tree 4b4b9e330b1e2aaa7b3769bc16097cfe82de0472
parent 12b638314be5e5174efb80a8ba8fe40f71083721
author Boris Zbarsky <bzbarsky@apple.com> 1689094404 -0400
committer GitHub <noreply@github.com> 1689094404 +0000
gpgsig -----BEGIN PGP SIGNATURE-----
 
 wsBcBAABCAAQBQJkrYkECRBK7hj4Ov3rIwAATaEIAKLBtaiWMdhmVhu1HWXRplpK
 sKLGrjTWiZKPO8D+6K1PcxQMutuvbgvFq3uAvMrA7OP0AGvuAFnhtzDDs5tJpRM1
 DS7CdEkscCQp9CjqgJ06WBKxnY4lF5P6nq96dHRK9AZo9CxPeot2tm7TzVUi5OcD
 xG7RhcLTvpzMy8uztHEVxMhAASRW1eu5qLqoRdiWKpooEtFpUHGRrS76FtvLuLMt
 cytEsXsayf4T7fHSYYPo6KcJ9m4ZnjURu4nhefxsM8LfKv2F8SPwQvqsmnkeDDsp
 BohPlC2pAbUQdYRKa7VSSLVA3T4pZsATN4Vgz9taOZTzWoAEwXkvFZz2BRQlZOI=
 =tcTP
 -----END PGP SIGNATURE-----
 

Fix exchange manager tests to handle losing timeslice better. (#27872)

We could get into a situation where we lost the timeslice after the SendMessage
call and before we asserted the response timeout had not happened yet, which
would cause the test to fail.

The changes here are:

1) Move the assert that we are not timed out to _before_ SendMessage().  This
   ensures that our state is correct up front, and generally nothing under
   SendMessage proper or sending the message should trigger a timeout per se.
2) Use more slack when waiting for the timeout, just in case.

Fixes https://github.com/project-chip/connectedhomeip/issues/27479