en Copyright 2015 Java http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#f7413c6a41ec258d5b922a501ffe84b626e09079 ld.so should stop calling msyscall(2), since it is fully neutered andabout to be removed. Please be careful building through this, you needa kernel at least March 29th or so to build through it, otherwise usesnapshots to cross over.ok various people List of files: /openbsd-src/libexec/ld.so/Makefile Fri, 05 Apr 2024 13:51:47 +0000 deraadt <deraadt@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#a09d28f1ea0e610cdfe26793c3fba26fd0f1a272 Stop initializing pinsyscall(SYS_execve in dynamic binaries that containa reference reaching the execve(2) stub. The new pinsyscalls(2) thatapplies to all system calls has made this redundant. List of files: /openbsd-src/libexec/ld.so/Makefile Fri, 19 Jan 2024 14:16:41 +0000 deraadt <deraadt@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#a346a8255fe7f006393c41b1a2263d4deb91d353 create a stub for pinsyscalls(2) List of files: /openbsd-src/libexec/ld.so/Makefile Wed, 20 Dec 2023 14:54:29 +0000 deraadt <deraadt@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#f81a9d33b8af5278286a7c09df952bc2d3108dff zap stray backslashok deraadt@ List of files: /openbsd-src/libexec/ld.so/Makefile Sat, 08 Jul 2023 11:03:45 +0000 jasper <jasper@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#a4ba1e578abc32528c379f6c94024325cd810809 Lookup the start,len of the "execve" stub in libc.so, and tell the kernelwhere it is with pinsyscall(). In non-static binaries, kernel execve(2)will now insist upon being called from that small region. Arriving froma different region smells like ROP methodology, and the process is killed. List of files: /openbsd-src/libexec/ld.so/Makefile Sat, 18 Feb 2023 01:22:50 +0000 deraadt <deraadt@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#baca2ca69e4060baafc40e8655b1c9fa68b70613 teach ld.so how to call the mimmutable() system callok kettenis List of files: /openbsd-src/libexec/ld.so/Makefile Sat, 05 Nov 2022 18:44:09 +0000 deraadt <deraadt@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#1d60349d0b961891264d426ffe1c0ced24b2374c Roll the syscalls that have an off_t argument to remove the explicit padding.Switch libc and ld.so to the generic stubs for these calls.WARNING: reboot to updated kernel before installing libc or ld.so!Time for a story...When gcc (back in 1.x days) first implemented long long, it didn't (always)pass 64bit arguments in 'aligned' registers/stack slots, with the result thatargument offsets didn't match structure offsets. This affected the nine systemcalls that pass off_t arguments: ftruncate lseek mmap mquery pread preadv pwrite pwritev truncateTo avoid having to do custom ASM wrappers for those, BSD put an explicit padargument in so that the off_t argument would always start on a even slot andthus be naturally aligned. Thus those odd wrappers in lib/libc/sys/ that use__syscall() and pass an extra '0' argument.The ABIs for different CPUs eventually settled how things should be passed oneach and gcc 2.x followed them. The only arch now where it helps is landisk,which needs to skip the last argument register if it would be the first half ofa 64bit argument. So: add new syscalls without the pad argument and on landiskdo that skipping directly in the syscall handler in the kernel. Keep compatsupport for the existing syscalls long enough for the transition.ok deraadt@ List of files: /openbsd-src/libexec/ld.so/Makefile Thu, 23 Dec 2021 18:50:31 +0000 guenther <guenther@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#bde8342fa15b1aa7e2c82b6c0ee725bd661853a6 Use AFLAGS when building syscall stubs. Drop AINC wich isn't used.ok deraadt@ List of files: /openbsd-src/libexec/ld.so/Makefile Sat, 26 Jun 2021 14:46:48 +0000 kettenis <kettenis@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#93a664f21378b90da9c7200b1020553be1df08d4 It is not always clear what ld.so was backed up to ld.so.backup, andbetter that folk doing development in here use their own cp tooling. List of files: /openbsd-src/libexec/ld.so/Makefile Mon, 02 Dec 2019 17:19:25 +0000 deraadt <deraadt@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#29ab13eb17d9e2d6e30c1ca48e4fee1f71c2a2fc As additional paranoia, make a copy of system ld.so into obj/ld.so.backupWe don't want to CLEANFILES this one. On occasion this comes in useful. List of files: /openbsd-src/libexec/ld.so/Makefile Fri, 29 Nov 2019 19:30:56 +0000 deraadt <deraadt@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#e50417250fb1d11dd90538c8f1a10723b2e4b3af Repurpose the "syscalls must be on a writeable page" mechanism toenforce a new policy: system calls must be in pre-registered regions.We have discussed more strict checks than this, but none satisfy thecost/benefit based upon our understanding of attack methods, anywayslet's see what the next iteration looks like.This is intended to harden (translation: attackers must put extraeffort into attacking) against a mixture of W^X failures and JIT bugswhich allow syscall misinterpretation, especially in environments withpolymorphic-instruction/variable-sized instructions. It fits in a bitwith libc/libcrypto/ld.so random relink on boot and no-restart-at-crashbehaviour, particularily for remote problems. Less effective once on-hostsince someone the libraries can be read.For static-executables the kernel registers the main program'sPIE-mapped exec section valid, as well as the randomly-placed sigtramppage. For dynamic executables ELF ld.so's exec segment is alsolabelled valid; ld.so then has enough information to register libc'sexec section as valid via call-once msyscall(2)For dynamic binaries, we continue to to permit the main program execsegment because "go" (and potentially a few other applications) haveembedded system calls in the main program. Hopefully at least go getsfixed soon.We declare the concept of embedded syscalls a bad idea for numerousreasons, as we notice the ecosystem has many ofstatic-syscall-in-base-binary which are dynamically linked againstlibraries which in turn use libc, which contains another set ofsyscall stubs. We've been concerned about adding even one additionalsyscall entry point... but go's approach tends to double the entry-pointattack surface.This was started at a nano-hackathon in Bob Beck's basement 2 weeksago during a long discussion with mortimer trying to hide from the SSLscream-conversations, and finished in more comfortable circumstancesnext to a wood-stove at Elk Lakes cabin with UVM scream-conversations.ok guenther kettenis mortimer, lots of feedback from othersconversations about go with jsing tb sthen List of files: /openbsd-src/libexec/ld.so/Makefile Fri, 29 Nov 2019 06:34:44 +0000 deraadt <deraadt@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#17fa88719dad59793b58356460b07659ff5d59db For more archs, ld.so itself only needs/uses the arch's "just add load offset"'relative' relocation. Take advantage of that to simplify ld.so's self-reloccode: * give the exceptional archs (hppa and mips64) copies of the current boot.c as boot_md.c * teach the Makefile to use boot_md.c when present * reduce boot.c down to the minimum necessary to handle just relative reloc * teach the Makefile to fail if the built ld.so has other types of relocsok visa@ kettenis@ List of files: /openbsd-src/libexec/ld.so/Makefile Sun, 20 Oct 2019 03:44:49 +0000 guenther <guenther@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#8f3b0808dbeea07d6b28e21f0f947ced6b8eca98 Suppress DWARF2 warnings on clang archs by building with -gdwarf-4ok deraadt@, kettenis@ List of files: /openbsd-src/libexec/ld.so/Makefile Sat, 03 Aug 2019 19:36:55 +0000 guenther <guenther@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#6eb4f976a0f7775de1d6294cab149fddd37b7a5b ld.so's $ORIGIN handling when argv[0] is a relative path was broken bythe change in __getcwd(2)'s return value. Fix it by switching to the__realpath(2) syscall, eliminating the ld.so copy of realpath().problem caught by regress and noted by bluhm@ok deraadt@ List of files: /openbsd-src/libexec/ld.so/Makefile Sun, 14 Jul 2019 03:23:12 +0000 guenther <guenther@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#c0197e40dd4094880fe205f9c12d3e12e14e7ff3 ld.so boot cleanup support: - put functions and data which are only used before calling the executable's start function into their own page-aligned segments for unmapping (only done on amd64, arm64, armv7, powerpc, and sparc64 so far) - pass .init_array and .preinit_array functions an addition argument which is a callback to get a structure which includes a function that frees the boot text and data - sometimes delay doing RELRO processing: for a shared-object marked DF_1_INITFIRST do it after the object's .init_array, for the executable do it after the .preinit_array - improve test-ld.so to link against libpthread and trigger its initialization latelibc changes to use this will come laterok kettenis@ List of files: /openbsd-src/libexec/ld.so/Makefile Fri, 10 May 2019 13:29:21 +0000 guenther <guenther@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#92ffaf13ffd174051475dab79f500436b1aeb294 make ld.so build reproducible by including a FILE symbol and by removing-x from ld; this is only used for syspatch creation so builds are stillrandomized just as beforeok guenther@ List of files: /openbsd-src/libexec/ld.so/Makefile Wed, 29 Aug 2018 15:56:46 +0000 robert <robert@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#544a27041b454910782e3cf3b3673938dffaa3b6 provide ffs, gcc generates calls to it, even when __builtin_ffs() is used.ok deraadt@ List of files: /openbsd-src/libexec/ld.so/Makefile Tue, 30 Jan 2018 20:15:25 +0000 otto <otto@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#4bb83bacbcdac4b7d90409410727ef942dbedce3 also cleanout ld.so.a List of files: /openbsd-src/libexec/ld.so/Makefile Fri, 08 Dec 2017 05:30:16 +0000 deraadt <deraadt@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#8fb1a259ab728b7a817ee2830af1a85293090f6c Consolidate lib.so.*.a, ld.so.a and the kernel relink kit intoone location under /usr/share/relink.Be more specific in src/etc/rc reorder_libs() what filesystemsneed r/w remount and ensure that their mount state is restored.Idea and positive feedback from deraadt@OK aja@ tb@ List of files: /openbsd-src/libexec/ld.so/Makefile Sun, 05 Nov 2017 10:29:24 +0000 rpe <rpe@openbsd.org> http://src.rcs.uwaterloo.ca:8080/history/openbsd-src/libexec/ld.so/Makefile#9e688d4960a3c649013fe7fec1384fb3aaba8291 Place all .o files, Symbols.map, and the test program into /usr/libdata/ld.so.a.My work is done; the next piece is coming from rpe.ok visa kettenis List of files: /openbsd-src/libexec/ld.so/Makefile Sun, 27 Aug 2017 22:18:08 +0000 deraadt <deraadt@openbsd.org>