Commit Graph

232 Commits

Author SHA1 Message Date
Oliver Schneider
dc1b0fb436 Compiles again 2017-03-21 13:53:55 +01:00
Scott Olson
0ca0676abd Merge pull request #146 from oli-obk/peephole1
[Peephole Optimization 1/n] Don't allocate for structs with a single primval field
2017-03-14 09:57:02 -07:00
Oliver Schneider
8405770b51 Rustup to rustc 1.17.0-nightly (be760566c 2017-02-28) 2017-03-02 13:11:33 +01:00
Oliver Schneider
8878a4030a rustup to rustc 1.17.0-nightly (60a0edc6c 2017-02-26) 2017-02-28 12:35:00 +01:00
Oliver Schneider
2282e6b582
represent single field structs as their single field 2017-02-24 11:04:13 +01:00
Oliver Schneider
529efc51e8
detect memory leaks 2017-02-24 10:41:15 +01:00
Scott Olson
0f714b72a0 Formatting. 2017-02-10 13:35:45 -08:00
Scott Olson
6ffd7005c1 Cache string and bytestring literal allocs. 2017-02-10 13:35:33 -08:00
Oliver Schneider
333264c956 clarify comment on drop glue 2017-02-10 08:13:18 +01:00
Oliver Schneider
e58f750a49 refactor drop glue 2017-02-09 19:27:07 +01:00
Oliver Schneider
4beb774caa don't mark the zst allocation as static 2017-02-08 17:24:20 +01:00
Oliver Schneider
080d3e4355 properly prevent recursive statics from marking each other 2017-02-08 16:27:28 +01:00
Oliver Schneider
fbfd2d4bca re-add spaces before static kind 2017-02-08 09:17:48 +01:00
Oliver Schneider
98cda6cb07 freeze -> static 2017-02-07 20:28:54 +01:00
Oliver Schneider
01ac19d358 fix static mut accidental dealloc or freeze 2017-02-07 19:20:16 +01:00
Oliver Schneider
45df853da7 cleanup buggy closure dispatch code 2017-02-07 14:22:20 +01:00
Scott Olson
9f7ca351e0 Dump fn ptr ABIs. 2017-02-07 01:57:42 -08:00
Scott Olson
097db58f30 Print fn name and type in dump_allocs. 2017-02-06 12:48:10 -08:00
Scott Olson
b36d8085a7 Fix duplicate allocation printing. 2017-02-06 09:26:01 -08:00
Scott Olson
7acbf7ef74 Cleanup: EvalResult<'a, ()> becomes EvalResult<'a>. 2017-02-04 13:09:10 -08:00
Oliver Schneider
c47c325252 remove duplicate alignment checks 2017-01-31 10:56:37 +01:00
Oliver Schneider
148c6de507 fix some unaligned reads 2017-01-31 10:36:46 +01:00
Oliver Schneider
96607d4593 document our packed struct strategy 2017-01-31 10:36:27 +01:00
Oliver Schneider
5e34740ab6 implement packed struct field access 2017-01-30 09:44:52 +01:00
Scott Olson
53fa985fc4 Update for changes in rustc. 2017-01-16 19:37:53 -08:00
Scott Olson
ac2bf50f9d Use the new field init shorthand. 2017-01-16 18:45:30 -08:00
Oliver Schneider
3a658e09e8
rustup (i128) 2017-01-12 08:28:42 +01:00
Scott Olson
b233ada529 Change Option<Value> to Value, using ByVal(Undef).
This job isn't quite finished because it caused me to discover bugs
related to reading `ByVal(Undef)` when a `ByValPair` is expected, e.g.
for a fat pointer. This wasn't a problem with the `None` of
`Option<Value>`, but I realized an equivalent bug existed even then,
since you could transmute a `u64` like `ByVal(Bytes(42))` to a fat
pointer type on 32-bit targets.

Likewise, you could transmute a fat pointer to `u64` and get panics
related to expecting `ByVal` but finding `ByValPair`, so the problem
goes both ways.
2016-12-18 20:59:01 -08:00
Scott Olson
4fe41ad8d5 Refactor PrimVal::bits() out of existence. 2016-12-17 03:09:57 -08:00
Scott Olson
29e690fe14 Handle writing undefined PrimVals and mask properly. 2016-12-17 01:36:02 -08:00
Scott Olson
b36a83171b Turn PrimVal into an enum including Undefined.
This is step 1 of a refactoring to fix #95. The `Undefined` variant is
so far unused and the old `bits` and `relocation` fields are emulated
with two new temporary methods. There should be no functional change due
to this commit.
2016-12-16 22:03:34 -08:00
Scott Olson
6ec3d65068 Revert "Revert "rustup to rustc 1.15.0-dev (ace092f56 2016-12-13)"" 2016-12-16 17:10:16 -08:00
Scott Olson
0deabf9c00 Revert "rustup to rustc 1.15.0-dev (ace092f56 2016-12-13)" 2016-12-15 01:16:06 -08:00
Oliver Schneider
fd0c21eeee
check that the null terminator is defined and not part of a pointer 2016-12-15 09:58:41 +01:00
Oliver Schneider
69fa3ebff6
rustup to rustc 1.15.0-dev (ace092f56 2016-12-13) (always_encode_mir) 2016-12-14 17:06:23 +01:00
Scott Olson
1af63171f8 Split primval into operator and value. 2016-12-10 16:23:07 -08:00
Scott Olson
5ce6514f23 Dump allocations within PrimVal pointers. 2016-12-07 22:00:46 -08:00
Scott Olson
49f784a3e4 Remove PrimValKind field from PrimVal. 2016-11-26 23:21:20 -08:00
Oliver Schneider
0039ebc940
replace most uses of usize with u64 so the host architecture isn't exposed anymore 2016-11-18 12:55:14 +01:00
Oliver Schneider
986b3a07c2
layout computation can fail, make it fail with a miri error 2016-11-18 10:38:07 +01:00
Oliver Schneider
fd6a90860c
simplify dumping of pointers to the zst or never alloc 2016-11-18 10:36:01 +01:00
Oliver Schneider
11a0594a1d
address comments 2016-11-18 10:35:41 +01:00
Oliver Schneider
51ff9fdaf6
deallocate all locals on function exit and transitively freeze constants through pointers 2016-11-18 10:05:55 +01:00
Oliver Schneider
a5aafbdfbf
rustup 2016-11-17 11:31:28 +01:00
Oliver Schneider
fd68670c0a
merge closures and function and implement some closure vtable cases 2016-11-15 17:19:37 +01:00
Oliver Schneider
f77a0ab10b
fix writing int->ptr transmuted primvals to memory 2016-11-15 15:18:49 +01:00
Oliver Schneider
14ff6411f0
make sure ByVal pointers act just like ByRef to a pointer 2016-11-15 14:12:49 +01:00
Oliver 'ker' Schneider
1c40fb0da1 report the bad integer size instead of just the fact that it is bad 2016-11-11 13:08:14 +01:00
Oliver 'ker' Schneider
921f5af1fe ensure that integers cast to pointers will never point at a valid alloc, not even the zst alloc 2016-11-10 19:20:11 +01:00
Scott Olson
e7bcf35f8a Simplify PrimValKind and remove a horrible hack.
This takes the `AllocId` out of PrimValKind, replacing it with a
`relocation` field on `PrimVal`, which is closer to an earlier design
for `PrimVal` I discussed with @eddyb.

This commit prepares the code for removing the `PrimValKind` from
`PrimVal` and making them more pure bitbags. The only code dealing with
`PrimValKind` will be code making decisions like "what kind of operation
do I need to do on these bits", like operators and casting. Transmutes
of `PrimVal`s will become true no-ops, not even adjusting a `kind`
field.

This commit also removes my horrible `value_to_primval` hack that made
an allocation for every `ByVal` passed in, so it could use `read_value`
to get a `PrimVal` with the right kind. Now I just compute the
`PrimValKind` from the `Ty` and re-tag the `PrimVal`.

The code got slightly messier in some areas here, but I think a _lot_ of
code will simplify in obvious ways once I remove the `kind` field from
`PrimVal`.

Gosh, if my commit messages aren't turning into essays these days.
2016-10-21 03:17:53 -06:00
Scott Olson
330be7766f Represent PrimVals as "bitbags".
Now instead of holding a native type based on the tag, all PrimVals
store a u64 (the `bits`), along with a `kind` corresponding to the
variant as it would be in the old PrimVal representation.

This commit makes no major optimizations and attempts to not change any
behaviour. There will be commits to follow that make use of this
representation to eliminate unnecessary allocation hacks like in
`value_to_primval`.

A number of places could be even more cleaned up after this commit,
particularly in `cast.rs`.
2016-10-20 04:42:19 -06:00
Scott Olson
197e89bbb0 Refactor alloc_ptr. 2016-10-16 00:12:11 -06:00
Scott Olson
49e6c57ef9 Do not pre-allocate local variables.
Thanks to the `Value` locals refactoring, now primitive locals (ints,
floats, chars, bools, and the like) will not require `Allocation`s at
all, and locals that are never initialized at all because of conditional
control flow won't be wasting memory.
2016-10-15 23:31:42 -06:00
Oliver Schneider
de38015e47 rustup 2016-10-01 15:33:07 +02:00
Oliver Schneider
5d1080d0ca
refactor Lvalue and PrimVal::{SlicePtr, VtablePtr} into Value::ByValPair 2016-09-26 17:49:30 +02:00
Oliver Schneider
fcbaf990f2
check Pointer::to_int for non-integer pointers 2016-09-26 11:37:23 +02:00
Oliver Schneider
2282289ad5
refactor away intermediate allocations, stage1 2016-09-23 10:27:14 +02:00
Oliver Schneider
38748fa615
refactor away IntegerPtr 2016-09-22 15:47:16 +02:00
Oliver Schneider
875a4542f9
remove the ZST allocation and abort all zero byte writes/reads 2016-09-22 15:22:00 +02:00
Oliver Schneider
0690a26ddf
make Memory::dump use trace! instead of println! 2016-09-22 13:01:08 +02:00
Oliver Schneider
0f578f0d2e
fully implement size_of_val and add various tests that now succeed 2016-09-22 13:00:43 +02:00
Scott Olson
5b012edc7a Rename AbstractPtr to Ptr. 2016-09-21 23:23:50 -06:00
Scott Olson
678b9ca328 Print "(immutable)" when dumping allocations. 2016-09-19 04:10:51 -06:00
Scott Olson
c679c71def Freeze static memory of string constants. 2016-09-19 04:10:18 -06:00
Scott Olson
f731766805 Fix allocation of fn items by allowing ZST alignment to be 0. 2016-09-13 21:31:12 -06:00
Oliver Schneider
c57233abca
needless clones 2016-09-13 13:03:54 +02:00
Oliver Schneider
5c47e3dbd8
only allow the modification of static mut or statics with interior mutability 2016-09-11 11:02:45 +02:00
Oliver Schneider
0e58c2a31b
document the fields of Allocation 2016-09-11 11:02:45 +02:00
Oliver Schneider
78bef956c9
don't load memory as mutable if not necessary 2016-09-11 11:02:45 +02:00
Oliver Schneider
00c551c5f0
implement calling methods through trait objects 2016-09-09 12:51:14 +02:00
Oliver Schneider
0f177fdecf
implement more casts 2016-09-07 18:34:59 +02:00
Oliver Schneider
cd91f9feee
replace all unreachable! and panic! calls with bug! 2016-09-06 16:16:49 +02:00
Oliver Schneider
f8cfc387fd
address nits 2016-07-25 12:30:35 +02:00
Oliver Schneider
ec897f9156
don't allow runtime-aligning of memory 2016-07-22 16:35:39 +02:00
Oliver Schneider
613d15c672
clippy 2016-07-07 13:19:17 +02:00
Oliver Schneider
4c258d1ed2
Merge remote-tracking branch 'origin/master' into alignment 2016-07-07 13:17:08 +02:00
Oliver Schneider
9669acc3a3
Merge remote-tracking branch 'origin/master' into the_outer_limits 2016-07-07 11:27:49 +02:00
Oliver Schneider
5381981446
shrink_to_fit some vectors to prevent interpreted code from passing the memory limits 2016-07-07 11:21:18 +02:00
Oliver Schneider
8d3817cfc6
use usize instead of u64 for memory limits 2016-07-07 11:20:46 +02:00
Oliver Schneider
51ce4a2584
use byteorder's write_f{32,64} instead of transmuting 2016-07-06 11:51:32 +02:00
Oliver Schneider
1bd8e04228
check alignment in various places 2016-07-06 11:12:44 +02:00
Oliver Schneider
50987e3697
some methods to check pointers for correct alignment 2016-07-06 10:58:51 +02:00
Oliver Schneider
082effb3ee
align allocations in the worst possible way 2016-07-05 14:27:27 +02:00
Oliver Schneider
756fbcce48
add a memory limit 2016-07-05 10:47:10 +02:00
Oliver Schneider
4b831569f6
implement floats by running the ops on the host architecture 2016-07-05 09:08:24 +02:00
Scott Olson
a7cc77a010 Compare against ZST_ALLOC_ID in points_to_zst. 2016-07-01 16:40:52 -06:00
Oliver Schneider
3d9588332f
address comments 2016-07-01 13:09:40 +02:00
Oliver Schneider
594f1d79da
optimize all ZST allocations into one single allocation 2016-07-01 13:09:17 +02:00
Scott Olson
756d73b3ca Remove filling drop to prep for elaborated drops. 2016-06-30 21:30:03 -06:00
Oliver Schneider
b91338b220
things priroda needs to be public or changed 2016-06-30 11:29:25 +02:00
Oliver 'ker' Schneider
37287d2a5d use free methods instead of traits 2016-06-25 16:50:33 +02:00
Oliver Schneider
0288486b73
use target byte order 2016-06-23 15:16:25 +02:00
Oliver Schneider
86040c0d29
simplify write_ptr 2016-06-23 13:04:05 +02:00
Oliver Schneider
4c7aae73bc
adjust all pointer_size checks to use the method 2016-06-23 10:02:39 +02:00
Oliver Schneider
d13153c424
add a pointer_size method to Memory for easy access 2016-06-23 09:59:16 +02:00
Oliver Schneider
205a988c1b
improve rustdoc rendering 2016-06-23 09:40:01 +02:00
Oliver Schneider
055b6a8d38
store full TargetDataLayout in Memory instead of just pointer size 2016-06-23 09:36:37 +02:00
Oliver Schneider
7a9272c8e1
no need to cast chars as u32 before casting to u64 2016-06-21 09:43:45 +02:00
Oliver Schneider
2dbd30fa51
implement char handling 2016-06-21 09:35:20 +02:00
Oliver Schneider
68469be89b
rename function cache member 2016-06-20 16:52:53 +02:00