<montjoie>
probably the driver will load and fail self tests
<sigmaris>
cool, I'll give it a try today after work
<sigmaris>
one thing I noticed was it's enabling both crypto0 and crypto1, ISTR reading something about one of them being reserved for secure world usage by TF-A
<sigmaris>
however, modprobe tcrypt results in a lot of warnings/failed tests: https://termbin.com/zk29
lurchi_ is now known as lurchi__
<montjoie>
sigmaris: normal for the second register fail, only one "algo name" could exists
lurchi__ is now known as lurchi_
<montjoie>
sigmaris: could you try to disable crypto0 and see what happen ?
<sigmaris>
ok, btw I'm looking at the TRM and wondering if the addresses in reg are correct - my reading is that crypto0 is at 0xff8b0000 and crypto1 is at 0xff8b8000
<sigmaris>
looks like the addresses at the bottom of each block is the starting address of the block
<sigmaris>
so at 0xff8c0000 is actually INTMEM0
<montjoie>
right, I took bad ones
<sigmaris>
I'll fix my dts and test again...
<sigmaris>
with crypto0 at 0xff8b0000 (disabled) and crypto1 at 0xff8b8000 the driver loads OK but there's still warnings from tcrypt: https://termbin.com/qyfi
<montjoie>
there are 3 resets, so perhaps it will really need them
<montjoie>
but there are not DMA timeout, so irq fired
<montjoie>
tomorow I will add a patch for handling all resets
<montjoie>
thanks for the test sigmaris
<sigmaris>
no problem, happy to test again if you have updates