Received-SPF: pass (google.com: domain of linux-kernel+bounces-121540-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:4601:e00::3 as permitted sender) client-ip=2604:1380:4601:e00::3;
Message-ID: <c63150d8-a479-4d75-a826-3f87b1a792b9@sifive.com>
Date: Wed, 27 Mar 2024 10:32:54 -0500
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH 0/7] riscv: Various text patching improvements
To: Andrea Parri <parri.andrea@gmail.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>, linux-riscv@lists.infradead.org,
 linux-kernel@vger.kernel.org, Ard Biesheuvel <ardb@kernel.org>,
 Daniel Borkmann <daniel@iogearbox.net>, Jason Baron <jbaron@akamai.com>,
 Josh Poimboeuf <jpoimboe@kernel.org>, Peter Zijlstra <peterz@infradead.org>,
 Steven Rostedt <rostedt@goodmis.org>, bpf@vger.kernel.org,
 Alexandre Ghiti <alexghiti@rivosinc.com>, =?UTF-8?B?QmrDtnJuIFTDtnBlbA==?=
 <bjorn@rivosinc.com>
References: <20240212025529.1971876-1-samuel.holland@sifive.com>
 <ZctnfZWWO3HCiXe5@andrea>
Content-Language: en-US
From: Samuel Holland <samuel.holland@sifive.com>
In-Reply-To: <ZctnfZWWO3HCiXe5@andrea>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit

Hi Andrea,

On 2024-02-13 6:58 AM, Andrea Parri wrote:
> On Sun, Feb 11, 2024 at 06:55:11PM -0800, Samuel Holland wrote:
>> Here are a few changes to minimize calls to stop_machine() and
>> flush_icache_*() in the various text patching functions, as well as
>> to simplify the code.
>>
>>
>> Samuel Holland (7):
>>   riscv: jump_label: Batch icache maintenance
>>   riscv: jump_label: Simplify assembly syntax
>>   riscv: kprobes: Use patch_text_nosync() for insn slots
>>   riscv: Simplify text patching loops
>>   riscv: Pass patch_text() the length in bytes
>>   riscv: Use offset_in_page() in text patching functions
>>   riscv: Remove extra variable in patch_text_nosync()
> 
> This does look like a nice clean-up.  Just curious (a "teach me"-like question),
> how did you test these changes? kselftests, micro-benchmarks, other?

For all my patches, I do boot testing on various physical boards (Unmatched, D1,
some internal hardware), plus some standard internal workloads. For
performance-related patches, I run microbenchmarks or whole-system benchmarks
(e.g. UnixBench). For this series specifically, I didn't do any extra
benchmarking, as all of the functional changes should be as fast or faster by
virtue of simply doing less work.

> BTW, I recall a parallel work from Alex and Bjorn [1] that might have some minor
> conflict with these changes; + both of them to Cc: for further sync.

As suggested by Alex, v2 of this series will be based on the latest version of
that patch.

Regards,
Samuel

> 
>   Andrea
> 
> [1] https://lore.kernel.org/lkml/20240206204607.527195-1-alexghiti@rivosinc.com/