sshd (openssh 3.7.1p1) dies during login on Solaris 8 system with SRM installed

Wiliams, Brian Brian.Williams at NBNZ.CO.NZ
Thu Sep 25 12:16:12 EST 2003


I have compiled ssh 3.7.1p1 using gcc and am trying to get it to run on our
Solaris 8 systems running Sun's SRM system.
With existing users it is fine, but with a new user, the user can not ssh in
on the first login, they get the message from SRM that no lnode has been
created.

I put sshd in debug and found that it SEG's here:

debug3: mm_sshpam_free_ctx: waiting for MONITOR_ANS_PAM_FREE_CTX
debug3: monitor_read: checking request 52
debug3: mm_answer_pam_free_ctx
debug3: mm_request_receive_expect entering: type 53
debug3: mm_request_receive entering
debug3: mm_request_send entering: type 53
debug2: monitor_read: 52 used once, disabling now
debug3: mm_do_pam_account entering
debug3: mm_request_receive_expect entering: type 44
debug3: mm_request_receive entering
debug3: mm_request_send entering: type 44
debug3: mm_request_receive_expect entering: type 45
debug3: mm_request_receive entering
debug1: Calling cleanup 0x4af30(0x0)
Bus Error


I trussed the process and see :

1061:   time()                                          = 1064443974
1061:   sigaction(0x0000000C, 0xFFBEEDD8, 0xFFBEEE58)   = 0
1061:       new: hand = 0xFF19E92C mask = 0 0 0 0 flags = 0x0012
1061:       old: hand = 0x00000000 mask = 0 0 0 0 flags = 0x0000
1061:   sys#177(0x0000007B, 0xFFBEEEFC, 0xFEE090A0)     = 0x00000000
[0x0000007B]
1061:   sigaction(0x0000000C, 0xFFBEEDD8, 0xFFBEEE58)   = 0
1061:       new: hand = 0x00000000 mask = 0 0 0 0 flags = 0x0012
1061:       old: hand = 0xFF19E92C mask = 0 0 0 0 flags = 0x0012
1061:   door(3, 0xFFBEE9E8)                             = 0
1061:   door(3, 0xFFBEE9D0)                             = 0
1061:   open(0xFF0C4D94, 0)                             = 9
1061:        0xFF0C4D94: "/etc/passwd"
1061:   fstat64(9, 0xFFBEE630)                          = 0
1061:       d=0x0154000A i=197963 m=0100444 l=1  u=0     g=3     sz=1462
1061:           at = Sep 25 10:52:54 NZST 2003  [ 1064443974 ]
1061:           mt = Sep 25 10:45:07 NZST 2003  [ 1064443507 ]
1061:           ct = Sep 25 10:45:07 NZST 2003  [ 1064443507 ]
1061:       bsz=8192  blks=4     fs=ufs
1061:   ioctl(9, 0x00005401, 0xFFBEE5BC)                Err#25 ENOTTY
1061:   read(9, 0x00160CCC, 8192)                       = 1462
1061:        0x00160CCC: " r o o t : x : 0 : 1 : S"..
1061:   llseek(9, 0xFFFFFFFFFFFFFFBF, 1)                = 1397
1061:   close(9)                                        = 0
1061:   sys#177(0x00000079, 0x000023B8, 0x001608C8)     Err#2 ENOENT
1061:       Incurred fault #5, FLTACCESS  %pc = 0xFEE129C0
1061:         siginfo: SIGBUS BUS_ADRALN addr=0xFF1BFAE6
1061:       Received signal #10, SIGBUS [default]
1061:         siginfo: SIGBUS BUS_ADRALN addr=0xFF1BFAE6
1061:           *** process killed ***
1062:   read(9, 0xFFBEF320, 4)                          = 0
1062:   write(2, 0xFFBEEDD0, 38)                        = 38
1062:        0xFFBEEDD0: " d e b u g 1 :   C a l l"..
1062:   shutdown(6, 2, 1)                               = 0
1062:   close(6)                                        = 0
1062:   llseek(0, 0, 1)                                 = 209524
1062:   _exit(255)

does any one know why sshd is getting the bus error?

anyone seen something similar? 

could this be similar to the sparcv9 issue?

Thanks
Brian

Brian Williams
Contractor
National Bank of New Zealand
Phone: IN 247451


This communication is confidential and may contain privileged material.
If you are not the intended recipient you must not use, disclose, copy or retain it.
If you have received it in error please immediately notify me by return email
and delete the emails.
Thank you.




More information about the openssh-unix-dev mailing list