Wesley Wiser
4e58e2e3a3
Work around for #64506
2019-09-27 20:11:13 -04:00
Wesley Wiser
bc17936c8a
[const-prop] Replace eval_place() with use of InterpCx
2019-09-27 20:11:12 -04:00
varkor
bea3d67c77
Rename subst::Kind to subst::GenericArg
2019-09-26 11:48:05 +01:00
Tyler Mandry
dbce655eda
Rollup merge of #64503 - RalfJung:miri-retag, r=oli-obk
...
rename Allocation::retag -> with_tags_and_extra
This is more consistent with `Pointer::with_tag`. Also, "retag" is a [term from Stacked Borrows](https://github.com/rust-lang/unsafe-code-guidelines/blob/master/wip/stacked-borrows.md#retagging ) with a [corresponding Machine hook](https://doc.rust-lang.org/nightly/nightly-rustc/rustc_mir/interpret/machine/trait.Machine.html#method.retag ), and this function has nothing to do with that other use of the term.
r? @oli-obk
2019-09-17 21:27:23 -07:00
Tyler Mandry
f8b6e26cfb
Rollup merge of #64541 - RalfJung:miri-errors, r=oli-obk
...
document Miri error categories
r? @oli-obk
2019-09-17 14:10:55 -07:00
Oliver Scherer
0de9485038
Get rid of special const intrinsic query in favour of const_eval
2019-09-17 16:31:55 +02:00
Ralf Jung
e8d858ef3c
rename Allocation::retag -> with_tags_and_extra
2019-09-17 09:29:06 +02:00
Ralf Jung
daafeb35b7
document Miri error categories
2019-09-17 08:58:51 +02:00
Mazdak Farrokhzad
45baedb14b
Rollup merge of #64116 - ZeGentzy:patch-1, r=oli-obk
...
Fix minor typo in docs.
This comment is prolly refering to the comment in [const_eval_raw_provider](https://github.com/rust-lang/rust/pull/64116/files#diff-e0b58bb6712edaa8595ad7237542c958R616 ), not itself.
2019-09-14 16:42:20 +02:00
Hal Gentz
094af9743b
Update value.rs
2019-09-13 20:37:38 -06:00
Alexander Regueiro
fd48ca20d3
Apply suggestions from code review
...
Co-Authored-By: Mazdak Farrokhzad <twingoow@gmail.com>
2019-09-07 16:27:20 +01:00
Alexander Regueiro
c1d29ee3c7
Aggregation of cosmetic changes made during work on REPL PRs: librustc
2019-09-07 16:27:19 +01:00
Mazdak Farrokhzad
afb6a7002d
Rollup merge of #64202 - alexreg:rush-pr-1, r=Centril
...
Fixed grammar/style in some error messages
Factored out from hacking on rustc for work on the REPL.
r? @Centril
2019-09-06 19:00:49 +02:00
Oliver Scherer
39bfb3626c
Fix miri
2019-09-06 11:10:53 +02:00
Alexander Regueiro
022d9c8eb5
Fixed grammar/style in error messages and reblessed tests.
2019-09-06 03:46:08 +01:00
Hal Gentz
7589d3a400
Update other doc.
...
Value was renamed to Operand in ad2de8b4eefe50b4eb1d
2019-09-03 01:54:48 -06:00
bors
b50520835f
Auto merge of #63561 - HeroicKatora:alloc-private-bytes, r=oli-obk
...
Make Allocation::bytes private
Fixes #62931 .
Direct immutable access to the bytes is still possible but redirected through the new method `raw_bytes_with_undef_and_ptr`, similar to `get_bytes_with_undef_and_ptr` but without requiring an interpretation context and not doing *any* relocation or bounds checks. The `size` of the allocation is stored separately which makes access as `Size` and `usize` more ergonomic.
cc: @RalfJung
2019-09-02 22:07:40 +00:00
Andreas Molzer
f3c435eb78
Reorder AllocationDefinedness members
...
This improves the clarity of the documentation a bit since they can
reference each other when reading the member docs in sequence.
2019-08-31 21:21:29 +02:00
Andreas Molzer
823c3b9845
Improve documentation around allocation accessors
2019-08-31 17:19:01 +02:00
Andreas Molzer
bee2d3748e
Move relocation range copies into allocation
2019-08-30 04:17:18 +02:00
Andreas Molzer
6fe31fefd8
Make allocation relocation field private
2019-08-29 20:02:21 +02:00
Andreas Molzer
85d6b7b9d3
Address naming and comments from reviews
2019-08-28 03:58:42 +02:00
Andreas Molzer
7b941e368f
Expose encapsulated undef mask as immutable
2019-08-21 10:55:16 +02:00
Andreas Molzer
d8c5bc7ec6
Replace usage of alloc.bytes in interpret
...
There is now a dedicate `len` method which avoids the need to access the
bytes. Access the length as `Size` can also be done by a direct member.
The constructors guarantee that these representations are convertable.
Access which relies on the bytes, such as snapshot, can use direct raw
access by reference as it does not care about undef and relocations or
properly checks them seperately.
2019-08-21 10:55:16 +02:00
bors
cdff918955
Auto merge of #63670 - Dante-Broggi:patch-2, r=Centril
...
Size has a ::zero
2019-08-19 05:12:58 +00:00
Dante-Broggi
d64f06ce31
size has a zero
2019-08-17 16:09:49 -04:00
Dante-Broggi
a7c34f1ce9
fix typos
2019-08-17 15:36:28 -04:00
Andreas Molzer
98cff69289
Move copy of undef_mask into allocation
...
This also means that the compressed representation chosen may be
optimized together with any changes to the undef_mask.
2019-08-17 19:58:15 +02:00
Andreas Molzer
2228b3f086
Derive HashStable for Allocation
...
Requires a manual implementation for Relocations since dereferencing to
SortedMap is not always implemented but that one is far more trivial.
Added fields could otherwise be silently forgotten since private fields
make destructing outside the module possible only with `..` pattern
which would then also be applicable to newly added public fields.
2019-08-17 19:58:15 +02:00
Andreas Molzer
9b9eecf964
Store allocation size, make bytes, undef_mask private
...
Direct access to the bytes was previously a problem (#62931 ) where
components would read their contents without properly checking
relocations and/or definedness.
Making bytes private instead of purely renaming them also helps in
allowing amendments to their allocation scheme (such as eliding
allocation for undef of constant regions).
2019-08-17 18:43:02 +02:00
Mazdak Farrokhzad
c358476c1b
Rollup merge of #63075 - RalfJung:deref-checks, r=oli-obk
...
Miri: Check that a ptr is aligned and inbounds already when evaluating `*`
This syncs Miri with what the Nomicon and the Reference say, and resolves https://github.com/rust-lang/miri/issues/447 .
Also this would not have worked without https://github.com/rust-lang/rust/pull/62982 due to new cycles. ;)
r? @oli-obk
2019-08-14 22:56:20 +02:00
Ralf Jung
440a5c8100
rename RUST_CTFE_BACKTRACE to RUSTC_CTFE_BACKTRACE
2019-08-10 16:37:40 +02:00
bors
5421d94960
Auto merge of #63152 - estebank:big-array, r=oli-obk
...
Always error on `SizeOverflow` during mir evaluation
Fix #55878 , fix #25116 .
r? @oli-obk
2019-08-07 04:28:08 +00:00
bors
4be0675589
Auto merge of #63079 - RalfJung:ctfe-no-align, r=oli-obk
...
CTFE: simplify ConstValue by not checking for alignment
I hope the test suite actually covers the problematic cases here?
r? @oli-obk
Fixes https://github.com/rust-lang/rust/issues/61952
2019-08-05 08:16:51 +00:00
Esteban Küber
bdd79b849e
tweak output and tests
2019-08-04 12:23:05 -07:00
Ralf Jung
b9d4c759f3
AssumptionNotHeld is used only once in Miri and never caught... remove from enum
2019-08-04 11:59:14 +02:00
Ralf Jung
18daa766f0
move AssumptionNotHeld to UB
2019-08-04 11:21:41 +02:00
Esteban Küber
2c5684208c
avoid mutable state and override main message
2019-08-03 13:37:44 -07:00
Ralf Jung
89a370db0f
add variant for experimental UB (like Stacked Borrows)
2019-08-03 20:36:05 +02:00
Ralf Jung
fe08d9e2d2
also add macros for free-form error messages
2019-08-03 20:35:52 +02:00
Ralf Jung
1e24c73ae5
dedup free-form Unsupported errors; add macros for free-form UB and Unsupported cases
2019-08-02 23:41:24 +02:00
Ralf Jung
1b132a2f41
be less British
2019-08-02 23:31:21 +02:00
Ralf Jung
e590b849b8
CTFE: simplify Value type by not checking for alignment
2019-08-02 23:02:53 +02:00
Esteban Küber
d3da411a08
Nicer labels for type layout errors
2019-08-02 12:33:58 -07:00
Esteban Küber
58bd8786e8
Do not lint on SizeOverflow, always error
2019-08-02 12:31:31 -07:00
Esteban Küber
83b5eb9615
Always error on SizeOverflow during mir evaluation
2019-08-02 12:30:46 -07:00
Saleem Jaffer
00d32e8687
code review fixes
2019-08-01 12:13:49 +05:30
Saleem Jaffer
0c4513e8ed
code review fixes
2019-08-01 09:49:01 +05:30
Saleem Jaffer
c17d11fb39
code review fixes
2019-07-31 18:14:49 +05:30
Saleem Jaffer
152f0d347e
code review fixes
2019-07-31 12:48:54 +05:30
