[dm-devel] [PATCH][RFC] dm: Do not open log and cow device read-write for read-only mappings

[Milan Broz]

On 02/15/2011 05:12 PM, Tejun Heo wrote:
> On Tue, Feb 15, 2011 at 05:05:48PM +0100, Milan Broz wrote:
>> On 02/15/2011 04:50 PM, Tejun Heo wrote:
>>>> That's why I'm arguing EACCES is not a good error to return and EROFS is
>>>> more appropriate.
>>>
>>> Frankly, I don't really mind one way or the other but EROFS isn't
>>> usually used in those areas.  It might make sense for this use case
>>> and then there will be cases it just feels awkward.  This being a dm
>>> thing, wouldn't it be just better to let dm massage the return value?
>>
>> It is not DM thing. That code was checking for generic block device.
>> No DM there (it was from cryptsetup code but not related to DM part).
> 
> Hmmm... I'm confused now.  Where was that -EROFS from then?  I don't
> recall changing -EROFS to -EACCES.  What did I miss?

Well, I am also not sure about that.

But the problem is that read-write open fails now while it worked before.
(TBH I have no idea when that EROFS fallback worked - because the code
opened device RW, issued EROGET ioctl and set read-only... for years.)

Anyway I think EROFS is used on block devices, just grep kernel source.

Milan