From: Josh Poimboeuf <jpoimboe-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
To: Jessica Yu <jeyu-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
Cc: Rusty Russell <rusty-8n+1lVoiYb80n/F98K4Iww@public.gmane.org>,
Seth Jennings <sjenning-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>,
Jiri Kosina <jikos-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>,
Vojtech Pavlik <vojtech-IBi9RG/b67k@public.gmane.org>,
Jonathan Corbet <corbet-T1hC0tSOHrs@public.gmane.org>,
Miroslav Benes <mbenes-AlSwsSmVLrQ@public.gmane.org>,
linux-api-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
live-patching-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
x86-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-s390-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-doc-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Subject: Re: livepatch: reuse module loader code to write relocations
Date: Mon, 11 Jan 2016 21:05:52 -0600 [thread overview]
Message-ID: <20160112030552.GC17874@treble.redhat.com> (raw)
In-Reply-To: <20160111223513.GA10796-N0bYjD2NfQ6k4hzjq3hgyGTy53QMssKEsp+A89P3RPuQWHG76I6BsA@public.gmane.org>
On Mon, Jan 11, 2016 at 05:35:13PM -0500, Jessica Yu wrote:
> +++ Josh Poimboeuf [11/01/16 15:33 -0600]:
> >On Fri, Jan 08, 2016 at 02:28:22PM -0500, Jessica Yu wrote:
> >>Reuse module loader code to write relocations, thereby eliminating the need
> >>for architecture specific relocation code in livepatch. Namely, we reuse
> >>apply_relocate_add() in the module loader to write relocations instead of
> >>duplicating functionality in livepatch's klp_write_module_reloc(). To apply
> >>relocation sections, remaining SHN_LIVEPATCH symbols referenced by relocs
> >>are resolved and then apply_relocate_add() is called to apply those
> >>relocations.
> >>
> >>In addition, remove x86 livepatch relocation code. It is no longer needed
> >>since symbol resolution and relocation work have been offloaded to module
> >>loader.
> >>
> >>Signed-off-by: Jessica Yu <jeyu-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> >>---
> >> arch/x86/include/asm/livepatch.h | 2 -
> >> arch/x86/kernel/Makefile | 1 -
> >> arch/x86/kernel/livepatch.c | 70 ---------------------------
> >> include/linux/livepatch.h | 33 +++++--------
> >> kernel/livepatch/core.c | 101 +++++++++++++++++++--------------------
> >> 5 files changed, 62 insertions(+), 145 deletions(-)
> >> delete mode 100644 arch/x86/kernel/livepatch.c
> >>
> >>diff --git a/arch/x86/include/asm/livepatch.h b/arch/x86/include/asm/livepatch.h
> >>index 19c099a..7312e25 100644
> >>--- a/arch/x86/include/asm/livepatch.h
> >>+++ b/arch/x86/include/asm/livepatch.h
> >>@@ -33,8 +33,6 @@ static inline int klp_check_compiler_support(void)
> >> #endif
> >> return 0;
> >> }
> >>-int klp_write_module_reloc(struct module *mod, unsigned long type,
> >>- unsigned long loc, unsigned long value);
> >>
> >> static inline void klp_arch_set_pc(struct pt_regs *regs, unsigned long ip)
> >> {
> >>diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile
> >>index b1b78ff..c5e9a5c 100644
> >>--- a/arch/x86/kernel/Makefile
> >>+++ b/arch/x86/kernel/Makefile
> >>@@ -67,7 +67,6 @@ obj-$(CONFIG_X86_MPPARSE) += mpparse.o
> >> obj-y += apic/
> >> obj-$(CONFIG_X86_REBOOTFIXUPS) += reboot_fixups_32.o
> >> obj-$(CONFIG_DYNAMIC_FTRACE) += ftrace.o
> >>-obj-$(CONFIG_LIVEPATCH) += livepatch.o
> >> obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o
> >> obj-$(CONFIG_FTRACE_SYSCALLS) += ftrace.o
> >> obj-$(CONFIG_X86_TSC) += trace_clock.o
> >>diff --git a/arch/x86/kernel/livepatch.c b/arch/x86/kernel/livepatch.c
> >>deleted file mode 100644
> >>index 92fc1a5..0000000
> >>--- a/arch/x86/kernel/livepatch.c
> >>+++ /dev/null
> >>@@ -1,70 +0,0 @@
> >>-/*
> >>- * livepatch.c - x86-specific Kernel Live Patching Core
> >>- *
> >>- * Copyright (C) 2014 Seth Jennings <sjenning-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
> >>- * Copyright (C) 2014 SUSE
> >>- *
> >>- * 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 2
> >>- * 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 <http://www.gnu.org/licenses/>.
> >>- */
> >>-
> >>-#include <linux/module.h>
> >>-#include <linux/uaccess.h>
> >>-#include <asm/elf.h>
> >>-#include <asm/livepatch.h>
> >>-
> >>-/**
> >>- * klp_write_module_reloc() - write a relocation in a module
> >>- * @mod: module in which the section to be modified is found
> >>- * @type: ELF relocation type (see asm/elf.h)
> >>- * @loc: address that the relocation should be written to
> >>- * @value: relocation value (sym address + addend)
> >>- *
> >>- * This function writes a relocation to the specified location for
> >>- * a particular module.
> >>- */
> >>-int klp_write_module_reloc(struct module *mod, unsigned long type,
> >>- unsigned long loc, unsigned long value)
> >>-{
> >>- size_t size = 4;
> >>- unsigned long val;
> >>- unsigned long core = (unsigned long)mod->core_layout.base;
> >>- unsigned long core_size = mod->core_layout.size;
> >>-
> >>- switch (type) {
> >>- case R_X86_64_NONE:
> >>- return 0;
> >>- case R_X86_64_64:
> >>- val = value;
> >>- size = 8;
> >>- break;
> >>- case R_X86_64_32:
> >>- val = (u32)value;
> >>- break;
> >>- case R_X86_64_32S:
> >>- val = (s32)value;
> >>- break;
> >>- case R_X86_64_PC32:
> >>- val = (u32)(value - loc);
> >>- break;
> >>- default:
> >>- /* unsupported relocation type */
> >>- return -EINVAL;
> >>- }
> >>-
> >>- if (loc < core || loc >= core + core_size)
> >>- /* loc does not point to any symbol inside the module */
> >>- return -EINVAL;
> >>-
> >>- return probe_kernel_write((void *)loc, &val, size);
> >>-}
> >>diff --git a/include/linux/livepatch.h b/include/linux/livepatch.h
> >>index a882865..2f12ce7 100644
> >>--- a/include/linux/livepatch.h
> >>+++ b/include/linux/livepatch.h
> >>@@ -65,27 +65,8 @@ struct klp_func {
> >> };
> >>
> >> /**
> >>- * struct klp_reloc - relocation structure for live patching
> >>- * @loc: address where the relocation will be written
> >>- * @sympos: position in kallsyms to disambiguate symbols (optional)
> >>- * @type: ELF relocation type
> >>- * @name: name of the referenced symbol (for lookup/verification)
> >>- * @addend: offset from the referenced symbol
> >>- * @external: symbol is either exported or within the live patch module itself
> >>- */
> >>-struct klp_reloc {
> >>- unsigned long loc;
> >>- unsigned long sympos;
> >>- unsigned long type;
> >>- const char *name;
> >>- int addend;
> >>- int external;
> >>-};
> >>-
> >>-/**
> >> * struct klp_object - kernel object structure for live patching
> >> * @name: module name (or NULL for vmlinux)
> >>- * @relocs: relocation entries to be applied at load time
> >> * @funcs: function entries for functions to be patched in the object
> >> * @kobj: kobject for sysfs resources
> >> * @mod: kernel module associated with the patched object
> >>@@ -95,7 +76,6 @@ struct klp_reloc {
> >> struct klp_object {
> >> /* external */
> >> const char *name;
> >>- struct klp_reloc *relocs;
> >> struct klp_func *funcs;
> >>
> >> /* internal */
> >>@@ -123,6 +103,19 @@ struct klp_patch {
> >> enum klp_state state;
> >> };
> >>
> >>+/*
> >>+ * Livepatch-specific symbols and relocation
> >>+ * sections are prefixed with a tag:
> >>+ * .klp.rel. for relocation sections
> >>+ * .klp.sym. for livepatch symbols
> >>+ */
> >>+#define KLP_TAG_LEN 9
> >>+/*
> >>+ * Livepatch-specific bits for specifying symbol
> >>+ * positions in the Elf_Sym st_other field
> >>+ */
> >>+#define KLP_SYMPOS(o) (o >> 2) & 0xff
> >>+
> >
> >Can st_value be used instead? I think we ended up deciding that would
> >be better:
> >
> > https://lkml.kernel.org/g/20151210213328.GA6553-N0bYjD2NfQ6k4hzjq3hgyGTy53QMssKEsp+A89P3RPuQWHG76I6BsA@public.gmane.org
> >
> >Because:
> >
> >- st_value is easily viewable in readelf
> >- st_other has some arch-specific uses
> >
> >And another reason not previously discussed:
> >
> >- st_other is an unsigned char, which limits sympos to values < 64
>
> I originally wanted to encode the symbol position in st_value, but
> I've discovered that since st_value is overwritten once the
> symbol is (first) resolved, we no longer have the symbol position if we
> need to resolve the symbols again (so we wouldn't be able to resolve
> them). This could happen when we patch a module that loads and
> unloads more than once for example.
Ah, good point.
> I chose st_other since it isn't touched in s390x kernel code nor in
> x86 kernel code (it does get used in the x86 userspace reloc tool in
> arch/x86/tools, which is why I left the first two bits alone for
> ELF_ST_VISIBILITY, and the rest had undefined usage). However this
> isn't the best approach and I'm afraid of stepping on other arch's
> toes in future patches. Perhaps there is another field where we can
> stuff the sympos in? st_size for instance doesn't seem to be touched
> anywhere in the module loader. I don't know if I'd want to stuff
> sympos in the sym name, there's enough going on there..
I think st_other still isn't going to work because of possible arch
conflicts and because of its small size (6 bits) otherwise.
st_size could work. It doesn't _seem_ to be used by the module code,
though it's hard to confirm that for all arches. But it makes me a
little nervous to override that field. For example, might it confuse
some user-space tools out there, or the linker?
Maybe encoding it in the sym name is the safest bet. Why not, we've
already got a bunch of other stuff there anyway :-) One idea would be
to append it with ',<sympos>' to be consistent with the sysfs entries:
.klp.sym.vmlinux.symname,1
--
Josh
next prev parent reply other threads:[~2016-01-12 3:05 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-01-08 19:28 [RFC PATCH v3 0/6] (mostly) Arch-independent livepatch Jessica Yu
2016-01-08 19:28 ` [RFC PATCH v3 1/6] Elf: add livepatch-specific Elf constants Jessica Yu
2016-01-08 19:28 ` [RFC PATCH v3 2/6] module: preserve Elf information for livepatch modules Jessica Yu
[not found] ` <1452281304-28618-3-git-send-email-jeyu-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2016-01-11 1:25 ` Rusty Russell
2016-01-14 4:47 ` Jessica Yu
[not found] ` <20160114044718.GC980-N0bYjD2NfQ6k4hzjq3hgyGTy53QMssKEsp+A89P3RPuQWHG76I6BsA@public.gmane.org>
2016-01-14 20:28 ` Rusty Russell
2016-01-08 19:28 ` [RFC PATCH v3 3/6] module: s390: keep mod_arch_specific " Jessica Yu
2016-01-08 19:28 ` [RFC PATCH v3 4/6] livepatch: reuse module loader code to write relocations Jessica Yu
[not found] ` <1452281304-28618-5-git-send-email-jeyu-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2016-01-11 16:56 ` Petr Mladek
2016-01-11 20:53 ` Josh Poimboeuf
2016-01-11 21:33 ` Josh Poimboeuf
[not found] ` <20160111213349.GB17874-8wJ5/zUtDR0XGNroddHbYwC/G2K4zDHf@public.gmane.org>
2016-01-11 22:35 ` Jessica Yu
[not found] ` <20160111223513.GA10796-N0bYjD2NfQ6k4hzjq3hgyGTy53QMssKEsp+A89P3RPuQWHG76I6BsA@public.gmane.org>
2016-01-12 3:05 ` Josh Poimboeuf [this message]
[not found] ` <20160112030552.GC17874-8wJ5/zUtDR0XGNroddHbYwC/G2K4zDHf@public.gmane.org>
2016-01-12 9:12 ` Petr Mladek
2016-01-14 5:07 ` Jessica Yu
2016-01-12 16:40 ` [RFC PATCH v3 4/6] " Miroslav Benes
2016-01-14 3:49 ` Jessica Yu
2016-01-14 9:04 ` Miroslav Benes
2016-01-13 9:19 ` [RFC PATCH v3 4/6] " Miroslav Benes
2016-01-13 9:31 ` Jiri Kosina
[not found] ` <alpine.LNX.2.00.1601131012340.10011-ztGlSCb7Y1iN3ZZ/Hiejyg@public.gmane.org>
2016-01-13 18:39 ` Jessica Yu
2016-01-14 9:10 ` Miroslav Benes
2016-01-08 19:28 ` [RFC PATCH v3 5/6] samples: livepatch: mark as livepatch module Jessica Yu
[not found] ` <1452281304-28618-1-git-send-email-jeyu-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2016-01-08 19:28 ` [RFC PATCH v3 6/6] Documentation: livepatch: outline the Elf format of a " Jessica Yu
2016-01-12 12:09 ` Petr Mladek
2016-01-12 14:45 ` Josh Poimboeuf
[not found] ` <20160112120951.GO731-KsEp0d+Q8qECVLCxKZUutA@public.gmane.org>
2016-01-14 5:04 ` Jessica Yu
-- strict thread matches above, loose matches on Subject: below --
2016-03-16 19:47 [PATCH v5 0/6] (mostly) Arch-independent livepatch Jessica Yu
2016-03-16 19:47 ` [PATCH v5 4/6] livepatch: reuse module loader code to write relocations Jessica Yu
2016-03-21 13:55 ` Miroslav Benes
2016-03-21 19:18 ` Jessica Yu
2016-03-21 19:24 ` Josh Poimboeuf
2016-03-21 21:16 ` Jiri Kosina
2016-03-21 21:34 ` Josh Poimboeuf
[not found] ` <20160321213445.e64jic2uoc7tdtbl-8wJ5/zUtDR0XGNroddHbYwC/G2K4zDHf@public.gmane.org>
2016-03-21 22:02 ` Jiri Kosina
2016-03-22 19:00 ` Jessica Yu
2016-03-21 16:31 ` [PATCH v5 4/6] " Petr Mladek
2016-03-21 16:46 ` Josh Poimboeuf
2016-03-21 17:36 ` Josh Poimboeuf
2016-03-21 18:07 ` Jessica Yu
2016-02-04 1:11 [RFC PATCH v4 0/6] (mostly) Arch-independent livepatch Jessica Yu
2016-02-04 1:11 ` [RFC PATCH v4 4/6] livepatch: reuse module loader code to write relocations Jessica Yu
2016-02-08 20:26 ` Josh Poimboeuf
[not found] ` <20160208202606.GD23106-8wJ5/zUtDR0XGNroddHbYwC/G2K4zDHf@public.gmane.org>
2016-02-10 0:56 ` Jessica Yu
2016-02-09 14:01 ` [RFC PATCH v4 4/6] " Petr Mladek
2016-02-10 1:21 ` Jessica Yu
2015-12-01 4:21 [RFC PATCH v2 0/6] (mostly) Arch-independent livepatch Jessica Yu
2015-12-01 4:21 ` [RFC PATCH v2 4/6] livepatch: reuse module loader code to write relocations Jessica Yu
[not found] ` <1448943679-3412-5-git-send-email-jeyu-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2015-12-01 8:43 ` Jessica Yu
2015-12-08 18:38 ` [RFC PATCH v2 4/6] " Josh Poimboeuf
[not found] ` <20151208183844.GC14846-8wJ5/zUtDR0XGNroddHbYwC/G2K4zDHf@public.gmane.org>
2015-12-09 19:10 ` Jessica Yu
[not found] ` <20151209191013.GA25387-N0bYjD2NfQ6k4hzjq3hgyGTy53QMssKEsp+A89P3RPuQWHG76I6BsA@public.gmane.org>
2015-12-10 14:28 ` Josh Poimboeuf
2015-12-10 21:33 ` Jessica Yu
[not found] ` <20151210213328.GA6553-N0bYjD2NfQ6k4hzjq3hgyGTy53QMssKEsp+A89P3RPuQWHG76I6BsA@public.gmane.org>
2015-12-10 21:41 ` Josh Poimboeuf
[not found] ` <20151210214115.GD4934-8wJ5/zUtDR0XGNroddHbYwC/G2K4zDHf@public.gmane.org>
2015-12-10 22:07 ` Jessica Yu
2015-12-16 5:40 ` Jessica Yu
2015-12-16 12:59 ` Miroslav Benes
2015-12-16 19:14 ` Jessica Yu
[not found] ` <20151216054048.GA28258-N0bYjD2NfQ6k4hzjq3hgyGTy53QMssKEsp+A89P3RPuQWHG76I6BsA@public.gmane.org>
2015-12-17 15:45 ` Petr Mladek
2015-12-21 5:57 ` Jessica Yu
2015-11-10 4:45 [RFC PATCH 0/5] Arch-independent livepatch Jessica Yu
2015-11-10 4:45 ` [RFC PATCH 3/5] livepatch: reuse module loader code to write relocations Jessica Yu
2015-11-11 14:30 ` Miroslav Benes
[not found] ` <alpine.LNX.2.00.1511111519250.12467-ztGlSCb7Y1iN3ZZ/Hiejyg@public.gmane.org>
2015-11-11 20:07 ` Jessica Yu
[not found] ` <20151111200732.GB30025-N0bYjD2NfQ6k4hzjq3hgyGTy53QMssKEsp+A89P3RPuQWHG76I6BsA@public.gmane.org>
2015-11-12 15:27 ` Miroslav Benes
2015-11-12 17:40 ` Josh Poimboeuf
[not found] ` <20151112174032.GG4038-M1RQNqTfbHcHO9+bouH2LOBSVt8DKT+lrE5yTffgRl4@public.gmane.org>
2015-11-12 20:22 ` Jessica Yu
[not found] ` <20151112202243.GC5841-N0bYjD2NfQ6k4hzjq3hgyGTy53QMssKEsp+A89P3RPuQWHG76I6BsA@public.gmane.org>
2015-11-12 20:32 ` Josh Poimboeuf
[not found] ` <20151112203221.GM4038-M1RQNqTfbHcHO9+bouH2LOBSVt8DKT+lrE5yTffgRl4@public.gmane.org>
2015-11-13 7:15 ` Jessica Yu
2015-11-13 13:51 ` Miroslav Benes
[not found] ` <alpine.LNX.2.00.1511121601370.9440-ztGlSCb7Y1iN3ZZ/Hiejyg@public.gmane.org>
2015-11-12 19:14 ` Jessica Yu
[not found] ` <20151112191458.GB5841-N0bYjD2NfQ6k4hzjq3hgyGTy53QMssKEsp+A89P3RPuQWHG76I6BsA@public.gmane.org>
2015-11-12 20:35 ` Jessica Yu
2015-11-11 15:22 ` [RFC PATCH 3/5] " Petr Mladek
[not found] ` <20151111152256.GP2599-KsEp0d+Q8qECVLCxKZUutA@public.gmane.org>
2015-11-11 18:27 ` Jessica Yu
[not found] ` <20151111182718.GA30025-N0bYjD2NfQ6k4hzjq3hgyGTy53QMssKEsp+A89P3RPuQWHG76I6BsA@public.gmane.org>
2015-11-12 9:16 ` Petr Mladek
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20160112030552.GC17874@treble.redhat.com \
--to=jpoimboe-h+wxahxf7alqt0dzr+alfa@public.gmane.org \
--cc=corbet-T1hC0tSOHrs@public.gmane.org \
--cc=jeyu-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
--cc=jikos-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org \
--cc=linux-api-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=linux-doc-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=linux-s390-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=live-patching-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=mbenes-AlSwsSmVLrQ@public.gmane.org \
--cc=rusty-8n+1lVoiYb80n/F98K4Iww@public.gmane.org \
--cc=sjenning-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
--cc=vojtech-IBi9RG/b67k@public.gmane.org \
--cc=x86-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).