Rob Norris
Rob Norris
> I'm going to strongly disagree that "it just fails to unlock the datasets and you could find a way to extract and feed in the raw key" is a...
I'm not trying to relitigate anything; I'm still fairly new here, and I'm happy to accept that my understanding is wrong. My understanding comes from reading docs and code and...
Thanks for the feedback. I'm kind of indifferent to `--force=...` just for this PR, but I can certainly see the utility of a general facility to be able to express...
Alright, latest push sets up `--force` as a set of comma-separated flags, and adds two: `vdevs`, which is the old `-f` behaviour, and `layout`, which is the override flag for...
> We should also include [dd108f5](https://github.com/openzfs/zfs/commit/dd108f5d73e241a7a3fa836e5c3322a10d135a5c), which fixes the static analyzer complaint about what should be a real bug that occurs after this is merged. Good call! I've picked that...
@behlendorf done, thanks.
> Could you rebase the thing? It's needed for the new testing workflow... So we see, it doesn't break sth. I will approve this next week, when the testings are...
> Short writes may occur. Yeah, fair. I've pushed something that should be a little more resilient.
> Don't use arithmetic operation on a `void *` directly, cast it to `char *` first, like: Yeah, silly me. > Consider using some C99 features to reduce scpoe of...
> I think the code means short writes not reads. Sigh, I'm really having a day of it today. Thanks, fixed. > May be it is better to have a...