Hi Daniel, > -----Original Message----- > From: Daniel Baluta [mailto:daniel.baluta@xxxxxxxxx] > Sent: Wednesday, November 8, 2017 12:30 AM > To: David Frey <DFrey@xxxxxxxxxxxxxxxxxx>; contact@xxxxxxxxxxxxxxxxxxx > Cc: linux-iio@xxxxxxxxxxxxxxx > Subject: Re: bmi160 drivers > > Hi David, > > + contact@xxxxxxxxxxxxxxxxxxx > > On Tue, Nov 7, 2017 at 8:49 AM, David Frey <DFrey@xxxxxxxxxxxxxxxxxx> > wrote: > > Hi, > > > > I was working with a vendor kernel based on an older version of Linux > > (3.14) and I backported the bmi160 driver that I found in a recent > > stable version of Linux. That worked fine. More recently, I picked > > up another slightly newer vendor kernel (3.18) for a slightly > > different product and I noticed that it already included a bmi160 > > driver. I was surprised to discover that it was a totally different > > driver. The driver appears to have been written by BoschSensortec > > (who makes the bmi160 hardware) and is available on github here: > > https://github.com/BoschSensortec/drivers > > > > The driver in the regular linux kernel (originally written by Daniel Baluta and > submitted to this mailing list) is around 800 lines of .c and .h files. The Bosch > driver is around 37000 lines of .c and .h files. I'm not sure how the drivers > compare in terms of features. It seems that the Bosch driver is a bit older > (based on the commit logs on github). So I'm wondering: > > > > Why were two drivers written? > > Because at the moment I wrote the driver and sent it upstream there was no > driver from BoshSensortec. Ahh ok, I guess my interpretation of the timing of the two drivers was off. > I think the driver wrote by BoschSensortec supports all the features of the > chip (including hardware fifo), but it also has a lot of non standard attributes > (so not easy acceptable upstream). > > I was planning at some point to implement hw fifo support but got derailed > with other things. > > > Was there a discussion about which driver to include in the Linux > > kernel? (I couldn't find any discussion on this mailing list) > > No, AFAIK. BoshSensortec never sent a request for their driver to be merged > upstream. > > I am not sure about their code license? > > https://github.com/BoschSensortec/drivers/commit/092cab48836cf0ca9ad3 > bb67cf86f5b7aa239d95#diff-a87e4e16f3f8950aafedf0b7b32c613f It seems that they updated their repository to indicate that it is GPL2 licensed. > but perhaps it worths someone looking at enhancing the upstream driver. I > will try to see in the next few weeks if we can work with BoschSensortec for > them to upstream their missing parts from upstream. Thanks for your work on this driver! Hopefully Bosch will see the value in upstreaming their work. > > Hope this helps. > > Daniel ��.n��������+%������w��{.n�����{��(��)�� jg��� �����ݢj����G���� ���j:+v���w�m������w�������h�����٥
