Hi,
Customer would like to know if there are any
unprotected bits (no parity detection) in the ARM L1 program or data cache in
the Keystone II 66AKE05.
Best,
Rick Nardone
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.
Hi,
Customer would like to know if there are any
unprotected bits (no parity detection) in the ARM L1 program or data cache in
the Keystone II 66AKE05.
Best,
Rick Nardone
Rick,
Short answer is no, but more details can be found on the wiki http://processors.wiki.ti.com/index.php/Keystone_Error_Detection_and_Correction_EDC_ECC#ARM-A15_Error_Detection_and_Correction_.28ECC.29_Keystone_Support
Note that the GHB and Indirect Predictor RAMs for the L1 not being protected would only result as a predictor miss if it were to be corrupted from something like an SER. Where the data is stored is protected.
Best Regards,
Chad
Paul,
As mentioned the only bits in the L1 that are not protected are used for prediction (i.e. predicting what needs to be accessed next so it can grab the data early.) If this 'fails' it does not result in any use/loss/corruption of actual data. It just means that they access may be slow for that one prediction. No worse than the prediction being wrong which is going to happen pretty much every time you access data that goes outside the current access pattern.
This should not be calculated in SEE.
Best Regards,
Chad