Export occasionally crash

[wallace11]

Hey!
Continuing our adventure from a few months back in #111, I noticed that with v0.9.0 I sometimes get an error when running an export command:

dsc export --all --target .

The error I get with RUST_BACKTRACE=full:

thread 'main' panicked at 'overflow when subtracting durations', library/core/src/time.rs:988:31
stack backtrace:
   0:           0x93c4fc - <unknown>
   1:           0x9803cc - <unknown>
   2:           0x933ea3 - <unknown>
   3:           0x93ee22 - <unknown>
   4:           0x93ea05 - <unknown>
   5:           0x93f473 - <unknown>
   6:           0x93f190 - <unknown>
   7:           0x93c9a4 - <unknown>
   8:           0x93eec9 - <unknown>
   9:           0x414ce1 - <unknown>
  10:           0x97eb81 - <unknown>
  11:           0x97eb2b - <unknown>
  12:           0x414bd6 - <unknown>
  13:           0x982cb7 - <unknown>
  14:           0x42e590 - <unknown>
  15:           0x457b52 - <unknown>
  16:           0x51fc64 - <unknown>
  17:           0x4961bd - <unknown>
  18:           0x417fa0 - <unknown>
  19:           0x415d73 - <unknown>
  20:           0x415e69 - <unknown>
  21:           0x93bc40 - <unknown>
  22:           0x418222 - <unknown>

Since it has to do with the time library, I'm guessing it can explain why it only happens occasionally and it cannot be reproduced reliably.

[eikek]

Thank you for the report! This looks like a bug when working with durations. Unfortunately the stacktrace is not very informative. Since it is the export command, I suspect the issue within the session.rs near_expiry() function.

Did you see any pattern to trigger this? Which OS are you running? If you can reproduce it again, you could run with the -vvv flag to have dsc log more.