Yes within the TF A ecosystem there exist artifacts such as hashes and objects which play a crucial role in facilitating attestation subsequent to a successful boot process.
Snow Services in O'Fallon IL

I'm glad to see that CMake is addressing this common challenge for developers. Thank you for your interest in getting over it and for your ongoing efforts to simplify the construction process!

It seems like a valid point regarding the use of dmb st in sprt_queue.c. Your analysis on the necessity of memory barriers in single CPU scenarios versus multi CPU/Thread scenarios adds depth to the discussion.
Professional Plumbing Services in Atlanta GA

To overcome this problem, it is necessary to determine the cause of the lack of EOI processing for the RAS interrupt and correct the error in the corresponding software or hardware. This often requires thorough examination and analysis of the source code, interrupt handlers and related systems. aa route planner

I solved this issue by myself.
The assert state works normally.

Sorry if it wasn't clear in my original answer, the SAVE_KEYS=1 option (and friends) must be passed on the command line when you build the firmware, not the cert_create tool itself. The tool has no built-in knowledge of which keys it should use, instead it is told so when it is invoked.

Hi @sandrine-bailleux-arm, where do the PEM files go? Because I still can't find them.

Thanks a lot @sandrine-bailleux-arm for the input. I will try it now.

By default, the cert_create tool creates temporary keys that it keeps in RAM just to sign the certificates. These keys are not stored in files on the disk and are thus discarded after the tool exits.
If you want to save them, please have a look at the SAVE_KEYS build option. In your case, adding SAVE_KEYS=1 NON_TRUSTED_WORLD_KEY=ntw.key BL31_KEY=bl31.key to your command line should do what you want. You'll get the private keys in PEM format I believe, from which you can generate the associated public keys using the openssl tool (or equivalent) if needed.

HI Rickdic,
Could you please send this query to the TF-A mailing list ?

I'm not too much versed into RAS error handling, so please take my explanation below with care.
I suggest you send the question to the TF-A ML to get more sensible insights.

I'll post this question to mailing list as suggested.

I am not aware of any specific reason for LOG_LEVEL values being multiple of 10's. I guess at the time we thought we should leave room in between values, just in case we'd like to add more intermediate values in the future. In the end, I think it proved unnecessary but it stayed like that. I don't foresee the need for more log levels today so IMHO it would be OK to change their values to 1,2,3 and so on, as you suggested.

If we set the LOG macros to 1,2,3 so on, we can avoid the mod and division operation when printing the msg to console.

In T498#6134, @semihalf-biernacki-radoslaw wrote:

Hi Haojian,

I have the same problem.
Can you please share the solution you got from maintainer?

Hi Haojian,

Hi Sandrine,

Could you please elaborate on your issue? How are you trying to configure your email address in Gerrit? Are you doing it from the user settings page? Do you have no email address registered on Gerrit at all or are you trying to add a secondary one? When is the error 422 showing?

Hi David,

Hi there!