Re: Using regmap_update_bits to update a write only register

[Lars-Peter Clausen <lars@metafoo.de>]

On 03/06/2015 06:36 PM, Lars-Peter Clausen wrote:
> On 03/06/2015 06:26 PM, Daniel Baluta wrote:
> [...]
>> I can reproduce the problem with:
>>
>> static struct reg_default xxx_reg_defaults[] = {
>>          { XXX_REG_CTRL0,  0x00 },
>>          { XXX_REG_CTRL1,  0x00 },
>>          { XXX_REG_STATUS, 0x00 },
>> };
>>
>> but, not if the reg default definition is:
>>
>> static struct reg_default xxx_reg_defaults[] = {
>>          { XXX_REG_STATUS, 0x00 },
>>          { XXX_REG_CTRL0,  0x00 },
>>          { XXX_REG_CTRL1,  0x00 },
>>    };
>>
>> Is this normal?
>
> That's a rhetorical question, right?
>
> It might be that there is a bug when growing a rbblock to the left. It
> probably went unnoticed because everybody has their reg defaults ordered in
> ascending order.
>
> Try to put a few debug printks into regcache_rbtree_write() and
> regcache_rbtree_insert_to_block() to see what exactly is going on when a new
> register is inserted into the block. How do base_reg and top_reg change.

Try to replace the bitmap_shift_right() in regcache_rbtree_insert_to_block() 
with shift bitmap_shift_left(), I suspect that will do the trick.

- Lars