Hi Vikas,
On 24 November 2014 at 17:40, vikasm vikas.manocha@st.com wrote:
Hi Simon,
On 11/24/2014 07:51 AM, Simon Glass wrote:
Hi Vikas,
On 18 November 2014 at 11:59, vikasm vikas.manocha@st.com wrote:
Thanks Simon,
On 11/17/2014 09:30 PM, Simon Glass wrote:
Hi Vikas,
On 18 November 2014 00:17, Vikas Manocha vikas.manocha@st.com wrote:
This patchset fixes the pl01x driver esp for pl011 baudrate & line control.
Vikas Manocha (5): serial: pl01x: pass pl01x_type to set baudrate serial: pl01x: fix pl011 baud rate configuration serial: pl01x: move all line control at same place serial: pl01x: disable as per type of pl01x serial: pl01x: avoid pl01x type check two times
drivers/serial/serial_pl01x.c | 46 +++++++++++++++++++++-------------------- 1 file changed, 24 insertions(+), 22 deletions(-)
Great to see this series. If you have not tested on p1010 I can do this in a week or so (am travelling). At least a few of these looks like bug fixes to my recent refactor so I could bring them through the DM tree if no one else picks them up.
I was able to test it only for pl011, don't have any board with pl010.
Sadly I was wrong, I only have pl011 also. It makes me wonder how it worked so nicely for me without your patches.
Please find below the possible explanation for all patches:
serial: pl01x: pass pl01x_type to set baudrate
- This patch only affect if DM is not defined.
If you are using driver model for pl011, pl011 will work without this patch.
serial: pl01x: fix pl011 baud rate configuration
- If pl011 is already configured to correct baudrate or LCR register is
being written later.
serial: pl01x: move all line control at same place
- this patch does not affect the functionality.
serial: pl01x: disable as per type of pl01x
- pl011 might work without disabling it properly at start.
pl010 control register(disable bit) location is reserved in case of pl011.
serial: pl01x: avoid pl01x type check two times
- does not affect the functionality.
I didn't mean for you to go to that much trouble, but thank you.
Raspberry Pi is running without driver model (I tested it both ways) so I guess I'm just lucky (or actually unlucky since I might have noticed these problems).
When Stephen ACKs my Raspberry Pi driver rmodel series then I'll pull this series in to DM.
Regards, Simon