From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 13435 invoked by alias); 13 Jun 2018 21:41:58 -0000 Mailing-List: contact libc-alpha-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: libc-alpha-owner@sourceware.org Received: (qmail 13416 invoked by uid 89); 13 Jun 2018 21:41:57 -0000 Authentication-Results: sourceware.org; auth=none X-Virus-Found: No X-Spam-SWARE-Status: No, score=-2.6 required=5.0 tests=BAYES_00,RCVD_IN_DNSWL_LOW,SPF_PASS,URIBL_RED autolearn=ham version=3.3.2 spammy= X-HELO: mx0a-001b2d01.pphosted.com From: Tulio Magno Quites Machado Filho To: Joseph Myers , libc-alpha@sourceware.org Cc: Subject: Re: [PATCH 5/9] ldbl-128ibm-compat: Provide ISO C functions not provided by the _Float128 API In-Reply-To: References: <20180606223909.16675-1-tuliom@linux.ibm.com> <20180606223909.16675-6-tuliom@linux.ibm.com> User-Agent: Notmuch/0.25 (http://notmuchmail.org) Emacs/25.3.1 (x86_64-redhat-linux-gnu) Date: Wed, 13 Jun 2018 21:41:00 -0000 MIME-Version: 1.0 Content-Type: text/plain X-TM-AS-GCONF: 00 x-cbid: 18061321-0068-0000-0000-0000030891B0 X-IBM-SpamModules-Scores: X-IBM-SpamModules-Versions: BY=3.00009185; HX=3.00000241; KW=3.00000007; PH=3.00000004; SC=3.00000265; SDB=6.01046535; UDB=6.00535987; IPR=6.00825511; MB=3.00021630; MTD=3.00000008; XFM=3.00000015; UTC=2018-06-13 21:41:52 X-IBM-AV-DETECTION: SAVI=unused REMOTE=unused XFE=unused x-cbparentid: 18061321-0069-0000-0000-000044ACC0CC Message-Id: <87602mcqjo.fsf@linux.ibm.com> X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:,, definitions=2018-06-13_09:,, signatures=0 X-SW-Source: 2018-06/txt/msg00368.txt.bz2 Joseph Myers writes: > On Wed, 6 Jun 2018, Tulio Magno Quites Machado Filho wrote: > >> These functions do not exist in the float128 api, but they need to be >> provided for long double (*l api) because they are part of ISO C. > > They're not part of ISO C; they are part of *the long double API provided > by glibc* (and thus need to be provided for any alternative long double > format just like any other semi-obsolescent but still-available long > double interface, e.g. qecvt). Indeed. > Is it possible to provide these particular functions in the API without > adding extra ABIs for them? For example, if the relevant case in math.h > had #defines of __dremieee128 or __ieee128_dreml or whatever it ends up > being to the corresponding name for remainder, and likewise for gamma / > lgamma, would that work correctly with __REDIRECT to get calls to dreml > and gammal working for -mabi=ieeelongdouble without needing to add two new > ABIs and two new source files? Yes, it is. I'm dropping this patch. -- Tulio Magno