[Patch] Fix a (long standing) bug in stringbuf::pbackfail

[Paolo Carlini]

Hi,

today I noticed this long standing bug (already in 3.3, at least): for
some reason, we weren't enforcing some sensible words in the standards.
Took the occasion to add missing tests for the virtual.

Tested x86-linux, will commit later today barring objections...

Paolo.

//////////////
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: CL_pbackfail
URL: <http://gcc.gnu.org/pipermail/libstdc++/attachments/20041001/809c6c53/attachment.ksh>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: patch_pbackfail.diff
URL: <http://gcc.gnu.org/pipermail/libstdc++/attachments/20041001/809c6c53/attachment-0001.ksh>