From: David Laight <David.Laight@ACULAB.COM>
To: 'Andy Lutomirski' <luto@amacapital.net>, Ingo Molnar <mingo@kernel.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>,
Linus Torvalds <torvalds@linux-foundation.org>,
Thomas Gleixner <tglx@linutronix.de>, X86 ML <x86@kernel.org>,
LKML <linux-kernel@vger.kernel.org>,
Borislav Petkov <bp@alien8.de>
Subject: RE: [RFC][PATCH 01/22] x86 user stack frame reads: switch to explicit __get_user()
Date: Sun, 29 Mar 2020 17:41:12 +0000 [thread overview]
Message-ID: <489c9af889954649b3453e350bab6464@AcuMS.aculab.com> (raw)
In-Reply-To: <CALCETrX=nXN14fqu-yEMGwwN-vdSz=-0C3gcOMucmxrCUpevdA@mail.gmail.com>
From: Andy Lutomirski
> Sent: 29 March 2020 17:51
..
> My incliniation is to just get rid of the __get_user()-style APIs.
> There shouldn't be any __get_user() calls that can't be directly
> replaced by get_user(), and a single integer comparison is not that
> expensive. On SMAP systems, the speedup of __get_user vs get_user is
> negligible.
On x86-64 (at least) __get_user() is inlined but get_user() isn't.
Since get_user() has to return two values, one will always be
a (usually) on-stack real memory location rather than a register.
For frequently use code paths this may be measurable.
I'm thinking of things like epoll_wait() writing out events.
(although that is a put_user() loop...)
It may be worth implementing get_user() as an inline
function that writes the result of access_ok() to a
'by reference' parameter and then returns the value
from an 'real' __get_user() function.
The compiler will then optimise away the memory reference.
David
-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)
next prev parent reply other threads:[~2020-03-29 17:41 UTC|newest]
Thread overview: 71+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-03-23 18:36 [RFC][PATCHSET] x86 uaccess cleanups Al Viro
2020-03-23 18:37 ` [RFC][PATCH 01/22] x86 user stack frame reads: switch to explicit __get_user() Al Viro
2020-03-23 18:37 ` [RFC][PATCH 02/22] x86 kvm page table walks: " Al Viro
2020-03-23 18:38 ` [RFC][PATCH 03/22] x86: switch sigframe sigset handling to explict __get_user()/__put_user() Al Viro
2020-03-23 18:38 ` [RFC][PATCH 04/22] x86: get rid of small constant size cases in raw_copy_{to,from}_user() Al Viro
2020-03-23 18:38 ` [RFC][PATCH 05/22] vm86: get rid of get_user_ex() use Al Viro
2020-03-23 18:38 ` [RFC][PATCH 06/22] x86: get rid of get_user_ex() in ia32_restore_sigcontext() Al Viro
2020-03-23 18:38 ` [RFC][PATCH 07/22] x86: get rid of get_user_ex() in restore_sigcontext() Al Viro
2020-03-23 18:38 ` [RFC][PATCH 08/22] x86: kill get_user_{try,catch,ex} Al Viro
2020-03-23 18:38 ` [RFC][PATCH 09/22] x86: switch save_v86_state() to unsafe_put_user() Al Viro
2020-03-23 18:38 ` [RFC][PATCH 10/22] x86: switch setup_sigcontext() " Al Viro
2020-03-23 18:38 ` [RFC][PATCH 11/22] x86: switch ia32_setup_sigcontext() " Al Viro
2020-03-23 18:38 ` [RFC][PATCH 12/22] x86: get rid of put_user_try in {ia32,x32}_setup_rt_frame() Al Viro
2020-03-23 18:38 ` [RFC][PATCH 13/22] x86: ia32_setup_sigcontext(): lift user_access_{begin,end}() into the callers Al Viro
2020-03-23 18:53 ` Linus Torvalds
2020-03-23 21:42 ` Al Viro
2020-03-23 18:38 ` [RFC][PATCH 14/22] x86: ia32_setup_frame(): consolidate uaccess areas Al Viro
2020-03-23 18:38 ` [RFC][PATCH 15/22] x86: ia32_setup_rt_frame(): " Al Viro
2020-03-23 18:38 ` [RFC][PATCH 16/22] x86: get rid of put_user_try in __setup_rt_frame() (both 32bit and 64bit) Al Viro
2020-03-23 18:38 ` [RFC][PATCH 17/22] x86: setup_sigcontext(): list user_access_{begin,end}() into callers Al Viro
2020-03-23 18:56 ` Linus Torvalds
2020-03-23 18:38 ` [RFC][PATCH 18/22] x86: __setup_frame(): consolidate uaccess areas Al Viro
2020-03-23 18:38 ` [RFC][PATCH 19/22] x86: __setup_rt_frame(): " Al Viro
2020-03-23 18:38 ` [RFC][PATCH 20/22] x86: x32_setup_rt_frame(): " Al Viro
2020-03-23 18:38 ` [RFC][PATCH 21/22] x86: unsafe_put_... macros for sigcontext and sigmask Al Viro
2020-03-23 18:38 ` [RFC][PATCH 22/22] kill uaccess_try() Al Viro
2020-03-24 15:15 ` [RFC][PATCH 01/22] x86 user stack frame reads: switch to explicit __get_user() Peter Zijlstra
2020-03-28 10:48 ` Ingo Molnar
2020-03-28 11:59 ` Al Viro
2020-03-29 9:26 ` Ingo Molnar
2020-03-29 16:50 ` Andy Lutomirski
2020-03-29 17:05 ` Linus Torvalds
2020-03-29 17:41 ` David Laight [this message]
2020-03-29 17:56 ` Linus Torvalds
2020-03-29 18:03 ` David Laight
2020-03-29 18:16 ` Linus Torvalds
2020-03-29 18:32 ` David Laight
2020-03-29 18:55 ` Linus Torvalds
2020-03-29 21:21 ` Andy Lutomirski
2020-03-29 22:06 ` Linus Torvalds
2020-03-29 22:12 ` Linus Torvalds
2020-03-29 18:16 ` Al Viro
2020-03-29 18:19 ` Linus Torvalds
2020-03-29 17:57 ` Al Viro
2020-03-30 15:54 ` David Laight
2020-03-23 19:16 ` [RFC][PATCHSET] x86 uaccess cleanups Linus Torvalds
2020-03-27 2:24 ` [RFC][PATCHSET v2] " Al Viro
2020-03-27 2:26 ` Al Viro
2020-03-27 2:30 ` Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 01/22] x86 user stack frame reads: switch to explicit __get_user() Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 02/22] x86 kvm page table walks: " Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 03/22] x86: switch sigframe sigset handling to explict __get_user()/__put_user() Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 04/22] x86: get rid of small constant size cases in raw_copy_{to,from}_user() Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 05/22] vm86: get rid of get_user_ex() use Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 06/22] x86: get rid of get_user_ex() in ia32_restore_sigcontext() Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 07/22] x86: get rid of get_user_ex() in restore_sigcontext() Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 08/22] x86: kill get_user_{try,catch,ex} Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 09/22] x86: switch save_v86_state() to unsafe_put_user() Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 10/22] x86: switch setup_sigcontext() " Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 11/22] x86: switch ia32_setup_sigcontext() " Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 12/22] x86: get rid of put_user_try in {ia32,x32}_setup_rt_frame() Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 13/22] x86: ia32_setup_sigcontext(): lift user_access_{begin,end}() into the callers Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 14/22] x86: ia32_setup_frame(): consolidate uaccess areas Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 15/22] x86: ia32_setup_rt_frame(): " Al Viro
2020-03-27 2:31 ` [RFC][PATCH v2 16/22] x86: get rid of put_user_try in __setup_rt_frame() (both 32bit and 64bit) Al Viro
2020-03-27 2:32 ` [RFC][PATCH v2 17/22] x86: setup_sigcontext(): list user_access_{begin,end}() into callers Al Viro
2020-03-27 2:32 ` [RFC][PATCH v2 18/22] x86: __setup_frame(): consolidate uaccess areas Al Viro
2020-03-27 2:32 ` [RFC][PATCH v2 19/22] x86: __setup_rt_frame(): " Al Viro
2020-03-27 2:32 ` [RFC][PATCH v2 20/22] x86: x32_setup_rt_frame(): " Al Viro
2020-03-27 2:32 ` [RFC][PATCH v2 21/22] x86: unsafe_put-style macro for sigmask Al Viro
2020-03-27 2:32 ` [RFC][PATCH v2 22/22] kill uaccess_try() Al Viro
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=489c9af889954649b3453e350bab6464@AcuMS.aculab.com \
--to=david.laight@aculab.com \
--cc=bp@alien8.de \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@amacapital.net \
--cc=mingo@kernel.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--cc=viro@zeniv.linux.org.uk \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).