Andrew Jeffery

Honestly, I'm not sure. I hadn't heard of Google's Colab before, and I haven't touched fpos in about 6 years now. The Makefile did the right things back when I...

Hi, sorry, this issue slipped through the cracks. This looks like a logic bug in the probe subcommand, it currently it requires that we successfully probe the soc before listing...

Ah, no, it's just the intent is poorly documented. I remember now. `--list-interfaces` is listing the bridges we know the SoC supports, and it varies from SoC-to-SoC. To determine which...

Okay, some immediate thoughts are there are at least three layers at which we should be locking. We need locking over: 1. Bridge operations 2. The fingerprinting operation, 3. Use...

actually, we could probably get away without the fingerprint lock, that's not actually achieving much as the operations are all reads.

Hello! Thanks for the interest. I'll share a bit of background too: * I started culvert internally at IBM as a demonstration of the issues outlined in [CVE-2019-6260](https://nvd.nist.gov/vuln/detail/CVE-2019-6260) prior to...

Quite a bit to work through there :) I took a look at the fwupd codebase after you opened the issue and I tend to agree the way forward is...

> I'm less keen on this; various vendors have tried to convince us to exec random binaries for enumeration, firmware writing and security validation, and I think in all honesty...

Okay, I've left a bunch of comments on the google doc. Let me know what you think.

> I think a good starting point would be to add the checks for the devmem bridge, e.g. Read DWORD HICRB (0x100h) and check that bit 6 is set --...