Canberk Topal

This does not take into account the type of memory request. That means we won't differentiate between seeing an integrity error on the memory interface while doing a store or...

While simulating with this version I found a minor mismatch regarding two back to back integrity errors (with load instructions). RTL side does not take the faulty data from memory...

Test pass rate for this memory test is now around 60%. We hit internal NMI coverpoints though. I suspect almost all failures are related with the issue that I mentioned...

Thanks @hcallahan-lowrisc for having a look, I updated the commit messages for almost all of them and dropped that one unnecessary commit about catching external RVFI signals -that one should...

I think this PR increases our test time for mem_error_test (after changing it with the new sequence we probably are injecting way more memory errors) and tips it over 30...

@GregAC can you have a look at https://github.com/lowRISC/ibex/pull/1811/commits/66d8d0fc0b087fa48f74cf8e0d343b582e2ad8e9 to see if this makes sense?

>remaining 2022-09-20 3

> remaining 2022-10-14 1

Okay I think I understood what's going on. The first problem is about how we let RVFI know that there is a problem with the issued instruction. However, when I...

In OTBN, this checking of responses to integrity failures is done as this: 1. Read scrambled data from RAM using `encrypt_sram_addr` and `decrypt_sram_data` from `sram_scrambler_pkg`. 2. Apply 1 or 2...