Re: [PATCH 2/2] x86, boot: Optimize the elf header handling.
From: Eric W. Biederman
Date: Sun Jul 01 2012 - 14:25:59 EST
"H. Peter Anvin" <hpa@xxxxxxxxx> writes:
> On 07/01/2012 10:09 AM, Eric W. Biederman wrote:
>> I am suspecting something similar. At the moment I am trying
>> to remember how to get early printk working in misc.c so I can
>> poke around a bit more.
>> The theory that I was working on, that used to be true, and
>> seems to be true of everything except the percpu section is
>> that objcopy just does the right thing when creating vmlinux.bin.
>> vmlinux.bin being what we compress.
>> I don't see anything obviously wrong with the headers of vmlinux.bin
>> although we still unnecessarily have section headers in that file.
> We leave the section headers in for the benefit of Xen IIRC.
Well I see what is going on now.
In the vmlinux the paddrs and the poffsests stop tracking when they
come to the percpu PT_LOAD segment.
It looks like we give the percpu segment a 4K alignment in memory
and a 2M alignment in the file.
It looks like something is causing ld to compute an unfortunate physical
load address for the percpu section.
At first glace I would say that is a bug. Either we need the alignment
or we don't. But if we need the alignment we certainly need it to
be respected for the real memory locations of the init symbols..
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/