Re: [uml-devel] The UML TODO list

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 454-5900

Hello,

Either of the new virtual ethernet drivers should fix this.

Jeff has both version in his possesion, I believe he will release one or =
the
other, or a combination of both in the next release.

Jim

On Wed, Feb 21, 2001 at 12:14:10PM -0800, Matt Clay wrote:
> Jeff,
>=20
> Ok, here's a more complete report on my test results.  After running th=
e
> Apache/PHP/MySQL test on the UML built from CVS sources I downloaded
> yesterday, I get the following error on the UML console:
>=20
> Kernel panic: Segfault with no mm
> In interrupt handler - not syncing
>=20
> I was running the debugger, here's where it stopped (at which point I
> got a backtrace), hopefully this helps more than my previous report:
>=20
> Program received signal SIGPIPE, Broken pipe.
> 0x100b8ca4 in __libc_write ()
> (gdb) backtrace
> #0  0x100b8ca4 in __libc_write ()
> #1  0xcb in ?? ()
> #2  0x100b086e in uml_net_user_tx (fd=3D13, buffer=3D0x5ef4686a "=FE=FD=
",
> len=3D1514)
>     at eth_user.c:132
> #3  0x100b0382 in uml_net_start_xmit (skb=3D0x5ef7d89c, dev=3D0x505b2e0=
0)
>     at eth_kern.c:260
> #4  0x1007ef0a in qdisc_restart (dev=3D0x505b2e00) at sch_generic.c:95
> #5  0x1007b2ee in dev_queue_xmit (skb=3D0x5ef7d89c)
>     at /home/uml/umlTest/linux-2.4.1/include/net/pkt_sched.h:439
> #6  0x10084932 in ip_output (skb=3D0x5ef7d89c) at ip_output.c:172
> #7  0x10084ce2 in ip_queue_xmit (skb=3D0x5ef7d89c) at ip_output.c:316
> #8  0x10090b5e in tcp_transmit_skb (sk=3D0x5ed379a0, skb=3D0x5ef7d89c)
>     at tcp_output.c:288
> #9  0x100911a3 in tcp_write_xmit (sk=3D0x5ed379a0) at tcp_output.c:492
> #10 0x1008efda in __tcp_data_snd_check (sk=3D0x5ed379a0, skb=3D0x5eda68=
9c)
>     at tcp_input.c:2944
> #11 0x1008f30a in tcp_rcv_established (sk=3D0x5ed379a0, skb=3D0x5efeac5=
4,
>     th=3D0x5ef02834, len=3D20) at tcp_input.c:2953
> #12 0x10095bc5 in tcp_v4_do_rcv (sk=3D0x5ed379a0, skb=3D0x5efeac54)
>     at tcp_ipv4.c:1557
> #13 0x10095f63 in tcp_v4_rcv (skb=3D0x5efeac54, len=3D20) at tcp_ipv4.c=
:1658
> #14 0x10082778 in ip_local_deliver (skb=3D0x5efeac54) at ip_input.c:254
> #15 0x10082a90 in ip_rcv (skb=3D0x5efeac54, dev=3D0x505b2e00, pt=3D0x10=
12f80c)
>     at ip_input.c:363
> #16 0x1007b78a in net_rx_action (h=3D0x1013b850) at dev.c:1403
> #17 0x1000f165 in do_softirq () at softirq.c:78
> #18 0x100a65bd in do_IRQ (irq=3D4, user_mode=3D0) at irq.c:330
> #19 0x100a6cc9 in sigio_handler (sig=3D29) at irq_user.c:61
> #20 0x100b1fa8 in __restore ()
>     at ../sysdeps/unix/sysv/linux/i386/sigaction.c:125
>=20
> Let me know if there's anything else I can do to help track this down.=20
> Also, if you'd like to see this for yourself, I'll make it available fo=
r
> you to login and check it out.
>=20
> Jeff Dike wrote:
> >=20
> > ma...@co... said:
> > > I ran my tests again on our Apache/PHP/MySQL setup using your 2.4.1
> > > kernel binary.  The system still died, but it did manage to last
> > > longer than before.
> >=20
> > I know of two crashes that may be affecting you.  One is in CVS, not =
in 2.4.1.
> >  The other I haven't fixed yet.
> >=20
> > So, either try the latest CVS or hold off until 2.4.2.
> >=20
> >                                 Thanks, Jeff
> >=20
> > _______________________________________________
> > User-mode-linux-devel mailing list
> > Use...@li...
> > http://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel
>=20
> --=20
>=20
>  - Matt Clay
>  - Cowboyz.com
>  - (503) 241-1990
>=20
> _______________________________________________
> User-mode-linux-devel mailing list
> Use...@li...
> http://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

--=20
James R. Leu