[U-Boot] commit 9a9d66f5ef break U-Boot Env

[Heiko Schocher]

Hello Horatiu,

Am 18.01.2019 um 10:18 schrieb Horatiu Vultur:
> Hi Heiko,
> 
> The 01/18/2019 09:34, Heiko Schocher wrote:
>> Hello Horatiu,
>>
>> just tried current mainline U-Boot on imx6ull based board with
>> Env in SPI NOR. Your commit "9a9d66f5eff0"
>>
>> env: add spi_flash_read_env function
>>
>> breaks reading Env from SPI NOR in my case...
>>
>> If I interprete your patch correct, you only read until the 2 0x0
>> in the env ... but for the env crc, the hole CONFIG_ENV_SIZE
>> must be read ... else missing for crc calculation ...
>>
>> Hmm... following patch fixes for me the issue:
>>
>> hs at threadripper:git  [master] $ git diff env/sf.c
>> diff --git a/env/sf.c b/env/sf.c
>> index b3dec82c35..dc4b56d4a4 100644
>> --- a/env/sf.c
>> +++ b/env/sf.c
>> @@ -98,7 +104,7 @@ static int spi_flash_read_env(struct spi_flash *flash, u32 offset, size_t len,
>>          u32 addr = 0;
>>          u32 page_size = flash->page_size;
>>
>> -       memset(buf, 0x0, len);
>> +       memset(buf, 0xff, len);
>>          for (int i = 0; i < len / page_size; ++i) {
>>                  int ret = spi_flash_read(flash, offset, page_size,
>>                                           &((char *)buf)[addr]);
>>
>> But if I make a bigger ENV_SIZE this may break again...
> 
> That's correct, the buf should be initialized to 0xff.
> 
> But I don't see why increasing the ENV_SIZE would break it again. Can
> you explain it why? Thank you.

No, you are correct. CONFIG_ENV_SIZE is passed to spi_flash_read_env()
so this should be fine.

I send a patch...

bye,
Heiko
-- 
-- 
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-52   Fax: +49-8142-66989-80   Email: hs at denx.de