Re: Latest brk patchset

[Yinghai Lu]

H. Peter Anvin wrote:
> Yinghai Lu wrote:
>> H. Peter Anvin wrote:
>>> On this subject, what is the point of the 64 K "slop space", and if it
>>> is necessary, why not just put it as a RESERVE_BRK() somewhere (*with* a
>>>  significant comment as to its necessity) instead of putting it as a
>>> hack in the linker script?
>>>
>>
>> 32bit already have that
>>
>> +INIT_MAP_SIZE = (PAGE_TABLE_SIZE + ALLOCATOR_SLOP) * PAGE_SIZE_asm
>> +RESERVE_BRK(pagetables, INIT_MAP_SIZE)
>>
>> in head_32.S
>>
>> so need to update vmlinux_32.lds.S
>>
> 
> What about 64 bits?  Also, do we really need the "allocator slop", or
> can we push that into the user location, which is where it really
> belongs?  What does use this?

64 bit does not has this problem.
32 bit we already removed bootmem there, that is not needed. 

ALLOCATOR_SLOP is needed...

+ *     (KERNEL_IMAGE_SIZE/4096) / 1024 pages (worst case, non PAE)
+ *     (KERNEL_IMAGE_SIZE/4096) / 512 + 4 pages (worst case for PAE)


YH
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/