It seems that there is not any patch regarding u-boot (or TI firmware) running at "wrong frequency" (1.25 GHz) on 1 GHz parts.
Are there any news?
Thank you.
Best regards,
Emanuele
This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
It seems that there is not any patch regarding u-boot (or TI firmware) running at "wrong frequency" (1.25 GHz) on 1 GHz parts.
Are there any news?
Thank you.
Best regards,
Emanuele
I would say that by using >=09.01.00.004 ti-linux-firmware I get rid of these messages:
cpufreq: cpufreq_online: CPU0: Running at unlisted initial frequency: 1250000 KHz, changing to: 1000000 KHz
So seems to be fixed. Can you confirm?
Which of these fixes, reported in the commit message, is related to this problem (may be the second last?)?
- Additional Klocwork MISRA-C issues addressed
- am62ax: LPM DeepSleep support
- am62x: LPM Partial IO mode support
- am62x/am62ax/am62px: ARM PLL initialization skipped in pm init
- am62px: MCU_OBSCLK_MUX bitfield issue fix
I assume that actually the clock is correct before linux starts.
But when is it set correctly?
Before or after dram_init in R5 SPL?
At which frequency was (before this firmware) running the dram_init function before this firmware?
Hi Emanuele,
The fix involves both Firmware update and u-boot project update.
On the DM firmware side, below change is needed:
- am62x/am62ax/am62px: ARM PLL initialization skipped in pm init
On the u-boot project, below patch is needed:
- https://git.ti.com/cgit/ti-u-boot/ti-u-boot/commit/?h=ti-u-boot-2023.04&id=ded83c57a2cfe9ec29628f52734c8f92ff07a9db
Regards,
Vishal
Thanks a lot Vishal,
I'll be back soon confirming that this resolve our "issue". In first instance, our issue, was to run AM62x to "correct" frequency... and we hope that should help us also about this issue: e2e.ti.com/.../4827908
Hello Hong,
now we get rid of cpufreq messages from linux kernel.
About the other issue linked it doesn't solve. But, at least, this reduces what we are searching for.
Thanks,
Emanuele