Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Tue, 29 Dec 2020 21:06:54 +0100
From:   Borislav Petkov <bp@alien8.de>
To:     Masami Hiramatsu <mhiramat@kernel.org>
Cc:     Andy Lutomirski <luto@amacapital.net>, X86 ML <x86@kernel.org>,
        LKML <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v1 03/19] x86/insn: Add an insn_decode() API
Message-ID: <20201229200654.GF29947@zn.tnic>
References: <20201223174233.28638-1-bp@alien8.de>
 <20201223174233.28638-4-bp@alien8.de>
 <20201228101510.49082d470ed328d81486ef04@kernel.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <20201228101510.49082d470ed328d81486ef04@kernel.org>
Precedence: bulk

On Mon, Dec 28, 2020 at 10:15:10AM +0900, Masami Hiramatsu wrote:
> BTW, insn_decode() can return -EINVAL if !insn_complete(), is that OK?

It does with this change. Or are you asking whether it returning -EINVAL
in that case is ok?

I don't see why not - this way callers can differentiate where it failed
- at fetching bytes with -ENODATA or it wasn't decoded completely -
-EINVAL.

> I think tools clone code must not use INSN_MODE_KERN because the tools may
> not use kernel Kconfig.
> 
> Hmm, this may be better to make a different patch to introduce a NOSYNC tag
> for sync checker in the tools. Something like;

I'd actually prefer this:

diff --git a/tools/arch/x86/include/asm/insn.h b/tools/arch/x86/include/asm/insn.h
index f8772b371452..545320c67855 100644
--- a/tools/arch/x86/include/asm/insn.h
+++ b/tools/arch/x86/include/asm/insn.h
@@ -98,8 +98,6 @@ extern int insn_get_length(struct insn *insn);
 enum insn_mode {
        INSN_MODE_32,
        INSN_MODE_64,
-       /* Mode is determined by the current kernel build. */
-       INSN_MODE_KERN,
        INSN_NUM_MODES,
 };
 

so that when a tool does use INSN_MODE_KERN, it would fail building:

In file included from util/intel-pt-decoder/intel-pt-insn-decoder.c:15:
util/intel-pt-decoder/../../../arch/x86/lib/insn.c: In function ‘insn_decode’:
util/intel-pt-decoder/../../../arch/x86/lib/insn.c:751:11: error: ‘INSN_MODE_KERN’ undeclared (first use in this function); did you mean ‘INSN_MODE_64’?
  751 |  if (m == INSN_MODE_KERN)
      |           ^~~~~~~~~~~~~~
      |           INSN_MODE_64
util/intel-pt-decoder/../../../arch/x86/lib/insn.c:751:11: note: each undeclared identifier is reported only once for each function it appears in
  LD       arch/perf-in.o
util/intel-pt-decoder/intel-pt-insn-decoder.c: In function ‘intel_pt_get_insn’:
util/intel-pt-decoder/intel-pt-insn-decoder.c:163:37: error: ‘INSN_MODE_KERN’ undeclared (first use in this function); did you mean ‘INSN_MODE_64’?
  163 |  ret = insn_decode(&insn, buf, len, INSN_MODE_KERN);
      |                                     ^~~~~~~~~~~~~~
      |                                     INSN_MODE_64

Thx.

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette