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;
MIME-Version: 1.0
References: <20210725104618.365790-1-geert@linux-m68k.org>
In-Reply-To: <20210725104618.365790-1-geert@linux-m68k.org>
From:   Arnd Bergmann <arnd@arndb.de>
Date:   Sun, 25 Jul 2021 17:26:46 +0200
Message-ID: <CAK8P3a2qD3LFic8L_3-KCnuVL6J8M_SLFsdj0c-n258MsUPKuQ@mail.gmail.com>
Subject: Re: [PATCH] m68k: Fix asm register constraints for atomic ops
To:     Geert Uytterhoeven <geert@linux-m68k.org>
Cc:     Greg Ungerer <gerg@uclinux.org>, Will Deacon <will@kernel.org>,
        Peter Zijlstra <peterz@infradead.org>,
        Boqun Feng <boqun.feng@gmail.com>,
        Brendan Jackman <jackmanb@google.com>,
        kernel test robot <lkp@intel.com>,
        Arnd Bergmann <arnd@arndb.de>,
        Alexander Viro <viro@zeniv.linux.org.uk>,
        linux-m68k <linux-m68k@lists.linux-m68k.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Sun, Jul 25, 2021 at 12:46 PM Geert Uytterhoeven
<geert@linux-m68k.org> wrote:
>
> Depending on register assignment by the compiler:
>
>     {standard input}:3084: Error: operands mismatch -- statement `andl %a1,%d1' ignored
>     {standard input}:3145: Error: operands mismatch -- statement `orl %a1,%d1' ignored
>     {standard input}:3195: Error: operands mismatch -- statement `eorl %a1,%d1' ignored
>
> Indeed, the first operand must not be an address register.  Fix this by
> adjusting the register constraint from "g" (general purpose register) to
> "d" (data register).
>
> Fixes: e39d88ea3ce4a471 ("locking/atomic, arch/m68k: Implement atomic_fetch_{add,sub,and,or,xor}()")
> Fixes: d839bae4269aea46 ("locking,arch,m68k: Fold atomic_ops")
> Reported-by: kernel test robot <lkp@intel.com>
> Reported-by: Arnd Bergmann <arnd@arndb.de>
> Reported-by: Alexander Viro <viro@zeniv.linux.org.uk>
> Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
> ---
> Personally, I have never seen this failure in an 68020+ build, but I can
> reproduce it on Coldfire with the config provided by lkp (with bogus
> CONFIG_RMW_INSNS=y).

This fixed the problem for me in the 68020 build with all compiler versions.

Tested-by: Arnd Bergmann <arnd@arndb.de>