about summary refs log tree commit diff
path: root/tvix/eval/src
AgeCommit message (Collapse)AuthorFilesLines
2022-10-16 r/5150 refactor(tvix/eval): unify compile_lambda() with thunk()Adam Joseph1-61/+59
This resolves a TODO. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: If4d2124648ac88094e547e1ad7f1b446feb26182 Reviewed-on: https://cl.tvl.fyi/c/depot/+/7010 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-16 r/5149 fix(tvix/eval): more faithfully serialise ast::LiteralVincent Ambo1-10/+9
The previous serialisation format kind of lost the information about what AST node we're dealing with (e.g. `1234` would serialise to an AST with a literal `1234`). That's great for pretty-printing the _code_, but we explicitly want to serialise how rnix-parser parses something. To that end, literals are now instead serialised into a structure like all the other ones (`kind: literal` and appropriate value fields). Change-Id: I586c95d7db41820b8ec43565ba4016ed3834d1b5 Reviewed-on: https://cl.tvl.fyi/c/depot/+/7030 Autosubmit: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI Reviewed-by: j4m3s <james.landrein@gmail.com> Reviewed-by: Adam Joseph <adam@westernsemico.com>
2022-10-16 r/5148 feat(tvix/eval): implement builtins.partitionJames Landrein3-0/+39
Change-Id: I8b591f3057c68c1542046fc5a771973f2238c9df Reviewed-on: https://cl.tvl.fyi/c/depot/+/7020 Autosubmit: j4m3s <james.landrein@gmail.com> Reviewed-by: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-16 r/5147 fix(tvix/eval): resolve home relative paths at runtimesterni3-12/+31
Home relative paths depend on the environment to be resolved. We have elected to do everything that depends on the environment, e.g. resolving SPATH expressions using NIX_PATH, at runtime, so tvix evaluation would continue to behave correctly even if we separated the compilation and execution phases more, e.g. via serializing bytecode. Then the value of HOME, NIX_PATH etc. could reasonably change in the time until execution, yielding wrong results if the resolution results were cached in the bytecode. We also take the opportunity to fix the broken path concatenation previously found in the compiler, fixing b/205. Another thing we could consider is emitting a warning for home relative path literals, as they are by nature relatively fragile. One sideeffect of this change is that home path resolution errors become catchable which is not the case in C++ Nix. This will need to be fixed up in a subsequent change. Change-Id: I30bd69b575667c49170a9fdea23a020565d0f9ec Reviewed-on: https://cl.tvl.fyi/c/depot/+/7024 Autosubmit: sterni <sternenseemann@systemli.org> Reviewed-by: Adam Joseph <adam@westernsemico.com> Tested-by: BuildkiteCI
2022-10-16 r/5146 refactor(tvix/eval): make OpFindFile use internal UnresolvedPathsterni3-12/+22
To assert that OpFindFile is only emitted for specially compiled SPATH expressions, as well as make sure it doesn't accidentally operate on “ordinary values”, introduce an UnresolvedPath internal value. If OpFindFile sees a non-UnresolvedPath value, it'll crash. Note that this change is not done purely for OpFindFile: We may want to compile SPATH expressions as function calls to __findFile (like C++ Nix does) in the future, so the UnresolvedPath value would definitely need to be an ordinary string again then. Rather, this change is done in preparation for resolving home dir relative paths at runtime (since they depend on the environment) for which we'll need a similar mechanism to OpFindFile. Change-Id: I6acf287f35197cd9e13377079f972b9d36e5b22e Reviewed-on: https://cl.tvl.fyi/c/depot/+/7023 Autosubmit: sterni <sternenseemann@systemli.org> Reviewed-by: Adam Joseph <adam@westernsemico.com> Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-16 r/5143 feat(tvix/eval): add an AST pretty-printing moduleVincent Ambo3-1/+499
This implements serde::Serialize for the rnix AST through a wrapper type, and exposes a function for serialising the AST into a (pretty-printed JSON) string representation. This can be used to debug issues with the AST, and to display an AST reprsentation in tools like tvixbolt. Serialize is implemented manually because we don't own any of the structs and the way to traverse them is not easily derived automatically, and this is quite verbose. We might be able to condense it a little bit, but at the same time it's also fairly straightforward. Change-Id: I922df43cfc25636f3c8baee7944c75ade516055c Reviewed-on: https://cl.tvl.fyi/c/depot/+/6943 Autosubmit: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI Reviewed-by: Adam Joseph <adam@westernsemico.com> Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-16 r/5142 feat(tvix/eval): remove Clone instance from Chunk and LambdaAdam Joseph2-2/+2
Lambda has a quite large and variable-sized runtime representation, unlike Rc<Lambda>. It would be easy to accidentally call clone() on this and create input-dependent performance regressions. Nothing in the codebase is currently using Lambda.clone(). Let's remove the derived instance. If it's really needed it is very easy to add it back in, but whoever does that will have to look at the struct they're adding Clone to, which will hopefully prompt them to think about whether or not that's really what they want to do. Change-Id: I7806a741862ab4402229839ce3f4ea75aafd6d12 Reviewed-on: https://cl.tvl.fyi/c/depot/+/7029 Autosubmit: Adam Joseph <adam@westernsemico.com> Reviewed-by: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-16 r/5141 fix(tvix/eval): don't coerce variable name to stringsterni2-3/+4
Change-Id: I8aa878dee009901feb453c489ce37c12fa3a31a8 Reviewed-on: https://cl.tvl.fyi/c/depot/+/7026 Autosubmit: sterni <sternenseemann@systemli.org> Reviewed-by: Adam Joseph <adam@westernsemico.com> Reviewed-by: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-16 r/5140 refactor(tvix/eval) remove Value::DynamicUpvalueMissingAdam Joseph2-10/+1
I believe this variant is left over from a previous implementation. If not, please let me know. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I02a3bf2f63794d09e96a5a92a034c0ad3d1ff221 Reviewed-on: https://cl.tvl.fyi/c/depot/+/7027 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-15 r/5137 feat(tvix/eval): Implement builtins.intersectAttrsGriffin Smith3-0/+19
Change-Id: Iaba9bcfa19f283cd0c1931be2f211e2528a1a940 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6998 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su> Reviewed-by: kanepyork <rikingcoding@gmail.com>
2022-10-15 r/5136 feat(tvix/eval): Initial impl of builtins.matchGriffin Smith3-0/+50
Implement an *initial* version of builtins.match, using the rust `regex` crate for regular expressions. The rust regex crate definitely has different semantics than nix's regular expressions - but we'd like to see how far we can get before the incompatibility starts to matter. This consciously leaves out any sort of memo for compiled regular expressions (which upstream nix also has) for the sake of expediency - in the future we should implement that so we don't have to compile the same regular expression multiple times. Change-Id: I5b718635831ec83397940e417a9047c4342b6fa1 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6989 Tested-by: BuildkiteCI Reviewed-by: Adam Joseph <adam@westernsemico.com> Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-15 r/5135 feat(tvix/eval): Implement builtins.fromJSONGriffin Smith8-4/+107
Using `serde_json` for parsing JSON here, plus an `impl FromJSON for Value`. The latter is primarily to stay "dependency light" for now - likely going with an actual serde `Deserialize` impl in the future is going to be way better as it allows saving significantly on intermediary allocations. Change-Id: I152a0448ff7c87cf7ebaac927c38912b99de1c18 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6920 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-15 r/5133 fix(tvix/eval): bring foldl' strictness in line with C++ Nixsterni6-1/+10
Working on https://github.com/NixOS/nix/pull/7158, I discovered that C++ Nix actually is strict in the accumulator, just not in the first value. This seems due to the fact that in the C++ evaluator, function calls don't seem to be thunked unconditionally and foldl' just elects not to wrap it in a thunk (don't quote me on this summary, even though it seems to line up with the code for primop_foldlStrict and testable behavior). It doesn't seem worth it to risk breaking the odd Nix expression just to be strict in one more value per invocation of foldl' (i.e. the initial accumulator value `nul`), so let's match the existing C++ Nix behavior here. Change-Id: If59e62271a90d97cb440f0ca72a58ec7840d1690 Reviewed-on: https://cl.tvl.fyi/c/depot/+/7022 Autosubmit: sterni <sternenseemann@systemli.org> Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-15 r/5132 feat(tvix/eval): implement builtins.dirOfAdam Joseph4-0/+27
This commit causes the test eval-okay-builtins.nix to pass. It also adds tests/tvix_tests/eval-okay-dirof.nix which has better coverage than the nix tests for this builtin. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I71d96b48680696fd6e4fea3a9861742b35cfaa66 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6987 Reviewed-by: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-14 r/5131 docs(tvix/eval) vm: explain VM::{frames,stack,with_stack}Adam Joseph1-2/+12
Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I94ba31ae25c1ff744f929a722c76a0c33cc361ff Reviewed-on: https://cl.tvl.fyi/c/depot/+/7016 Reviewed-by: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-14 r/5130 feat(tvix/eval): implement builtins.toPathAdam Joseph1-6/+9
This commit implements builtins.toPath. Like OP_ADD, it currently does not handle string contexts. This commit allows the tests::nix_eval_okay_src_tests_nix_tests_eval_okay_pathexists_nix test to pass. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: Iadd4f7605f8f297adbd0dba187b8481c21370b6e Reviewed-on: https://cl.tvl.fyi/c/depot/+/6996 Reviewed-by: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-14 r/5129 feat(tvix/eval): eliminate the only `unsafe` in the codebaseAdam Joseph1-4/+2
Maybe I misunderstood this part of the code, but the use of `unsafe` appears unnecessary here? In any event it is the one and only `unsafe` in the codebase. Hopefully getting to "no `unsafe` anywhere" is worth the extra never-taken branch caused by unwrap() instead of unwrap_unchecked()? Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I33fbd5aad9d8307ea82c24b6220412783e1973c6 Reviewed-on: https://cl.tvl.fyi/c/depot/+/7011 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-14 r/5128 refactor(tvix/eval): remove OpResolveWithOrUpvalueAdam Joseph2-20/+0
The case branch in vm.rs for OpResolveWithOrUpvalue is unreachable/deadcode. I believe this opcode is unnecessary, since it should always be statically detectable (at parse-time) whether a reference is to an upvalue (i.e. enclosing binding); otherwise, and only then, is with-resolution applicable. Perhaps I've misunderstood how with-resolution works. If so, please explain it to me and -1/-2 this CL. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I4a90b9eb6cb3396df92a6a943d42ecc301871ba0 Reviewed-on: https://cl.tvl.fyi/c/depot/+/7009 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-14 r/5127 refactor(tvix/eval): order builtins alphabeticallyJames Landrein1-45/+45
This makes it easier to compare currently implemented ones with the full list. Change-Id: Ibaffd99d05afa15fc9ab644fd101afa24fc7a1b2 Reviewed-on: https://cl.tvl.fyi/c/depot/+/7008 Tested-by: BuildkiteCI Autosubmit: j4m3s <james.landrein@gmail.com> Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-13 r/5125 feat(tvix/eval): implement builtins.baseNameOfAdam Joseph3-0/+16
This commit implements builtins.baseNameOf and adds a test case eval-okay-basenameof.nix to the test suite. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: Ib8bbafba2ac9ca0e1d3dc5e844167f94890d9fee Reviewed-on: https://cl.tvl.fyi/c/depot/+/6997 Tested-by: BuildkiteCI Reviewed-by: grfn <grfn@gws.fyi>
2022-10-13 r/5124 fix(tvix/eval): parseDrvName should not coerce, and xfail testAdam Joseph2-2/+3
builtins.parseDrvName should not coerce its argument to a string. This commit fixes that oversight in my previous commit, and adds an xfail test to cover this condition. Thanks to @sterni for noticing this. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I76bc78f1a82e1e08fe5c787c563a221d55de2639 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6991 Tested-by: BuildkiteCI Reviewed-by: sterni <sternenseemann@systemli.org>
2022-10-13 r/5123 feat(tvix/eval): enable the XFAIL testsAdam Joseph2-14/+45
This commit adds support for running the "expected failure" tests in both the nix and tvix test suites. I have disabled the eval-fail-blackhole.nix test because it gets stuck running forever. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: Iba75ce6c8f2becab3c834fcfdd9f4fdc5a4bdb9f Reviewed-on: https://cl.tvl.fyi/c/depot/+/6990 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su> Reviewed-by: grfn <grfn@gws.fyi>
2022-10-13 r/5122 test(tvix/eval): update for cppnix PR #7149Adam Joseph1-0/+3
cppnix has merged #7149, so let's update our copy of their tests: https://github.com/NixOS/nix/pull/7149 Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I0be3bf9da937abd24102e1997daa2087ecfafd98 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6992 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-13 r/5121 fix(tvix/eval): fix Compiler::new on wasmVincent Ambo1-1/+3
This path normalisation business causes runtime panics on WebAssembly because those operations are unsupported. Maybe this shouldn't be happening in the compiler anyways, not sure, but for now this commit adds a workaround based on the target to disable the normalisation if we're compiling for wasm. Change-Id: I908a84fbdffc3401f8d443e2c73ec673e9f397ff Reviewed-on: https://cl.tvl.fyi/c/depot/+/7004 Autosubmit: tazjin <tazjin@tvl.su> Reviewed-by: sterni <sternenseemann@systemli.org> Tested-by: BuildkiteCI
2022-10-13 r/5120 fix(tvix/eval): src/compiler: ensure root_dir is absoluteAdam Joseph1-8/+16
Cppnix immediately absolutizes pathnames at parse time; if you write `./foo`, it is immediately converted to `$(pwd)/foo` and manipulated as an absolute path at all times. To avoid having to introduce filesystem access operations in the implementation of otherwise-pure builtins, let's guarantee that the `root_dir` of the VM is always an absolute path. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I7cbbae2cba4b2716ff3f5ff7c9ce0ad529358c8a Reviewed-on: https://cl.tvl.fyi/c/depot/+/6995 Reviewed-by: grfn <grfn@gws.fyi> Tested-by: BuildkiteCI
2022-10-13 r/5119 refactor(tvix/eval): factor out all calls to canon_pathAdam Joseph4-5/+18
Right now we're pretending that the Rust library path_clean does the same thing that cppnix's canonPath() does. This is not true. It's close enough for the test suite, but may come back to bite us. Let's create our own canon_path() function and call that in all the places where we intend to match the behavior of cppnix's canonPath(). That way when we fix this we can fix it once, in one place. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: Ia6f9577f62f49ef352ff9cfa5efdf37c32d31b11 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6993 Tested-by: BuildkiteCI Reviewed-by: grfn <grfn@gws.fyi>
2022-10-13 r/5118 fix(tvix/eval): nix_search_path.rs: use /etc instead of /binAdam Joseph1-2/+2
On Debian machines (and most FHS distros) /bin is now a symlink to /usr/bin, which causes this test to fail. Let's use /etc as a root-relative test case, since it is less likely to be a symlink. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I1c94de0d2a242394182442fe1c895dc17eb04a4a Reviewed-on: https://cl.tvl.fyi/c/depot/+/6994 Tested-by: BuildkiteCI Reviewed-by: grfn <grfn@gws.fyi>
2022-10-13 r/5117 feat(tvix/eval): allow to disable warningsAdam Joseph2-7/+28
The nix_tests test suite produces lots of warnings. We can't fix these, since they are kept in sync with upstream, so there's little point in cluttering up the console with them every time the tests are run. Let's add a clap flag "warnings" and TVIX_WARNINGS environment variable. The default is "true". The test runner overrides this default and mutes the warnings. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I4b065f96fe15838afcca6970491a54e248ae4df7 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6985 Reviewed-by: tazjin <tazjin@tvl.su> Reviewed-by: grfn <grfn@gws.fyi> Tested-by: BuildkiteCI
2022-10-12 r/5115 test(tvix/eval): add eval-okay-parsedrvname.{exp,nix}Adam Joseph2-0/+13
Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I59254a0ac513b3d2a6b2184554358bb6eba2ba99 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6949 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-12 r/5114 feat(tvix/eval): builtins: implement parseDrvNameAdam Joseph1-0/+25
This commit passes nix_eval_okay_src_tests_nix_tests_eval_okay_versions_nix. See also: https://github.com/NixOS/nix/pull/7149 Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I24605c2a0cd0da434f37f6c518f20693bfa1b799 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6913 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-12 r/5113 refactor(tvix/eval) s/NixPath/NixSearchPath/Adam Joseph5-46/+49
Since NixString is the Rust type for nix strings, people might mistake NixPath for the Rust type of nix paths, which it is not. Let's call it NixSearchPath instead. Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: Ib2ea155c4b27cb90d6180a04ea7b951d86607373 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6927 Reviewed-by: kanepyork <rikingcoding@gmail.com> Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-12 r/5111 fix(tvix/eval): remove overlapping instancesAdam Joseph1-12/+0
I broke the build. I didn't understand that before hitting "submit" you need to re-test your changes on latest HEAD (and that CI doesn't do that for you); I failed to re-test cl/6912 following the merge of cl/6907. This commit fixes the build by removing the overlapping instances. Change-Id: I2a720d2c60cc7103b350f78102a8998f93bac828 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6965 Tested-by: BuildkiteCI Autosubmit: Adam Joseph <adam@westernsemico.com> Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-12 r/5110 feat(tvix/eval): From<Into<NixString>> for ValueAdam Joseph1-0/+9
Signed-off-by: Adam Joseph <adam@westernsemico.com> Change-Id: I27615baf4104ab96600708b12a780599a3714bae Reviewed-on: https://cl.tvl.fyi/c/depot/+/6912 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-12 r/5109 feat(tvix/eval): From<Utf8Error> for ErrorKindAdam Joseph1-0/+7
Change-Id: I074d9e862fbdd4e78e443cbaf0e77c7ffe825a10 Signed-off-by: Adam Joseph <adam@westernsemico.com> Reviewed-on: https://cl.tvl.fyi/c/depot/+/6911 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-11 r/5105 fix(tvix/eval/builtins): force acc not list element in foldl'sterni3-1/+10
When investigating discrepancies between foldl' in tvix and C++ Nix, I discovered that C++ Nix's foldl' doesn't seem to be strict at all. Since this seemed wrong, I looked into Haskell's foldl' implementation which doesn't force the list elements (`val` in our code), but the accumulation value (`res` in our code). You can look at the code here: https://hackage.haskell.org/package/base-4.17.0.0/docs/src/GHC.List.html#foldl%27 This actually makes a lot of sense: If `res` is not forced after each application of `op`, we'll end up thunks nested as deeply as the list is long, potentially taking up a lot of space. This can be limited by forcing the `res` thunk before applying `op` again (and creating a new thunk). I've also PR-ed an equivalent change for C++ Nix at https://github.com/NixOS/nix/pull/7158. Since this is not merged nor backported to our Nix 2.3 fork, I've not copied the eval fail test yet, since it wouldn't when checking our tests against C++ Nix in depot. Change-Id: I34edf6fc3031fc1485c3e714f2280b4fba8f004b Reviewed-on: https://cl.tvl.fyi/c/depot/+/6947 Autosubmit: sterni <sternenseemann@systemli.org> Reviewed-by: grfn <grfn@gws.fyi> Tested-by: BuildkiteCI
2022-10-11 r/5103 feat(tvix/eval): Allow directly evaluating an expr via mainGriffin Smith1-0/+7
This *maybe* should do something to check that we don't pass both a file and an expr, but for now this is useful enough to cut corners (plus we're probably due for a CLI revamp eventually anyway). Change-Id: Id44357074150b336b6215ba596cc52d01d037dbd Reviewed-on: https://cl.tvl.fyi/c/depot/+/6941 Reviewed-by: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-11 r/5102 feat(tvix/eval): Implement builtins.removeAttrsGriffin Smith4-1/+31
Change-Id: I28910991a0108436a42ac7bf3458f9180a44154e Reviewed-on: https://cl.tvl.fyi/c/depot/+/6928 Reviewed-by: Adam Joseph <adam@westernsemico.com> Reviewed-by: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-11 r/5101 feat(tvix/eval): Implement builtins.traceGriffin Smith1-0/+12
This is currently implemented with a simple println inline, but in the future we could hook into this via something pluggable on the VM. Change-Id: Idd9cc3b34aa13d6ebc64c02aade81ecdf439656a Reviewed-on: https://cl.tvl.fyi/c/depot/+/6938 Autosubmit: grfn <grfn@gws.fyi> Reviewed-by: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-11 r/5100 fix(tvix/eval): Pop frames even if running op failsGriffin Smith3-269/+285
Previously, the VM assumed that if an error was returned from `run()`, the evaluation was "finished" and the state of the VM didn't matter. This used to be a reasonable assumption, but now that we've got `tryEval` around we need to actually make sure that we clean up after ourselves if we're about to return an error. Specifically, if the *last* instruction in an evaluation frame returns an error, we previously wouldn't pop that evaluation frame, which could cause all sorts of bizarre errors based on what happened to be in the stack at the time. This commit splits out a `run_op` method from `VM::run`, and uses that to check the evaluation frame return condition even if the op we're running is about to return an error, and pop the evaluation frame if we're at the last instruction. Change-Id: Ib40649d8915ee1571153cb71e3d76492542fc3d7 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6940 Tested-by: BuildkiteCI Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-11 r/5099 feat(tvix/eval): observe stack after exiting call frames/builtinsVincent Ambo2-8/+22
Change-Id: I1937d37551503a0b6bb0ac899d067302e4791e5f Reviewed-on: https://cl.tvl.fyi/c/depot/+/6939 Reviewed-by: grfn <grfn@gws.fyi> Autosubmit: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-11 r/5098 fix(tvix/eval): implement functor calling for non-tail-callsVincent Ambo1-0/+15
Change-Id: I1113921c46010021b61c412d74d60193c156e0f1 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6937 Reviewed-by: grfn <grfn@gws.fyi> Autosubmit: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-10 r/5097 fix(tvix/eval): Actually trace spans for thunksGriffin Smith4-22/+34
Currently, the span on *all* thunk force errors is the span at which the thunk is forced, which for recursive thunk forcing ends up just being the same span over and over again. This changes the span on thunk force errors to be the span at which point the thunk is *created*, which is a bit more helpful (though the printing atm is a little... crowded). To make this work, we have to thread through the span at which a thunk is created into a field on the thunk itself. Change-Id: I81474810a763046e2eb3a8f07acf7d8ec708824a Reviewed-on: https://cl.tvl.fyi/c/depot/+/6932 Autosubmit: grfn <grfn@gws.fyi> Reviewed-by: Adam Joseph <adam@westernsemico.com> Reviewed-by: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-10 r/5096 feat(tvix/eval): Implement builtins.getEnvGriffin Smith1-1/+7
Change-Id: I09f512a60989a37184e73e521d4a3aa23f33a1a8 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6922 Autosubmit: grfn <grfn@gws.fyi> Tested-by: BuildkiteCI Reviewed-by: Adam Joseph <adam@westernsemico.com> Reviewed-by: kanepyork <rikingcoding@gmail.com>
2022-10-10 r/5095 feat(tvix/eval): Implement builtins.readFileGriffin Smith4-1/+16
Change-Id: If3fd0b087009a2bfbad8bb7aca0aa20de906eb12 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6921 Tested-by: BuildkiteCI Reviewed-by: Adam Joseph <adam@westernsemico.com> Reviewed-by: kanepyork <rikingcoding@gmail.com> Autosubmit: grfn <grfn@gws.fyi> Reviewed-by: tazjin <tazjin@tvl.su>
2022-10-10 r/5094 refactor(tvix/eval): after calling, the caller has to popVincent Ambo3-21/+29
Previously the various call functions either returned `EvalResult<()>` or `EvalResult<Value>`, which was confusing. Now only vm::call_with returns a Value directly, and other parts of the API just leave the stack top in the post-call state. This makes it easier to reason about what's going on in non-tail-call cases (which are making a comeback). Change-Id: I264ffc683a11aca72dd06e2220a5ff6e7c5fc2b0 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6936 Tested-by: BuildkiteCI Reviewed-by: grfn <grfn@gws.fyi>
2022-10-10 r/5092 fix(tvix/eval): format nested compiler errors in fancy outputVincent Ambo1-10/+22
Instead of just printing the number of errors (useless!) actually emit separate diagnostics for each nested error. Change-Id: I97b53c3276c906af5def89077b5b6ba6ec108b37 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6933 Reviewed-by: grfn <grfn@gws.fyi> Autosubmit: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI
2022-10-10 r/5091 fix(tvix/eval): path resolution errors are catchableGriffin Smith3-2/+3
Despite this not being documented, `tryEval` is empirically able to catch errors caused by a <...> path not resolving (and nixpkgs depends on this). Change-Id: Ia3b78a2d9d2d0c603aba829518b351102dc55396 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6926 Reviewed-by: sterni <sternenseemann@systemli.org> Autosubmit: grfn <grfn@gws.fyi> Tested-by: BuildkiteCI
2022-10-10 r/5090 fix(tvix/eval): Thunk `if` exprGriffin Smith3-1/+10
Since the body of an `if` expr can refer to deferred upvalues, it needs to be thunked so when we actually compile those deferred upvalues we have something for the finalize op to point at. Without this all sorts of weird things can happen due to the finalize op being run in the wrong lambda context, up to and including a panic. Change-Id: I040d5e1a7232fd841cfa4953539898fa49cbbb83 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6929 Reviewed-by: sterni <sternenseemann@systemli.org> Reviewed-by: tazjin <tazjin@tvl.su> Autosubmit: grfn <grfn@gws.fyi> Tested-by: BuildkiteCI
2022-10-10 r/5089 feat(tvix/eval): Implement builtins.tryEvalGriffin Smith4-0/+32
With asserts compiled using conditional jumps, this ends up being quite straightforward - the only real tricky bit is that we have to know whether an error can or can't be handled. Change-Id: I75617da73b7a9c5cdd888c0e26ae81d2c5c0d714 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6924 Reviewed-by: sterni <sternenseemann@systemli.org> Autosubmit: grfn <grfn@gws.fyi> Tested-by: BuildkiteCI
2022-10-10 r/5088 feat(tvix/eval): Implement builtins.pathExistsGriffin Smith3-6/+10
Change-Id: Ife8a690e9036868964771893ab29a9ae3a2d2365 Reviewed-on: https://cl.tvl.fyi/c/depot/+/6919 Reviewed-by: tazjin <tazjin@tvl.su> Tested-by: BuildkiteCI