https://gcc.gnu.org/bugzilla/show_bug.cgi?id=19495
--- Comment #28 from Jonathan Wakely ---
(In reply to Jonathan Wakely from comment #27)
> But the alignment problem still seems to be present in the COW std::string.
But I think that's covered by PR 8670, which is still open.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=19495
--- Comment #27 from Jonathan Wakely ---
(In reply to Ben Elliston from comment #26)
> This test now passes on powerpc*-linux-gnu.
I wonder how ... the "fix" got reverted, and we still use an allocator of char
to create the storage for the COW
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=19495
Bug 19495 depends on bug 24882, which changed state.
Bug 24882 Summary: [meta-bug] Non-refcounted, moveable basic_string
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=24882
What|Removed |Added
--- Comment #26 from bje at gcc dot gnu dot org 2009-05-14 02:54 ---
This test now passes on powerpc*-linux-gnu.
--
bje at gcc dot gnu dot org changed:
What|Removed |Added
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-05-28
21:57 ---
Subject: Bug 19495
CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED] 2005-05-28 21:57:03
Modified files:
libstdc++-v3 : ChangeLog
--- Additional Comments From cvs-commit at gcc dot gnu dot org 2005-05-18
22:11 ---
Subject: Bug 19495
CVSROOT:/cvs/gcc
Module name:gcc
Changes by: [EMAIL PROTECTED] 2005-05-18 22:11:24
Modified files:
libstdc++-v3 : ChangeLog
--
What|Removed |Added
Severity|normal |enhancement
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19495
--- Additional Comments From ncm at cantrip dot org 2005-04-01 13:24
---
Subject: Re: basic_string::_M_rep() can produce an unnaturally aligned pointer
to _Rep
On Fri, Apr 01, 2005 at 11:42:27AM -, pcarlini at suse dot de wrote:
What|Removed
--- Additional Comments From pcarlini at suse dot de 2005-04-01 13:31
---
Ok, my change was only dictated by consistency, and the original idea of using
enhancement is not mine ;) Let's remove enhancement from both. By the way,
I really noticed yesterday for the first time that
--- Additional Comments From pcarlini at suse dot de 2005-02-09 09:33
---
Will be trivially fixed in v7...
--
What|Removed |Added
Status|ASSIGNED
--- Additional Comments From pinskia at gcc dot gnu dot org 2005-02-09
19:38 ---
*** Bug 19867 has been marked as a duplicate of this bug. ***
--
What|Removed |Added
--- Additional Comments From janis at gcc dot gnu dot org 2005-02-10 00:27
---
I've made execution of ext/array_allocator/2.cc XFAIL for powerpc*-*-linux*.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19495
--
What|Removed |Added
Severity|normal |enhancement
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19495
--- Additional Comments From ncm-nospam at cantrip dot org 2005-01-21
15:22 ---
This is a real bug, but easily fixed, and (I think) without breaking ABI.
The problem is in basic_string.h, where it says
struct _Rep : _Rep_base
{
// Types:
typedef typename _Alloc::template
--- Additional Comments From ncm-nospam at cantrip dot org 2005-01-21
15:37 ---
Hmm, it's a little more complicated than I said, although it might be
academic. There's an implicit assumption in the code that any type
on which basic_string might be instantiated has no stricter
--- Additional Comments From pcarlini at suse dot de 2005-01-21 15:41
---
Thanks Nathan, I will implement what you are suggesting. The last issue,
actually
is filed as libstdc++/8670 and in the audit trail we agreed to fix it using a
suited __attribute__(aligned), which however,
--- Additional Comments From pcarlini at suse dot de 2005-01-21 15:43
---
By the way, I understand that tr1/aligned_storage co facilities cannot be
directly used for implementing the current std, since we cannot pollute the
std namespace.
--
--
What|Removed |Added
Severity|enhancement |normal
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=19495
--- Additional Comments From ncm-nospam at cantrip dot org 2005-01-21
16:39 ---
I agree that 8670 is a separate bug.
The referenced test 2.cc can be made to fail more reliably with
the following changes:
First, leave enough space for alignment adjustments, even on 128-bit
machines:
-
--- Additional Comments From pcarlini at suse dot de 2005-01-19 09:09
---
Does this behaviour seem a little bit unusual to you? You said: You
cannot create a basic_stringchar object in memory aligned one.
That is rather counter-intuitive to a user of libstdc++ who has no
--- Additional Comments From pcarlini at suse dot de 2005-01-19 10:05
---
Well, I can see that basic_string default allocator, std::allocatorCharT
according to the standard shall return memory only aligned as CharT requests,
not more; whereas our implementation of std::allocator
--- Additional Comments From pcarlini at suse dot de 2005-01-19 11:51
---
For 4.0, besides the documentation issue, I think we should change
ext/array_allocator to use tr1::type_traits::aligned_storage, finally
available. I will post a patch ASAP...
--
What|Removed
--- Additional Comments From gdr at integrable-solutions dot net
2005-01-19 15:57 ---
Subject: Re: basic_string::_M_rep() can produce an unnaturally aligned pointer
to _Rep
pcarlini at suse dot de [EMAIL PROTECTED] writes:
| Does this behaviour seem a little bit unusual to you?
--- Additional Comments From gdr at integrable-solutions dot net
2005-01-19 16:02 ---
Subject: Re: basic_string::_M_rep() can produce an unnaturally aligned pointer
to _Rep
pcarlini at suse dot de [EMAIL PROTECTED] writes:
| Well, I can see that basic_string default allocator,
--- Additional Comments From pcarlini at suse dot de 2005-01-20 00:34
---
Hi Gaby,
Yes. Basically, we need to have tha aligned attribute work correctly.
Agreed, in principle: indeed, we are filing together and taking care of
many PRs in this area.
| is not present in the original
--- Additional Comments From gdr at integrable-solutions dot net
2005-01-20 04:08 ---
Subject: Re: basic_string::_M_rep() can produce an unnaturally aligned pointer
to _Rep
pcarlini at suse dot de [EMAIL PROTECTED] writes:
| This is way I'm proposing, *for 4.0, only for 4.0*, to
--- Additional Comments From pcarlini at suse dot de 2005-01-18 09:17
---
This is essentially a duplicate of libstdc++/8670: we'll try to look again into
it: perhaps we can fix it now, but __attribute__(align(K)) is still very weak,
see c++/19163 and c++/17743.
*** This bug has been
--- Additional Comments From pcarlini at suse dot de 2005-01-18 09:35
---
Actually, sorry, but I'm not sure this is really the same of libstdc++/8670
(which is *not* about basic_string instantiated for plain char). The problem
here seems that the alignment requirements of basic_string
--- Additional Comments From pcarlini at suse dot de 2005-01-18 09:45
---
In other terms, as far as this PR is concerned, basic_string seems ok, just
you cannot create a basic_stringchar object in memory aligned one, as happens
in ext/array_allocator/2.cc, which therefore should be
--- Additional Comments From pcarlini at suse dot de 2005-01-18 09:53
---
Benjamin, can you please have a look?
--
What|Removed |Added
CC|
--- Additional Comments From bje at au1 dot ibm dot com 2005-01-19 06:05
---
Subject: Re: basic_string::_M_rep() can produce an unnaturally aligned pointer
to _Rep
On Tue, Jan 18, 2005 at 09:45:48AM -, pcarlini at suse dot de wrote:
In other terms, as far as this PR is
--- Additional Comments From amodra at bigpond dot net dot au 2005-01-19
06:28 ---
The reason for the bus error is the __exchange_and_add decrement of _M_refcount.
On powerpc, lwarx and stwcx. must have an aligned effective address.
--
What|Removed
32 matches
Mail list logo
