Re: libstdc++/6671: ext search path

Re: libstdc++/6671: ext search path

[paolo]

Synopsis: ext search path

State-Changed-From-To: open->analyzed
State-Changed-By: paolo
State-Changed-When: Sun May 19 08:48:29 2002
State-Changed-Why:
    Hi.
    I see two issues here:
    1- If all is needed is a shorter access to the old STL extensions (those exisiting in ext when the doc you mention was written, i.e., hash_map, hash_set, rope, slist) then the solution is simply using -idirafter instead of -I (see libstdc++/6720). This allows to write safely <hash_map> (and still <iostream>, of course, as in your testcase).
    2- OTOH, in this case <ext/memory>, f.i., must still be written in full in order to disambiguate it from <memory>.
    In order to solve the latter problem, the only solution I see is renaming the new extension bits to something different.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=6671

Re: libstdc++/6671: ext search path

[paolo]

Synopsis: ext search path

Responsible-Changed-From-To: unassigned->pme
Responsible-Changed-By: paolo
Responsible-Changed-When: Mon May 20 14:19:29 2002
Responsible-Changed-Why:
    .
State-Changed-From-To: analyzed->closed
State-Changed-By: paolo
State-Changed-When: Mon May 20 14:19:29 2002
State-Changed-Why:
    Fixed with:
    http://gcc.gnu.org/ml/gcc-patches/2002-05/msg01633.html

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=6671

libstdc++/6671: ext search path

[morandini]

>Number:         6671
>Category:       libstdc++
>Synopsis:       ext search path
>Confidential:   no
>Severity:       non-critical
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          doc-bug
>Submitter-Id:   net
>Arrival-Date:   Thu May 16 01:06:00 PDT 2002
>Closed-Date:
>Last-Modified:
>Originator:     Marco Morandini
>Release:        gcc version 3.1
>Organization:
>Environment:
Suse linux 7.3
>Description:
libstdc++, 5.4 Extensions and Backward Compatibility
Although you can specify -I options to make the preprocessor search the g++-v3/ext.....

This is wrong, because <ext/memory> will get loaded instead of <memory>, and <ext/memory> does not include  ../memory
Also, note that other files (ext/algorithm, ext/functional, ext/iterator, ext/numeric) suffer for a similar problem: for example, <ext/algorithm> include <algorithm>, and this will never be found with -I.../ext
>How-To-Repeat:
g++ -c -Igcc_include_path/g++-v3/ext pippo.C
>Fix:
a) fix the doc imposing the use of #include <ext/....>
b) rename ext/memory (and the other files)
c) Is there a way to include "../memory" if and only if ext/ is in the include search path?
>Release-Note:
>Audit-Trail:
>Unformatted:
----gnatsweb-attachment----
Content-Type: text/plain; name="pippo.C"
Content-Disposition: inline; filename="pippo.C"

#include <iostream>

int main(void) {
	return 0;
};