From: Peter Zijlstra <peterz@infradead.org>
To: Alexandre Chartre <alexandre.chartre@oracle.com>
Cc: x86@kernel.org, linux-kernel@vger.kernel.org,
jpoimboe@redhat.com, jthierry@redhat.com, tglx@linutronix.de
Subject: Re: [PATCH V3 6/9] objtool: Report inconsistent stack changes in alternative
Date: Thu, 16 Apr 2020 16:18:43 +0200 [thread overview]
Message-ID: <20200416141843.GP20730@hirez.programming.kicks-ass.net> (raw)
In-Reply-To: <20200414103618.12657-7-alexandre.chartre@oracle.com>
On Tue, Apr 14, 2020 at 12:36:15PM +0200, Alexandre Chartre wrote:
> To allow a valid stack unwinding, an alternative should have code
> where the same stack changes happens at the same places as in the
> original code. Add a check in objtool to validate that stack changes
> in alternative are effectively consitent with the original code.
This thing is completely buggered, it warns all over the place, even for
obviously correct alternatives like:
0000000000000310 <return_to_handler>:
310: 48 83 ec 18 sub $0x18,%rsp
314: 48 89 04 24 mov %rax,(%rsp)
318: 48 89 54 24 08 mov %rdx,0x8(%rsp)
31d: 48 89 ef mov %rbp,%rdi
320: e8 00 00 00 00 callq 325 <return_to_handler+0x15>
321: R_X86_64_PLT32 ftrace_return_to_handler-0x4
325: 48 89 c7 mov %rax,%rdi
328: 48 8b 54 24 08 mov 0x8(%rsp),%rdx
32d: 48 8b 04 24 mov (%rsp),%rax
331: 48 83 c4 18 add $0x18,%rsp
335: ff e7 jmpq *%rdi
337: 90 nop
338: 90 nop
339: 90 nop
Where 335 has two alternatives:
0: e9 00 00 00 00 jmpq 5 <.altinstr_replacement+0x5>
1: R_X86_64_PLT32 __x86_retpoline_rdi-0x4
and
5: 0f ae e8 lfence
8: ff e7 jmpq *%rdi
And it then comes back with:
defconfig-build/arch/x86/kernel/ftrace_64.o: warning: objtool: .entry.text+0x335: error in alternative
defconfig-build/arch/x86/kernel/ftrace_64.o: warning: objtool: .altinstr_replacement+0x5: in alternative 2
defconfig-build/arch/x86/kernel/ftrace_64.o: warning: objtool: .altinstr_replacement+0x8: misaligned alternative state change
which is just utter crap, JMP has no (CFI) state change.
next prev parent reply other threads:[~2020-04-16 14:33 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-04-14 10:36 [PATCH V3 0/9] objtool changes to check retpoline code Alexandre Chartre
2020-04-14 10:36 ` [PATCH V3 1/9] objtool: is_fentry_call() crashes if call has no destination Alexandre Chartre
2020-05-01 18:22 ` [tip: objtool/core] " tip-bot2 for Alexandre Chartre
2020-04-14 10:36 ` [PATCH V3 2/9] objtool: Allow branches within the same alternative Alexandre Chartre
2020-04-14 10:36 ` [PATCH V3 3/9] objtool: Add support for intra-function calls Alexandre Chartre
2020-04-14 12:07 ` Julien Thierry
2020-04-16 12:12 ` Miroslav Benes
2020-05-01 18:22 ` [tip: objtool/core] " tip-bot2 for Alexandre Chartre
2020-04-14 10:36 ` [PATCH V3 4/9] objtool: Handle return instruction with intra-function call Alexandre Chartre
2020-04-14 13:44 ` Julien Thierry
2020-04-14 10:36 ` [PATCH V3 5/9] objtool: Add return address unwind hints Alexandre Chartre
2020-04-14 16:16 ` Peter Zijlstra
2020-04-14 16:40 ` Alexandre Chartre
2020-04-14 17:56 ` Peter Zijlstra
2020-04-14 18:31 ` Alexandre Chartre
2020-04-14 18:42 ` Peter Zijlstra
2020-04-14 19:27 ` Alexandre Chartre
2020-04-14 19:48 ` Peter Zijlstra
2020-04-14 10:36 ` [PATCH V3 6/9] objtool: Report inconsistent stack changes in alternative Alexandre Chartre
2020-04-14 15:35 ` Julien Thierry
2020-04-14 22:41 ` kbuild test robot
2020-04-14 23:09 ` kbuild test robot
2020-04-16 14:18 ` Peter Zijlstra [this message]
2020-04-16 14:43 ` Alexandre Chartre
2020-04-14 10:36 ` [PATCH V3 7/9] x86/speculation: Change __FILL_RETURN_BUFFER to work with objtool Alexandre Chartre
2020-04-14 10:36 ` [PATCH V3 8/9] x86/speculation: Add return address unwind hints to retpoline and RSB stuffing Alexandre Chartre
2020-04-14 10:36 ` [PATCH V3 9/9] x86/speculation: Annotate intra-function calls Alexandre Chartre
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=20200416141843.GP20730@hirez.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=alexandre.chartre@oracle.com \
--cc=jpoimboe@redhat.com \
--cc=jthierry@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=tglx@linutronix.de \
--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 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).