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:   Fri, 22 Jan 2021 15:29:21 +0000
From:   Qais Yousef <qais.yousef@arm.com>
To:     Masami Hiramatsu <mhiramat@kernel.org>
Cc:     Catalin Marinas <catalin.marinas@arm.com>,
        Will Deacon <will@kernel.org>,
        Jean-Philippe Brucker <jean-philippe@linaro.org>,
        linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] arm64: kprobes: Fix Uexpected kernel BRK exception at EL1
Message-ID: <20210122152921.gnevsymmd4mckkee@e107158-lin>
References: <20210122110909.3324607-1-qais.yousef@arm.com>
 <20210122223601.4b3b7e01c6ec583c0439f1c9@kernel.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <20210122223601.4b3b7e01c6ec583c0439f1c9@kernel.org>
Precedence: bulk

On 01/22/21 22:36, Masami Hiramatsu wrote:
> > Further analysis showed that kcb->kprobe_status is set to
> > KPROBE_REENTER when the error occurs. By teaching
> > kprobe_breakpoint_ss_handler() to handle this status I can no  longer
> > reproduce the problem.
> 
> Very good catch! Yes, this missed the reentered kprobe case.
> 
> Acked-by: Masami Hiramatsu <mhiramat@kernel.org>

Thanks!

> 
> > 
> > Fixes: ba090f9cafd5 ("arm64: kprobes: Remove redundant kprobe_step_ctx")
> > Signed-off-by: Qais Yousef <qais.yousef@arm.com>
> > ---
> > 
> > Another change in behavior I noticed is that before ba090f9cafd5 ("arm64:
> > kprobes: Remove redundant kprobe_step_ctx") if 'cur' was NULL we wouldn't
> > return DBG_HOOK_ERROR, but after the change we do.
> 
> It should not happen, since the KPROBES_BRK_SS_IMM must be used only for
> kprobes's second break which must happen on the trampoline instruction
> buffer, which must set current kprobes before execution.

I see. Thanks for the explanation!

Cheers

--
Qais Yousef