there's no reason a store should not open a connection during init if it
needs to open any connection to function. delaying connection setup like
this makes openStore less deterministic and graceful fallback impossible
this caused test failures where the old behaviour was required to ensure
test output stability. because of course something like that must happen
Change-Id: I946adddee026f1f4c74b4699d730b4d1ac9a5072
adding logs to a daemon store can fail before the entire request has
been read, leaving non-command data in socket buffers. if the daemon
does not terminate immediately after the failure it will try to read
the remaining data as commands, which is very likely to not work and
cause unbounded memory allocation instead. this doesn't give clients
an attack vector they did not have before though, they could've just
as well sent malformed commands without a bad preceding AddBuildLog.
Change-Id: I86c22890eace19164d932cbd342ac9f52cee6531
we don't support range-based iteration yet because it's a huge hassle to
make c++ and rust iteration styles meet in any reasonable way. this is a
good start to make rust collections *actually* usable though, so here we
go. `into_iter` is not explicitly supported, but can be used regardless.
Change-Id: Iebb76c9409059bc81623e5be81ac2656b1ec5138
frames manually added with `addErrorContext` are generally a lot more
useful/informative to the average user than other frames, esp. in
the module system, which can create much better error messages than
we can.
however, before this change, frames from addErrorContext were truncated
by default if they weren't in the first 3 frames, so they were basically
useless (with `--show-trace`, you're dredging through 250 frames of
module shenanigans just to spot one singular line).
this change also removes the frames for _the call to_ `addErrorContext`,
which is just pure noise.
the way this change is done hopefully leaves a bit of space for future
similar changes to error printing, by introducing a new `TraceKind` enum
that can be used to categorize traces (i haven't done that in this CL
because that would be a pretty herculean task, given all the calls to
`BaseError::addTrace` in the codebase, and we probably want to be
careful about what categories we choose). the actual printing code could
definitely be improved tho... (e.g. by iterating twice through the trace
stack instead to first pick out the most important traces and _then_
printing less "important" traces if there's space left)
Change-Id: I52acc52f231991a9f2309d9cecae362397c6888c
this makes using the zero-copy pipes a lot more ergonomic. error
handling is unfortunately not trivial and duplicating it for all
pipe users would thus be a bad idea. we're also not oblivious to
the fact that this is a `sourceToSink`, but it's async this time
around. (at least we don't need terrible stackful coroutines..?)
Change-Id: I1ba59f27183988ad68e7f88d102935d005690f43
rpc will often need to transfer data received on push-bashed interfaces
to consumers that expect input stream sources, which are pull-based. we
want to avoid copies for performance reasons (since kj overhead as kind
of on the high side for us already), so we'll use a shared-buffer class
that behaves much like a mutex. we *don't* use mutexes because ours are
cross-thread-capable and thus require syscalls for all wakeups they do.
Change-Id: I3b14925f5d9f5e07ea2cafdf00a88a64f79e4742
Many issues stems from `<nixpkgs>` not resolving anywhere or having the
wrong version, let's make it awfully obvious again here.
Towards #230.
Change-Id: If630e9616566e77143f028cdcfe0b7f00a1d486d
Signed-off-by: Raito Bezarius <raito@lix.systems>
without maxprocesses we can utilize core-rich systems better, and test
timeouts are reportedly to be too low to run on small systems as well.
fixes#890
Change-Id: I89386b89fcd69ef4bf77ecf0a49c0b85e1f17c3d
this fixes a large portion of tests currently marked no_daemon. most of
them only needed to set some trusted settings, which is easily done now
Change-Id: Id5a5ee94951cdc92bddd2264c738ca4f98980c8b
attrpathsSuperset got renamed to preEval in nixpkgs[1] breaking Hydra's
evaluation. Updating nixpkgs to make sure this is consistent now, no
matter if flake-inputs or release.nix is being used.
[1] See commit 19a31658dc39324c9acf8d81198cd7137bdc1e92.
Change-Id: I69ab3f894534b1d89b626ddd4feda8ea502b69b3
this is a bit of a hack, but since zngur cannot handle multiple trait
implementations per type yet we will have to commit to singles types.
Change-Id: I60e7b96bbeaa9fb87cf43662d4a9a5d44116bf47
note: only two of the tests actually needed migrating, as the others
were already covered (e.g. by lang)
Change-Id: I30abb2bd728ae2b144ad76566096a216aed3b8f8
treating a file with an empty or zero hash as locked is not helpful.
these are placeholdes for "hash is not known", thus treating them as
a valid lock makes them completely useless (and confusing to users).
fixes#1233
Change-Id: If42b47281e6973fc86662b973db69e26f6346f5a
using result types in capnp is fraught. while it makes some sense for
interfaces that need neither streaming nor pipelining and can provide
much better error fidelity there it's also fundamentally incompatible
with those that *do* need streaming or pipelining: streams will never
be stopped early unless an exception flies, and pipelines cannot look
through result types. likely the best thing we can do is to encode an
error for transport in the capnp/kj exception description strings. :(
Change-Id: Icb7d16238fa9a7aaf92c00363f7be4076ac02a61
This is the real deal. The new tests have surfaced one bug in the test
suite itself, and one in the RPC legacy protocol wrapper.
Co-authored-by: eldritch horrors <pennae@lix.systems>
Change-Id: Id75b671c10b45655d3627abd88563e2d41719446
All tests should have been clear here, but alas
Co-authored-by: eldritch horrors <pennae@lix.systems>
Change-Id: Id642be1f6e0fb714b844ee5348dc096866cf3906
This commit enables the parametrization for "legacy-combined" protocol.
All failing tests with that have been either fixed or disabled. Notably:
- A couple of tests (e.g. involving builders) could be made to work with
a non-local store, but this would require some refactoring to testlib in
order to make certain configuration settings generic over local and
remote operations. We've disabled those for now, in order to make
progress
- The store tests all run locally only because of their nature
- The flakes tests also all run locally only because my energy for
fixing them is limited (and it's a *lot* of test failures, with probably
little overall benefit in terms of test coverage)
Change-Id: I56fa249a64f7c17c952f688ec89e9687f2a13f12
This first geralizes the `nix` fixture with `pytest_generate_tests` to
iterate over all protocols (unless they are marked with `no_daemon`),
though for now the list of protocols is set to be empty. The lang tests
are all tagged with `no_daemon` because they are mostly pure and running
them multiple times would be wasteful.
Co-authored-by: rootile <lix@rootile.de>
Co-authored-by: piegames <git@piegames.de>
Change-Id: Ib407edb420ba4bf434cacf9563a71f5ae6fa8eef
they're not doing anything, and they're slightly in the way of reworking
the rust build infra. we'll add them back once that rework is completed.
Change-Id: Idde8ba3585fd08b89b4b4298349f8f58a29a9661
When removing the f1 repl characterization test suit, these files were
kept on accident, despite them being dead code now
Change-Id: I13b2c5a96004787c26fe7209e4cb3aeedd8ada15
This reverts commit 0d6b372c19.
Reason for revert: In the time between this getting +2 and it being merged, all repl tests and its syntax have been migrated to f2, making these changes null and to dead code
Change-Id: I00f5237bf3f2ca3a60cc1584f67fb67196805e7e
Gosh. Why was there never a README in here explaining what it does
without requiring people to read C++? There was a half-assed grammar in
a comment, but it was very half-assed, so let's replace it with a
better-written one.
I'm documenting this so that neither I nor anyone else have to read C++
while rewriting the parser in Python.
See: https://git.lix.systems/lix-project/lix/issues/1196
Change-Id: Ia34ea7b02f109eb7753833e1c03d4b5d1d3c3b75
When extracting a global pyproject.toml from the f2 one, some f2
specific things slipped through and ended up in the global one.
This commit moves those parts back where they belong
Change-Id: I95706fabec487410717fb362c95bfee900cf5b8c
if the netrcFile setting has been changed tell curl that the file is required, so
fetching fails if the file does not exist
if the caFile setting has been specified, check if the file exists and throw an
error if is it non-existent
Refs: #1106
Change-Id: Icb9330a7a715175d35b9ed894ed945f0fd4d7061
rootile