From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-wr1-f51.google.com (mail-wr1-f51.google.com [209.85.221.51]) by sourceware.org (Postfix) with ESMTPS id 8570B3858C2D; Fri, 13 May 2022 08:04:33 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.4.1 sourceware.org 8570B3858C2D Authentication-Results: sourceware.org; dmarc=none (p=none dis=none) header.from=palves.net Authentication-Results: sourceware.org; spf=pass smtp.mailfrom=gmail.com Received: by mail-wr1-f51.google.com with SMTP id h16so9632589wrb.2; Fri, 13 May 2022 01:04:33 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:message-id:date:mime-version:user-agent:subject :content-language:to:cc:references:from:in-reply-to :content-transfer-encoding; bh=I9i6f7mqwQljMKyvxnTi/EjQGg6OTSHpsavWgNRN39Q=; b=578FWPWKq0jLBHqBiIpStB8aElbpakTTjOTwYcmWEjZz8e+RRk8qlG2xl24o3PTuZo qj8ljBinPGkBOSztRFRQRloVFP3y264mG1gYW0SLdQsNr16a6hxnsgqyDxsl3+V+By/S 5ht7maN9JuK5enSQDWu/IOHcV4DjizS9RvrnktGoryK1YnIWZ77kVDHIGc7N3NgFLFZv z7mc8wgzuVy8Ixa1uHnHSy01zDB+0RPM4Q9eWjGSxK/6HTFIirbct9E+9XYpT1jqAo9A f1c1whLsmOO79Mis3DrPhDZceGx3+HnJoGQGBfoNdxcOpSFp+NHeg/j2IiVJc8BhSXWp KhXg== X-Gm-Message-State: AOAM533nmS3qfp7iksQrtndS2j2s6MgS0K2lFULSH8XZkpwa//z7s7hr pniNcVqkf/HZGhkjzrkbqbIbao9y1bw= X-Google-Smtp-Source: ABdhPJyrNHSMy8+ZYGCgmBXWXbRiR7ikZ3/fQH/wHWVFunJLTC1A2oxLhjfiP8h31RfEzMBUnOefjg== X-Received: by 2002:adf:fc11:0:b0:20c:5fc2:2ee0 with SMTP id i17-20020adffc11000000b0020c5fc22ee0mr2786635wrr.225.1652429072209; Fri, 13 May 2022 01:04:32 -0700 (PDT) Received: from ?IPV6:2001:8a0:f924:2600:209d:85e2:409e:8726? ([2001:8a0:f924:2600:209d:85e2:409e:8726]) by smtp.gmail.com with ESMTPSA id x20-20020adfbb54000000b0020c5253d90dsm1472636wrg.89.2022.05.13.01.04.30 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Fri, 13 May 2022 01:04:31 -0700 (PDT) Message-ID: <71eeb967-80c9-63ef-ed59-9e7a758c58f8@palves.net> Date: Fri, 13 May 2022 09:04:29 +0100 MIME-Version: 1.0 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101 Thunderbird/91.9.0 Subject: Re: [PATCH] Fix "gdb --write" with core files Content-Language: en-US To: Alan Modra Cc: gdb-patches@sourceware.org, binutils@sourceware.org References: <20220512144915.3270761-1-pedro@palves.net> From: Pedro Alves In-Reply-To: Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit X-Spam-Status: No, score=-11.3 required=5.0 tests=BAYES_00, FREEMAIL_FORGED_FROMDOMAIN, FREEMAIL_FROM, GIT_PATCH_0, HEADER_FROM_DIFFERENT_DOMAINS, KAM_DMARC_STATUS, KAM_SHORT, NICE_REPLY_A, RCVD_IN_DNSWL_NONE, RCVD_IN_MSPIKE_H2, SPF_HELO_NONE, SPF_PASS, TXREP, T_SCC_BODY_TEXT_LINE autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on server2.sourceware.org X-BeenThere: gdb-patches@sourceware.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Gdb-patches mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 13 May 2022 08:04:38 -0000 On 2022-05-13 03:57, Alan Modra wrote: > On Thu, May 12, 2022 at 03:49:15PM +0100, Pedro Alves wrote: >> That happens because _bfd_elf_set_section_contents does this to seek >> to the section's offset: >> >> pos = hdr->sh_offset + offset; >> if (bfd_seek (abfd, pos, SEEK_SET) != 0 >> || bfd_bwrite (location, count, abfd) != count) >> return false; >> >> ... and 'hdr->sh_offset' is zero, so we seek to just OFFSET, which is >> incorrect. The reason 'hdr->sh_offset' is zero is that >> kernel-generated core files normally don't even have a section header >> table (gdb-generated ones do, but that's more an accident than a >> feature), and indeed elf_core_file_p doesn't even try to read sections >> at all: > > I'm a little concerned about your patch keying off both_direction. In > particular it seems odd to change bfd behaviour when reading depending > on whether the file was opened "r" or "r+". Fair point. > > Did you try changing > pos = hdr->sh_offset + offset; > to > pos = section->filepos + offset; > in _bfd_elf_set_section_contents? I think that should work for you, > and not break ELF targets. > Ah, I had assumed using hdr->sh_offset was important in write_direction somehow. That works, and then we could just defer to _bfd_generic_set_section_contents, which does exactly the same thing. It's like as if in C++ you override a method to do extra things, and then call the base class's implementation. WDYT of this version, then? I've retested gas/binutils/ld/gdb on x86_64 GNU/Linux. Thanks! >From a9935ffa786d4aefd7bd12f4ad23ace87ad5d4a1 Mon Sep 17 00:00:00 2001 From: Pedro Alves Date: Wed, 11 May 2022 14:20:15 +0100 Subject: [PATCH] Fix "gdb --write" with core files If you load a core file into GDB with the --write option, or "set write on" (equivalent), and then poke memory expecting it to patch the core binary, you'll notice something odd -- the write seems to succeed, but in reality, it doesn't. The value you wrote doesn't persist. Like so: $ gdb -q --write -c testsuite/outputs/gdb.base/patch/gcore.test [New LWP 615986] Core was generated by `/home/pedro/gdb/build/gdb/testsuite/outputs/gdb.base/patch/patch'. Program terminated with signal SIGTRAP, Trace/breakpoint trap. #0 0x0000555555555131 in ?? () (gdb) p *(unsigned char *)0x0000555555555131 = 1 $1 = 1 '\001' (gdb) p *(unsigned char *)0x0000555555555131 $2 = 185 '\271' (gdb) Diffing hexdumps of before/after patching, reveals that a "0x1" was actually written somewhere in the file. The problem is that the "0x1" was written at the wrong offset in the file... That happens because _bfd_elf_set_section_contents does this to seek to the section's offset: pos = hdr->sh_offset + offset; if (bfd_seek (abfd, pos, SEEK_SET) != 0 || bfd_bwrite (location, count, abfd) != count) return false; ... and 'hdr->sh_offset' is zero, so we seek to just OFFSET, which is incorrect. The reason 'hdr->sh_offset' is zero is that kernel-generated core files normally don't even have a section header table (gdb-generated ones do, but that's more an accident than a feature), and indeed elf_core_file_p doesn't even try to read sections at all: /* Core files are simply standard ELF formatted files that partition the file using the execution view of the file (program header table) rather than the linking view. In fact, there is no section header table in a core file. The process status information (including the contents of the general register set) and the floating point register set are stored in a segment of type PT_NOTE. We handcraft a couple of extra bfd sections that allow standard bfd access to the general registers (.reg) and the floating point registers (.reg2). */ bfd_cleanup elf_core_file_p (bfd *abfd) Changing _bfd_elf_set_section_contents from: pos = hdr->sh_offset + offset; to: pos = section->filepos + offset; fixes it. If we do that however, the tail end of _bfd_elf_set_section_contents ends up as a copy of _bfd_generic_set_section_contents, so just call the latter, thus eliminating some duplicate code. New GDB testcase included, which exercises both patching an executable and patching a core file. Patching an executable already works without this fix, because in that case BFD reads in the sections table. Still, we had no testcase for that yet. In fact, we have no "set write on" testcases at all, this is the first one. Tested on x86-64 GNU/Linux, gdb, ld, binutils, and gas. Change-Id: I0f49f58b48aabab2e269f2959b8fd8a7fe36fdce --- bfd/elf.c | 9 +-- gdb/testsuite/gdb.base/patch.c | 26 ++++++++ gdb/testsuite/gdb.base/patch.exp | 106 +++++++++++++++++++++++++++++++ 3 files changed, 134 insertions(+), 7 deletions(-) create mode 100644 gdb/testsuite/gdb.base/patch.c create mode 100644 gdb/testsuite/gdb.base/patch.exp diff --git a/bfd/elf.c b/bfd/elf.c index f046994e3a8..c493aa5b172 100644 --- a/bfd/elf.c +++ b/bfd/elf.c @@ -9404,7 +9404,6 @@ _bfd_elf_set_section_contents (bfd *abfd, bfd_size_type count) { Elf_Internal_Shdr *hdr; - file_ptr pos; if (! abfd->output_has_begun && ! _bfd_elf_compute_section_file_positions (abfd, NULL)) @@ -9457,12 +9456,8 @@ _bfd_elf_set_section_contents (bfd *abfd, return true; } - pos = hdr->sh_offset + offset; - if (bfd_seek (abfd, pos, SEEK_SET) != 0 - || bfd_bwrite (location, count, abfd) != count) - return false; - - return true; + return _bfd_generic_set_section_contents (abfd, section, + location, offset, count); } bool diff --git a/gdb/testsuite/gdb.base/patch.c b/gdb/testsuite/gdb.base/patch.c new file mode 100644 index 00000000000..c75f7800c94 --- /dev/null +++ b/gdb/testsuite/gdb.base/patch.c @@ -0,0 +1,26 @@ +/* Copyright 2022 Free Software Foundation, Inc. + + This file is part of GDB. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 3 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program. If not, see . */ + +/* Value != 0 so it doesn't ends up in .bss, which we wouldn't be able + to patch. */ +int extern_global = 1; + +int +main() +{ + return 0; +} diff --git a/gdb/testsuite/gdb.base/patch.exp b/gdb/testsuite/gdb.base/patch.exp new file mode 100644 index 00000000000..96d8f2bcc68 --- /dev/null +++ b/gdb/testsuite/gdb.base/patch.exp @@ -0,0 +1,106 @@ +# Copyright 2022 Free Software Foundation, Inc. + +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 3 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program. If not, see . + +standard_testfile + +if {[build_executable "failed to prepare" $testfile $srcfile debug]} { + return -1 +} + +# Check that we can patch an executable. + +with_test_prefix "exec" { + clean_restart + + gdb_test_no_output "set write on" + + gdb_load $binfile + + gdb_test "p extern_global" " = 1" "read original value" + gdb_test "p extern_global = 2" " = 2" "modify value" + gdb_test "p extern_global" " = 2" "value modified" + + clean_restart $binfile + + gdb_test "p extern_global" " = 2" "value modified persisted" +} + +# Check that we can patch a core file. + +# Generate a core file. +with_test_prefix "gcore" { + clean_restart $binfile + + if {![runto_main]} { + return + } + + # Extract the value at PC, and add 1, letting it wrap if + # necessary. This is the value we will poke into memory. + set poke_value "" + gdb_test_multiple "p/x (*(unsigned char *) \$pc) + 1" "compute poke value" { + -re -wrap " = ($hex)" { + set poke_value $expect_out(1,string) + pass $gdb_test_name + } + } + if {$poke_value == ""} { + return + } + + set corefile [standard_output_file gcore.test] + set core_supported [gdb_gcore_cmd "$corefile" "save a corefile"] + if {!$core_supported} { + return + } +} + +# Load it into GDB with "set write on", and change the instruction at +# PC. +with_test_prefix "load core, write on" { + # Don't load a binary, we want to make sure we're patching the + # core, not the executable. + clean_restart + + gdb_test_no_output "set write on" + + set core_loaded [gdb_core_cmd "$corefile" "re-load generated corefile"] + if { $core_loaded == -1 } { + # No use proceeding from here. + return + } + + gdb_test "p/x *(unsigned char *) \$pc = $poke_value" \ + " = $poke_value" \ + "poke value" + gdb_test "p/x *(unsigned char *) \$pc" \ + " = $poke_value" \ + "value modified" +} + +# Re-load it into a new GDB session, now with "set write off", and +# confirm the value change persisted. +with_test_prefix "re-load modified core" { + # Don't load a binary, we want to make sure we've patched the + # core, not the executable. + clean_restart + + set core_loaded [gdb_core_cmd "$corefile" "re-load generated corefile"] + gdb_assert { $core_loaded != -1 } "core re-loaded" + + gdb_test "p/x *(unsigned char *) \$pc" \ + " = $poke_value" \ + "value modified persisted" +} base-commit: 845cbaa9ffbfd6a1f7976a6c7f3e4461e4d41993 -- 2.36.0