Re: [PATCH] x86: Merge setup_32/64.c into setup.c

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

On Wed, Jun 25, 2008 at 8:36 AM, Ingo Molnar <mingo@elte.hu> wrote:
>
> * Yinghai Lu <yhlu.kernel@gmail.com> wrote:
>
>>
>> Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com>
>>
>> ---
>>  arch/x86/kernel/Makefile   |    2
>>  arch/x86/kernel/setup.c    |  676 ++++++++++++++++++++++++++++++++++++++++++++-
>>  arch/x86/kernel/setup_32.c |  543 ------------------------------------
>>  arch/x86/kernel/setup_64.c |  381 -------------------------
>>  include/asm-x86/setup.h    |    2
>>  5 files changed, 670 insertions(+), 934 deletions(-)
>
> very nice!
>
> could we please split this up into several, gradual steps that bring
> setup_32.c and setup_64.c to exactly the same content - where the final
> patch just renames arch/x86/kernel/setup_32.c to arch/x86/kernel/setup.c
> and deletes arch/x86/kernel/setup_64.c ?

OK,  someone (Mike Triavis) already stole setup.c for X86_NUMA.

You need to change that setup.c to other name. or split it away...

>
> it would still result in exactly the same end result - but is much more
> bisectable (and reviewable, etc.). A bit like how arch/x86/mm/ioremap.c
> or arch/x86/mm/pageattr.c was unified.
>
> maybe it can be done in less than 10 patches - but it guess it should be
> rather something in the neighborhood of 20 patches. Changes like this:
>
> -       /*
> -        * NOTE: before this point _nobody_ is allowed to allocate
> -        * any memory using the bootmem allocator.  Although the
> -        * allocator is now initialised only the first 8Mb of the kernel
> -        * virtual address space has been mapped.  All allocations before
> -        * paging_init() has completed must use the alloc_bootmem_low_pages()
> -        * variant (which allocates DMA'able memory) and care must be taken
> -        * not to exceed the 8Mb limit.
> -        */
>
> should be in a separate patch - you removed this restriction from the
> x86 architecture via your earlier patches, so the removal of the comment
> comes from that (and deserves a separate commit), not purely from the
> mechanic unification in the end.

OK.

>
> other changes like this:
>
> -#ifdef CONFIG_X86_FIND_SMP_CONFIG
> +#if defined(CONFIG_X86_FIND_SMP_CONFIG) && defined(CONFIG_X86_32) || \
> +    defined(CONFIG_X86_MPPARSE) && defined(CONFIG_X86_64)
>
> here you fix coding style:
>
> -        /*
> -         * Parse SRAT to discover nodes.
> -         */
> -        acpi_numa_init();
> +       /*
> +        * Parse SRAT to discover nodes.
> +        */
> +       acpi_numa_init();
>

OK

> etc. etc. You do a lot more very useful work than the plain single
> unification patch you sent tells us - all these changes should be split
> up.

YH