{"id":2226054,"url":"http://patchwork.ozlabs.org/api/patches/2226054/?format=json","web_url":"http://patchwork.ozlabs.org/project/gcc/patch/orwlxzse5e.fsf_-_@lxoliva.fsfla.org/","project":{"id":17,"url":"http://patchwork.ozlabs.org/api/projects/17/?format=json","name":"GNU Compiler Collection","link_name":"gcc","list_id":"gcc-patches.gcc.gnu.org","list_email":"gcc-patches@gcc.gnu.org","web_url":null,"scm_url":null,"webscm_url":null,"list_archive_url":"","list_archive_url_format":"","commit_url_format":""},"msgid":"","list_archive_url":null,"date":"2026-04-22T06:27:41","name":"libstdc++: simd: use preprocessor to test for 64-bit long double [PR124657]","commit_ref":null,"pull_url":null,"state":"new","archived":false,"hash":"4a0ac7847ae419038d201a7be5c0703a5d6ac5f7","submitter":{"id":74937,"url":"http://patchwork.ozlabs.org/api/people/74937/?format=json","name":"Alexandre Oliva","email":"oliva@adacore.com"},"delegate":null,"mbox":"http://patchwork.ozlabs.org/project/gcc/patch/orwlxzse5e.fsf_-_@lxoliva.fsfla.org/mbox/","series":[{"id":500926,"url":"http://patchwork.ozlabs.org/api/series/500926/?format=json","web_url":"http://patchwork.ozlabs.org/project/gcc/list/?series=500926","date":"2026-04-22T06:27:41","name":"libstdc++: simd: use preprocessor to test for 64-bit long double [PR124657]","version":1,"mbox":"http://patchwork.ozlabs.org/series/500926/mbox/"}],"comments":"http://patchwork.ozlabs.org/api/patches/2226054/comments/","check":"pending","checks":"http://patchwork.ozlabs.org/api/patches/2226054/checks/","tags":{},"related":[],"headers":{"Return-Path":"","X-Original-To":["incoming@patchwork.ozlabs.org","gcc-patches@gcc.gnu.org"],"Delivered-To":["patchwork-incoming@legolas.ozlabs.org","gcc-patches@gcc.gnu.org"],"Authentication-Results":["legolas.ozlabs.org;\n\tdkim=pass (2048-bit key;\n secure) header.d=adacore.com header.i=@adacore.com header.a=rsa-sha256\n header.s=google header.b=b8Dlakbb;\n\tdkim-atps=neutral","legolas.ozlabs.org;\n spf=pass (sender SPF authorized) smtp.mailfrom=gcc.gnu.org\n (client-ip=38.145.34.32; helo=vm01.sourceware.org;\n envelope-from=gcc-patches-bounces~incoming=patchwork.ozlabs.org@gcc.gnu.org;\n receiver=patchwork.ozlabs.org)","sourceware.org;\n\tdkim=pass (2048-bit key,\n secure) header.d=adacore.com header.i=@adacore.com header.a=rsa-sha256\n header.s=google header.b=b8Dlakbb","sourceware.org; dmarc=pass (p=quarantine dis=none)\n header.from=adacore.com","sourceware.org; spf=pass smtp.mailfrom=adacore.com","server2.sourceware.org;\n arc=none smtp.remote-ip=74.125.82.54"],"Received":["from vm01.sourceware.org (vm01.sourceware.org [38.145.34.32])\n\t(using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)\n\t key-exchange x25519 server-signature ECDSA (secp384r1) server-digest SHA384)\n\t(No client certificate requested)\n\tby legolas.ozlabs.org (Postfix) with ESMTPS id 4g0q7F6Z6Tz1yCv\n\tfor ; Wed, 22 Apr 2026 16:28:56 +1000 (AEST)","from vm01.sourceware.org (localhost [127.0.0.1])\n\tby sourceware.org (Postfix) with ESMTP id 00B324BA901A\n\tfor ; Wed, 22 Apr 2026 06:28:53 +0000 (GMT)","from mail-dl1-f54.google.com (mail-dl1-f54.google.com\n [74.125.82.54])\n by sourceware.org (Postfix) with ESMTPS id 3F57C4BA901A\n for ; Wed, 22 Apr 2026 06:27:57 +0000 (GMT)","by mail-dl1-f54.google.com with SMTP id\n a92af1059eb24-12c8cc7a77eso1957014c88.1\n for ; Tue, 21 Apr 2026 23:27:57 -0700 (PDT)","from free.home ([2804:14c:4d1:41a6::2000])\n by smtp.gmail.com with ESMTPSA id\n a92af1059eb24-12c8caecebfsm12348662c88.5.2026.04.21.23.27.54\n (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256);\n Tue, 21 Apr 2026 23:27:55 -0700 (PDT)","from livre (livre.home [172.31.160.2])\n by free.home (8.15.2/8.15.2) with ESMTPS id 63M6Rfh0032463\n (version=TLSv1.3 cipher=TLS_AES_256_GCM_SHA384 bits=256 verify=NOT);\n Wed, 22 Apr 2026 03:27:42 -0300"],"DKIM-Filter":["OpenDKIM Filter v2.11.0 sourceware.org 00B324BA901A","OpenDKIM Filter v2.11.0 sourceware.org 3F57C4BA901A"],"DMARC-Filter":"OpenDMARC Filter v1.4.2 sourceware.org 3F57C4BA901A","ARC-Filter":"OpenARC Filter v1.0.0 sourceware.org 3F57C4BA901A","ARC-Seal":"i=1; a=rsa-sha256; d=sourceware.org; s=key; t=1776839277; cv=none;\n b=ECHbd4jOrmsqQP1e8ecmbeulXNzFoY+7+XsmVstqTpFRwgqlzo40ZaoY9z9rJPnBqIBiOSC98BzqcOogoFFv/Uq8Z1wAmvnsY2krXdiZyeajT2odebr13AaPCJNXbyw2hrCUu/8mEE9MR+z9tj2n6I084pLyNl/wgcjYs29WlHc=","ARC-Message-Signature":"i=1; a=rsa-sha256; d=sourceware.org; s=key;\n t=1776839277; c=relaxed/simple;\n bh=FFk6+f/QKJrCc8IxdUIAxybKRkI2jCR1QO4xBstjNkA=;\n h=DKIM-Signature:From:To:Subject:Date:Message-ID:MIME-Version;\n b=VECJWB1ST7Fo/73vrb52qObO68WRKnE2KXBdPja8KJprLtTjxXg/SYw6Wwv6GpYP64TMfe5I92O15EUyeXxGVgyDQMbeYNVNAGJVSlBEW6VO28BTTpqfuLu47UNBx1zKCrrUgHxLnANxKNGH9dKOQzXGNERf2IRRfOmrfEGnprc=","ARC-Authentication-Results":"i=1; server2.sourceware.org","DKIM-Signature":"v=1; a=rsa-sha256; c=relaxed/relaxed;\n d=adacore.com; s=google; t=1776839276; x=1777444076; darn=gcc.gnu.org;\n h=mime-version:user-agent:message-id:in-reply-to:date:references\n :organization:subject:cc:to:from:from:to:cc:subject:date:message-id\n :reply-to; bh=Hpu75Yy1pkEj+sOxssox+2nT3MT7vHDHCKgDLe/JCz4=;\n b=b8DlakbblqKTBRWwnRaldyDP/wTrQTLXSLtHdXX/JuDQShCNRsye3cQ6mDV5oX6eRf\n XXhURuibTVsKk4mJIciy9mMq+WWis/PTpBkMRoP+xMAsS0e6yXUg2tq0IDLYaE1u3DNW\n vnP3tHd1ieeiW6okuJvyk7mGFK1ZVCIQVjCtBxG8vjY1dYesX6KOGa22p7lTefSH8Dwc\n f2AJ7qoHRJKJPGv5g1w/wsmVGYS5JDXlAj1boq/nDkn6WXtGz2lkZTc2nnT11vjYL+XA\n 30YOtZwdoXDzN/aNzuZO4yCa7PpGi+YHE0zc1N/W41d7mh1/WwO3cj27rQtxQv405i/q\n PtRw==","X-Google-DKIM-Signature":"v=1; a=rsa-sha256; c=relaxed/relaxed;\n d=1e100.net; s=20251104; t=1776839276; x=1777444076;\n h=mime-version:user-agent:message-id:in-reply-to:date:references\n :organization:subject:cc:to:from:x-gm-gg:x-gm-message-state:from:to\n :cc:subject:date:message-id:reply-to;\n bh=Hpu75Yy1pkEj+sOxssox+2nT3MT7vHDHCKgDLe/JCz4=;\n b=M4fFpb1tcA9DsViNBGAfzv0PqLKuAJ07YXzei4/ueyLRcev5jpyfI+y4P5eagvlWrT\n txOq/tg9jxpIEJ7Hrh6i/i3WGBY0j4LIPz6ikN0MNMdPKYBz61QJzRKYFn/wNtcxUre/\n YcKRK6eydOMPiROPh5HcNmWuP8sMGE1GMABWWxB/aZMuP0WywtsJVMqiIA/jDz5pD7NH\n IUze575fPRtGBEuS6o1BKbaDDkduABYdbP8FIum1qJmJGYuJY4JU6fj2Be4vFcgyGQM3\n SclF1wapUHrb1W0voCAtR9L+paKwupMaoePy8TcGzjzLvaBpUh6Ix9ylVprIab7PkeP1\n q9wA==","X-Forwarded-Encrypted":"i=1;\n AFNElJ+dm3RuOzYkfNRxhA/VpaeNEbsr42ChAV9qbGKziy0tFiWSz35qFD3deW9WO7WBh2WYgcLK2CBIj72iqg==@gcc.gnu.org","X-Gm-Message-State":"AOJu0YxAHKqnzfKKpkwNDJUKTUvSrhjRBjkyy8A4eMjdomK1HuF6igDT\n PJmx+CPMKITRcC4PvAb6+b6EmRnMNBdnLM2HhrF4wvJwAd2aCoYIwSBlh7nS7gSe+g==","X-Gm-Gg":"AeBDiesD+Owg+pR/3TkEAxT2Eg2G1yy0atF8Zo7V+FzkF6r02xZg3Nqul+XAQgEjM4G\n Efv87EYOz/OdaHR9Bw14teKzCCGptb3a6AHUKJ5v8NIiFgDdS6oAbwCZlIgmE1cJFfAHmPgGZMA\n 8/JGW18JHxJy2wkabUa7/kvzyOkLfwiggEsjjr/FFBKeniOp3TS1NAeyI+XxiTRvz5/kS1IycMc\n 5wqBktRNGFNg4UmpdK+fydU9i6ANBnR9WEGUNFJxu9zMl1HNPjRHOit+N246Vj0yjVrWp0dHBCz\n IfOKYlIbz5/xIu8T9pucV5Xe8AAhf0Pwh9fMaYmVIHc7kNJzXQkgMaML8MwdnO0z6Vqh4mx2Y1N\n 3kuaOtakjtKKVhQJhqc+UsAIzpJ8lmo01vd6Semfv4DdxwbOVbdKA0SfcfqoqoMMjyG4kjxR1k5\n d9ZFjO1HWtOQ==","X-Received":"by 2002:a05:7022:fa2:b0:129:1d25:f1da with SMTP id\n a92af1059eb24-12c73ae91a5mr10134874c88.3.1776839275742;\n Tue, 21 Apr 2026 23:27:55 -0700 (PDT)","From":"Alexandre Oliva ","To":"Jonathan Wakely ","Cc":"Matthias Kretz , libstdc++@gcc.gnu.org,\n Jonathan Wakely , gcc-patches\n ","Subject":"[PATCH] libstdc++: simd: use preprocessor to test for 64-bit long\n double [PR124657]","Organization":"Free thinker, does not speak for AdaCore","References":"\n \n \n <47992676.fMDQidcC6G@vir-laptop> \n \n ","Date":"Wed, 22 Apr 2026 03:27:41 -0300","In-Reply-To":"\n \n (Jonathan Wakely's message of \"Tue, 21 Apr 2026 09:27:36 +0100\")","Message-ID":"","User-Agent":"Gnus/5.13 (Gnus v5.13) Emacs/27.1 (gnu/linux)","MIME-Version":"1.0","Content-Type":"text/plain","X-Scanned-By":"MIMEDefang 2.84","X-BeenThere":"gcc-patches@gcc.gnu.org","X-Mailman-Version":"2.1.30","Precedence":"list","List-Id":"Gcc-patches mailing list ","List-Unsubscribe":",\n ","List-Archive":"","List-Post":"","List-Help":"","List-Subscribe":",\n ","Errors-To":"gcc-patches-bounces~incoming=patchwork.ozlabs.org@gcc.gnu.org"},"content":"On Apr 21, 2026, Jonathan Wakely wrote:\n\n> #if __LDBL_MANT_DIG__ == __DBL_MANT_DIG__\n\nI tried to apply the same technique to simd.h, on top of\nhttps://gcc.gnu.org/pipermail/gcc-patches/2026-April/712616.html, but it\ndoesn't save any instantiations, neither on x86 nor on ppc. It seems to\nmake things generally uglier, but it might offer a little compile-time\nbenefit on both, so I left it as a separate patch... WDYT?\n\nAlso regstrapped on x86_64-linux-gnu and powerpc64le-linux-gnu. I also\ntested pr109261_constexpr_simd.cc with -mlong-double-64 on both. Ok for\nstage1?\n\n\nWhen long double is as wide as double on x86 or powerpc\n(-mlong-double-64), we remap it to double in __intrinsic_type.\n\nUse the preprocessor to compare __L?DBL_MANT_DIG instead of\nfloating-type type sizes on x86 SSE and PowerPC AltiVec.\n\n\nfor libstdc++-v3/ChangeLog\n\n\tPR libstdc++/124657\n\t* include/experimental/bits/simd.h [_GLIBCXX_SIMD_HAVE_SSE]\n\t(__intrinsic_type::_S_is_ldouble): New. Use all over.\n\t(__intrinsic_type::type): Use preprocessor tests.\n\t[__ALTIVEC__] (__intrinsic_type): Likewise.\n---\n libstdc++-v3/include/experimental/bits/simd.h | 25 ++++++++++++++++++++-----\n 1 file changed, 20 insertions(+), 5 deletions(-)","diff":"diff --git a/libstdc++-v3/include/experimental/bits/simd.h b/libstdc++-v3/include/experimental/bits/simd.h\nindex 8c30c8af26881..177d3598b54b5 100644\n--- a/libstdc++-v3/include/experimental/bits/simd.h\n+++ b/libstdc++-v3/include/experimental/bits/simd.h\n@@ -2469,16 +2469,24 @@ template <>\n template \n struct __intrinsic_type<_Tp, _Bytes, enable_if_t<__is_vectorizable_v<_Tp> && _Bytes <= 64>>\n {\n+ static constexpr bool _S_is_ldouble = is_same_v<_Tp, long double>;\n+\n // allow _Tp == long double with -mlong-double-64\n- static_assert(!(is_same_v<_Tp, long double>\n-\t\t && sizeof(long double) > sizeof(double)),\n+#if __LDBL_MANT_DIG != __DBL_MANT_DIG\n+ static_assert(!_S_is_ldouble,\n \t\t \"no __intrinsic_type support for long double on x86\");\n+#endif\n \n static constexpr size_t _S_VBytes = _Bytes <= 16 ? 16 : _Bytes <= 32 ? 32 : 64;\n \n using type [[__gnu__::__vector_size__(_S_VBytes)]]\n = conditional_t, long long int,\n-\t\t conditional_t, double, _Tp> >;\n+#if __LDBL_MANT_DIG != __DBL_MANT_DIG\n+\t\t _Tp\n+#else\n+\t\t conditional_t<_S_is_ldouble, double, _Tp>\n+#endif\n+\t\t >;\n };\n #endif // _GLIBCXX_SIMD_HAVE_SSE\n \n@@ -2582,12 +2590,17 @@ template \n static constexpr bool _S_is_ldouble = is_same_v<_Tp, long double>;\n \n // allow _Tp == long double with -mlong-double-64\n- static_assert(!(_S_is_ldouble && sizeof(long double) > sizeof(double)),\n+#if __LDBL_MANT_DIG != __DBL_MANT_DIG\n+ static_assert(!_S_is_ldouble,\n \t\t \"no __intrinsic_type support for 128-bit floating point on PowerPC\");\n+#endif\n \n #ifndef __VSX__\n static_assert(!(is_same_v<_Tp, double>\n-\t\t || (_S_is_ldouble && sizeof(long double) == sizeof(double))),\n+#if __LDBL_MANT_DIG == __DBL_MANT_DIG\n+\t\t || _S_is_ldouble\n+#endif\n+\t\t ),\n \t\t \"no __intrinsic_type support for 64-bit floating point on PowerPC w/o VSX\");\n #endif\n \n@@ -2595,9 +2608,11 @@ template \n {\n if constexpr (is_floating_point_v<_Tp>)\n \t{\n+#if __LDBL_MANT_DIG == __DBL_MANT_DIG\n \t if constexpr (_S_is_ldouble)\n \t return double {};\n \t else\n+#endif\n \t return _Tp {};\n \t}\n else if constexpr (is_signed_v<_Tp>)\n","prefixes":[]}