Hi, in preparation for the next hardware version (expected to be with TILER and without ssd2858), I have updated the device tree file names a little.
We now have two major Pyra device tree binaries:
a) omap5-letux-cortex15+pyra.dtb expects a display board with ssd2858 rotator chip and BOE panel that will be programmed through the panel-mipi-debug driver by the /root/ssd2858 script
b) new: omap5-letux-cortex15-v5.1+pyra-v5.1.dtb expects a display board with bridged ssd2858, i.e. a non-rotating board to use the TILER for rotation. It directly probes the boe-w677l panel driver and does not need a shell script to start.
Despite its name currently also works on the 5.0 prototype boards (but this may change. Hence the differentiating name scheme).
There are some other pyra related .dtb files but you can ignore them since they are for my hardware experiments and old prototypes and will be retired in the future.
How to use: either edit uEnv.txt on the boot partition or rename/copy the file so that it has the default name "omap5-letux-cortex15+pyra.dtb"
Patch behind this change:
http://git.goldelico.com/?p=gta04-kernel.git;a=blob;f=arch/arm/boot/dts/omap...
It is (will be - upload to server is not yet finished) included in
http://download.goldelico.com/gta04-kernel/letux-4.8-rc1/device-trees.tbz
BR, Nikolaus
Hi,
On 11/08/16 13:25, H. Nikolaus Schaller wrote:
Hi, in preparation for the next hardware version (expected to be with TILER and without ssd2858), I have updated the device tree file names a little.
Did you test that TILER will work fine enough? Is there some issue with ssd2858?
There's also errata i878 - "MPU Lockup With Concurrent DMM and EMIF Accesses" which you could hit there.
Tomi
Hi Tomi,
Am 11.08.2016 um 13:41 schrieb Tomi Valkeinen tomi.valkeinen@ti.com:
Hi,
On 11/08/16 13:25, H. Nikolaus Schaller wrote:
Hi, in preparation for the next hardware version (expected to be with TILER and without ssd2858), I have updated the device tree file names a little.
Did you test that TILER will work fine enough?
Matthijs is still testing and fixing things. He has it working on the HDMI output but has no't tested on the LCD port.
Is there some issue with ssd2858?
Yes, we are still stuck with converting the panel-mipi-debug based driver into a really working driver. There is no difference in the DSS, DSI, ssd2858 registers, but we get a black picture and slightly different timing with oscilloscope. So it must be some initialization sequence issue we have not understood yet.
If the TILER works well enough, we won't need the ssd2858 any more.
There's also errata i878 - "MPU Lockup With Concurrent DMM and EMIF Accesses" which you could hit there.
Hm. I have no errata document for that on OMAP5. The highest number I can find is ca. i820.
BR, Nikolaus
-
H. Nikolaus Schaller -
Tomi Valkeinen