Calling set_sys_clock_khz(438000, false) hangs the Pico with 1.3V #102

New issue

Closed

opened 2021-02-08 14:22:45 -06:00 by Hermann-SW · 4 comments

Hermann-SW commented

2021-02-08 14:22:45 -06:00

(Migrated from github.com)

For code that determines maximal working overclocking for a single Pico I have currently to start with highest value of 437999 because of this bug.
Just did a run with 1.2V and it kept going down from 437999 until it reached 404000 and hangs.

For code that determines maximal working overclocking for a single Pico I have currently to start with highest value of 437999 because of this bug. Just did a run with 1.2V and it kept going down from 437999 until it reached 404000 and hangs.

lurch commented

2021-02-08 15:49:39 -06:00

(Migrated from github.com)

I guess you ought to expect undefined behaviour (e.g. crashes or hangs) at such high overclocks?
Or have I misunderstood and you're actually reporting something else?

I guess you ought to expect undefined behaviour (e.g. crashes or hangs) at such high overclocks? Or have I misunderstood and you're actually reporting something else?

Hermann-SW commented

2021-02-08 15:57:38 -06:00

(Migrated from github.com)

I expected to see whether setting and reading gpio works, not crashes.
I identified the reason for the crashes, I forgot to call set_sys_clock_48mhz() before doing printf() to USB minicom. So the reason for the hang seems to be printing at very high clocks to USB subsystem. Going down to lower clocks before printing avoids the hang.

I expected to see whether setting and reading gpio works, not crashes. I identified the reason for the crashes, I forgot to call set_sys_clock_48mhz() before doing printf() to USB minicom. So the reason for the hang seems to be printing at very high clocks to USB subsystem. Going down to lower clocks before printing avoids the hang.

Wren6991 commented

2021-02-09 08:22:12 -06:00

(Migrated from github.com)

I have currently to start with highest value of 437999 because of this bug.

What is the bug? You have described the hardware not functioning correctly at 3x its rated speed, and yes, this is why the rated speed is 3x lower than this. Please explain the software problem you would like to see fixed.

> I have currently to start with highest value of 437999 because of this bug. What is the bug? You have described the hardware not functioning correctly at 3x its rated speed, and yes, this is why the rated speed is 3x lower than this. Please explain the **software** problem you would like to see fixed.

aallan commented

2021-02-09 08:53:09 -06:00

(Migrated from github.com)

The rated maximum clock for RP2040 is 133MHz. If you're overclocking you're entirely on your own. Unexpected behaviour like crashes should be expected.