topic Re: P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error in P-Series

P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

mlipczynski — Wed, 10 May 2023 09:25:10 GMT

CPU:P2020

OS: linux-2.6.35

Error information output from linux as below:

kernel: Machine check in kernel mode

kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

We have this problem on 2 from few hundred boards. CPU freq is in spec, it is 1000 MHz.

What's wrong with this CPU, and how to resolve this problem?

Re: P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

yipingwang — Tue, 16 May 2023 03:00:18 GMT

1- does the problem follow the part or board if you do swap test?
2- have you checked [supply voltages, clock clean signals, , temperature, etc] and compared it with our RD?
3- is this an old or new design? if it is old design but new build was the cpu from new batch?

Re: P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

mlipczynski — Tue, 16 May 2023 12:41:53 GMT

1 - no swap test, not possible to be done

2 - not yet

3 - it is old design and cpu is from old batch, this device was working correctly few years, problems started now

Re: P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

yipingwang — Fri, 19 May 2023 07:56:43 GMT

Would you please provide more information regarding the following questions from AE team?

1. Why not possible to do swap test?

2. Also is the failure happening to a new manufactured board or to an already existing board at their customer site?

Re: P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

mlipczynski — Fri, 19 May 2023 11:05:39 GMT

1. I don't have proper tools for it.

2. It is already existing board at their customer site.

Re: P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

mlipczynski — Fri, 19 May 2023 11:11:41 GMT

1. I don't have proper tools to do it.

2. It is already existing board at their customer site.

Re: P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

yipingwang — Tue, 23 May 2023 05:41:50 GMT

I just discussed with the AE team, please refer to the following update.

Did the problem happen to only two boards?

How often did the failure occur?

how long did the board work ok before it started to fail?

Also would you please contract some 3rd party to do swap test for you?

Re: P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

mlipczynski — Tue, 23 May 2023 09:08:50 GMT

1. No tools for that.

2. Only to already existing board at their customer site.

Re: P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

yipingwang — Wed, 24 May 2023 02:37:19 GMT

Would you please answer the following questions?

1. Did the problem happen to only two boards?

2. How often did the failure occur?

3. how long did the board work ok before it started to fail?

4. Also would you please contract some 3rd party to do swap test for you?

Re: P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

mlipczynski — Wed, 24 May 2023 08:21:05 GMT

1. I don't have proper tools for swap test.

2. It is happening to an already existing board at their customer site.

Re: P2020 kernel: Caused by (from MCSR=10000000): Data Cache Parity Error

yipingwang — Wed, 24 May 2023 09:05:54 GMT

Why do you always repeat the same content?

Would you please provide answers for the 4 questions which I asked previously?