From mboxrd@z Thu Jan  1 00:00:00 1970
From: mathieu.desnoyers@efficios.com (Mathieu Desnoyers)
Date: Thu, 28 Jun 2018 16:50:40 -0400 (EDT)
Subject: [PATCH 3/3] rseq/selftests: Add support for arm64
In-Reply-To: <20180628164700.GD10751@arm.com>
References: <1529949285-11013-1-git-send-email-will.deacon@arm.com>
 <1529949285-11013-4-git-send-email-will.deacon@arm.com>
 <501929863.3051.1529950210436.JavaMail.zimbra@efficios.com>
 <20180626151427.GF23375@arm.com>
 <1763491947.3520.1530029512923.JavaMail.zimbra@efficios.com>
 <20180628164700.GD10751@arm.com>
Message-ID: <176714835.9396.1530219040151.JavaMail.zimbra@efficios.com>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

----- On Jun 28, 2018, at 12:47 PM, Will Deacon will.deacon at arm.com wrote:

> Hi Mathieu,
> 
> On Tue, Jun 26, 2018 at 12:11:52PM -0400, Mathieu Desnoyers wrote:
>> ----- On Jun 26, 2018, at 11:14 AM, Will Deacon will.deacon at arm.com wrote:
>> > On Mon, Jun 25, 2018 at 02:10:10PM -0400, Mathieu Desnoyers wrote:
>> >> I notice you are using the instructions
>> >> 
>> >>   adrp
>> >>   add
>> >>   str
>> >> 
>> >> to implement RSEQ_ASM_STORE_RSEQ_CS(). Did you compare
>> >> performance-wise with an approach using a literal pool
>> >> near the instruction pointer like I did on arm32 ?
>> > 
>> > I didn't, no. Do you have a benchmark to hand so I can give this a go?
>> 
>> see tools/testing/selftests/rseq/param_test_benchmark --help
>> 
>> It's a stripped-down version of param_test, without all the code for
>> delay loops and testing checks.
>> 
>> Example use for counter increment with 4 threads, doing 5G counter
>> increments per thread:
>> 
>> time ./param_test_benchmark -T i -t 4 -r 5000000000
> 
> Thanks. I ran that on a few arm64 systems I have access to, with three
> configurations of the selftest:
> 
> 1. As I posted
> 2. With the abort signature and branch in-lined, so as to avoid the CBNZ
>   address limitations in large codebases
> 3. With both the abort handler and the table inlined (i.e. the same thing
>   as 32-bit).
> 
> There isn't a reliably measurable difference between (1) and (2), but I take
> between 12% and 27% hit between (2) and (3).

Those results puzzle me. Do you have the actual code snippets of each
implementation nearby ?

Thanks,

Mathieu

> 
> So I'll post a v2 based on (2).
> 
> Will

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