intermittent RTIO PLL lock failure #66

New Issue

sb10q · 2020-07-20T13:54:22+08:00

sb10q commented

2020-07-20 13:54:22 +08:00

When using internal clock...

Core 0 panic at runtime/src/main.rs:125:9: RTIO PLL failed to lock

Happens rarely (few % of the boots).

When using internal clock... ```text Core 0 panic at runtime/src/main.rs:125:9: RTIO PLL failed to lock ``` Happens rarely (few % of the boots).

sb10q commented

2020-07-20 14:50:24 +08:00

Poster

d65e893d1c might fix it.
Let me know if you still see the problem.

The delay while holding the reset is most likely not necessary, according to the datasheet a 5ns long reset pulse is sufficient.

Could be incorrect timer operation with 1ms delay (@astro)?

https://git.m-labs.hk/M-Labs/artiq-zynq/commit/d65e893d1c7b16a833f093e155a141b6fc1bd207 might fix it. Let me know if you still see the problem. The delay while holding the reset is most likely not necessary, according to the datasheet a 5ns long reset pulse is sufficient. Could be incorrect timer operation with 1ms delay (@astro)?

astro commented

2020-07-23 04:46:54 +08:00

d65e893d1c might fix it. Let me know if you still see the problem.

Still seeing it every boot.

Could be incorrect timer operation with 1ms delay (@astro)?

It was delaying "up to" instead of "at least" the specified milliseconds. I'm pushing a correction but that is still not fixing this issue.

> d65e893d1c might fix it. Let me know if you still see the problem. Still seeing it *every* boot. > Could be incorrect timer operation with 1ms delay (@astro)? It was delaying "up to" instead of "at least" the specified milliseconds. I'm pushing a correction but that is still not fixing this issue.

pca006132 commented

2020-07-23 09:45:53 +08:00

This does not happen for my board (connected to zeus). Is it possible that this is somewhat related to the hardware?

sb10q commented

2020-07-25 11:16:41 +08:00

Poster

I'm now resetting the PLL again and retrying when it fails to lock. Does that help?

astro commented

2020-07-27 07:43:15 +08:00

I'm now resetting the PLL again and retrying when it fails to lock. Does that help?

No. It is now retrying twice per second without end.

>I'm now resetting the PLL again and retrying when it fails to lock. Does that help? No. It is now retrying twice per second without end.

sb10q commented

2020-07-27 08:21:58 +08:00

Poster

Are you using the internal RTIO clock as you should? Do you have a mismatch between gateware and firmware?

sb10q commented

2020-08-04 14:04:46 +08:00

Poster

@pca006132 have you seen any RTIO PLL problems (including retries) since the timer was fixed? Seems to me everything is working correctly now.
We can keep the retry code as it may be useful in case the user forgets to connect the external clock before boot.

@pca006132 have you seen any RTIO PLL problems (including retries) since the timer was fixed? Seems to me everything is working correctly now. We can keep the retry code as it may be useful in case the user forgets to connect the external clock before boot.

pca006132 commented

2020-08-04 14:11:39 +08:00

@pca006132 have you seen any RTIO PLL problems (including retries) since the timer was fixed? Seems to me everything is working correctly now.
We can keep the retry code as it may be useful in case the user forgets to connect the external clock before boot.

No, seems working correctly.

> @pca006132 have you seen any RTIO PLL problems (including retries) since the timer was fixed? Seems to me everything is working correctly now. > We can keep the retry code as it may be useful in case the user forgets to connect the external clock before boot. No, seems working correctly.

sb10q closed this issue

2020-08-04 14:12:48 +08:00

Sign in to join this conversation.