Re: 2.6.24-rc1 does not boot on SGI

[Ricardo Mendoza <ricmm@kanux.com>]

Giuseppe Sacco wrote:
> I reply to my own message, providing more details, hoping that anyone
> here could give a hint or the solution.
> 
> During bootup on ip32, since 2.4.24-rc1, the system loop printing a
> message about unexpected interrupt #13. (see transcript below.)
> 
> I enabled more debug in the kernel, and studied the code. What I
> understood is that interrupt #13 from CRIME means that the system should
> check on MACEISA for the real interrupt.
> 
> The interrupt start appearing just after executing the code
> psmouse_init() that enable ps2 drivers for keyboard and mouse. Keyboard
> interrupt #49 is enabled first and mouse interrupt #51 is enabled later.
> 
> When initialising the keyboard interrupt (it is a MACEISA interrupt),
> the interrupt start appearing, so I am pretty sure that interrupt #13 is
> related to the keyboard interrupt.
> 
> When the system receive interrupt #13, it correctly detect it is a
> MACEISA interrupt, and check for mace->perif.ctrl.istat value. The
> problem seems to be that this value is zero instead of having bit #9 on
> (that would mean, interrupt #49, keyboard).
> 
> So, either the interrupt #49 is not correctly enabled, or maceisa
> interrupt aren't correctly checked.
> 
> Does this description ring a bell to anyone?
> 
> Bye,
> Giuseppe
> 
> Calling initcall 0xffffffff80496ca0: serport_init+0x0/0x48()
> initcall 0xffffffff80496ca0: serport_init+0x0/0x48() returned 0.
> initcall 0xffffffff80496ca0 ran for 0 msecs: serport_init+0x0/0x48()
> Calling initcall 0xffffffff80496ce8: maceps2_init+0x0/0xe0()
> initcall 0xffffffff80496ce8: maceps2_init+0x0/0xe0() returned 0.
> initcall 0xffffffff80496ce8 ran for 1 msecs: maceps2_init+0x0/0xe0()
> Calling initcall 0xffffffff80496dc8: serio_raw_init+0x0/0x18()
> initcall 0xffffffff80496dc8: serio_raw_init+0x0/0x18() returned 0.
> initcall 0xffffffff80496dc8 ran for 1 msecs: serio_raw_init+0x0/0x18()
> Calling initcall 0xffffffff80496f40: mousedev_init+0x0/0xd0()
> mice: PS/2 mouse device common for all mice
> initcall 0xffffffff80496f40: mousedev_init+0x0/0xd0() returned 0.
> initcall 0xffffffff80496f40 ran for 16 msecs: mousedev_init+0x0/0xd0()
> Calling initcall 0xffffffff80497010: atkbd_init+0x0/0x18()
> initcall 0xffffffff80497010: atkbd_init+0x0/0x18() returned 0.
> initcall 0xffffffff80497010 ran for 0 msecs: atkbd_init+0x0/0x18()
> Calling initcall 0xffffffff80497028: psmouse_init+0x0/0x90()
> maceisa enable: 49
> crime_int 00000020 enabled
> *irq 13, crime_int=00002000, crime_mask=003003a0, mace_int=00000000*
> irq 13, desc: ffffffff80448390, depth: 1, count: 0, unhandled: 0
> ->handle_irq():  ffffffff80065eb0, handle_bad_irq+0x0/0x2c0
> ->chip(): ffffffff8043e320, 0xffffffff8043e320
> ->action(): 0000000000000000
>   IRQ_DISABLED set

I recommend you pull latest git. Looks like some issue that Ralf and I
fixed a few weeks ago.


     Ricardo