Simulate file locks in blockio-sim
#415
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The simulation piggybacks on existing `HasFS` functionality for opening files in different `OpenMode`s. The simulation does not simulate inter-process locking.
jorisdral
force-pushed
the
jdral/blockio-sim-filelock
branch
from
dba71e5
to
8eb3c61
Compare
With the nice addition that we now assert that there are no open file handles at the end of each iteration of the tests!
jorisdral
force-pushed
the
jdral/blockio-sim-filelock
branch
from
8eb3c61
to
3e19457
Compare
jorisdral
force-pushed
the
jdral/blockio-sim-filelock
branch
from
3e19457
to
63d7868
Compare
It's better but not perfect. It passes the test, but would fail if there were genuine concurrency. Doing it properly would need an implementation in fs-api and fs-sim. Fortunately we do not need a concurrency safe implementation (for now).
Follow a more normal convention, and allow running individual tests more easily in GHCi via quickCheck $ theProperty
jorisdral
commented
Oct 4, 2024
| -- | ||
| -- We implement this using the content of the lock file. The content is a | ||
| -- counter, positive for readers and negaive (specifically -1) for writers. | ||
| -- There can be any number of readers, but only one writer. |
There was a problem hiding this comment.
Suggested change
| -- There can be any number of readers, but only one writer. | |
| -- There can be any number of readers, but only one writer. Writers can not coexist with readers. |
| -- | ||
| -- Warning: This implementation is not robust under concurrent use (because | ||
| -- operations on files are not atomic) but should be ok for casual use. A | ||
| -- proper implementation would need to be part of the underlying 'HasFs' |
There was a problem hiding this comment.
Suggested change
| -- proper implementation would need to be part of the underlying 'HasFs' | |
| -- proper implementation would need to be part of the underlying 'HasFS' |
Comment on lines
+44
to
+47
| -- Warning: This implementation is not robust under concurrent use (because | ||
| -- operations on files are not atomic) but should be ok for casual use. A | ||
| -- proper implementation would need to be part of the underlying 'HasFs' | ||
| -- implementations. |
|
|
||
| readCount :: Handle h -> m Int | ||
| readCount h = do | ||
| content <- API.hGetSome hfs h 10 |
There was a problem hiding this comment.
We should use hGetAllAt or something similar, because hGetSome can do partial reads
| writeCount :: Handle h -> Int -> m () | ||
| writeCount h n = do | ||
| API.hSeek hfs h AbsoluteSeek 0 | ||
| _ <- API.hPutSome hfs h (BS.pack (show n)) |
There was a problem hiding this comment.
We should hPutAllStrict or something similar, because hPutSome can do partial writes
Comment on lines
+38
to
+48
| -- | Lock files are reader\/writer locks. | ||
| -- | ||
| -- We implement this using the content of the lock file. The content is a | ||
| -- counter, positive for readers and negaive (specifically -1) for writers. | ||
| -- There can be any number of readers, but only one writer. | ||
| -- | ||
| -- Warning: This implementation is not robust under concurrent use (because | ||
| -- operations on files are not atomic) but should be ok for casual use. A | ||
| -- proper implementation would need to be part of the underlying 'HasFs' | ||
| -- implementations. | ||
| -- |
There was a problem hiding this comment.
We should also maybe warn that the simulation only defines how tryLockFiles on the same path interact. A regular hOpen on the same path will still work
jorisdral
requested review from
dcoutts,
mheinzel and
recursion-ninja
as code owners
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What's neat is that we can now find out in our
StateMachinetests whether our implementation is leaking file descriptors!TODO: the test I added shows that my implementation of file locks is not right just yet... Two
ExclusiveLocks work fine, butSharedLocks don't work. I was assuming that mocked file handles don't allow concurrent read and write locks, but it seems they do, so we can't just piggyback on theOpenModes