This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: (libstdc++-v3) PATCH: Fix trivial typo in C++ library test case
- From: Loren James Rittle <rittle at latour dot rsch dot comm dot mot dot com>
- To: libstdc++ at gcc dot gnu dot org
- Date: Tue, 11 Mar 2003 02:41:27 -0600 (CST)
- Subject: Re: (libstdc++-v3) PATCH: Fix trivial typo in C++ library test case
- References: <200303110800.h2B80txB071230@latour.rsch.comm.mot.com>
- Reply-to: rittle at labs dot mot dot com
In article <200303110800 dot h2B80txB071230 at latour dot rsch dot comm dot mot dot com>,
Loren James Rittle <rittle at latour dot rsch dot comm dot mot dot com> writes:
> The test case was actually looking to see if
> ctype<char>::classic_table() returned non-0 and was likely propagated
> instead of testing proper propagation.
Additional commentary for the libstdc++-v3 list only. If the user is
always allowed to obtain a non-0 mask table from ctype<char>::table()
and then access it directly, then we should add another VERIFY:
VERIFY ( tmp );
The fix to this problem would be fairly annoying and hit about half
the config/os/.../ctype_noninline.h files.
ctype<char>::classic_table() would need to reverse-engineer a table
from many is* calls or, more reasonably, we could pre-build a copy for
a port and store it in the library.
Regards,
Loren