TVars were not updated atomically, which was ok when each thread got its
own External that was the only thing using these TVars. But, with the
async extension, several External instances can share the same var, so
it needs to be a TMVar to avoid read/write conflicts.
In particular, this makes PREPARE only be sent once.
Moving jobid generation to the git-annex side lets it be simplified a
lot.
Note that it will also be possible to generate one jobid per connection,
rather than a new job per request. That will make overflow not an issue,
and will avoid some work, and will simplify some of the code.
Receive loop looks right. Still need the send loop.
And, a complication is that some messages git-annex
sends need to be wrapped in REPLY_ASYNC, while others
do not. So will probably need to split externalSend
into two.
Now that I've started implementation, I see it's really necessary that
every message the special remote sends use the protocol, otherwise
nasty edge cases abound.
Since an external process can be in the middle of some operation when an
async exception is received, it has to be shut down then. Using
cleanupProcess will close its IO handles and send it a SIGTERM.
If a special remote choses to catch SIGTERM, it's fine for it to do some
cleanup then, but until it finishes, git-annex will be blocked waiting
for it. If a special remote blocked SIGTERM, it would cause a hang.
Mentioned in docs.
Also, in passing, fixed a FD leak, it was not closing the error handle
when shutting down the external. In practice that didn't matter before because
it was only run when git-annex was itself shutting down, but now that it
can run on exception, it would have been a problem.
Apparently some externals use GETCONFIG name, and the documentation as
written can be read to allow it, although IIRC it was not really my
intent to. If someone does do this, they should not include it in
LISTCONFIGS, since git-annex already tracks its own config settings.
Special remote programs that use GETCONFIG/SETCONFIG are recommended
to implement it.
The description is not yet used, but will be useful later when adding a way
to make initremote list all accepted configs.
configParser now takes a RemoteConfig parameter. Normally, that's not
needed, because configParser returns a parter, it does not parse it
itself. But, it's needed to look at externaltype and work out what
external remote program to run for LISTCONFIGS.
Note that, while externalUUID is changed to a Maybe UUID, checkExportSupported
used to use NoUUID. The code that now checks for Nothing used to behave
in some undefined way if the external program made requests that
triggered it.
Also, note that in externalSetup, once it generates external,
it parses the RemoteConfig strictly. That generates a
ParsedRemoteConfig, which is thrown away. The reason it's ok to throw
that away, is that, if the strict parse succeeded, the result must be
the same as the earlier, lenient parse.
initremote of an external special remote now runs the program three
times. First for LISTCONFIGS, then EXPORTSUPPORTED, and again
LISTCONFIGS+INITREMOTE. It would not be hard to eliminate at least
one of those, and it should be possible to only run the program once.
Made responses to git-annex requests be listed under each request.
This did lead to a little duplication since some replies are used for 2
requests, but it also makes it much clearer and easier to see how the
protocol works.
And, it makes each request self-contained, so they can be split out into
separate pages.
Avoid a warning message when renameExport is not supported, and just
fallback to deleting with a subsequent re-upload. Especially needed for
importtree remotes, where renameExport needs to be disabled.
This changes the external special remote protocol, but in a
backwards-compatible way. A reply of UNSUPPORTED-REQUEST to an older
version of git-annex will cause it to make renameExport return False.
Note that I tried an evil remote that lists ImportLocations with
../../../ in them and indeed this resulted in git blowing up and the
import failing, and not writing outside the repo.
Had to add two more API calls to override export APIs that are not safe
for use in combination with import.
It's unfortunate that removeExportDirectory is documented to be allowed
to remove non-empty directories. I'm not entirely sure why it's that
way, my best guess is it was intended to make it easy to implement with
just rm -rf.
Finishes the start made in 983c9d5a53, by
handling the case where `transfer` fails for some other reason, and so the
ReadContent callback does not get run. I don't know of a case where
`transfer` does fail other than the locking dealt with in that commit, but
it's good to have a guarantee.
StoreContent and StoreContentTo had a similar problem.
Things like `getViaTmp` may decide not to run the transfer action.
And `transfer` could certianly fail, if another transfer of the same
object was in progress. (Or a different object when annex.pidlock is set.)
If the transfer action was not run, the content of the object would
not all get consumed, and so would get interpreted as protocol commands,
which would not go well.
My approach to fixing all of these things is to set a TVar only
once all the data in the transfer is known to have been read/written.
This way the internals of `transfer`, `getViaTmp` etc don't matter.
So in ReadContent, it checks if the transfer completed.
If not, as long as it didn't throw an exception, send empty and Invalid
data to the callback. On an exception the state of the protocol is unknown
so it has to raise ProtoFailureException and close the connection,
same as before.
In StoreContent, if the transfer did not complete
some portion of the DATA has been read, so the protocol is in an unknown
state and it has to close the conection as well.
(The ProtoFailureMessage used here matches the one in Annex.Transfer, which
is the most likely reason. Not ideal to duplicate it..)
StoreContent did not ever close the protocol connection before. So this is
a protocol change, but only in an exceptional circumstance, and it's not
going to break anything, because clients already need to deal with the
connection breaking at any point.
The way this new behavior looks (here origin has annex.pidlock = true so will
only accept one upload to it at a time):
git annex copy --to origin -J2
copy x (to origin...) ok
copy y (to origin...)
Lost connection (fd:25: hGetChar: end of file)
This work is supported by the NIH-funded NICEMAN (ReproNim TR&D3) project.