When taking the address of an unsized field we generate a rvalue datum
for the field and then convert it to an lvalue datum. At that point,
cleanup is scheduled for the field, leading to multiple drop calls.
The problem is that we generate an rvalue datum for the field, since the
pointer does not own the data and there's already cleanup scheduled
elsewhere by the true owner. Instead, an lvalue datum must be created.
Thanks to @eddyb for identifying the underlying cause and suggesting the
correct fix.
Fixes#25549Fixes#25515
Using regular pointer arithmetic to iterate collections of zero-sized types
doesn't work, because we'd get the same pointer all the time. Our
current solution is to convert the pointer to an integer, add an offset
and then convert back, but this inhibits certain optimizations.
What we should do instead is to convert the pointer to one that points
to an i8*, and then use a LLVM GEP instructions without the inbounds
flag to perform the pointer arithmetic. This allows to generate pointers
that point outside allocated objects without causing UB (as long as you
don't dereference them), and it wraps around using two's complement,
i.e. it behaves exactly like the wrapping_* operations we're currently
using, with the added benefit of LLVM being able to better optimize the
resulting IR.
* segfault due to not copying drop flag when coercing
* fat pointer casts
* segfault due to not checking drop flag properly
* debuginfo for DST smart pointers
* unreachable code in drop glue
There were still some mentions of `~[T]` and `~T`, mostly in comments and debugging statements. I tried to do my best to preserve meaning, but I might have gotten some wrong-- I'm happy to fix anything :)
An automated script was run against the `.rs` and `.md` files,
subsituting every occurrence of `task` with `thread`. In the `.rs`
files, only the texts in the comment blocks were affected.
Also remove comments that reference the unique_type_id HEAP_VEC_BOX
metadata, which was removed in 3e62637 and the unique_type_id GC_BOX
metadata, which was removed in 8a91d33.
The former stopped making sense when we started interning substs and made
TraitRef a 2-word copy type, and I'm moving the latter into an arena as
they live as long as the type context.
These are useful when you want to catch the signals, like when you're making a kernel, or if you just don't want the overhead. (I don't know if there are any of the second kind of people, I don't think it's a good idea, but hey, choice is good).
Inspect enum discriminant *after* calling its destructor
Includes some drive-by cleanup (e.g. changed some field and method names to reflect fill-on-drop; added comments about zero-variant enums being classified as `_match::Single`).
Probably the most invasive change was the expansion of the maps `available_drop_glues` and `drop_glues` to now hold two different kinds of drop glues; there is the (old) normal drop glue, and there is (new) drop-contents glue that jumps straight to dropping the contents of a struct or enum, skipping its destructor.
* For all types that do not have user-defined Drop implementations, the normal glue is generated as usual (i.e. recursively dropping the fields of the data structure).
(And this actually is exactly what the newly-added drop-contents glue does as well.)
* For types that have user-defined Drop implementations, the "normal" drop glue now schedules a cleanup before invoking the `Drop::drop` method that will call the drop-contents glue after that invocation returns.
Fix#23611.
----
Is this a breaking change? The prior behavior was totally unsound, and it seems unreasonable that anyone was actually relying on it.
Nonetheless, since there is a user-visible change to the language semantics, I guess I will conservatively mark this as a:
[breaking-change]
(To see an example of what sort of user-visible change this causes, see the comments in the regression test.)
Closes#17841.
The majority of the work should be done, e.g. trait and inherent impls, different forms of UFCS syntax, defaults, and cross-crate usage. It's probably enough to replace the constants in `f32`, `i8`, and so on, or close to good enough.
There is still some significant functionality missing from this commit:
- ~~Associated consts can't be used in match patterns at all. This is simply because I haven't updated the relevant bits in the parser or `resolve`, but it's *probably* not hard to get working.~~
- Since you can't select an impl for trait-associated consts until partway through type-checking, there are some problems with code that assumes that you can check constants earlier. Associated consts that are not in inherent impls cause ICEs if you try to use them in array sizes or match ranges. For similar reasons, `check_static_recursion` doesn't check them properly, so the stack goes ka-blooey if you use an associated constant that's recursively defined. That's a bit trickier to solve; I'm not entirely sure what the best approach is yet.
- Dealing with consts associated with type parameters will raise some new issues (e.g. if you have a `T: Int` type parameter and want to use `<T>::ZERO`). See rust-lang/rfcs#865.
- ~~Unused associated consts don't seem to trigger the `dead_code` lint when they should. Probably easy to fix.~~
Also, this is the first time I've been spelunking in rustc to such a large extent, so I've probably done some silly things in a couple of places.
Kudos to dotdash for tracking down this fix.
Presumably the use of `ByRef` was because this value is a reference to
the drop-flag; but an Lvalue will serve just as well for that. dotdash
argues:
> since the drop_flag is in its "final home", Lvalue seems to be the
> correct choice.
That is, scheduled drops are executed in reverse order, so for
correctness, we *schedule* the lifetime end before we schedule the
drop, so that when they are executed, the drop will be executed
*before* the lifetime end.
These new intrinsics are comparable to `atomic_signal_fence` in C++,
ensuring the compiler will not reorder memory accesses across the
barrier, nor will it emit any machine instructions for it.
Closes#24118, implementing RFC 888.
This addresses to-do in my code, and simplifies this method a lot to boot.
(The necessary enum dispatch has now effectively been shifted entirely
into the scheduled cleanup code for the enum contents.)
without invoking the Drop::drop implementation.
This is necessary for dealing with an enum that switches own `self` to
a different variant while running its destructor.
Fix#23611.
(This may not be the *best* fix, compared to e.g. returning
`_match::NoBranch` from `trans_switch` on a zero-variant enum. But it
is one of the *simplest* fixes available.)
This commit removes all the old casting/generic traits from `std::num` that are
no longer in use by the standard library. This additionally removes the old
`strconv` module which has not seen much use in quite a long time. All generic
functionality has been supplanted with traits in the `num` crate and the
`strconv` module is supplanted with the [rust-strconv crate][rust-strconv].
[rust-strconv]: https://github.com/lifthrasiir/rust-strconv
This is a breaking change due to the removal of these deprecated crates, and the
alternative crates are listed above.
[breaking-change]
Loading from and storing to small aggregates happens by casting the
aggregate pointer to an appropriately sized integer pointer to avoid
the usage of first class aggregates which would lead to less optimized
code.
But this means that, for example, a tuple of type (i16, i16) will be
loading through an i32 pointer and because we currently don't provide
alignment information LLVM assumes that the load should use the ABI
alignment for i32 which would usually be 4 byte alignment. But the
alignment requirement for the (i16, i16) tuple will usually be just 2
bytes, so we're overestimating alignment, which invokes undefined
behaviour.
Therefore we must emit appropriate alignment information for
stores/loads through such casted pointers.
Fixes#23431
I doubt this PR is ready to merge as-is, for a couple reasons:
* There are no tests for this change. I'm not sure how to add tests for this change, as it modifies the C ABI for a cross-compilation target. Anecdotally, I have an iOS library I've been working on, and before this change, it crashes running on an arm64 device due to bad calling conventions (a simplified example is in #24154), and after this change, it runs correctly.
* This is my first foray into LLVM. I did my best to reimplement what Clang does for AArch64 codegen (https://github.com/llvm-mirror/clang/blob/master/lib/CodeGen/TargetInfo.cpp), particularly in `ABIInfo::isHomogeneousAggregate`, `AArch64ABIInfo::isHomogeneousAggregateBaseType`, and `AArch64ABIInfo::isHomogeneousAggregateSmallEnough`, but I'm not confident I got a complete translation, particularly because Clang includes a lot of checks that I don't believe are necessary for rustc.
Fixes#24154.
We provide tools to tell what exact symbols to emit for any fn or static, but
don’t quite check if that won’t cause any issues later on. Some of the issues
include LLVM mangling our names again and our names pointing to wrong locations,
us generating dumb foreign call wrappers, linker errors, extern functions
resolving to different symbols altogether (`extern {fn fail();} fail();` in some
cases calling `fail1()`), etc.
Before the commit we had a function called `note_unique_llvm_symbol`, so it is
clear somebody was aware of the issue at some point, but the function was barely
used, mostly in irrelevant locations.
Along with working on it I took liberty to start refactoring trans/base into
a few smaller modules. The refactoring is incomplete and I hope I will find some
motivation to carry on with it.
This is possibly a [breaking-change] because it makes dumbly written code
properly invalid.
This fixes all those issues about incorrect use of #[no_mangle] being not reported/misreported/ICEd by the compiler.
NB. This PR does not attempt to tackle the parallel codegen issue that was mentioned in https://github.com/rust-lang/rust/pull/22811, but I believe it should be very straightforward in a follow up PR by modifying `trans::declare::get_defined_value` to look at all the contexts.
cc @alexcrichton @huonw @nrc because you commented on the original RFC issue.
EDIT: wow, this became much bigger than I initially intended.
Implements an intrinsic for extracting the value of the discriminant
enum variant values. For non-enum types, this returns zero, otherwise it
returns the value we use for discriminant comparisons. This means that
enum types that do not have a discriminant will also work in this
arrangement.
This is (at least part of) the work on Issue #24263
In addition to being nicer, this also allows you to use `sum` and `product` for
iterators yielding custom types aside from the standard integers.
Due to removing the `AdditiveIterator` and `MultiplicativeIterator` trait, this
is a breaking change.
[breaking-change]
We provide tools to tell what exact symbols to emit for any fn or static, but
don’t quite check if that won’t cause any issues later on. Some of the issues
include LLVM mangling our names again and our names pointing to wrong locations,
us generating dumb foreign call wrappers, linker errors, extern functions
resolving to different symbols altogether (extern {fn fail();} fail(); in some
cases calling fail1()), etc.
Before the commit we had a function called note_unique_llvm_symbol, so it is
clear somebody was aware of the issue at some point, but the function was barely
used, mostly in irrelevant locations.
Along with working on it I took liberty to start refactoring trans/base into
a few smaller modules. The refactoring is incomplete and I hope I will find some
motivation to carry on with it.
This is possibly a [breaking-change] because it makes dumbly written code
properly invalid.
This PR solves #21559 by making sure that unreachable if-expressions are not further translated.
Could someone who knows their way around `trans` take a look at the changes in `controlflow.rs`? I'm not sure if any other code relies on any side-effects of translating unreachable things.
cc @nikomatsakis @nrc @eddyb
const_eval : add overflow-checking for {`+`, `-`, `*`, `/`, `<<`, `>>`}.
One tricky detail here: There is some duplication of labor between `rustc::middle::const_eval` and `rustc_trans::trans::consts`. It might be good to explore ways to try to factor out the common structure to the two passes (by abstracting over the particular value-representation used in the compile-time interpreter).
----
Update: Rebased atop #23841Fix#22531Fix#23030Fix#23221Fix#23235
Add option-returning variants to `const_to_int`/`const_to_uint` that
never assert fail. (These will be used for overflow checking from
rustc_trans::trans::consts.)
