<mszyprow>
(although it still work fine with 0 - IRQ_TYPE_LEVEL_NONE)
<mszyprow>
the warning was added in v4.18-rc1 or v4.17-rc1
<mszyprow>
just to force everyone to fix dts to use proper values
<mszyprow>
use "interrupts = <GIC_SPI 219 IRQ_TYPE_LEVEL_HIGH>, <GIC_SPI 74 IRQ_TYPE_LEVEL_HIGH>, <GIC_SPI 117 IRQ_TYPE_LEVEL_HIGH>;" for mali
<memeka>
so <0 219 0> should be <GIC_SPI 219 IRQ_TYPE_LEVEL_HIGH
<memeka>
yes :)
<memeka>
you were faster :))
<memeka>
mszyprow: also:
<memeka>
[ 2.552373] s5p-mfc 11000000.codec: Direct firmware load for s5p-mfc-v8.fw failed with error -2
<memeka>
[ 2.560912] s5p_mfc_load_firmware:73: Firmware is not present in the /lib/firmware directory nor compiled in kernel
<memeka>
i remember when this patch was added :|
<memeka>
firmware actually is there
<memeka>
but rootfs is not mounted at that point i think
<memeka>
so it's a stupid message
<mszyprow>
that warning comes because rootfs is not yet mounted
<mszyprow>
also on my todo list to convert that to use deferred probe
<mszyprow>
i was surprised that kernel firmware loader doesn't support deferred probe and don't check if rootfs is mounted...
<memeka>
:) thx the intrerrupt spec did eliminate the warnings
<memeka>
mszyprow: last thing :)
<memeka>
so when exporting buffers to DRM
<memeka>
should i use sizeimage or bytesused? is there a problem if i use sizeimage?
<mszyprow>
buffer can be larger than the size of the real video data
<memeka>
(i don't get any error in ssh playing videos on DRM on any of the 2 planes now ... but i can't see if there is an actual image :D)
<memeka>
yeah so sizeimage should be ok
<mszyprow>
so it is better to use the same size as the one used for allocating the buffer
<memeka>
yup
<mszyprow>
it will be rounded up to page size anyway
<memeka>
mszyprow: on primary plane, CPU usage is ~17% playing 1080p video; on overlay it's ~26% same file
<memeka>
so either one of them is not showing anything on screen
<memeka>
or the overlay is using more CPU because it flickers and shows the console too behind
Vasco_O is now known as Vasco
genii has joined #linux-exynos
nighty- has joined #linux-exynos
<willmore>
mszyprow, it may be that the firmware loader doesn't wait because the firmware may be built into the kernel and may be needed to mount root--maybe net booting over a USB network connection that needs a firmware?
<willmore>
Or I'm worng. That's likely, too.
Vasco is now known as Vasco_O
nighty- has quit [Quit: Disappears in a puff of smoke]
mszyprow_ has joined #linux-exynos
mszyprow has quit [Read error: Connection reset by peer]
<memeka>
mszyprow_ page fault still occurs on 4.18