Re: [PATCH 2/2] x86 boot: early_res_to_e820

["Yinghai Lu" <yhlu.kernel@gmail.com>]

On Thu, Jun 26, 2008 at 12:27 AM, Huang, Ying <ying.huang@intel.com> wrote:
> On Thu, 2008-06-26 at 00:13 -0700, Yinghai Lu wrote:
>> On Wed, Jun 25, 2008 at 11:33 PM, Huang, Ying <ying.huang@intel.com> wrote:
>> > This patch reserves some memory areas of early_res in E820 as
>> > E820_RESVD_KERN. This is used mainly on i386, for some memory area
>> > from boot-loader and in highmem, such as linked list of setup_data,
>> > EFI memory map, etc.
>> >
>> > This patch is based on latest x86/master branch of git-x86 tree and
>> > has been tested on i386 and x86_64 platform.
>> >
>> > Signed-off-by: Huang Ying <ying.huang@intel.com>
>> >
>> > ---
>> >  arch/x86/kernel/e820.c     |   22 ++++++++++++++++++++++
>> >  arch/x86/kernel/setup_32.c |    2 ++
>> >  include/asm-x86/e820.h     |    1 +
>> >  3 files changed, 25 insertions(+)
>> >
>> > --- a/arch/x86/kernel/e820.c
>> > +++ b/arch/x86/kernel/e820.c
>> > @@ -844,6 +844,28 @@ void __init early_res_to_bootmem(u64 sta
>> >        }
>> >  }
>> >
>> > +void __init early_res_to_e820(u64 start, u64 end)
>> > +{
>> > +       int i, changed = 0;
>> > +       u64 final_start, final_end;
>> > +       for (i = 0; i < MAX_EARLY_RES && early_res[i].end; i++) {
>> > +               struct early_res *r = &early_res[i];
>> > +               final_start = max(start, r->start);
>> > +               final_end = min(end, r->end);
>> > +               if (final_start >= final_end)
>> > +                       continue;
>> > +               printk(KERN_INFO "  early res: %d [%llx-%llx] %s\n", i,
>> > +                       final_start, final_end - 1, r->name);
>> > +               e820_update_range(final_start, final_end - final_start,
>> > +                                 E820_RAM, E820_RESVD_KERN);
>> > +               changed = 1;
>> > +       }
>> > +       if (changed) {
>> > +               printk(KERN_INFO "update e820 for early_res\n");
>> > +               update_e820();
>> > +       }
>> > +}
>> > +
>> >  /* Check for already reserved areas */
>> >  static inline int __init bad_addr(u64 *addrp, u64 size, u64 align)
>> >  {
>> > --- a/include/asm-x86/e820.h
>> > +++ b/include/asm-x86/e820.h
>> > @@ -96,6 +96,7 @@ extern void reserve_early(u64 start, u64
>> >  extern void reserve_early_overlap_ok(u64 start, u64 end, char *name);
>> >  extern void free_early(u64 start, u64 end);
>> >  extern void early_res_to_bootmem(u64 start, u64 end);
>> > +extern void early_res_to_e820(u64 start, u64 end);
>> >  extern u64 early_reserve_e820(u64 startt, u64 sizet, u64 align);
>> >
>> >  extern unsigned long e820_end_of_ram(void);
>> > --- a/arch/x86/kernel/setup_32.c
>> > +++ b/arch/x86/kernel/setup_32.c
>> > @@ -447,6 +447,8 @@ void __init setup_arch(char **cmdline_p)
>> >         acpi_numa_init();
>> >  #endif
>> >
>> > +       early_res_to_e820(max_low_pfn<<PAGE_SHIFT, -1);
>> > +
>> >        initmem_init(0, max_pfn);
>> >
>> >  #ifdef CONFIG_ACPI_SLEEP
>> >
>> > --
>>
>> this one is not needed.
>
> For example, if EFI memory map is allocated in highmem area by
> boot-loader, it need to be reserved in kernel. But on i386, bootmem can
> only reserve memory area below max_low_pfn. So I think this is needed.
>
> In general, I think it is needed to provide a reservation mechanism for
> highmem area on i386. Because some memory area from BIOS or boot-loader
> may in highmem area.

I mean call reserve_setup_data right after finish_e820_parsing,  and
it could call e820_update_range directly.

YH