2.4 kernel doesn't allocate resource for my PCI Board
Gabriel Paubert
paubert at iram.es
Fri Oct 13 01:01:23 EST 2000
On Thu, 12 Oct 2000, Xavier Grave wrote:
> On Thu, 12 Oct 2000, Gabriel Paubert wrote:
> > Could you try my kernel and patches on ftp://vlab1.iram.es/pub/linux-2.4 ?
>
> When I try the zImage-2.4.0-test7 on my MVME2303 i have the following
> error message :
> pc_keyb: controller jammed (0xFF).
> pc_keyb: controller jammed (0xFF).
> pc_keyb: controller jammed (0xFF).
Ok, the bootloader should give up if there is no keyboard. Strange thing
is that I am convinced I have tested it on an MVME2400 (no keyboard
either). Try with CONFIG_VT off first (this should remove any attempt
to use the keyboard from the bootloader)...
> I have patched the test10 kernet with your patch but I have the following
> compile error :
> ppc_ksyms.c:289: `int_control' undeclared here (not in a function)
> ppc_ksyms.c:289: initializer element is not constant
> ppc_ksyms.c:289: (near initialization for `__ksymtab_int_control.value')
>
> In fact after a search in .c and .h file I only have this :
> ./arch/ppc/kernel/apus_setup.c: int_control.int_sti = __no_use_sti;
> ./arch/ppc/kernel/apus_setup.c: int_control.int_cli = __no_use_cli;
> ./arch/ppc/kernel/apus_setup.c: int_control.int_save_flags =
> __no_use_save_flags;
> ./arch/ppc/kernel/apus_setup.c: int_control.int_restore_flags =
> __no_use_restore_flags;
> ./arch/ppc/kernel/pmac_pic.c: int_control.int_set_lost =
> __no_use_set_lost;
> ./arch/ppc/kernel/ppc_ksyms.c:EXPORT_SYMBOL(int_control);
>
> ./include/asm-ppc/hw_irq.h:struct int_control_struct
> ./include/asm-ppc/hw_irq.h:extern struct int_control_struct int_control;
Hmmm, I suspect that you had rejects in te patch in this case. Indeed, I
do not have int_control because I consider this utterly stupid. Calling a
subroutine to mask/unmask interrupts is an incredible waste and
furthermore is absolutely unnecessary whatever the configuration of your
system is.
The same holds for the replacement of eieio by sync in all in/out
instructions, or the emulation of mfpvr, there are simply a few things
whih will never make it to my tree, ever.
Regards,
Gabriel.
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
More information about the Linuxppc-dev
mailing list