Matt Johnston
Matt Johnston
I think I've managed to resolve the stuck read in #37 . It was a stupid mistake I didn't notice in testing, somehow sunsetc never hits it. For NoRoom I...
Would using `maitake-sync` require programs set `panic="abort"`? That might be a problem for `std` programs using parts of `embassy-sync` (crates designed to work both `std` and `no_std`). Maybe that's fine...
I can't reproduce a crash here (testing on x86-64). The regression commit doesn't change much when DROPBEAR_SNTRUP761 is disabled. You could try setting `KEXHASHBUF_MAX_INTS` to the prior value, though if...
> Hi, I'm facing the same issue. I'm using key type ECDSA with a key size of 521 bytes. I'll investigate more and try to reproduce here. @SebastianKonplan what platform...
Thanks, this should be fixed now. It wasn't platform dependent, I just didn't have private ecdsa keys being loaded here when I tested.
It's on my TODO list as `-Q`
Dropbear should check if the input is a TTY or not, I'll fix that. As a workaround `-T` argument will work.
It would be useful to add, I'd be happy to take patches or may implement it myself (no timeframe).
> CI failure looks to me like a flaky test? Yep sorry.