Rust interface to libnitrokey https://git.ireas.org/nitrokey-rs/atom?h=v0.4.0-alpha.3 2019-07-16T08:41:13Z 2019-07-16T08:41:13Z Robin Krahl robin.krahl@ireas.org 2019-07-16T08:40:32Z urn:sha1:f150d59410eefdec2ae69b2422906a3d1d88aa07 2019-07-16T08:38:57Z Robin Krahl robin.krahl@ireas.org 2019-07-16T08:36:50Z urn:sha1:5a0d9b44263caf38958080573cb0ae0e5d57f980 Previously, we were using a development version of nitrokey-test that was compatible with nitrokey 0.4. This patch updates nitrokey-test to version 0.3, which includes the required changes. 2019-07-15T06:23:25Z Robin Krahl robin.krahl@ireas.org 2019-07-15T06:23:25Z urn:sha1:8161edda54f39f1c8ac0f4cee7472a62ee6f5942 2019-07-15T06:18:51Z Robin Krahl robin.krahl@ireas.org 2019-07-15T06:18:05Z urn:sha1:34efcfadf1436102e42144f710edabaa2c4b55cd 2019-07-09T11:37:42Z Robin Krahl robin.krahl@ireas.org 2019-07-09T11:34:53Z urn:sha1:fe8161ba51d65382a20650a75f06e1cc7b753e68 Due to a timing issue, some calls to the build_aes_key function may fail after a factory reset. As a workaround for this firmware bug, we check the user retry count before building the aes key in the factory_reset test. For details, see the upstream issue: https://github.com/Nitrokey/nitrokey-pro-firmware/issues/57 2019-07-09T11:07:52Z Robin Krahl robin.krahl@ireas.org 2019-07-09T11:03:01Z urn:sha1:a0ebd37765027121b85dfd6b78b453a50adc69c9 To avoid a ConcurrentAccessError, we have to use the Device::into_manager function instead of calling take to obtain a Manager instance. 2019-07-09T10:55:30Z Robin Krahl robin.krahl@ireas.org 2019-07-09T10:51:55Z urn:sha1:4dc5b7e73ef348f18b4d2b032dc9a27059f4c47f This patch updates the nitrokey-test dependency to a new development version that uses force_take instead of take to get a Manager instance. If a test fails, the thread panics, leading to a poisoned cache – yet this should not affect the other test cases. Therefore we want to ignore the poisoned caches. 2019-07-09T10:47:26Z Robin Krahl robin.krahl@ireas.org 2019-07-09T10:44:45Z urn:sha1:5e8f0fbaf6df0cb919e4b02401cc21d5280bf09c The take and take_blocking functions return a PoisonError if the cache is poisoned, i. e. if a thread panicked while holding the manager. This is a sensible default behaviour, but for example during testing, one might want to ignore the poisoned cache. This patch adds the force_take function that unwraps the PoisonError and returns the cached Manager even if the cache was poisoned. 2019-07-09T10:23:14Z Robin Krahl robin.krahl@ireas.org 2019-07-09T10:22:08Z urn:sha1:88b32f5c2187e59fece93cd245aeadb4e5f9e04a During the connection manager refactoring, we temporarily used deprecated methods. This is no longer the case, so we can remove the allow(deprecated) attribute. 2019-07-09T08:49:16Z Robin Krahl robin.krahl@ireas.org 2019-07-09T08:49:16Z urn:sha1:7ce751225f12c295c6e33dd46cfb5dcb88f8fbb3 Refactor the connection management to prevent multiple device connections at the same time. RFC: https://lists.sr.ht/~ireas/nitrokey-rs-dev/%3C20190126174327.tbuyk2s535kfiqm4%40localhost%3E