comp.lang.ada
 help / color / mirror / Atom feed
From: Victor Porton <porton@narod.ru>
Subject: Re: Allocators design flaw
Date: Sat, 14 Oct 2017 18:18:46 +0300
Date: 2017-10-14T18:18:46+03:00	[thread overview]
Message-ID: <ort9sv$1oaf$1@gioia.aioe.org> (raw)
In-Reply-To: ort6qu$1iul$2@gioia.aioe.org

Dmitry A. Kazakov wrote:

> On 2017-10-14 16:03, Victor Porton wrote:
>> Dmitry A. Kazakov wrote:
>> 
>>> On 2017-10-14 04:53, Victor Porton wrote:
>>>> It is impossible to properly implement an allocator through a C
>>>> function (such as raptor_alloc_memory() from Raptor C library) which
>>>> allocates a struct and returns the pointer to the allocated struct.
>>>>
>>>> It is because RM13.11(21.5/3) "The Alignment parameter is a nonzero
>>>> integral multiple of D'Alignment..."
>>>>
>>>> (If it were "The Alignment parameter is equal to D'Alignment", then we
>>>> would be able just to check (in Allocate procedure implementation) that
>>>>
>>>> pragma Assert(Dummy_Record'Alignment mod Alignment = 0);
>>>> -- where Dummy_Record is an arbitrary C-convention record
>>>> -- (as all C records have the same alignment reqs)
>>>>
>>>> So Alignment parameter may be arbitrarily big and the C function
>>>> alignment may not conform to it.
>>>
>>> Usually allocators return addresses already rounded and there is nothing
>>> to worry about.
>>>
>>>> Let us think how to work around (in Ada 202x) of this design flaw.
>>>
>>> If any it is _alloc_memory() flaw, not Ada's.
>>>
>>> Add max alignment + log max alignment - 1 to the desired size. Add log
>>> max alignment to the returned address and round to the required
>>> alignment. Place the offset to original address in front (log alignment
>>> length). Return the rounded address. When freed use the stored offset to
>>> get the original address.
>> 
>> As far as I understand, it will not work, because the C library I am
>> writing bindings for may try to free an object allocated by me (or I my
>> need to free an object allocated by the library).
> 
> The last sentence describes freeing memory, i.e. for Deallocate.

I again state that we need the C function *_free() to free memory. But it is 
impossible in your scenario, because you change the pointer to point to 
another byte of memory.

-- 
Victor Porton - http://portonvictor.org


  reply	other threads:[~2017-10-14 15:18 UTC|newest]

Thread overview: 36+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-10-14  2:53 Allocators design flaw Victor Porton
2017-10-14  7:27 ` Dmitry A. Kazakov
2017-10-14 13:52   ` Victor Porton
2017-10-14 14:25     ` Dmitry A. Kazakov
2017-10-14 14:03   ` Victor Porton
2017-10-14 14:26     ` Dmitry A. Kazakov
2017-10-14 15:18       ` Victor Porton [this message]
2017-10-14 15:44         ` Dmitry A. Kazakov
2017-10-14 16:42           ` Victor Porton
2017-10-14 16:13     ` Simon Wright
2017-10-14 16:38       ` Victor Porton
2017-10-14 14:12   ` Victor Porton
2017-10-14 14:20     ` Victor Porton
2017-10-14 14:24       ` Victor Porton
2017-10-14 14:36         ` Dmitry A. Kazakov
2017-10-14 15:17           ` Victor Porton
2017-10-14 15:51             ` Dmitry A. Kazakov
2017-10-14 16:34               ` Victor Porton
2017-10-14 17:14                 ` Dmitry A. Kazakov
2017-10-14 17:24                   ` Victor Porton
2017-10-14 18:08                     ` Dmitry A. Kazakov
2017-10-14 14:28     ` Dmitry A. Kazakov
2017-10-14 15:14       ` Victor Porton
2017-10-14 15:42         ` Simon Wright
2017-10-14 16:29           ` Victor Porton
2017-10-14 20:07             ` Simon Wright
2017-10-14 21:26               ` Victor Porton
2017-10-21  1:42     ` Randy Brukardt
2017-10-14  8:02 ` Simon Wright
2017-10-14 13:59   ` Victor Porton
2017-10-14 14:35     ` Simon Wright
2017-10-14 15:11       ` Victor Porton
2017-10-14 15:56         ` Simon Wright
2017-10-14 16:22           ` Victor Porton
2017-10-29 16:01           ` David Thompson
2017-10-14 14:11 ` Victor Porton
replies disabled

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox