RE: [SSI] SSI and ypbind

[Walker, B. J <bru...@hp...>]

Martin,
   Excellent bug report - thanks.  Initial look, based your strace,
is that ypbind is mucking with rt signals, which we discovered
overlap with the two signal we introduced (SIGCLUSTER and=20
SIGMIGRATE).  John Byrne will hopefully look at what we can do in
the short term.  There is a strategy to eliminate the 2 new signals, if
we must do that.  Maybe we can move those signals away from RTMIN.

bruce=20

-----Original Message-----
From: Martin H=F8y [mailto:mar...@pv...]
Sent: Wednesday, December 11, 2002 8:56 AM
To: ssi...@li...
Subject: [SSI] SSI and ypbind



I'm having trouble getting ypbind to work with the (binary) SSI kernel.

The machine is running RedHat 7.3 with all updates, and when booting=20
with the standard RedHat kernel (2.4.18-18.7.x), the ypbind process
is able to connect to the ypserver, and everything works fine.

When booting with 2.4.18-18.7.x_0.7.5r1ssi, it fails to connect.
Running in debug mode, the following is printed:

[root@test root]# ypbind -debug
parsing config file
Trying entry: domain fysikalsk server phys
parsed domain 'fysikalsk' server 'phys'
add_server() domain: fysikalsk, host: phys, nobroadcast, slot: 0
[Welcome to ypbind-mt, version 1.10]

ypbind-mt already running (pid 66823) - exiting


While that is not true. It seems that somehow ypbind-mt gets confused
by some of the process magic that SSI performs. Has anyone seen this
problem before?

The interesting part from an strace of the above command looks like:

write(2, "[Welcome to ypbind-mt, version 1"..., 37[Welcome to ypbind-mt,
version 1.10]
) =3D 37
write(2, "\n", 1
)                       =3D 1
rt_sigprocmask(SIG_BLOCK, [HUP INT QUIT SEGV TERM CHLD], NULL, 8) =3D 0
pipe([3, 4])                            =3D 0
clone(child_stack=3D0x8053900,
flags=3DCLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND) =3D 66888
write(4, " \371\377\277\5\0\0\0\214r\2@\\\220\4\0100\371\377\277"...,
148) =3D 148
rt_sigprocmask(SIG_SETMASK, NULL, [HUP INT QUIT SEGV TERM CHLD RTMIN],
8) =3D 0
write(4, "\240`\3@\0\0\0\0\0\0\0\0\200\243\4\10\0\0\0\0\7D\1\200"...,
148) =3D 148
rt_sigprocmask(SIG_SETMASK, NULL, [HUP INT QUIT SEGV TERM CHLD RTMIN],
8) =3D 0
rt_sigsuspend([HUP INT QUIT SEGV TERM CHLD] <unfinished ...>
--- SIGRTMIN (Real-time signal 0) ---
<... rt_sigsuspend resumed> )           =3D -1 EINTR (Interrupted system =
call)
sigreturn()                             =3D ? (mask now [HUP INT QUIT =
SEGV
TERM CHLD RTMIN])
rt_sigprocmask(SIG_SETMASK, NULL, [HUP INT QUIT SEGV TERM CHLD RTMIN],
8) =3D 0
rt_sigsuspend([HUP INT QUIT SEGV TERM CHLD]ypbind-mt already running
(pid 66889) - exiting
 <unfinished ...>
--- SIGRT_1 (Real-time signal 1) ---
<... rt_sigsuspend resumed> )           =3D -1 EINTR (Interrupted system =
call)
wait4(66888, NULL, __WCLONE, NULL)      =3D 66888
_exit(1)                                =3D ?


Regards,

Martin
--=20
"Computer science is not about computers any more than astronomy
is about telescopes." -- EW Dijkstra


-------------------------------------------------------
This sf.net email is sponsored by:
With Great Power, Comes Great Responsibility=20
Learn to use your power at OSDN's High Performance Computing Channel
http://hpc.devchannel.org/
_______________________________________________
ssic-linux-devel mailing list
ssi...@li...
https://lists.sourceforge.net/lists/listinfo/ssic-linux-devel