linux-kernel - Re: Dealing with the NMI mess

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20150731080303.GA2128@nazgul.tnic>
Date:	Fri, 31 Jul 2015 10:03:03 +0200
From:	Borislav Petkov <bp@...en8.de>
To:	Andy Lutomirski <luto@...capital.net>
Cc:	Paolo Bonzini <pbonzini@...hat.com>,
	Peter Zijlstra <peterz@...radead.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	Willy Tarreau <w@....eu>, Steven Rostedt <rostedt@...dmis.org>,
	X86 ML <x86@...nel.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	Thomas Gleixner <tglx@...utronix.de>,
	Brian Gerst <brgerst@...il.com>
Subject: Re: Dealing with the NMI mess

On Thu, Jul 30, 2015 at 10:11:40PM -0700, Andy Lutomirski wrote:
> This instruction is awesome.  Binutils can disassemble it (it's called
> "icebp") but it can't assemble it.  KVM has special handling for it on
> VMX and actually reports it to QEMU on SVM (complete with a defined
> ABI).

Fun.

> We have an asm macro so we can assemble it for 32-bit but not
> 64-bit, despite the fact that it works on 64-bit.
> 
> The kernel instruction decoder can't decode it.

Yeah, the kernel insn decoder needs to be fixed. Even my decoder can
decode it:

$ echo "0xf1" | ./x86d -
0:       f1                      icebp

Big deal. :-)

Let's do some fun and games:

$ cat icebp.c
int main()
{
        asm volatile(".byte 0xf1");

        return 0;
}

$ gcc -Wall -o icebp{,.c}
$ objdump -d icebp

...

00000000004004ac <main>:
  4004ac:       55                      push   %rbp
  4004ad:       48 89 e5                mov    %rsp,%rbp
  4004b0:       f1                      icebp  
  4004b1:       b8 00 00 00 00          mov    $0x0,%eax
  4004b6:       5d                      pop    %rbp
  4004b7:       c3                      retq   
  4004b8:       90                      nop
...

$ ./icebp
Trace/breakpoint trap

^ this in qemu.

On baremetal it gets a SIGTRAP with TRAP_BRKPT. Looks like signal
handling knows about it...

$ strace /tmp/icebp
execve("/tmp/icebp", ["/tmp/icebp"], [/* 27 vars */]) = 0
brk(0)                                  = 0x1680000
access("/etc/ld.so.nohwcap", F_OK)      = -1 ENOENT (No such file or directory)
mmap(NULL, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f71e243d000
access("/etc/ld.so.preload", R_OK)      = -1 ENOENT (No such file or directory)
open("/etc/ld.so.cache", O_RDONLY|O_CLOEXEC) = 3
fstat(3, {st_mode=S_IFREG|0644, st_size=127070, ...}) = 0
mmap(NULL, 127070, PROT_READ, MAP_PRIVATE, 3, 0) = 0x7f71e241d000
close(3)                                = 0
access("/etc/ld.so.nohwcap", F_OK)      = -1 ENOENT (No such file or directory)
open("/lib/x86_64-linux-gnu/libc.so.6", O_RDONLY|O_CLOEXEC) = 3
read(3, "\177ELF\2\1\1\3\0\0\0\0\0\0\0\0\3\0>\0\1\0\0\0P\34\2\0\0\0\0\0"..., 832) = 832
fstat(3, {st_mode=S_IFREG|0755, st_size=1729984, ...}) = 0
mmap(NULL, 3836448, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 0x7f71e1e76000
mprotect(0x7f71e2015000, 2097152, PROT_NONE) = 0
mmap(0x7f71e2215000, 24576, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x19f000) = 0x7f71e2215000
mmap(0x7f71e221b000, 14880, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0x7f71e221b000
close(3)                                = 0
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f71e241c000
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f71e241b000
mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f71e241a000
arch_prctl(ARCH_SET_FS, 0x7f71e241b700) = 0
mprotect(0x7f71e2215000, 16384, PROT_READ) = 0
mprotect(0x7f71e243f000, 4096, PROT_READ) = 0
munmap(0x7f71e241d000, 127070)          = 0
--- SIGTRAP {si_signo=SIGTRAP, si_code=TRAP_BRKPT, si_pid=4195505, si_uid=0} ---
+++ killed by SIGTRAP +++
Trace/breakpoint trap

> Fortunately, it looks like the vm86 case is correct (or as correct as
> any of the vm86 junk can be), although I haven't tested it.  I bet
> that icebp is like int3 in that it punches through vm86 mode instead
> of sending #GP.

Yeah, INT 1. I wonder whether INT 1, i.e. CD imm8 does the same thing.

But why do you say it is special - it simply raises #DB, i.e. vector 1.
Web page seems to say so when interrupt redirection is disabled. It
sounds like a nice and quick way to generate a breakpoint. You can do
that with INT 01, i.e., the CD opcode, too.

If I'd had to guess, it isn't documented because of the proprietary ICE
aspect. And no one uses ICEs anymore so it is going to be forgotten with
people popping off and on and asking about the undocumented opcode.

-- 
Regards/Gruss,
    Boris.

ECO tip #101: Trim your mails when you reply.
--
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/