Re: c++/2166

Re: c++/2166

[Mark Mitchell]

The following reply was made to PR c++/2166; it has been noted by GNATS.

From: Mark Mitchell <mark@codesourcery.com>
To: Gabriel.Dos-Reis@cmla.ens-cachan.fr
Cc: nobody@gcc.gnu.org, gcc-gnats@gcc.gnu.org, carlo@alinoe.com
Subject: Re: c++/2166
Date: Wed, 18 Apr 2001 00:05:48 -0700

 >>>>> "Gabriel" == Gabriel Dos Reis <Gabriel.Dos-Reis@cmla.ens-cachan.fr> writes:
 
     Gabriel> The Standard requires that a member of an anonymous
     Gabriel> namespace be given a linker-visible unique name in a
     Gabriel> given translation unit so that two different translation
     Gabriel> units don't mess up.
 
 So, do you agree with me that things are OK, or with Carlo that there
 is a problem?
 
 Thanks,
 
 --
 Mark Mitchell                   mark@codesourcery.com
 CodeSourcery, LLC               http://www.codesourcery.com

Re: c++/2166

[Carlo Wood]

The following reply was made to PR c++/2166; it has been noted by GNATS.

From: Carlo Wood <carlo@alinoe.com>
To: Mark Mitchell <mark@codesourcery.com>
Cc: mmitchel@gcc.gnu.org, gcc-gnats@gcc.gnu.org, nobody@gcc.gnu.org
Subject: Re: c++/2166
Date: Wed, 18 Apr 2001 16:56:53 +0200

 On Tue, Apr 17, 2001 at 08:54:54PM -0700, Mark Mitchell wrote:
 > I don't think so.
 > 
 > I think that `test' has external linkage.  Therefore, using it in this
 > mangled name is safe.  You could only get a conflict if you had two
 > tests in two different translation units, both with external linkage.
 > And that would create a link error anyhow.
 
 Ah... Ok.  I am sorry, I really thought you meant that it was
 the variable 'test' that had to be unique itself.  It just appeared
 to me as a bug since the compiler constantly used long names
 including the input file name and random characters and then
 suddenly I ran into one that didn't contain any random characters
 at all etc.  But I guess you're right.
 
 Thanks for clarifying,
 
 -- 
 Carlo Wood <carlo@alinoe.com>

Re: c++/2166

[Gabriel Dos Reis]

The following reply was made to PR c++/2166; it has been noted by GNATS.

From: Gabriel Dos Reis <Gabriel.Dos-Reis@cmla.ens-cachan.fr>
To: Mark Mitchell <mark@codesourcery.com>
Cc: Gabriel.Dos-Reis@cmla.ens-cachan.fr, nobody@gcc.gnu.org,
        gcc-gnats@gcc.gnu.org, carlo@alinoe.com
Subject: Re: c++/2166
Date: 18 Apr 2001 10:07:09 +0200

 Mark Mitchell <mark@codesourcery.com> writes:
 
 | >>>>> "Gabriel" == Gabriel Dos Reis <Gabriel.Dos-Reis@cmla.ens-cachan.fr> writes:
 | 
 |     Gabriel> I ran several tests and I think the current scheme is not
 |     Gabriel> robust enought -- it seems to use very few characters ( <
 |     Gabriel> 7) to fake uniqueness.
 | 
 | THat should be enough.  Each character is chosen at random from 62
 | choices, or about 2^6.  So, that's (2^6)^7 = 2^42 possible choices,
 | plus the input file name.
 
 OK, so I think that should clarify things for Carlo.
 
 -- Gaby

Re: c++/2166

[Mark Mitchell]

The following reply was made to PR c++/2166; it has been noted by GNATS.

From: Mark Mitchell <mark@codesourcery.com>
To: Gabriel.Dos-Reis@cmla.ens-cachan.fr
Cc: nobody@gcc.gnu.org, gcc-gnats@gcc.gnu.org, carlo@alinoe.com
Subject: Re: c++/2166
Date: Wed, 18 Apr 2001 00:59:00 -0700

 >>>>> "Gabriel" == Gabriel Dos Reis <Gabriel.Dos-Reis@cmla.ens-cachan.fr> writes:
 
     Gabriel> I ran several tests and I think the current scheme is not
     Gabriel> robust enought -- it seems to use very few characters ( <
     Gabriel> 7) to fake uniqueness.
 
 THat should be enough.  Each character is chosen at random from 62
 choices, or about 2^6.  So, that's (2^6)^7 = 2^42 possible choices,
 plus the input file name.
 
 --
 Mark Mitchell                   mark@codesourcery.com
 CodeSourcery, LLC               http://www.codesourcery.com

Re: c++/2166

[Gabriel Dos Reis]

The following reply was made to PR c++/2166; it has been noted by GNATS.

From: Gabriel Dos Reis <Gabriel.Dos-Reis@cmla.ens-cachan.fr>
To: Mark Mitchell <mark@codesourcery.com>
Cc: Gabriel.Dos-Reis@cmla.ens-cachan.fr, nobody@gcc.gnu.org,
        gcc-gnats@gcc.gnu.org, carlo@alinoe.com
Subject: Re: c++/2166
Date: 18 Apr 2001 09:40:39 +0200

 Mark Mitchell <mark@codesourcery.com> writes:
 
 | >>>>> "Gabriel" == Gabriel Dos Reis <Gabriel.Dos-Reis@cmla.ens-cachan.fr> writes:
 | 
 |     Gabriel> The Standard requires that a member of an anonymous
 |     Gabriel> namespace be given a linker-visible unique name in a
 |     Gabriel> given translation unit so that two different translation
 |     Gabriel> units don't mess up.
 | 
 | So, do you agree with me that things are OK, or with Carlo that there
 | is a problem?
 
 Well, what I meant is that g++ has to ensure that the construct
 
 	namespace {
 		int test;
 	};
 
 can appear in different translation units (within the same program)
 and each translation unit have its own idea of `test' -- the different
 `test's should not cause any conflict at the linker level.  
 
 Carlo think that means the compiler should use the file-definition name
 as part of the mangled name of `test'.
 That is a way to achieve uniqueness. Another can be the time of
 tranlation.  Actually anything which can lead to uniqueness is good.
 
 I ran several tests and I think the current scheme is not robust
 enought -- it seems to use very few characters ( < 7) to fake
 uniqueness.
 
 I didn't run into *actual* problems (I ran only 5 tests) so I cannot
 definitely say that there IS a problem.  However,  I think there is a
 potential problem. 
 
 This issue deserve further investigation -- I would do that as soon as
 I have a chance.
 
 -- Gaby

Re: c++/2166

[Gabriel Dos Reis]

The following reply was made to PR c++/2166; it has been noted by GNATS.

From: Gabriel Dos Reis <Gabriel.Dos-Reis@cmla.ens-cachan.fr>
To: Mark Mitchell <mark@codesourcery.com>
Cc: nobody@gcc.gnu.org, gcc-gnats@gcc.gnu.org, carlo@alinoe.com
Subject: Re: c++/2166
Date: 18 Apr 2001 08:53:42 +0200

 Mark Mitchell <mark@codesourcery.com> writes:
 
 | The following reply was made to PR c++/2166; it has been noted by GNATS.
 | 
 | From: Mark Mitchell <mark@codesourcery.com>
 | To: carlo@alinoe.com
 | Cc: mmitchel@gcc.gnu.org, gcc-gnats@gcc.gnu.org, nobody@gcc.gnu.org
 | Subject: Re: c++/2166
 | Date: Tue, 17 Apr 2001 20:54:54 -0700
 | 
 |  >>>>> "Carlo" == Carlo Wood <carlo@alinoe.com> writes:
 |  
 |      Carlo> You misread my report.
 |  
 |  I don't think so.
 |  
 |  I think that `test' has external linkage. 
 
 Yes, it has.
 
 | ... Therefore, using it in this
 |  mangled name is safe.  You could only get a conflict if you had two
 |  tests in two different translation units, both with external linkage.
 |  And that would create a link error anyhow.
 
 The Standard requires that a member of an anonymous namespace be
 given a linker-visible unique name in a given translation unit so that
 two different translation units don't mess up.
 
 -- Gaby

Re: c++/2166

[Mark Mitchell]

The following reply was made to PR c++/2166; it has been noted by GNATS.

From: Mark Mitchell <mark@codesourcery.com>
To: carlo@alinoe.com
Cc: mmitchel@gcc.gnu.org, gcc-gnats@gcc.gnu.org, nobody@gcc.gnu.org
Subject: Re: c++/2166
Date: Tue, 17 Apr 2001 20:54:54 -0700

 >>>>> "Carlo" == Carlo Wood <carlo@alinoe.com> writes:
 
     Carlo> You misread my report.
 
 I don't think so.
 
 I think that `test' has external linkage.  Therefore, using it in this
 mangled name is safe.  You could only get a conflict if you had two
 tests in two different translation units, both with external linkage.
 And that would create a link error anyhow.
 
 --
 Mark Mitchell                   mark@codesourcery.com
 CodeSourcery, LLC               http://www.codesourcery.com

Re: c++/2166

[Carlo Wood]

The following reply was made to PR c++/2166; it has been noted by GNATS.

From: Carlo Wood <carlo@alinoe.com>
To: mmitchel@gcc.gnu.org
Cc: gcc-gnats@gcc.gnu.org, nobody@gcc.gnu.org
Subject: Re: c++/2166
Date: Wed, 18 Apr 2001 02:55:59 +0200

 Please look again at my bug report.
 'test' is not the variable that gives the problem,
 it only magically turns up in the anonymous name of
 the variable that this IS about: the one declared at
 the bottom of the source file:
 
   namespace {
     int anonymous;
   }
 
 You misread my report.
 
 Carlo
 
 On Mon, Apr 16, 2001 at 02:07:54AM -0000, mmitchel@gcc.gnu.org wrote:
 > Synopsis: More mangling bugs
 > 
 > State-Changed-From-To: open->closed
 > State-Changed-By: mmitchel
 > State-Changed-When: Sun Apr 15 19:07:54 2001
 > State-Changed-Why:
 >     Not a bug: `test' is an external symbol so there cannot
 >     be two object modules defining it.
 > 
 > http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=2166&database=gcc
 
 -- 
 Carlo Wood <carlo@alinoe.com>

Re: c++/2166

[mmitchel]

The following reply was made to PR c++/2166; it has been noted by GNATS.

From: mmitchel@gcc.gnu.org
To: carlo@alinoe.com, gcc-gnats@gcc.gnu.org, nobody@gcc.gnu.org
Cc:  
Subject: Re: c++/2166
Date: 16 Apr 2001 02:07:54 -0000

 Synopsis: More mangling bugs
 
 State-Changed-From-To: open->closed
 State-Changed-By: mmitchel
 State-Changed-When: Sun Apr 15 19:07:54 2001
 State-Changed-Why:
     Not a bug: `test' is an external symbol so there cannot
     be two object modules defining it.
 
 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=2166&database=gcc