Re: [PATCH 1/2] media: imx: Round line size to 4 bytes

[Dorota Czaplejewicz]

Hi,

On Wed, 14 Sep 2022 17:29:48 +0300
Dan Carpenter <dan.carpenter@xxxxxxxxxx> wrote:

> On Wed, Sep 14, 2022 at 01:58:08PM +0200, Dorota Czaplejewicz wrote:
> > This has been broken in e352833d, which does not actually copy the logic
> > from 056f9af9.
> >   
> 
> There is no commit 056f9af9.  Always the human readable subject for the
> patch.  Also use 12 characters and not 8 so we don't get two commits
> with the same hash.

Thanks. Unbeknownst to me, no one ever applied or rejected the patch series I sent and saw under 056f9af9.

I'll send a fresh series in a moment.

Cheers
--Dorota
> 
> Please add a Fixes tag.
> Fixes: e352833d32e6 ("media: staging: media: imx: imx7-media-csi: Import format helpers")
> 
> A Fixes tag can be used to automatically determine if a patch needs a
> backport or not.  It's just useful information as well.
> 
> > Section 13.7.6.13 "CSI Image Parameter Register" of the
> > i.MX 8M Quad Applications Processors Reference Manual
> > states that the line size should be divisible by 8 bytes.
> > However, the hardware also accepts sizes divisible by 4 bytes.
> > 
> > This patch accepts line sizes divisible 4-bytes in non-planar mode.
> > 
> > Signed-off-by: Dorota Czaplejewicz <dorota.czaplejewicz@xxxxxxx>
> > ---
> > Hi,
> > 
> > in recent kernels, I've noticed that the Librem 5 camera driver no longer works. The s5k3l6xx out of tree sensor driver uses sizes divisible by 4 and not by 8, for which I submitted the patch 056f9af9 . The new kernels no longer accept sizes divisible by 4.
> >   
> 
> This information needs to be in the commit message.  Otherwise the
> commit message just sounds like theory and abstraction which we do not
> care about.  We do care about real hardware which is not working and
> especially if it used to work.
> 
> I'm not really qualified to review the details of this patch.
> Laurent is best qualified to review this.  He'll get this message
> because he's subscribed to the lists but it's also good to CC him as
> well because he's probably drowning in email like the rest of us.  I've
> added him.
> 
> regards,
> dan carpenter
>

Attachment: pgppNqgrPvos9.pgp
Description: OpenPGP digital signature