From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 99823 invoked by alias); 14 Feb 2020 12:34:01 -0000 Mailing-List: contact binutils-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: binutils-owner@sourceware.org Received: (qmail 99815 invoked by uid 89); 14 Feb 2020 12:34:00 -0000 Authentication-Results: sourceware.org; auth=none X-Spam-SWARE-Status: No, score=-4.2 required=5.0 tests=AWL,BAYES_00,RCVD_IN_DNSWL_NONE,SPF_PASS autolearn=ham version=3.3.1 spammy=H*MI:sk:CAKdteO X-HELO: mail-vs1-f67.google.com Received: from mail-vs1-f67.google.com (HELO mail-vs1-f67.google.com) (209.85.217.67) by sourceware.org (qpsmtpd/0.93/v0.84-503-g423c35a) with ESMTP; Fri, 14 Feb 2020 12:33:59 +0000 Received: by mail-vs1-f67.google.com with SMTP id x123so6134879vsc.2 for ; Fri, 14 Feb 2020 04:33:59 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=mime-version:from:date:message-id:subject:to; bh=t55pBVvpQ1C/kNrtzHlJY3H/OCtqfqKmixC7FaSg2pM=; b=SGVhOz6aPkYwK8s3YFC60+ciw6WlsScAtc+LOtLopslU4kkPe2aYbvFAd331FfgQMX ai4uizVenM6SV90cirR6Zh3R3qncPsUcTMprkAoBH2fp3KV8FZHAYvsF0GqqJzL2Wgx1 kGxwiyCdR3FhtyMTxgzZLbBet/ABJDn/bk1adagkkWJnvWGfRPbsu2j6DVY9dfWvx9Ee 77Of9FfX5D+raX/ntzK3+6jr3t7CaAMHRwFX+j87uzaG96T+OD/Kbrh9yvFMapPHeXIM MX5WMeoxZHkIc1XS7foVH+A0wafQ6OpydfPPcNuRwCyoKq/vLGbx7tBVG0OyzmeELky0 mxyg== MIME-Version: 1.0 From: Christophe Lyon Date: Fri, 14 Feb 2020 12:34:00 -0000 Message-ID: Subject: [PATCH] Add support for non-contiguous memory regions To: binutils Content-Type: text/plain; charset="UTF-8" X-IsSubscribed: yes X-SW-Source: 2020-02/txt/msg00337.txt.bz2 Hi, This is a follow-up to https://sourceware.org/ml/binutils/2019-11/msg00402.html https://sourceware.org/ml/binutils/2020-01/msg00064.html and https://sourceware.org/ml/binutils/2020-02/msg00037.html Changes between v3 and v4: * I ran the tests on more targets (those that have xxx_build_one_stub: avr csky hppa mc68ch11 metag nios2 ppc aarch64 m68hc11 m68hc12) * this made me update the generic test (match addresses more lazily to cope with 64 bits outputs, ignore symbol table whose order depends on the target), and skip it for the targets that don't support -mlittle-endian. Is there a more generic way of doing this? (for sure, there are other targets that don't support that option but that I haven't tested) * I updated xxx_build_one_stub similarly to what I needed for arm, although I didn't write specific tests The changes between v2 and v3: * added some doc/NEWS * fixed bugs detected when running the testsuite with the option activated by default. In particular, I've noticed and documented the incompatibility with INSERT (or rather, it's likely to cause problems, but depends on the actual linker scripts contents), because my new option breaks the assumption of the parsing & processing order assumed by INSERT. * added new option --enable-non-contiguous-regions-warnings which activates some warnings to help understand why the behaviour changes in some cases There are 4 patches, to hopefully make review/comments easier; I can squash them at commit time if this is preferable. * patch1: is the main (code) patch * patch2: generic test * patch3: arm tests * patch4: powerpc test OK? Thanks, Christophe