libstdc++/92: Bug in istream::get(basic_streambuf*);

[brent@rcfile.org]

On 02 Oct 2000 at 05:17 (-0000), brent@rcfile.org wrote:
| The following reply was made to PR libstdc++/92; it has been noted by GNATS.
| 
| From: brent@rcfile.org
| To: bgarcia@laurelnetworks.com
| Cc: libstdc++-gnats@sourceware.cygnus.com
| Subject: Re: libstdc++/92: Bug in istream::get(basic_streambuf*);
| Date: Mon, 2 Oct 2000 01:20:11 -0400
| 
|  On 28 Sep 2000 at 17:23 (-0000), bgarcia@laurelnetworks.com wrote:
|  | >Environment:
|  | linux x86, libstdc++ v.2.90.8, gcc v 2.95.2
|  
|  verified that most recent cvs has this bug as well. thanks for the bug 
|  report. I'll provide a patch for this tomorrow.

patches attached. diffs made against each file where it lives.

testsuite failures with this patch applied:
-       -       22_locale/codecvt_unicode_char.cc
-       -       22_locale/codecvt_unicode_wchar_t.cc
-       -       22_locale/ctor_copy_dtor.cc
-       23_containers/map_operators.cc         (compilation failed)
-       23_containers/set_operators.cc         (compilation failed)
-       -       27_io/ios_base_callbacks.cc
-       -       27_io/ios_manip_fmtflags.cc

failures appear to all be unrelated to any code this patch affects.

ChangeLog entry:

2000-10-03  Brent Verner  <brent@rcfile.org>
	* bits/istream.tcc [basic_istream::get(basic_streambuf&)]:
    removed test for _M_gcount < in_avail(), as in_avail()
    only reports info for current buffer, causing method to
    return at end of buffer.
    http://sources.redhat.com/ml/libstdc++-prs/2000-q3/msg00041.html
  * testsuite/27_io/istream_unformatted.cc [test07()]: new test
  * testsuite/27_io/istream_unformatted-07.txt: new file

-- 
All opinions expressed are My own, unless otherwise attributed. In
presenting facts, I expressly reserve the right to be Wrong. Portions
of this message authored by Me are subject to the Free Thought License.