Chronos fails with `IndexError` in NBC about nim-chronos HOT 5 CLOSED

I was able to constantly reproduce this error with incorrect arguments, in this particular case readExactly() with nbytes < 0 could cause this problem.
In chronos we can only add check for arguments, but its impossible to fix on this side.
#128.

That makes sense, #128 should allows us to at least catch all the places where this might be happening.

from nim-chronos.

For the record, this issue is caused by exceptions leaking out of asyncCheck - this corrupts chronos' internal state - vacp2p/nim-libp2p#338 shows the exception site

from nim-chronos.

More info here status-im/nimbus-eth2#1518

from nim-chronos.

I was able to constantly reproduce this error with incorrect arguments, in this particular case readExactly() with nbytes < 0 could cause this problem.
In chronos we can only add check for arguments, but its impossible to fix on this side.
#128.

from nim-chronos.

I'm closing it because there no reports on an issue anymore, so maybe it was fixed not only in chronos but in nim-libp2p too.

from nim-chronos.