[RFC 4/5] bcm2835-dma: add support for slave_sg transfer mode

[Stephen Warren]

On 11/03/2014 03:07 PM, Lee Jones wrote:
> You're missing LKML and the DMA Maintainer.
> 
> Use ./scripts/get_maintainer.pl
> 
> On Wed, 29 Oct 2014, Piotr Król wrote:
> 
>> Add bcm2835_dma_prep_slave_sg to suport bcm2835-mmc DMA transfers, what
>> can improve throughput and system CPU load.
>>
>> Based on Gellert Weisz <gellert at raspberrypi.org> patch.
>>
>> Signed-off-by: Piotr Król <piotr.krol at 3mdeb.com>
> 
> I'll let you fight this out with Stephen, but personally I think you
> are right to remove the other Signed-off-bys _if_ you made significant
> changes before submitting.  If you would like to keep them, as it's
> 'the nice thing to do', you need to get them to review your changes
> and indicate that they are happy with them.
> The alternative is to add another tag between theirs and yours (very
> succinctly) describing your changes.

I don't believe there's any requirement at all to get the original patch
authors to OK the changes you made to the patch. It's fine to ask the
original author for another round of review though. It is certainly good
practice to describe the changes you made. If the changes are small, you
can as Lee says throw a line into the middle of the tag paragraph, like:

Signed-off-by: Random J Developer <random at developer.example.org>
[lucky: struct foo moved from foo.c to foo.h]
Signed-off-by: Lucky K Maintainer <lucky at maintainer.example.org>

... but if they're much more involved, I often just write a more
detailed description in the main body of the patch description.

If you did make *significant* changes, it probably is reasonable to
claim authorship of the patch and remove the original s-o-b lines. But a
"Based on work by ..." note would be useful.