This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: [libstdc++] Partial implementation of DR 60.
- From: Benjamin Kosnik <bkoz at redhat dot com>
- To: Phil Edwards <phil at jaj dot com>
- Cc: libstdc++ at gcc dot gnu dot org, gcc-patches at gcc dot gnu dot org
- Date: Mon, 21 Apr 2003 11:06:06 -0500
- Subject: Re: [libstdc++] Partial implementation of DR 60.
- References: <20030420225948.GA20240@disaster.jaj.com>
>DR 60 makes changes to which functions affect, or do not affect, subsequent
>calls to gcount(). Easily implemented as below. I'd actually noticed
>these back when doxygenating the I/O library, but never checked them in. (?)
>
>DR 60 also changes other things, not done here.
Ouch!!!! I had just changed some of this behavior when the testsuites
got re-packaged. I was sure that I had added tests for some of this
stuff. Perhaps not, if you are changing this behavior as well and not
noticing anything? Is there any chance I could get you to add some
simple tests for the functions you've modified? They don't have to be
anything more than one liners with a call to gcount.
At least for basic_istream/[putback,seekg,tellg].
Also, both seekg's should have _M_gcount = 0 removed.
>Should this go in 3.3? *shrug*
I think you might as well put this in both places (you can keep the test
cases only in mainline if you'd prefer, since merging them can be a
pain.)
thanks,
-benjamin