This commit splits ACL policies into more fine-grained capabilities.
This both drastically simplifies the checking code and makes it possible
to support needed workflows that are not possible with the previous
method. It is backwards compatible; policies containing a "policy"
string are simply converted to a set of capabilities matching previous
behavior.
Fixes#724 (and others).
The rollback manager was using a saved MountTable rather than the
current table, causing it to attempt to rollback unmounted mounts, and
never rollback new mounts.
In fixing this, it became clear that bad things could happen to the
mount table...the table itself could be locked, but the table pointer
(which is what the rollback manager needs) could be modified at any time
without locking. This commit therefore also returns locking to a mutex
outside the table instead of inside, and plumbs RLock/RUnlock through to
the various places that are reading the table but not holding a write
lock.
Both unit tests and race detection pass.
Fixes#771
You can now turn on and off the lease behavior in the generic backend by
using one of two factories. Core uses the normal one if it's not already
set, so unit tests can use the custom one and all stay working.
This also adds logic into core to check, when the response is coming
from a generic backend, whether that backend has leases enabled. This
adds some slight overhead.
fields in 'lookup-self'. Importantly, this also makes credential
backends use the SystemView per-backend TTL values and fixes unit tests
to expect this.
Fully fixes#527
In order to implement this efficiently, I have introduced the concept of
"singleton" backends -- currently, 'sys' and 'cubbyhole'. There isn't
much reason to allow sys to be mounted at multiple places, and there
isn't much reason you'd need multiple per-token storage areas. By
restricting it to just one, I can store that particular mount instead of
iterating through them in order to call the appropriate revoke function.
Additionally, because revocation on the backend needs to be triggered by
the token store, the token store's salt is kept in the router and
client tokens going to the cubbyhole backend are double-salted by the
router. This allows the token store to drive when revocation happens
using its salted tokens.
return a descriptive error indicating that the Secret cannot be returned
because when the token was revoked the secret was too. This prevents
confusion where credentials come back but cannot be used.
Fixes#615
