Hi Manuel
On 11/24/20 1:50 PM, Manuel Reis wrote:
To: Maxime Coquelin <mcoquelin.stm32@xxxxxxxxx>
To: Alexandre Torgue <alexandre.torgue@xxxxxx>
Cc: linux-kernel@xxxxxxxxxxxxxxx
CC: Michael Opdenacker <michael.opdenacker@xxxxxxxxxxx>
Hi there,
Mainline stable kernel 5.9.10 hangs indefinitely on a STM32MP157A-DK1 Discovery Kit board.
Built plain vanilla 5.9.10 stable kernel for multi_v7_defconfig (set compression to XZ) using arm-linux-gnueabi-gcc (Ubuntu 10.2.0-8ubuntu1) 10.2.0.
Downloaded it to the board memory via tftp running U-Boot v2020.07. After boot, kernel initiates and prints several messages until it hangs on:
[ 2.692879] stpmic1 1-0033: PMIC Chip Version: 0x10
[ 2.704158] vddcore: supplied by regulator-dummy
[ 2.710304] vdd_ddr: supplied by regulator-dummy
[ 2.716414] vdd: supplied by regulator-dummy
[ 2.722355] v3v3: supplied by regulator-dummy
[ 2.728033] v1v8_audio: supplied by v3v3
[ 2.734287] v3v3_hdmi: supplied by regulator-dummy
[ 2.741035] vtt_ddr: supplied by vdd_ddr
[ 2.743833] vdd_usb: supplied by regulator-dummy
[ 2.751332] vdda: supplied by regulator-dummy
[ 2.757371] v1v2_hdmi: supplied by v3v3
No other information or indication is given, even though I added kernel debugging features such as the ones in "Kernel hacking ->Debug Oops, Lockups and Hangs".
Any help would be appreciated. Let me know if I can provide any further information.
It has been introduced by commit aea6cb99703e ("regulator: resolve supply after creating regulator") and should fixed by this one:
cf1ad559a20d ("regulator: defer probe when trying to get voltage from unresolved supply").
Should be taken in stable tree.
regards
alex
Cheers,
Manuel