[Qemu-devel] Re: [PATCH] vvfat mbr fixes

[Lorenzo Campedelli <lorenzo.campedelli@tele2.it>]

Johannes Schindelin wrote:
> Hi,
> 
> On Mon, 24 Sep 2007, Ivan Kalvachev wrote:
> 
>> 2007/9/24, Johannes Schindelin <Johannes.Schindelin@gmx.de>:
>>
>>> On Mon, 24 Sep 2007, Ivan Kalvachev wrote:
>>>
>>>> I had a discussion with Johannes Schindelin over my patch, that I 
>>>> thought is on the maillist, but apparently it wasn't. I'm 
>>>> subscribed, so please don't send me mails directly, gmail web 
>>>> interface could be quite misleading.
>>>>
>>>> So here is the third revision of my patch. Changes include: using 
>>>> more structures instead of fixed byte locations. chs and nt_id. more 
>>>> detailed comments, function name shortened and if(lba) moved to ?: 
>>>> construct.
>>> Almost all my comments went unheeded.
>> I believe that I've answered and addressed all your comments.
> 
> Ooops.  I think I mixed up your patch with the other patch for vvfat that 
> floated around recently.  (Probably because the patch was not inlined...)
> 
> FWIW if we're talking about qemu_vvfat_mbr_v3.patch, I have no more 
> gripes.
> 
> Thanks,
> Dscho
> 
> 
> 
I think you were referring to the small patch I sent.
I actually gave up with it, as I don't see how to make
it in a clean way.

Honestly I found your suggestion to try to have it
less special-casing vvfat a bit puzzling...
vvfat is the only case in which there's any need to
override realpath() behaviour, so I tried to make it
as clear as possible.
Why is it better to affect code paths which don't need
any change?

Regards,
Lorenzo