[U-Boot] [PATCH v6] arm: ep9315: Return back Cirrus Logic EDB9315A board support

[sergey kostanbaev]

Hi Masahiro,

You were right, .vectors is needed now, but for the other reason. Although
the edb93xx board in uboot doesn't exploit any interrupts but the _start
symbol is needed to calculate the monitor size.
gd->mon_len = (ulong)&__bss_end - (ulong)_start;

Without .vectors _start is 0 and gd->mon_len is about 1.5GB.
Some time ago it worked fine without .vectors section at all.

Adding  .vectors to lds fixes mon_len to actual size and uboot works fine.

So I'll add this section and remove extra lds script.

Thank you for pointing out the problem.

Besrt Regards,
Sergey


On Fri, Jul 25, 2014 at 8:27 PM, sergey kostanbaev <
sergey.kostanbaev at gmail.com> wrote:

> Hi Masahiro,
>
> Sorry I missed you mail since I was on vacation.
>
> On Mon, Jul 7, 2014 at 2:57 PM, Masahiro Yamada <yamada.m at jp.panasonic.com
> > wrote:
>
>> I have two questions:
>>
>> [1] Is this board really working?
>>
>> This patch added a linker script
>> board/cirrus/edb93xx/u-boot.lds
>>
>> But '.vectors' section is missing from it.
>>
>> In my understanding, commit 41623c91b0 expects '.vectors'
>> section handling for all ARM linker scripts.
>>
>> On the time I made the patch it worked fine. I'll test version in master.
>
>
>
>>
>> [2] Why was board/cirrus/edb93xx/u-boot.lds added
>> instead of re-using (or modifying) arch/arm/cpu/arm920t/ep93xx/u-boot.lds
>> ?
>
>
>>
>> 'edb9315a' is the only one board with 'ep93xx' SoC.
>>
>> It means, if you add 'board/cirrus/edb93xx/u-boot.lds',
>> 'arch/arm/cpu/arm920t/ep93xx/u-boot.lds' is never used.
>>
>> Please delete either of them.
>>
>
> Yes, I missed it I'll delete unneeded file.
>
> Thank you for your remarks
>
>