uvos | we require the udev file (well it needs to be updated for the new driver) | 13:47 |
---|---|---|
uvos | the info in dts has no effect, as this driver dosent pass it on to userspace | 13:47 |
uvos | it lacks the reuqired support | 13:47 |
uvos | also all of this is irrelivant on n900 | 13:47 |
uvos | as we disable the accel to avoid h-d trying to rotate breaking x | 13:48 |
sicelo | what are you talking about? how is iio-sensor-proxy able to correctly tell the device orientation if userspace doesn't get those values? | 14:08 |
uvos | it cant from dts | 14:08 |
uvos | because the driver dose not pass via the sysfs intface (this is not implemented) | 14:08 |
sicelo | without the matrix in dts, the orientation reported by iio-sensor-proxy is wrong | 14:09 |
uvos | unless the driver was updated since i last looked | 14:09 |
uvos | if the featrue was later added then sure | 14:09 |
uvos | but when i set up iio-sensor-proxy for d4 the driver did not support the matrix at all | 14:09 |
uvos | this is why we have the udev rules | 14:09 |
uvos | they inject the matrix later in the same place | 14:10 |
uvos | (gets picked up by udev from the driver in the supported path) | 14:10 |
sicelo | when i added this, it worked fine with i-s-p. i use it in pmOS too, and there i don't have the udev rule | 14:10 |
uvos | well then support was added | 14:11 |
uvos | see https://www.kernel.org/doc/Documentation/ABI/testing/sysfs-bus-iio | 14:11 |
uvos | /sys/bus/iio/devices/iio:deviceX/mount_matrix was not supported by our diver at the time (5.9 or so) | 14:12 |
uvos | if it is now we can drop the rules, yes | 14:12 |
sicelo | https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/drivers/iio/accel/st_accel_core.c?id=3d8ad94bb175c2de7200569bb706d67c45903838 | 14:17 |
sicelo | yes, looks like it arrived in 5.13 | 14:17 |
uvos | sicelo: great, yeah we should drop all the rules then | 16:21 |
uvos | wait no, there was another issue | 16:22 |
uvos | on hildon expects normal orientaiton (y up) to be lanscape | 16:22 |
uvos | but on mapphones as they are native portrait y up is ofc protrait | 16:22 |
uvos | so this helps n900 only | 16:23 |
uvos | or we modify hildon to expect normal orientation to be portrait | 16:24 |
uvos | since pretty mutche every device ever made beside the n900 is going to be that way | 16:24 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!