From: Denys Vlasenko <vda.linux@googlemail.com>
To: Borislav Petkov <bp@alien8.de>
Cc: Denys Vlasenko <dvlasenk@redhat.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Oleg Nesterov <oleg@redhat.com>, "H. Peter Anvin" <hpa@zytor.com>,
Andy Lutomirski <luto@amacapital.net>,
Frederic Weisbecker <fweisbec@gmail.com>, X86 ML <x86@kernel.org>,
Alexei Starovoitov <ast@plumgrid.com>,
Will Drewry <wad@chromium.org>, Kees Cook <keescook@chromium.org>
Subject: Re: [PATCH 3/4] x86: open-code register save/restore in trace_hardirqs thunks
Date: Fri, 9 Jan 2015 19:54:08 +0100 [thread overview]
Message-ID: <CAK1hOcMsGJ7geRAzfazPLZitcO0o9XYRPJDYa8sJHR4i6DPL-g@mail.gmail.com> (raw)
In-Reply-To: <20150109121950.GD13637@pd.tnic>
Hi Borislav, thank you for testing and finding it.
On Fri, Jan 9, 2015 at 1:19 PM, Borislav Petkov <bp@alien8.de> wrote:
> Hmm, this patch breaks booting my kvm guest: it stops booting at some
> point and restarts itself after a couple of seconds.
>
> The monitor says rIP points to ffffffff8167ae30 which is this:
>
> ffffffff8167ae30 <async_page_fault>:
> ffffffff8167ae30: ff 15 fa 62 31 00 callq *0x3162fa(%rip) # ffffffff81991130 <pv_irq_ops+0x30>
> ffffffff8167ae36: 48 83 ec 78 sub $0x78,%rsp
> ffffffff8167ae3a: e8 d1 01 00 00 callq ffffffff8167b010 <error_entry>
> ffffffff8167ae3f: 48 89 e7 mov %rsp,%rdi
> ffffffff8167ae42: 48 8b 74 24 78 mov 0x78(%rsp),%rsi
> ffffffff8167ae47: 48 c7 44 24 78 ff ff movq $0xffffffffffffffff,0x78(%rsp)
> ffffffff8167ae4e: ff ff
> ffffffff8167ae50: e8 9b 9e 9c ff callq ffffffff81044cf0 <do_async_page_fault>
> ffffffff8167ae55: e9 76 02 00 00 jmpq ffffffff8167b0d0 <error_exit>
> ffffffff8167ae5a: 66 0f 1f 44 00 00 nopw 0x0(%rax,%rax,1)
I just looked at disassembly of thunk_64.o
before and after the patch. Here's what I see:
Before:
Disassembly of section .text:
0000000000000000 <restore-0x30>:
0: 48 83 ec 48 sub $0x48,%rsp
4: 48 89 7c 24 40 mov %rdi,0x40(%rsp)
9: 48 89 74 24 38 mov %rsi,0x38(%rsp)
e: 48 89 54 24 30 mov %rdx,0x30(%rsp)
13: 48 89 4c 24 28 mov %rcx,0x28(%rsp)
18: 48 89 44 24 20 mov %rax,0x20(%rsp)
1d: 4c 89 44 24 18 mov %r8,0x18(%rsp)
22: 4c 89 4c 24 10 mov %r9,0x10(%rsp)
27: 4c 89 54 24 08 mov %r10,0x8(%rsp)
2c: 4c 89 1c 24 mov %r11,(%rsp)
0000000000000030 <restore>:
30: 4c 8b 1c 24 mov (%rsp),%r11
34: 4c 8b 54 24 08 mov 0x8(%rsp),%r10
39: 4c 8b 4c 24 10 mov 0x10(%rsp),%r9
3e: 4c 8b 44 24 18 mov 0x18(%rsp),%r8
43: 48 8b 44 24 20 mov 0x20(%rsp),%rax
48: 48 8b 4c 24 28 mov 0x28(%rsp),%rcx
4d: 48 8b 54 24 30 mov 0x30(%rsp),%rdx
52: 48 8b 74 24 38 mov 0x38(%rsp),%rsi
57: 48 8b 7c 24 40 mov 0x40(%rsp),%rdi
5c: 48 83 c4 48 add $0x48,%rsp
60: c3 retq
After:
Disassembly of section .text:
0000000000000000 <restore>:
0: 4c 8b 1c 24 mov (%rsp),%r11
4: 4c 8b 54 24 08 mov 0x8(%rsp),%r10
9: 4c 8b 4c 24 10 mov 0x10(%rsp),%r9
e: 4c 8b 44 24 18 mov 0x18(%rsp),%r8
13: 48 8b 44 24 20 mov 0x20(%rsp),%rax
18: 48 8b 4c 24 28 mov 0x28(%rsp),%rcx
1d: 48 8b 54 24 30 mov 0x30(%rsp),%rdx
22: 48 8b 74 24 38 mov 0x38(%rsp),%rsi
27: 48 8b 7c 24 40 mov 0x40(%rsp),%rdi
2c: 48 03 24 25 48 00 00 add 0x48,%rsp
33: 00
34: c3 retq
IOW, my patch, on the level of generated assembly, results only in removal
of unreachable "SAVE_ARGS" thing.
I looked into git history all the way back to 2005. The part
+ /* SAVE_ARGS below is used only for the .cfi directives it contains. */
+ CFI_STARTPROC
+ SAVE_ARGS
+restore:
was there in the very first git commit.
I don't see how this SAVE_ARGS can affect anything. It *is* unreachable,
right?
Does kvm guest code really parse and use CFI data in its operation?
That's the only way the breakage can be explained.
In order to narrow it down, can you, instead of my patch, try
just deleting this one line, and see whether breakage appears?
Thanks!
--
vda
next prev parent reply other threads:[~2015-01-09 18:54 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-01-08 16:25 [PATCH 0/4] x86: entry.S cleanup Denys Vlasenko
2015-01-08 16:25 ` [PATCH 1/4] x86: entry_64.S: delete unused code Denys Vlasenko
2015-01-08 18:16 ` Borislav Petkov
2015-01-13 22:01 ` Andy Lutomirski
2015-01-08 16:25 ` [PATCH 2/4] x86: ia32entry.S: fix wrong symbolic constant usage: R11->ARGOFFSET Denys Vlasenko
2015-01-09 10:41 ` Borislav Petkov
2015-01-08 16:25 ` [PATCH 3/4] x86: open-code register save/restore in trace_hardirqs thunks Denys Vlasenko
2015-01-09 10:55 ` Borislav Petkov
2015-01-09 20:29 ` Denys Vlasenko
2015-01-10 13:52 ` Borislav Petkov
2015-01-09 12:19 ` Borislav Petkov
2015-01-09 18:54 ` Denys Vlasenko [this message]
2015-01-10 14:23 ` Borislav Petkov
2015-01-10 20:14 ` Denys Vlasenko
2015-01-10 20:17 ` Andy Lutomirski
2015-01-10 20:42 ` Borislav Petkov
2015-01-10 21:02 ` Andy Lutomirski
2015-01-10 21:09 ` Denys Vlasenko
2015-01-10 21:27 ` Linus Torvalds
2015-01-10 21:57 ` Denys Vlasenko
2015-01-10 20:43 ` Denys Vlasenko
2015-01-10 21:08 ` Linus Torvalds
2015-01-10 21:26 ` Borislav Petkov
2015-01-10 22:00 ` Borislav Petkov
2015-01-10 22:03 ` Denys Vlasenko
2015-01-10 22:04 ` Andy Lutomirski
2015-01-08 16:25 ` [PATCH 4/4] x86: entry_64.S: fold SAVE_ARGS_IRQ macro into its sole user Denys Vlasenko
2015-01-10 22:00 [PATCH 0/4 v2] x86: entry.S cleanup Denys Vlasenko
2015-01-10 22:00 ` [PATCH 3/4] x86: open-code register save/restore in trace_hardirqs thunks Denys Vlasenko
2015-01-10 22:07 ` Linus Torvalds
2015-01-10 22:35 ` Denys Vlasenko
2015-01-10 22:41 ` Borislav Petkov
2015-01-11 3:33 ` Denys Vlasenko
2015-01-11 10:54 ` Borislav Petkov
2015-01-11 23:06 ` Denys Vlasenko
2015-02-11 2:38 ` Andy Lutomirski
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=CAK1hOcMsGJ7geRAzfazPLZitcO0o9XYRPJDYa8sJHR4i6DPL-g@mail.gmail.com \
--to=vda.linux@googlemail.com \
--cc=ast@plumgrid.com \
--cc=bp@alien8.de \
--cc=dvlasenk@redhat.com \
--cc=fweisbec@gmail.com \
--cc=hpa@zytor.com \
--cc=keescook@chromium.org \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@amacapital.net \
--cc=oleg@redhat.com \
--cc=torvalds@linux-foundation.org \
--cc=wad@chromium.org \
--cc=x86@kernel.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.