alloc_system: don’t assume MIN_ALIGN for small sizes, fix #45955
The GNU C library (glibc) is documented to always allocate with an alignment
of at least 8 or 16 bytes, on 32-bit or 64-bit platforms:
https://www.gnu.org/software/libc/manual/html_node/Aligned-Memory-Blocks.html
This matches our use of `MIN_ALIGN` before this commit.
However, even when libc is glibc, the program might be linked
with another allocator that redefines the `malloc` symbol and friends.
(The `alloc_jemalloc` crate does, in some cases.)
So `alloc_system` doesn’t know which allocator it calls,
and needs to be conservative in assumptions it makes.
The C standard says:
https://port70.net/%7Ensz/c/c11/n1570.html#7.22.3
> The pointer returned if the allocation succeeds is suitably aligned
> so that it may be assigned to a pointer to any type of object
> with a fundamental alignment requirement
https://port70.net/~nsz/c/c11/n1570.html#6.2.8p2
> A fundamental alignment is represented by an alignment less than
> or equal to the greatest alignment supported by the implementation
> in all contexts, which is equal to `_Alignof (max_align_t)`.
`_Alignof (max_align_t)` depends on the ABI and doesn’t seem to have
a clear definition, but it seems to match our `MIN_ALIGN` in practice.
However, the size of objects is rounded up to the next multiple
of their alignment (since that size is also the stride used in arrays).
Conversely, the alignment of a non-zero-size object is at most its size.
So for example it seems ot be legal for `malloc(8)` to return a pointer
that’s only 8-bytes-aligned, even if `_Alignof (max_align_t)` is 16.
2017-11-20 15:30:04 +01:00
|
|
|
#![feature(allocator_api)]
|
2015-03-11 05:58:16 +01:00
|
|
|
#![feature(box_syntax)]
|
2017-07-15 03:54:17 +02:00
|
|
|
#![feature(drain_filter)]
|
2016-11-22 23:31:31 +01:00
|
|
|
#![feature(exact_size_is_empty)]
|
2019-06-21 03:52:38 +02:00
|
|
|
#![feature(option_flattening)]
|
2015-06-10 22:33:52 +02:00
|
|
|
#![feature(pattern)]
|
2018-12-19 09:38:15 +01:00
|
|
|
#![feature(repeat_generic_slice)]
|
2019-06-21 03:52:38 +02:00
|
|
|
#![feature(trusted_len)]
|
2018-03-08 15:36:43 +01:00
|
|
|
#![feature(try_reserve)]
|
2015-03-11 05:58:16 +01:00
|
|
|
#![feature(unboxed_closures)]
|
2019-08-09 00:33:57 +02:00
|
|
|
#![feature(associated_type_bounds)]
|
2015-07-11 13:34:57 +02:00
|
|
|
|
2016-09-29 02:23:36 +02:00
|
|
|
use std::hash::{Hash, Hasher};
|
|
|
|
use std::collections::hash_map::DefaultHasher;
|
2015-08-12 02:27:05 +02:00
|
|
|
|
2018-07-06 19:30:09 +02:00
|
|
|
mod arc;
|
2015-03-11 05:58:16 +01:00
|
|
|
mod binary_heap;
|
|
|
|
mod btree;
|
2016-11-04 02:07:00 +01:00
|
|
|
mod cow_str;
|
2015-03-11 05:58:16 +01:00
|
|
|
mod fmt;
|
alloc_system: don’t assume MIN_ALIGN for small sizes, fix #45955
The GNU C library (glibc) is documented to always allocate with an alignment
of at least 8 or 16 bytes, on 32-bit or 64-bit platforms:
https://www.gnu.org/software/libc/manual/html_node/Aligned-Memory-Blocks.html
This matches our use of `MIN_ALIGN` before this commit.
However, even when libc is glibc, the program might be linked
with another allocator that redefines the `malloc` symbol and friends.
(The `alloc_jemalloc` crate does, in some cases.)
So `alloc_system` doesn’t know which allocator it calls,
and needs to be conservative in assumptions it makes.
The C standard says:
https://port70.net/%7Ensz/c/c11/n1570.html#7.22.3
> The pointer returned if the allocation succeeds is suitably aligned
> so that it may be assigned to a pointer to any type of object
> with a fundamental alignment requirement
https://port70.net/~nsz/c/c11/n1570.html#6.2.8p2
> A fundamental alignment is represented by an alignment less than
> or equal to the greatest alignment supported by the implementation
> in all contexts, which is equal to `_Alignof (max_align_t)`.
`_Alignof (max_align_t)` depends on the ABI and doesn’t seem to have
a clear definition, but it seems to match our `MIN_ALIGN` in practice.
However, the size of objects is rounded up to the next multiple
of their alignment (since that size is also the stride used in arrays).
Conversely, the alignment of a non-zero-size object is at most its size.
So for example it seems ot be legal for `malloc(8)` to return a pointer
that’s only 8-bytes-aligned, even if `_Alignof (max_align_t)` is 16.
2017-11-20 15:30:04 +01:00
|
|
|
mod heap;
|
2015-03-11 05:58:16 +01:00
|
|
|
mod linked_list;
|
2018-07-06 19:30:09 +02:00
|
|
|
mod rc;
|
2015-03-11 05:58:16 +01:00
|
|
|
mod slice;
|
|
|
|
mod str;
|
|
|
|
mod string;
|
|
|
|
mod vec_deque;
|
|
|
|
mod vec;
|
2015-08-12 02:27:05 +02:00
|
|
|
|
|
|
|
fn hash<T: Hash>(t: &T) -> u64 {
|
2016-09-29 02:23:36 +02:00
|
|
|
let mut s = DefaultHasher::new();
|
2015-08-12 02:27:05 +02:00
|
|
|
t.hash(&mut s);
|
|
|
|
s.finish()
|
|
|
|
}
|
2017-08-21 16:15:02 +02:00
|
|
|
|
2017-09-10 18:13:19 +02:00
|
|
|
// FIXME: Instantiated functions with i128 in the signature is not supported in Emscripten.
|
|
|
|
// See https://github.com/kripken/emscripten-fastcomp/issues/169
|
|
|
|
#[cfg(not(target_os = "emscripten"))]
|
2017-08-21 16:15:02 +02:00
|
|
|
#[test]
|
|
|
|
fn test_boxed_hasher() {
|
|
|
|
let ordinary_hash = hash(&5u32);
|
|
|
|
|
|
|
|
let mut hasher_1 = Box::new(DefaultHasher::new());
|
|
|
|
5u32.hash(&mut hasher_1);
|
|
|
|
assert_eq!(ordinary_hash, hasher_1.finish());
|
|
|
|
|
2018-07-13 07:25:22 +02:00
|
|
|
let mut hasher_2 = Box::new(DefaultHasher::new()) as Box<dyn Hasher>;
|
2017-08-21 16:15:02 +02:00
|
|
|
5u32.hash(&mut hasher_2);
|
|
|
|
assert_eq!(ordinary_hash, hasher_2.finish());
|
|
|
|
}
|