The Compiler Query System is the key to our new demand-driven organization. The idea is pretty simple. You have various queries that compute things about the input -- for example, there is a query called type_of(def_id) that, given the def-id of some item, will compute the type of that item and return it to you.
Query execution is memoized -- so the first time you invoke a query, it will go do the computation, but the next time, the result is returned from a hashtable. Moreover, query execution fits nicely into incremental computation; the idea is roughly that, when you do a query, the result may be returned to you by loading stored data from disk (but that‘s a separate topic we won’t discuss further here).
The overall vision is that, eventually, the entire compiler control-flow will be query driven. There will effectively be one top-level query (“compile”) that will run compilation on a crate; this will in turn demand information about that crate, starting from the end. For example:
However, that vision is not fully realized. Still, big chunks of the compiler (for example, generating MIR) work exactly like this.
To invoke a query is simple. The tcx (“type context”) offers a method for each defined query. So, for example, to invoke the type_of query, you would just do this:
let ty = tcx.type_of(some_def_id);
Currently, cycles during query execution should always result in a compilation error. Typically, they arise because of illegal programs that contain cyclic references they shouldn't (though sometimes they arise because of compiler bugs, in which case we need to factor our queries in a more fine-grained fashion to avoid them).
However, it is nonetheless often useful to recover from a cycle (after reporting an error, say) and try to soldier on, so as to give a better user experience. In order to recover from a cycle, you don't get to use the nice method-call-style syntax. Instead, you invoke using the try_get method, which looks roughly like this:
use ty::query::queries; ... match queries::type_of::try_get(tcx, DUMMY_SP, self.did) { Ok(result) => { // no cycle occurred! You can use `result` } Err(err) => { // A cycle occurred! The error value `err` is a `DiagnosticBuilder`, // meaning essentially an "in-progress", not-yet-reported error message. // See below for more details on what to do here. } }
So, if you get back an Err from try_get, then a cycle did occur. This means that you must ensure that a compiler error message is reported. You can do that in two ways:
The simplest is to invoke err.emit(). This will emit the cycle error to the user.
However, often cycles happen because of an illegal program, and you know at that point that an error either already has been reported or will be reported due to this cycle by some other bit of code. In that case, you can invoke err.cancel() to not emit any error. It is traditional to then invoke:
tcx.sess.delay_span_bug(some_span, "some message")
delay_span_bug() is a helper that says: we expect a compilation error to have happened or to happen in the future; so, if compilation ultimately succeeds, make an ICE with the message "some message". This is basically just a precaution in case you are wrong.
So you may be wondering what happens when you invoke a query method. The answer is that, for each query, the compiler maintains a cache -- if your query has already been executed, then, the answer is simple: we clone the return value out of the cache and return it (therefore, you should try to ensure that the return types of queries are cheaply cloneable; insert a Rc if necessary).
If, however, the query is not in the cache, then the compiler will try to find a suitable provider. A provider is a function that has been defined and linked into the compiler somewhere that contains the code to compute the result of the query.
Providers are defined per-crate. The compiler maintains, internally, a table of providers for every crate, at least conceptually. Right now, there are really two sets: the providers for queries about the local crate (that is, the one being compiled) and providers for queries about external crates (that is, dependencies of the local crate). Note that what determines the crate that a query is targeting is not the kind of query, but the key. For example, when you invoke tcx.type_of(def_id), that could be a local query or an external query, depending on what crate the def_id is referring to (see the self::keys::Key trait for more information on how that works).
Providers always have the same signature:
fn provider<'cx, 'tcx>(tcx: TyCtxt<'cx, 'tcx, 'tcx>, key: QUERY_KEY) -> QUERY_RESULT { ... }
Providers take two arguments: the tcx and the query key. Note also that they take the global tcx (i.e., they use the 'tcx lifetime twice), rather than taking a tcx with some active inference context. They return the result of the query.
When the tcx is created, it is given the providers by its creator using the Providers struct. This struct is generate by the macros here, but it is basically a big list of function pointers:
struct Providers { type_of: for<'cx, 'tcx> fn(TyCtxt<'cx, 'tcx, 'tcx>, DefId) -> Ty<'tcx>, ... }
At present, we have one copy of the struct for local crates, and one for external crates, though the plan is that we may eventually have one per crate.
These Provider structs are ultimately created and populated by librustc_driver, but it does this by distributing the work throughout the other rustc_* crates. This is done by invoking various provide functions. These functions tend to look something like this:
pub fn provide(providers: &mut Providers) { *providers = Providers { type_of, ..*providers }; }
That is, they take an &mut Providers and mutate it in place. Usually we use the formulation above just because it looks nice, but you could as well do providers.type_of = type_of, which would be equivalent. (Here, type_of would be a top-level function, defined as we saw before.) So, if we want to add a provider for some other query, let's call it fubar, into the crate above, we might modify the provide() function like so:
pub fn provide(providers: &mut Providers) { *providers = Providers { type_of, fubar, ..*providers }; } fn fubar<'cx, 'tcx>(tcx: TyCtxt<'cx, 'tcx>, key: DefId) -> Fubar<'tcx> { .. }
NB. Most of the rustc_* crates only provide local providers. Almost all extern providers wind up going through the rustc_metadata crate, which loads the information from the crate metadata. But in some cases there are crates that provide queries for both local and external crates, in which case they define both a provide and a provide_extern function that rustc_driver can invoke.
So suppose you want to add a new kind of query, how do you do so? Well, defining a query takes place in two steps:
To specify the query name and arguments, you simply add an entry to the big macro invocation in mod.rs. This will probably have changed by the time you read this README, but at present it looks something like:
define_queries! { <'tcx>
/// Records the type of every item.
[] fn type_of: TypeOfItem(DefId) -> Ty<'tcx>,
...
}
Each line of the macro defines one query. The name is broken up like this:
[] fn type_of: TypeOfItem(DefId) -> Ty<'tcx>, ^^ ^^^^^^^ ^^^^^^^^^^ ^^^^^ ^^^^^^^^ | | | | | | | | | result type of query | | | query key type | | dep-node constructor | name of query query flags
Let's go over them one by one:
tcx.type_of(..)). Also used as the name of a struct (ty::query::queries::type_of) that will be generated to represent this query.DepNode variant, which can be added by modifying the define_dep_nodes! macro invocation in librustc/dep_graph/dep_node.rs.ty::query::keys::Key trait, which defines (for example) how to map it to a crate, and so forth.RefCell or other interior mutability and (b) be cheaply cloneable. Interning or using Rc or Arc is recommended for non-trivial data types.ty::steal::Steal type, which is used to cheaply modify MIR in place. See the definition of Steal for more details. New uses of Steal should not be added without alerting @rust-lang/compiler.So, to add a query:
define_queries! using the format above.provide method; or add a new one if needed and ensure that rustc_driver is invoking it.For each kind, the define_queries macro will generate a “query struct” named after the query. This struct is a kind of a place-holder describing the query. Each such struct implements the self::config::QueryConfig trait, which has associated types for the key/value of that particular query. Basically the code generated looks something like this:
// Dummy struct representing a particular kind of query: pub struct type_of<'tcx> { phantom: PhantomData<&'tcx ()> } impl<'tcx> QueryConfig for type_of<'tcx> { type Key = DefId; type Value = Ty<'tcx>; }
There is an additional trait that you may wish to implement called self::config::QueryDescription. This trait is used during cycle errors to give a “human readable” name for the query, so that we can summarize what was happening when the cycle occurred. Implementing this trait is optional if the query key is DefId, but if you don't implement it, you get a pretty generic error (“processing foo...”). You can put new impls into the config module. They look something like this:
impl<'tcx> QueryDescription for queries::type_of<'tcx> { fn describe(tcx: TyCtxt<'_, '_, '_>, key: DefId) -> String { format!("computing the type of `{}`", tcx.item_path_str(key)) } }