| Commit message (Collapse) | Author | Age | Files | Lines |
| ... | |
| |
|
|
|
|
|
|
| |
Slug is the general type. Instance kinds are slugs too. We need a
word for the slug-within-an-instance which identifies a storage
location for StorageHandle, or a raw subdirectory. Let's use "key".
Change the terminology (and variable names) everywhere.
|
| | |
|
| |
|
|
|
|
| |
Add comments where it's done automatically.
(There will be tests of the claims I'm making, in a later commit.)
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
This will be slightly less ergonomic at call sites but it is a more
orthogonal and logical API.
|
| |
|
|
|
|
|
| |
Purging depends on the current time.
We will want this to be able to come from Runtime but we don't want to
have those types in this API.
|
| |
|
|
| |
We purge per-kind.
|
| |
|
|
| |
This was probably the result of rebase conflicts during develoopment.
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
This crate has had new features added. It's 0.x. So bump in-tree
dependencies' references:
fs-mistrust
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bump the minor version of these crates, and update the in-tree
dependencies.
Recently published as fresh crates, let's just assume there are
breaking changes:
fslock-guard
test-temp-dir
Breaking API change affecting many many downstream crates:
tor-rtcompat
Downstream crates which we're (conservatively) assuming have
tor-rtcompat types in their APIs:
tor-rtmock
tor-log-ratelim
tor-rpcbase
tor-llcrypto
tor-protover
tor-bytes
tor-hscrypto
tor-hspow
tor-socksproto
tor-checkable
tor-cert
tor-linkspec
tor-cell
tor-proto
tor-netdoc
tor-consdiff
tor-netdir
tor-congestion
tor-persist
tor-chanmgr
tor-ptmgr
tor-guardmgr
tor-circmgr
tor-dirclient
tor-dirmgr
tor-keymgr
tor-hsclient
tor-hsservice
tor-hsrproxy
arti-client
arti-rpcserver
arti-config
arti-hyper
arti-bench
arti-testing
|
| |
|
|
| |
Discard hunks in examples/
|
| | |
|
| |
|
|
|
|
| |
And export the type, so callers don't need to depend directly on
fslock_guard; many callers will need to own the returned value, not do
anything else with it.
|
| | |
|
| |\
| |
| |
| |
| |
| |
| | |
tor-keymgr: Lowercase the role when building the ArtiPath.
Closes #1195
See merge request tpo/core/arti!1934
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| | |
The `slug` docs say slugs are not allowed to contain uppercase
characters, so let's return an error if they do.
Part of #1195
|
| | | |
|
| | |
| |
| |
| | |
This is now InstanceStateHandle::purge.
|
| | |
| |
| |
| |
| | |
We should add more tests for these error representations but we want
to rework these errors so let's not do all that now.
|
| | |
| |
| |
| |
| |
| | |
Apparently we need
https://github.com/dbrgn/tracing-test/pull/17
too.
|
| | |
| |
| |
| |
| |
| |
| |
| | |
We need this bugfix
https://github.com/dbrgn/tracing-test/pull/15
for `#[traced_test]` when `Result` is redefined.
Fixes the tests with minimal-versions.
|
| | | |
|
| | |
| |
| |
| | |
We use `/` to separate, not `+`.
|
| | | |
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| | |
This would fix the tests on Windows I think ? (Tests which don't seem
to run in routine CI AFAICT.)
|
| | |
| |
| |
| | |
As previously decided and documented.
|
| | |
| |
| |
| |
| |
| |
| | |
We're about to reuse this.
This changes the Resource in the error to be more accurate, too. This
will be more important as we add another call site.
|
| | |
| |
| |
| | |
We're going to want this some more.
|
| | | |
|
| | | |
|
| | | |
|
| | |
| |
| |
| | |
Remaining formatting churn.
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| | |
And also instance_peek_storage, which needs to share some code, so
determined the shape of this implementation.
Make StateDirectory have the (one) field we need.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
These are the straightforward methods which, given a CheckedDir,
perform the relevant operation.
Make InstanceStateHandle have the necessary fields.
Constructing an InstanceStateHandle is a bit complex and will come in
a moment.
|
| | |
| |
| |
| |
| |
| |
| |
| | |
In practice, allowing Bug here was quite inconvenient for everyone.
I think very few call sites will want to throw anything interesting.
If they *do* want to cause an error, without panicking, they could
write a syntactically invalid slug.
|
| | |
| |
| |
| |
| | |
state_dir doesn't support a "shared read-only" mode, so it needs an
error.
|
| | |
| |
| |
| |
| |
| |
| | |
Introduce a Bug variant.
Then, use it for conversions from BadSlug. See the comment for why I
think this is appropriate.
|
| | |
| |
| |
| |
| |
| |
| |
| | |
You don't need to consume a Mistrust to get a CheckedDir - you use a
Verifier and then something clones it for you. So StateDirectory::new
should take &Mistrust.
And instance_peek_storage needs T to be deserialisable!
|
| | |
| |
| |
| |
| | |
This is rather more serious than many deletions - it wipes out the
whole directory. And this name goes better with purge_instances.
|
| | |
| |
| |
| | |
The formatting will be normalised roughly as we go.
|
| | |
| |
| |
| | |
Now we get log messages about what we're reading/writing.
|
| |/ |
|