linux-kernel - Re: [patch V5 10/12] futex: Convert to get/put_user

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAHk-=wjZ39CSZwN3W6n5kSAqL-OhJghygh3-dRsrJKpOa9nTwg@mail.gmail.com>
Date: Tue, 28 Oct 2025 09:13:36 -0700
From: Linus Torvalds <torvalds@...ux-foundation.org>
To: Thomas Gleixner <tglx@...utronix.de>
Cc: Mathieu Desnoyers <mathieu.desnoyers@...icios.com>, LKML <linux-kernel@...r.kernel.org>, 
	Peter Zijlstra <peterz@...radead.org>, Darren Hart <dvhart@...radead.org>, 
	Davidlohr Bueso <dave@...olabs.net>, André Almeida <andrealmeid@...lia.com>, 
	kernel test robot <lkp@...el.com>, Russell King <linux@...linux.org.uk>, 
	linux-arm-kernel@...ts.infradead.org, x86@...nel.org, 
	Madhavan Srinivasan <maddy@...ux.ibm.com>, Michael Ellerman <mpe@...erman.id.au>, 
	Nicholas Piggin <npiggin@...il.com>, Christophe Leroy <christophe.leroy@...roup.eu>, 
	linuxppc-dev@...ts.ozlabs.org, Paul Walmsley <pjw@...nel.org>, 
	Palmer Dabbelt <palmer@...belt.com>, linux-riscv@...ts.infradead.org, 
	Heiko Carstens <hca@...ux.ibm.com>, Christian Borntraeger <borntraeger@...ux.ibm.com>, 
	Sven Schnelle <svens@...ux.ibm.com>, linux-s390@...r.kernel.org, 
	Andrew Cooper <andrew.cooper3@...rix.com>, David Laight <david.laight.linux@...il.com>, 
	Julia Lawall <Julia.Lawall@...ia.fr>, Nicolas Palix <nicolas.palix@...g.fr>, 
	Alexander Viro <viro@...iv.linux.org.uk>, Christian Brauner <brauner@...nel.org>, Jan Kara <jack@...e.cz>, 
	linux-fsdevel@...r.kernel.org
Subject: Re: [patch V5 10/12] futex: Convert to get/put_user_inline()

On Tue, 28 Oct 2025 at 08:56, Thomas Gleixner <tglx@...utronix.de> wrote:
>
> There was not justification for the open coded inline either and
> converting it to get/put must be a completely seperate change.

Actually, there's some justification in the original: see commit
43a43faf5376 ("futex: improve user space accesses") which talks about
the original impetus for it all: avoiding the very expensive barrier
in __get_user(), and how __get_user() itself couldn't be fixed.

So then it was converted to the modern user access helpers - including
address masking - and the inlining was mostly incidental to that, but
the commit message does point out that it actually makes the address
generation a bit cleaner in addition to avoiding the function call.

But I doubt that the extra instructions are all that noticeable.

That said - this code *is* in a very hot path on some loads, so it is
entirely possible that the inlining here is noticeable. I$ patterns in
particular can be a real thing.

(There was an additional issue of just making those user accesses -
get, put and cmpxchg - look a bit more similar)

           Linus