Re: [PATCH 0/8] linker-section array fix and clean ups

From: Jessica Yu <jeyu@kernel.org>
To: Johan Hovold <johan@kernel.org>
Cc: linux-kernel@vger.kernel.org,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Rob Herring <robh+dt@kernel.org>,
	Frank Rowand <frowand.list@gmail.com>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Arnd Bergmann <arnd@arndb.de>,
	Geert Uytterhoeven <geert@linux-m68k.org>,
	Dmitry Torokhov <dmitry.torokhov@gmail.com>,
	David Miller <davem@davemloft.net>,
	Jakub Jelinek <jakub@redhat.com>,
	Peter Zijlstra <peterz@infradead.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	Steven Rostedt <rostedt@goodmis.org>,
	Daniel Kurtz <djkurtz@chromium.org>,
	linux-arch@vger.kernel.org, linux-m68k@lists.linux-m68k.org
Subject: Re: [PATCH 0/8] linker-section array fix and clean ups
Date: Wed, 25 Nov 2020 15:51:20 +0100	[thread overview]
Message-ID: <20201125145118.GA32446@linux-8ccs> (raw)
In-Reply-To: <X7uRZUY+2L9Yg9wt@localhost>

+++ Johan Hovold [23/11/20 11:39 +0100]:
>On Fri, Nov 13, 2020 at 03:18:36PM +0100, Johan Hovold wrote:
>> On Wed, Nov 11, 2020 at 04:47:16PM +0100, Jessica Yu wrote:
>>
>> > Thanks for providing the links and references. Your explanation and
>> > this reply from Jakub [1] clarified things for me. I was not aware of
>> > the distinction gcc made between aligned attributes on types vs. on
>> > variables. So from what I understand now, gcc suppresses the
>> > optimization when the alignment is specified in the variable
>> > declaration, but not necessarily when the aligned attribute is just on
>> > the type.
>> >
>> > Even though it's been in use for a long time, I think it would be
>> > really helpful if this gcc quirk was explained just a bit more in the
>> > patch changelogs, especially since this is undocumented behavior.
>> > I found the explanation in [1] (as well as in your cover letter) to be
>> > sufficient. Maybe something like "GCC suppresses any optimizations
>> > increasing alignment when the alignment is specified in the variable
>> > declaration, as opposed to just on the type definition. Therefore,
>> > explicitly specify type alignment when declaring entries to prevent
>> > gcc from increasing alignment."
>>
>> Sure, I can try to expand the commit messages a bit.
>
>I've amended the commit messages of the relevant patches to make it more
>clear that the optimisation can be suppressed by specifying alignment
>when declaring variables, but without making additional claims about the
>type attribute. I hope the result is acceptable to you.
>
>Perhaps you can include a lore link to the patches when applying so that
>this thread can be found easily if needed.

Hi Johan,

Good idea, I've included a link to this thread for each patch.
I've queued up patches 3, 4, 6, 7, 8 for testing before pushing them
out to modules-next.

Thanks!

Jessica