Hi Dan,
On Wed, Sep 18, 2019 at 6:27 PM Dan Carpenter <[email protected]> wrote:
> Put it in drivers/staging/sdfat/.
It'll conflict with the current exfat staging drivers.
And moreover, I don't think it makes sense to use sdfat naming in mainline.
Samsung uses it since it handles all fat filesystems.
From what I can tell, that's not in mainline's interests:
https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging.git/commit/?h=staging-next&id=58985a9d2d03e977db93bf574a16162766a318fe
What I'm proposing is to remove the current exfat drivers and add
sdfat-based one(that I removed fat16/32 handlings and renamed to
exfat).
> But really we want someone from Samsung to say that they will treat
> the staging version as upstream.
Agreed.
Perhaps Namjae didn't pick up our questions with all those mails we
sent during last few days.
Maybe ping him again?
Thanks.
On Wed, Sep 18, 2019 at 06:53:49PM +0900, Ju Hyung Park wrote:
> Hi Dan,
>
> On Wed, Sep 18, 2019 at 6:27 PM Dan Carpenter <[email protected]> wrote:
> > Put it in drivers/staging/sdfat/.
>
> It'll conflict with the current exfat staging drivers.
Use Kconfig.
> And moreover, I don't think it makes sense to use sdfat naming in mainline.
The directory doesn't need to be permanent.
regards,
dan carpenter
On Wed, Sep 18, 2019 at 7:09 PM Dan Carpenter <[email protected]> wrote:
> Use Kconfig.
Not just that.
There are a lot of non-static functions that's not marked ex/sdfat-specific.
(which we would have to clean it up eventually)
Even with sdFAT base, there are some non-static functions named as exfat.
Figuring out a solution for this is pretty pointless imho when one of
the drivers will be dropped soon(ish) anyways.
Thanks.
On Wed, Sep 18, 2019 at 07:46:25PM +0900, Ju Hyung Park wrote:
> On Wed, Sep 18, 2019 at 7:09 PM Dan Carpenter <[email protected]> wrote:
> > Use Kconfig.
>
> Not just that.
> There are a lot of non-static functions that's not marked ex/sdfat-specific.
> (which we would have to clean it up eventually)
Then clean them up :)
> Even with sdFAT base, there are some non-static functions named as exfat.
Then just force both filesystems to only be built as a module and all
should be fine, right?
> Figuring out a solution for this is pretty pointless imho when one of
> the drivers will be dropped soon(ish) anyways.
Given we only have one filesytem that is submitted in patch form, I
think people are making a lot of noise over nothing :)
thanks,
greg k-h