Date: Sun, 20 Dec 2020 06:34:28 +0200 From: Konstantin Belousov <kostikbel@gmail.com> To: Ryan Libby <rlibby@freebsd.org> Cc: John Baldwin <jhb@freebsd.org>, src-committers <src-committers@freebsd.org>, svn-src-all <svn-src-all@freebsd.org>, svn-src-head <svn-src-head@freebsd.org>, Alex Richardson <arichardson@freebsd.org> Subject: Re: svn commit: r368789 - head/libexec/rtld-elf/rtld-libc Message-ID: <X97UVARK%2BfCjuLtt@kib.kiev.ua> In-Reply-To: <CAHgpiFz2ir10rbHWWF=oX%2BPkT%2Bb7nV99Wwzfj03oj6VkLV-5rg@mail.gmail.com> References: <202012190838.0BJ8cVJ3064816@repo.freebsd.org> <47e9db0f-14b4-0a0d-45c4-7e466e69711f@FreeBSD.org> <CAHgpiFz2ir10rbHWWF=oX%2BPkT%2Bb7nV99Wwzfj03oj6VkLV-5rg@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Dec 19, 2020 at 08:27:31PM -0800, Ryan Libby wrote: > On a tangent, it might be neat to split out an rtld_bootstrap > (everything through init_rtld()) so that only the bootstrap code needs > to be compiled and linked with no-SSP. I looked at this some but I > figured there might not be appetite for a bunch of reorganization of > rtld just to enable SSP. Anyway the bootstrap code would still need > these particular libcompiler_rt functions to be no-SSP, as they get used > in the printf procedure, which I am guessing the bootstrap may need. Just my opinion. SSP is not too useful itself, and I believe it is useless in rtld, simply due to what rtld typically executes (matching static strings). Rototiling early rtld stages to get another bullet on the list of doubtful features should not be done.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?X97UVARK%2BfCjuLtt>