Robert Watson writes:
>
> As of this afternoon, I seem to be having problems with machines hanging
> on incoming TCP connections. As soon as some process attempts to do
> useful TCP stuff (incoming telnet, ssh, et al), all process hang. If I
> break to ddb, a ps shows that telnetd has no wmesg. If I begin tapping on
> various virtual consoles, gradually those processes also show no wmesg.
> If I do a trace while in ddb, it whos that the active process is the
> telnetd, but that it is hung in write():
>
> ...
> --- interrupt, eip = 0xc0195337, esp = 0xc5ea8ddc, ebp = 0xc5ea8df0 ---
> sbappend(c554caf4,c057000) at sbappend+0x4b
> tcp_usr_send(c554ca80,0,c0578000,0,0) at tcp_usr_send+0xc0
> sosend(c554ca80,0,c5ea8edc,c0578000,0) at sosend+0x61b
> soo_write(c0c5f380,c5ea8edc,c0543780,0,c4d1f560) at soo_write+0x24
> dofilewrite(c4d1f560,c0c5f80,0,805b120,3) at dofilewrite+0xb1
> write(c4d1f560,c5ea8f80,3ffffffff,805bcc8) at write+0x33
> syscall2(2f,2f,2f,805bcc8,ffffffff) at syscall2+0x1f1
> Xint80x80_syscall() at Xint80x80_syscall+0x25
>
> So it looks like it might have to do with the recent socket buffer
> patches, et al, and isn't a result of my recent commits. However, I'm not
> in front of a machine with a serial debugger right now, so won't be able
> to follow up on this for a few hours. It seems to affect both machines
> I've tried it on. ex cpl returns 63109a.
>
> So as a word of warning: if you're thinking about cvsup'ing, don't :-)
Not a terribly good day to upgrade in -stable either -- RELENG4 is
currently broken with exactly the same symptoms. See my post in the
"4.1 STABLE broken since today!" thread on -stable..
Drew
------------------------------------------------------------------------------
Andrew Gallatin, Sr Systems Programmer http://www.cs.duke.edu/~gallatin
Duke University Email: [EMAIL PROTECTED]
Department of Computer Science Phone: (919) 660-6590
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message
