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;
From:   Serge Belyshev <belyshev@depni.sinp.msu.ru>
To:     Christophe Leroy <christophe.leroy@csgroup.eu>
Cc:     Andreas Schwab <schwab@linux-m68k.org>,
        Paul Mackerras <paulus@samba.org>,
        linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] powerpc/32s: Setup the early hash table at all time.
References: <b8f8101c368b8a6451844a58d7bd7d83c14cf2aa.1601566529.git.christophe.leroy@csgroup.eu>
        <87wnz8vizm.fsf@igel.home> <87y2jouw8k.fsf@mpe.ellerman.id.au>
        <87v9esaxlv.fsf@igel.home>
        <20201030140047.Horde.TJJqKGzG9vSGbMRNIj-MPg7@messagerie.c-s.fr>
        <87pn4zc0zl.fsf@igel.home>
        <1f8494cd-36db-e3a2-8ea4-28fb976468e7@csgroup.eu>
        <875z6mmfna.fsf@depni.sinp.msu.ru>
        <5acd7caf-99e9-9cb5-ed24-578d2e0a5ee1@csgroup.eu>
        <871rh9mu4e.fsf@depni.sinp.msu.ru> <87wnz1lbce.fsf@depni.sinp.msu.ru>
Date:   Wed, 04 Nov 2020 17:53:01 +0300
In-Reply-To: <87wnz1lbce.fsf@depni.sinp.msu.ru> (Serge Belyshev's message of
        "Wed, 04 Nov 2020 12:28:49 +0300")
Message-ID: <87pn4tkwc2.fsf@depni.sinp.msu.ru>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/28.0.50 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk

>>> To be sure we are not in front of a long lasting bug, could you try
>>> CONFIG_KASAN=y on v5.9 ?
>>
>> Indeed it started to fail somewhere between v5.6 and v5.7.
>>
>> v5.7 fails early with few messages on the console with reboot, v5.8 and
>> later hang right at bootloader.
>>
>> I'm bisecting now.
>
> (side note: I tried FB_OF=y instead of DRM_RADEON + DRM_FBDEV_* to speed
> up bisection and it turns out in that configuration KASAN never worked,
> down to commit 305d600123046, hanging right after bootloader or even
> with invalid access in the bootloader itself).

My bisection ended up nowhere (at net-next merge with 2k commits), and
given the above failure with unrelated configuration change, I conclude
that KASAN=y was always broken on this box.