Takao Indoh <indou.takao at jp.fujitsu.com> writes:
> On Tue, 14 Dec 2010 12:43:58 +0900, Kenji Kaneshige wrote:
>
>>Hi,
>>
>>I tested this patch on the system that has large amount of memory (1TB),
>>and I encountered the immediate system reset problem that happens every
>>time I modify the EFI boot entry using efibootmgr command. It seems that
>>triple fault happens due to the incorrect page table setup.
>>
>>> +void __init efi_pagetable_init(void)
>>> +{
>>(snip.)
>>> + pgd = efi_pgd + pgd_index(PAGE_OFFSET);
>>> + set_pgd(pgd, *pgd_offset_k(PAGE_OFFSET));
>>> + pgd = efi_pgd + pgd_index(__START_KERNEL_map);
>>> + set_pgd(pgd, *pgd_offset_k(__START_KERNEL_map));
>>> +}
>>
>>Maybe we need to map whole kernel address space. The problem doesn't
>>happen by modifying as follows.
>>
>> clone_pgd_range(efi_pgd + KERNEL_PGD_BOUNDARY,
>> swapper_pg_dir + KERNEL_PGD_BOUNDARY,
>>KERNEL_PGD_PTRS);
>
>
> Besides this bug, I'm thinking that we need global TLB flush after
> restoring cr3 because EFI code page is mapped with PAGE_KERNEL_EXEC.
>
> void efi_call_phys_epilog_in_physmode(void)
> {
> write_cr3(get_cpu_var(save_cr3));
> + if (cpu_has_pge)
> + __flush_tlb_global();
> local_irq_restore(get_cpu_var(efi_flags));
> }
>
> Somethinkg like this. Anybody comments?
If only one cpu runs efi we shouldn't need a global flush.
I presume you aren't modifying the kernel's global page table?
If we are giving the entire machine to efi then yes we would need
to set cr3 on all machines.
Eric
