Re: Mirage-net-xen and Generic Segmentation Offload, and question about genAI and code

[pierre.alain@xxxxxxx]

Dear Hannes,
Thank you for your reply.

3 août 2026, 16:38 hannes@xxxxxxxxxxx wrote:

>
> That sounds great - do you have some profiling what are the expensive bits 
> compared to the linux firewall? A factor of 2 is still very high IMHO - a 
> good selling point would be a factor of 0.2 ;)
>
That's just based on some quick experiments but so far I got up to 4Gb/s by 
using a different hypercall that vectorises grant copies. That has kept the 
firewall below 100℅ CPU (just under, at around 90℅), and it seems the next step 
is to implement multiqueue communication to remove stress from sys-net. So 
that's not a near future perspective :(
>
> From my point of view, the first item is crucial: be transparent (as you are) 
> about the usage thereof. Review all the changes in detail, and figure out 
> whether that is the way to go or there would be some other way (which also 
> means, be open for constructive feedback on PRs).
>
> The license issue is a real one, where I don't have any clue about. If the 
> first point - you reviewed and now own the code - is done extensively, it 
> shouldn't be a thing to re-implement the solution you saw with your own 
> fingers -- but that certainly won't solve the licensing in any way (since it 
> is not a clean-room implementation). I wonder what a good strategy is? Maybe 
> other people who use AI a lot have some ideas about it? All I can tell is 
> that last Friday in Germany some AI music generator lost in court against 
> real musicians -- so the AI generated music violates copyright. I believe the 
> same will be in the future being adapted to software and copyright.
>
> And obviously, using AI to find references, compare packets, ... is an ok 
> utilization - as long as you agree to the terms and impact (to climate, 
> politics, strengthening bigtech companies, likely more) of AI. [To make it 
> transparent, I never used AI, and I don't plan to use any for writing code. 
> If you're interested about some reasoning, I find the article 
> https://superrr.net/en/blog/about-ai-and-unlikelihood very interesting, with 
> lots of useful links. If you're more of an audio person, 
> https://techwontsave.us/episodes also has tons of interesting episodes.]
>
To be honest l'm not thrilled to have actually found that tool so helpful. I'm 
not happy with the climate impact, nor any of all impacts you mention. Nor I'm 
happy about the social impact : I feel as though I have missed out a good 
session of sharing and coding with peers as during my only mirage retreat. I 
wanted to understand the hype with them, but I think I was too naive :(

Now the code exist, and I'm really keen to spare anyone the cost to recode it 
under a clean-room protocol.

> I am happy to review your PRs, and am not so happy to review random 
> passing-by PRs which are heavily done by AI without any manual review (as 
> https://github.com/mirage/mirage/pull/1647).
>

Thank you for your review proposal, and I'll be glad to defend/adapt my code 
proposition :)

About the PR you mention, the PR text clearly feels AI-generated, perhaps the 
code too. Clearly it should have been disclosed if that was the case. On the 
"how to implement that" it seems straightforward enough to me to approve, but 
maybe it would have been better to decline due to AI usage.

Best
Pierre