• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 1752
  • Last Modified:

zlogin to solaris container causes segmentation fault and core dump

1. whenever I login to a zone either thru ssh or zlogin I get a segmentation fault
2. running sparc solaris 10 10/09 version with recommended patches dated 2010.03.07
3. machine is T6300 blade

here is my zone config export

# zonecfg -z zone1 export
create -b
set zonepath=/zones/zone1
set autoboot=false
set ip-type=shared
add inherit-pkg-dir
set dir=/lib
end
add inherit-pkg-dir
set dir=/platform
end
add inherit-pkg-dir
set dir=/sbin
end
add inherit-pkg-dir
set dir=/usr
end
add net
set address=10.20.11.105
set physical=e1000g0
set defrouter=10.20.11.254
end

any idea what is causing this?
0
edwinbmiller
Asked:
edwinbmiller
  • 3
  • 2
1 Solution
 
edwinbmillerAuthor Commented:
just found out the direct cause of the segmentation fault was the login trying to run the quota command from the /etc/profile

case "$0" in
-sh | -ksh | -jsh | -bash)

        if [ ! -f .hushlogin ]
        then
                /usr/sbin/quota

if I run the quota command within the container it will segment fault

turning on .hushlogin in the home directory keeps quota from running and no error occur during login

-----------------------------------

running as root gives same error, running quota from host os works fine.

truss of running quota from a zone

munmap(0xFF090000, 32768)                       = 0
brk(0x000249C8)                                 = 0
brk(0x000269C8)                                 = 0
open("/dev/zfs", O_RDWR)                        = 3
open("/etc/mnttab", O_RDONLY)                   = 4
open("/etc/dfs/sharetab", O_RDONLY)             = 5
open("/etc/mnttab", O_RDONLY)                   = 6
ioctl(6, MNTIOC_GETMNTENT, 0xFFBFF6D4)          = 0
    Incurred fault #6, FLTBOUNDS  %pc = 0xFF131D40
      siginfo: SIGSEGV SEGV_MAPERR addr=0x00000000
    Received signal #11, SIGSEGV [default]
      siginfo: SIGSEGV SEGV_MAPERR addr=0x00000000


0
 
Brian UtterbackPrinciple Software EngineerCommented:
Do you get a core file produced? If so, please run pstack on the core and post the results. Can you login okay if you comment out the quota call?
0
 
edwinbmillerAuthor Commented:
# pstack core
core 'core' of 27618:   quota
 ff131d40 strcmp   (7, 0, 0, 0, 246b8, 0) + 160
 000113e8 main     (1, ffbffc24, ffbffc2c, 24000, 0, 12c00) + 11c
 000111a0 _start   (0, 0, 0, 0, 0, 0) + 108
#
if i comment out the quota I login fine.
0
 
edwinbmillerAuthor Commented:
if i truss quota within the container

unmap(0xFEC38000, 65536)                       = 0
memcntl(0xFEC20000, 25572, MC_ADVISE, MADV_WILLNEED, 0, 0) = 0
close(3)                                        = 0
stat64("/lib/libxml2.so.2", 0xFFBFE950)         Err#2 ENOENT
munmap(0xFF090000, 32768)                       = 0
brk(0x000249C8)                                 = 0
brk(0x000269C8)                                 = 0
open("/dev/zfs", O_RDWR)                        = 3
open("/etc/mnttab", O_RDONLY)                   = 4
open("/etc/dfs/sharetab", O_RDONLY)             = 5
open("/etc/mnttab", O_RDONLY)                   = 6
ioctl(6, MNTIOC_GETMNTENT, 0xFFBFF714)          = 0
    Incurred fault #6, FLTBOUNDS  %pc = 0xFF131D40
      siginfo: SIGSEGV SEGV_MAPERR addr=0x00000000
    Received signal #11, SIGSEGV [default]
      siginfo: SIGSEGV SEGV_MAPERR addr=0x00000000
#
looks like it's having problems with /etc/mnttab
0
 
Brian UtterbackPrinciple Software EngineerCommented:
It looks like somehow your mnttab file got corrupted, which is very weird since it is not a regular file.  It would appear that the file system type field is screwed up somehow, most liekly on the very first entry.
0

Featured Post

Receive 1:1 tech help

Solve your biggest tech problems alongside global tech experts with 1:1 help.

  • 3
  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now