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;
Date:   Fri, 15 Jun 2018 13:37:19 -0400 (EDT)
From:   Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
To:     James Hogan <jhogan@kernel.org>
Cc:     Paul Burton <paul.burton@mips.com>,
        linux-mips <linux-mips@linux-mips.org>,
        Peter Zijlstra <peterz@infradead.org>,
        linux-kernel <linux-kernel@vger.kernel.org>,
        Boqun Feng <boqun.feng@gmail.com>,
        "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>,
        Ralf Baechle <ralf@linux-mips.org>
Message-ID: <1373980460.14412.1529084239034.JavaMail.zimbra@efficios.com>
In-Reply-To: <20180615105809.GB7603@jamesdev>
References: <20180614235211.31357-1-paul.burton@mips.com> <20180614235211.31357-5-paul.burton@mips.com> <20180615105809.GB7603@jamesdev>
Subject: Re: [PATCH 4/4] rseq/selftests: Implement MIPS support
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Thread-Topic: rseq/selftests: Implement MIPS support
Thread-Index: sJDkoLXztL0rd01o4PFqQYpfV3YGlA==
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

----- On Jun 15, 2018, at 6:58 AM, James Hogan jhogan@kernel.org wrote:

> On Thu, Jun 14, 2018 at 04:52:10PM -0700, Paul Burton wrote:
>> +#define __RSEQ_ASM_DEFINE_TABLE(version, flags,	start_ip,			\
> 
> Nit: technically all these \'s are on 81st column...
> 
>> +#define __RSEQ_ASM_DEFINE_ABORT(table_label, label, teardown,			\
>> +				abort_label, version, flags,			\
>> +				start_ip, post_commit_offset, abort_ip)		\
>> +		".balign 32\n\t"						\
> 
> ARM doesn't do this for DEFINE_ABORT. Is it intentional that we do for
> MIPS?

Given that include/uapi/linux/rseq.h declares struct rseq_cs as
__attribute__((aligned(4 * sizeof(__u64)))), and considering this
comment:

/*
 * struct rseq_cs is aligned on 4 * 8 bytes to ensure it is always
 * contained within a single cache-line. It is usually declared as
 * link-time constant data.
 */

The .balign 32 is the right thing to do here. I will add a .balign 32
to ARM selftests code as well.

Thanks,

Mathieu

> 
> Thanks
> James

-- 
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com