishimura wrote:
> I've switched dhclient(8) to dhcpcd(8).  The transition was flawless
> except of the issue.   My /etc/rc.conf does not set hostname variable
> to let it determined by dhclient(8).  It worked well for me so far,
> but
> dhcpcd(8) does not any more.  I would like to have the feature to be
> added in furture release.

dhcpcd(8) requests the DHCP host_name option by default in 
dhcpcd.conf(5).
The assignment of the hostname follows a similar ruleset as dhclient 
and can be found in /libexec/dhcpcd-hooks/30-hostname

Depending on the dhcpcd version you have installed, you can query the 
lease assigned by so

dhcpcd -U bge0

Replace bge0 with the interface dhcpcd worked on. It should show all 
the options received by dhcpcd that it can translate into useable shell 
variables.
This will include the hostname.

Thanks

Roy

(Continue reading)

Toru Nishimura | 2 May 2012 00:42

Re: dhcpcd(8) and automatic hostname

Roy Marples said;

> ...
> dhcpcd(8) requests the DHCP host_name option by default in 
> dhcpcd.conf(5).
> The assignment of the hostname follows a similar ruleset as dhclient 
> and can be found in /libexec/dhcpcd-hooks/30-hostname
> 
> Depending on the dhcpcd version you have installed, you can query the 
> lease assigned by so
> 
> dhcpcd -U bge0

Thanks for the detailed reply.  I should write my message clearer.  The issue
resides in /etc/rc.d/ structure, not dhcpcd(8) functionality itself.

/// short summary ///
When /etc/rc.conf does not have hostname= field, having
ifconfig_fxp0="dhcp" fails to assign hostname.

/// details ///
When /etc/rc.conf has dhclient="YES" and dhclient_flags="fxp0", dhclient
can assign hostname provided by DHCP server.  If /etc/rc.conf has
ifconfig_fxp0="dhcp" replacing the above two lines, /etc/rc emits the
following error message and the machine ends up with no hostname.

Starting Network.
/etc/rc: WARNING: $hostname not set.
...
Configuring network interface: fxp0
(Continue reading)

NetBSD source update | 2 May 2012 05:22
Picon

daily CVS update output


Updating src tree:
P src/crypto/external/bsd/openssh/dist/PROTOCOL.mux
P src/crypto/external/bsd/openssh/dist/auth-options.c
P src/crypto/external/bsd/openssh/dist/auth2-pubkey.c
P src/crypto/external/bsd/openssh/dist/auth2.c
P src/crypto/external/bsd/openssh/dist/authfile.c
P src/crypto/external/bsd/openssh/dist/channels.c
P src/crypto/external/bsd/openssh/dist/channels.h
P src/crypto/external/bsd/openssh/dist/clientloop.c
P src/crypto/external/bsd/openssh/dist/clientloop.h
P src/crypto/external/bsd/openssh/dist/compat.c
P src/crypto/external/bsd/openssh/dist/compat.h
P src/crypto/external/bsd/openssh/dist/dh.c
P src/crypto/external/bsd/openssh/dist/key.c
P src/crypto/external/bsd/openssh/dist/mac.c
P src/crypto/external/bsd/openssh/dist/misc.c
P src/crypto/external/bsd/openssh/dist/moduli.c
P src/crypto/external/bsd/openssh/dist/monitor.c
P src/crypto/external/bsd/openssh/dist/mux.c
P src/crypto/external/bsd/openssh/dist/packet.c
P src/crypto/external/bsd/openssh/dist/packet.h
P src/crypto/external/bsd/openssh/dist/readconf.c
P src/crypto/external/bsd/openssh/dist/readconf.h
P src/crypto/external/bsd/openssh/dist/roaming.h
P src/crypto/external/bsd/openssh/dist/roaming_client.c
P src/crypto/external/bsd/openssh/dist/roaming_common.c
P src/crypto/external/bsd/openssh/dist/scp.1
P src/crypto/external/bsd/openssh/dist/scp.c
P src/crypto/external/bsd/openssh/dist/servconf.c
(Continue reading)


Gmane