Issue with pkg upgrade on diskless workstation

[BERTRAND Joël]

BERTRAND Joël a écrit :
> Mark Linimon a écrit :
>> I am running a powerpc64 machine diskless but only with some awful hacks.
>> I can make them available if need be, but I hope that someone else has a
>> better recommendation for you.
>>
>> mcl
>>
>
> Before the last pkg binary upgrade, this workstation ran fine. What kind
> of hack do you use ?

	Some (bad) news.

	I have downloaded and built pkg from git repository. It runs better but 
is unable to achieve upgrade. Now, it stalls after it has downloaded all 
packages to upgrade:

# /usr/local/sbin/pkg upgrade
...

Number of packages to be removed: 3
Number of packages to be installed: 31
Number of packages to be upgraded: 254
Number of packages to be reinstalled: 45

The process will require 1 GiB more space.

Proceed with this action? [y/N]: y


	pkg uses 100% of a CPU core:

last pid:3557;load averages: 1.06, 1.14, 1.15 up 18+20:22:27  14:47:00
109 processes: 2 running, 105 sleeping, 2 zombie
CPU: 25.3% user,  0.0% nice,  0.6% system,  0.2% interrupt, 73.9% idle
Mem: 2465M Active, 3935M Inact, 1251M Wired, 743M Buf, 157M Free
Swap: 8192M Total, 10M Used, 8182M Free

   PID USERNAME THR PRI NICE   SIZE    RES STATE C   TIME    WCPU COMMAND
  3336 root       1 103    0  1029M   990M CPU3  3  67:09  98.60% pkg

	As I have written, this workstation is a diskless machine:

$ df -h
Filesystem                     Size    Used   Avail Capacity  Mounted on
192.168.10.128:/srv/pythagore  523G     53G    444G    11%    /
devfs                          1,0K    1,0K      0B   100%    /dev
procfs                         4,0K    4,0K      0B   100%    /proc
fdescfs                        1,0K    1,0K      0B   100%    /dev/fd
192.168.10.128:/home           3,6T    439G    3,0T    12%    /home

Mount options for / are nfsv3,tcp,soft,intr,rw,async,nolockd (nolockd is 
mandatory to avoid locking error on server side). For /home, 
nfsv3,tcp,soft,intr,rw,async.

	I suppose that is bug is NFS related, but I'm not able to found a 
workaround. gdb crashes when I try to attach it on pkg process :

root at pythagore:~ # gdb -p 3336
GNU gdb 6.1.1 [FreeBSD]
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain 
conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "amd64-marcel-freebsd".
Attaching to process 3336
/usr/src/gnu/usr.bin/gdb/libgdb/../../../../contrib/gdb/gdb/solib-svr4.c:1444: 
internal-error: legacy_fetch_link_map_offsets called without legacy 
link_map support enabled.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Quit this debugging session? (y or n)

	NFS server is a 7.0.2 NetBSD (that runs without any trouble).

	Best regards,

	JKB