Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Message-ID: <1544176449.1251.6.camel@mtksdccf07>
From:   Kathleen Chang <yt.chang@mediatek.com>
To:     <mingo@redhat.com>
CC:     <wsd_upstream@mediatek.com>
Content-Type: text/plain; charset="ISO-8859-1"
MIME-Version: 1.0
Subject: sched: panic() in schedule_bug
Date:   Fri, 7 Dec 2018 18:00:20 +0800
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

hi, 

[kernel-4.9, kernel-4.14] 


When preemption disable in schedule(), 
Current linux design will panic when panic_on_warn=1 

However, 
   default setting, panic_on_warn=0,  and 
   if panic_on_warn=1  , and call panic() after printing out WARN()
location.  It affects too much cases.  

Could panic directly without checking panic_on_warn? 
This is obviously incorrect behavior, and should be panic to prevent
misbehavior.  


        if (panic_on_warn)
                panic("scheduling while atomic\n");


thanks!  


=============================
kernel/sched/core.c

schedule_debug =>  when in atomic preempt off   => __schedule_bug 

====