Re: [PATCH] C++: "#define timeval rpl_timeval" -> typedef in GNULIB_NAMESPACE

Thanks again. I installed the attached. From b2090b3cb8b18e4bf83e4f46e34c90e4515dd328 Mon Sep 17 00:00:00 2001 From: Pedro Alves Date: Tue, 15 Nov 2016 02:36:00 + Subject: [PATCH] sys_time: add gnulib::timeval for C++ * lib/sys_time.in.h [__cplusplus && defined GNULIB_NAMESPACE]: Define "ti

Re: [PATCH] Fix real-floating argument functions in C++ mode

Thanks, I installed the attached; it's what you sent, except in git format-patch format, with the commit message equaling the ChangeLog entry (with one empty line appended after the first line, and without indentation), and with the first commit-message line trying to fit in 50 chars with a pre

Re: [PATCH] strftime: don't use __THROW

On 11/15/2016 08:07 AM, Bruno Haible wrote: I don't understand why you see this warning but the glibc developers haven't noticed it yet?? I was puzzled about that too. For what it's worth, those __THROWs were introduced to glibc in 2004 by Ulrich Drepper, here: https://sourceware.org/git/?p

Re: [PATCH] strftime: don't use __THROW

Hi Jim, > I've just pushed the attached, to avoid new warnings when attempting > to build coreutils with the latest from gnulib: > + strftime: don't use __THROW > + Each use of __THROW would provoke this from gcc-7-to-be: > + > + lib/strftime.c:371:1: warning: '__leaf__' attribute ha