Problem with de4x5 on Alpha?

Problem with de4x5 on Alpha?

[Nathan Poznick]

[-- Attachment #1: Type: text/plain, Size: 7108 bytes --]


I've been trying to get 2.6.0-test[6|7] to boot on my AS2100 (Sable),
and have run into a problem.  If I disable networking, I can get the
machine to boot.  However if the rc script to start networking runs, the
machine hangs on it.  It's using the de4x5 driver for the NICs, but only
one interface out of the 5 possible is ever brought up.  I've only got a
serial console on the thing, so I don't know if anything gets printed to
the screen when this happens to indicate something.  However, when it
hangs I've done an Alt-SysRq-t to dump a trace, and then run the trace
through ksymoops to try and get some idea of what's happening.  Below
are three traces from when it hangs;  the first two are from booting
with smp enabled, and the third one is from when I booted with 'nosmp'.
Is there any other information I can collect to help debug this?
(Everything works fine with 2.4.21)


Trace; fffffc000032afac <show_state+9c/110>
Trace; fffffc0000429878 <sysrq_handle_showstate+18/30>
Trace; fffffc0000429c04 <__handle_sysrq_nolock+d4/1c0>
Trace; fffffc0000429b08 <handle_sysrq+68/90>
Trace; fffffc0000420404 <kbd_keycode+444/4a0>
Trace; fffffc00004204ac <kbd_event+4c/b0>
Trace; fffffc000048ddac <input_event+16c/4d0>
Trace; fffffc0000490cf4 <atkbd_interrupt+1f4/500>
Trace; fffffc000031ca20 <timer_interrupt+a0/180>
Trace; fffffc0000318514 <handle_IRQ_event+74/e0>
Trace; fffffc0000495db4 <serio_interrupt+64/70>
Trace; fffffc0000319768 <handle_irq+108/220>
Trace; fffffc0000496cc8 <i8042_interrupt+178/230>
Trace; fffffc0000318514 <handle_IRQ_event+74/e0>
Trace; fffffc0000319768 <handle_irq+108/220>
Trace; fffffc00003f9590 <_outb+0/30>
Trace; fffffc0000327e34 <rebalance_tick+94/a0>
Trace; fffffc0000339e3c <do_timer+11c/120>
Trace; fffffc000031ca20 <timer_interrupt+a0/180>
Trace; fffffc0000318514 <handle_IRQ_event+74/e0>
Trace; fffffc0000319768 <handle_irq+108/220>
Trace; fffffc0000323354 <sable_lynx_srm_device_interrupt+44/60>
Trace; fffffc000031a034 <do_entInt+124/190>
Trace; fffffc0000313160 <ret_from_sys_call+0/10>
Trace; fffffc000046396c <de4x5_interrupt+ec/370>
Trace; fffffc0000319d10 <synchronize_irq+0/50>
Trace; fffffc0000319d40 <synchronize_irq+30/50>
Trace; fffffc0000318514 <handle_IRQ_event+74/e0>
Trace; fffffc0000319768 <handle_irq+108/220>
Trace; fffffc0000323354 <sable_lynx_srm_device_interrupt+44/60>
Trace; fffffc000031a034 <do_entInt+124/190>
Trace; fffffc0000313160 <ret_from_sys_call+0/10>
Trace; fffffc0000314fb0 <default_idle+0/10>
Trace; fffffc0000315018 <cpu_idle+58/80>
Trace; fffffc0000314fb0 <default_idle+0/10>
Trace; fffffc0000314fb0 <default_idle+0/10>
Trace; fffffc00003100e0 <rest_init+60/90>
Trace; fffffc000031001c <_stext+1c/20>



Trace; fffffc000032afac <show_state+9c/110>
Trace; fffffc0000429878 <sysrq_handle_showstate+18/30>
Trace; fffffc0000429c04 <__handle_sysrq_nolock+d4/1c0>
Trace; fffffc0000429b08 <handle_sysrq+68/90>
Trace; fffffc0000420404 <kbd_keycode+444/4a0>
Trace; fffffc00004204ac <kbd_event+4c/b0>
Trace; fffffc000048ddac <input_event+16c/4d0>
Trace; fffffc0000490cf4 <atkbd_interrupt+1f4/500>
Trace; fffffc0000495db4 <serio_interrupt+64/70>
Trace; fffffc0000496c08 <i8042_interrupt+b8/230>
Trace; fffffc0000496cc8 <i8042_interrupt+178/230>
Trace; fffffc0000318514 <handle_IRQ_event+74/e0>
Trace; fffffc0000327e34 <rebalance_tick+94/a0>
Trace; fffffc0000339e3c <do_timer+11c/120>
Trace; fffffc000031ca20 <timer_interrupt+a0/180>
Trace; fffffc0000318514 <handle_IRQ_event+74/e0>
Trace; fffffc0000319768 <handle_irq+108/220>
Trace; fffffc0000323354 <sable_lynx_srm_device_interrupt+44/60>
Trace; fffffc000031a034 <do_entInt+124/190>
Trace; fffffc0000313160 <ret_from_sys_call+0/10>
Trace; fffffc0000464cbc <set_multicast_list+10c/140>
Trace; fffffc000046396c <de4x5_interrupt+ec/370>
Trace; fffffc0000319d10 <synchronize_irq+0/50>
Trace; fffffc0000319d44 <synchronize_irq+34/50>
Trace; fffffc0000318514 <handle_IRQ_event+74/e0>
Trace; fffffc0000319768 <handle_irq+108/220>
Trace; fffffc0000323354 <sable_lynx_srm_device_interrupt+44/60>
Trace; fffffc000046bea8 <type4_infoblock+68/1e0>
Trace; fffffc000031a034 <do_entInt+124/190>
Trace; fffffc0000313160 <ret_from_sys_call+0/10>
Trace; fffffc0000464cbc <set_multicast_list+10c/140>
Trace; fffffc00004b18f4 <dev_activate+f4/150>
Trace; fffffc00004b1520 <pfifo_fast_init+0/40>
Trace; fffffc00004b184c <dev_activate+4c/150>
Trace; fffffc00004a24d8 <dev_open+d8/100>
Trace; fffffc00004a4604 <dev_change_flags+84/1f0>
Trace; fffffc00004f3e14 <devinet_ioctl+384/8f0>
Trace; fffffc00004f78b8 <inet_ioctl+e8/170>
Trace; fffffc00004988c8 <sock_ioctl+148/500>
Trace; fffffc0000390378 <sys_ioctl+188/3d0>
Trace; fffffc0000313144 <entSys+a4/c0>
Trace; fffffc00003130a0 <entSys+0/c0>



Trace; fffffc000032afac <show_state+9c/110>
Trace; fffffc0000429878 <sysrq_handle_showstate+18/30>
Trace; fffffc0000429c04 <__handle_sysrq_nolock+d4/1c0>
Trace; fffffc0000429b08 <handle_sysrq+68/90>
Trace; fffffc0000420404 <kbd_keycode+444/4a0>
Trace; fffffc00004204ac <kbd_event+4c/b0>
Trace; fffffc000048ddac <input_event+16c/4d0>
Trace; fffffc0000490cf4 <atkbd_interrupt+1f4/500>
Trace; fffffc0000495db4 <serio_interrupt+64/70>
Trace; fffffc0000496c08 <i8042_interrupt+b8/230>
Trace; fffffc0000496cc8 <i8042_interrupt+178/230>
Trace; fffffc0000318514 <handle_IRQ_event+74/e0>
Trace; fffffc0000464cbc <set_multicast_list+10c/140>
Trace; fffffc0000319768 <handle_irq+108/220>
Trace; fffffc00003f9590 <_outb+0/30>
Trace; fffffc0000327e34 <rebalance_tick+94/a0>
Trace; fffffc0000339e3c <do_timer+11c/120>
Trace; fffffc000031ca20 <timer_interrupt+a0/180>
Trace; fffffc0000318514 <handle_IRQ_event+74/e0>
Trace; fffffc0000319768 <handle_irq+108/220>
Trace; fffffc0000323354 <sable_lynx_srm_device_interrupt+44/60>
Trace; fffffc000031a034 <do_entInt+124/190>
Trace; fffffc0000313160 <ret_from_sys_call+0/10>
Trace; fffffc0000464cbc <set_multicast_list+10c/140>
Trace; fffffc000046396c <de4x5_interrupt+ec/370>
Trace; fffffc0000319d10 <synchronize_irq+0/50>
Trace; fffffc0000319d44 <synchronize_irq+34/50>
Trace; fffffc0000318514 <handle_IRQ_event+74/e0>
Trace; fffffc0000319768 <handle_irq+108/220>
Trace; fffffc0000323354 <sable_lynx_srm_device_interrupt+44/60>
Trace; fffffc000046bea8 <type4_infoblock+68/1e0>
Trace; fffffc000031a034 <do_entInt+124/190>
Trace; fffffc0000313160 <ret_from_sys_call+0/10>
Trace; fffffc0000464cbc <set_multicast_list+10c/140>
Trace; fffffc00004b18f4 <dev_activate+f4/150>
Trace; fffffc00004b1520 <pfifo_fast_init+0/40>
Trace; fffffc00004b184c <dev_activate+4c/150>
Trace; fffffc00004a24d8 <dev_open+d8/100>
Trace; fffffc00004a4604 <dev_change_flags+84/1f0>
Trace; fffffc00004f3e14 <devinet_ioctl+384/8f0>
Trace; fffffc00004f78b8 <inet_ioctl+e8/170>
Trace; fffffc00004988c8 <sock_ioctl+148/500>
Trace; fffffc0000390378 <sys_ioctl+188/3d0>
Trace; fffffc0000313144 <entSys+a4/c0>
Trace; fffffc00003130a0 <entSys+0/c0>


-- 
Nathan Poznick <kraken@drunkmonkey.org>

Speak not of my debts unless you mean to pay them. - George Herbert


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: Problem with de4x5 on Alpha?

[Marc Zyngier]

>>>>> "Nathan" == Nathan Poznick <kraken@drunkmonkey.org> writes:

Nathan> I've been trying to get 2.6.0-test[6|7] to boot on my AS2100
Nathan> (Sable),

[...]

Could you please try 2.6.0-test5 ? Or even better, 2.6.0-test7 with
the 2.6.0-test5 driver (revert de4x5.[ch] and the Space.c changes) ?

The usual debug informations would be helpful too...

	M.
-- 
Places change, faces change. Life is so very strange.

Re: Problem with de4x5 on Alpha?

[Nathan Poznick]

[-- Attachment #1: Type: text/plain, Size: 5403 bytes --]

Thus spake Marc Zyngier:
> Could you please try 2.6.0-test5 ? Or even better, 2.6.0-test7 with
> the 2.6.0-test5 driver (revert de4x5.[ch] and the Space.c changes) ?
> 
> The usual debug informations would be helpful too...

I reverted de4x5.[ch] and Space.c to their -test5 versions, and I get
the same effect.  If I disable the starting of networking, and boot, I
can then see this when I attempt to start it:

root@buddha:~# /etc/init.d/networking start
Setting up IP spoofing protection: rp_filter.
Configuring network interfaces... done.
root@buddha:~# eth1: media is 100Mb/s.

After that, it's locked up.  Sometimes it prints the media notice before
it hangs, and sometimes it doesn't.  I don't believe I've ever seen it
print that when I booted with nosmp (always hangs before it gets that
far.)  Also, when SMP is enabled, sometimes when I do the SysRq-t I see
this right before the trace starts:

smp_call_function_on_cpu: initial timeout -- trying long wait

And then down mixed in with the trace, I get a notice from a BUG_ON() at
line 878 of arch/alpha/kernel/smp.c  Decoded, it looks like this:

Kernel bug at arch/alpha/kernel/smp.c:878
CPU 1 ifup(102): Kernel Bug 1
pc = [<fffffc000031e6a4>]  ra = [<fffffc000031e624>]  ps = 0000    Not tainted
Using defaults from ksymoops -t elf64-alpha -a alpha
v0 = 0000000000000041  t0 = 0000000000000001  t1 = 0000000000000001
t2 = 0000000100036c4d  t3 = 0000000000000000  t4 = ffffffff00000000
t5 = 0000000000000001  t6 = 61635f706d733e33  t7 = fffffc001f41c000
a0 = fffffc00005e8b70  a1 = 0000000000002600  a2 = ffffffffffffffff
a3 = 000000000000000a  a4 = fffffc0000000008  a5 = fffffc001f41fa00
t8 = 0000000000000004  t9 = 000000000000000d  t10= 0000000000000000
t11= 000186f800000004  pv = fffffc00003f9590  at = fffffc000063f0e8
gp = fffffc0000631200  sp = fffffc001f41f968
Trace:fffffc000031e96c fffffc000031e8a0 fffffc0000367320
fffffc000032c470 fffffc00003859f4 fffffc00003860f0 fffffc00003acae0
fffffc00003ad430 fffffc00003ad454 fffffc00003869ac fffffc0000386e28
fffffc0000315930 fffffc0000313144 fffffc00003130a0
Code: 2ffe0000  47e10400  a59e0020  23de0060  6bfa8001  00000081
<0000036e> 00520922


>>RA;  fffffc000031e624 <smp_call_function_on_cpu+d4/200>

>>PC;  fffffc000031e6a4 <smp_call_function_on_cpu+154/200>   <=====

Trace; fffffc000031e96c <flush_tlb_mm+3c/100>
Trace; fffffc000031e8a0 <ipi_flush_tlb_mm+0/90>
Trace; fffffc0000367320 <exit_mmap+230/290>
Trace; fffffc000032c470 <mmput+c0/140>
Trace; fffffc00003859f4 <exec_mmap+114/180>
Trace; fffffc00003860f0 <flush_old_exec+690/a80>
Trace; fffffc00003acae0 <load_elf_binary+370/d30>
Trace; fffffc00003ad430 <load_elf_binary+cc0/d30>
Trace; fffffc00003ad454 <load_elf_binary+ce4/d30>
Trace; fffffc00003869ac <search_binary_handler+1bc/430>
Trace; fffffc0000386e28 <do_execve+208/290>
Trace; fffffc0000315930 <sys_execve+60/b0>
Trace; fffffc0000313144 <entSys+a4/c0>
Trace; fffffc00003130a0 <entSys+0/c0>

Code;  fffffc000031e68c <smp_call_function_on_cpu+13c/200>
0000000000000000 <_PC>:
Code;  fffffc000031e68c <smp_call_function_on_cpu+13c/200>
   0:   00 00 fe 2f       unop 
Code;  fffffc000031e690 <smp_call_function_on_cpu+140/200>
   4:   00 04 e1 47       mov  t0,v0
Code;  fffffc000031e694 <smp_call_function_on_cpu+144/200>
   8:   20 00 9e a5       ldq  s3,32(sp)
Code;  fffffc000031e698 <smp_call_function_on_cpu+148/200>
   c:   60 00 de 23       lda  sp,96(sp)
Code;  fffffc000031e69c <smp_call_function_on_cpu+14c/200>
  10:   01 80 fa 6b       ret
Code;  fffffc000031e6a0 <smp_call_function_on_cpu+150/200>
  14:   81 00 00 00       bugchk
Code;  fffffc000031e6a4 <smp_call_function_on_cpu+154/200>   <=====
  18:   6e 03 00 00       call_pal     0x36e   <=====
Code;  fffffc000031e6a8 <smp_call_function_on_cpu+158/200>
  1c:   22 09 52 00       call_pal     0x520922


This is what lspci shows.  The 21040 is built into the I/O board, but is
only 10mbit so I don't use it.  The 4 21142/43 entries are the ports on
a 4-port DE500-BA card.  I'm using the first port on that card for my
connection.  I've tried swapping around and using the 21040, or a
different port on the other card, but it doesn't seem to change
anything.

00:00.0 Ethernet controller: Digital Equipment Corporation DECchip 21040 [Tulip] (rev 23)
00:01.0 SCSI storage controller: LSI Logic / Symbios Logic 53c810 (rev 02)
00:02.0 Non-VGA unclassified device: Intel Corp. 82375EB (rev 04)
00:06.0 RAID bus controller: Mylex Corporation DAC960P (rev 02)
00:07.0 PCI bridge: Digital Equipment Corporation DECchip 21050 (rev 02)
00:08.0 PCI bridge: Digital Equipment Corporation DECchip 21152 (rev 03)
01:00.0 SCSI storage controller: QLogic Corp. ISP1020 Fast-wide SCSI (rev 02)
02:00.0 Ethernet controller: Digital Equipment Corporation DECchip 21142/43 (rev 41)
02:01.0 Ethernet controller: Digital Equipment Corporation DECchip 21142/43 (rev 41)
02:02.0 Ethernet controller: Digital Equipment Corporation DECchip 21142/43 (rev 41)
02:03.0 Ethernet controller: Digital Equipment Corporation DECchip 21142/43 (rev 41)

If there's any specific information that I'm not presenting that would
be useful, please just let me know what things to gather up.


-- 
Nathan Poznick <kraken@drunkmonkey.org>

"Ouch!! I landed on my eight-sided dice!" -Joel (as a human sacrifice).
#301


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]