From: Josh Poimboeuf <jpoimboe@kernel.org>
To: x86@kernel.org
Cc: linux-kernel@vger.kernel.org,
Thomas Gleixner <tglx@linutronix.de>,
Borislav Petkov <bp@alien8.de>,
Peter Zijlstra <peterz@infradead.org>,
Pawan Gupta <pawan.kumar.gupta@linux.intel.com>,
Waiman Long <longman@redhat.com>,
Dave Hansen <dave.hansen@linux.intel.com>,
Ingo Molnar <mingo@redhat.com>,
Linus Torvalds <torvalds@linux-foundation.org>,
Michael Ellerman <mpe@ellerman.id.au>,
linuxppc-dev@lists.ozlabs.org,
Andrew Cooper <andrew.cooper3@citrix.com>,
Mark Rutland <mark.rutland@arm.com>,
"Kirill A . Shutemov" <kirill@shutemov.name>
Subject: [PATCH v3 2/6] x86/uaccess: Avoid barrier_nospec() in 64-bit __get_user()
Date: Mon, 28 Oct 2024 18:56:15 -0700 [thread overview]
Message-ID: <82cbb9983fef5ecf6f1cb33661e977172d40a7e6.1730166635.git.jpoimboe@kernel.org> (raw)
In-Reply-To: <cover.1730166635.git.jpoimboe@kernel.org>
The barrier_nospec() in 64-bit __get_user() is slow. Instead use
pointer masking to force the user pointer to all 1's if a previous
access_ok() mispredicted true for an invalid address.
Note that for safety on some AMD CPUs, this relies on recent commit
86e6b1547b3d ("x86: fix user address masking non-canonical speculation
issue").
Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
---
arch/x86/lib/getuser.S | 24 ++++++++++++++++++++----
1 file changed, 20 insertions(+), 4 deletions(-)
diff --git a/arch/x86/lib/getuser.S b/arch/x86/lib/getuser.S
index 4357ec2a0bfc..998d5be6b794 100644
--- a/arch/x86/lib/getuser.S
+++ b/arch/x86/lib/getuser.S
@@ -112,8 +112,12 @@ EXPORT_SYMBOL(__get_user_8)
/* .. and the same for __get_user, just without the range checks */
SYM_FUNC_START(__get_user_nocheck_1)
- ASM_STAC
+#ifdef CONFIG_X86_64
+ check_range size=1
+#else
ASM_BARRIER_NOSPEC
+#endif
+ ASM_STAC
UACCESS movzbl (%_ASM_AX),%edx
xor %eax,%eax
ASM_CLAC
@@ -122,8 +126,12 @@ SYM_FUNC_END(__get_user_nocheck_1)
EXPORT_SYMBOL(__get_user_nocheck_1)
SYM_FUNC_START(__get_user_nocheck_2)
- ASM_STAC
+#ifdef CONFIG_X86_64
+ check_range size=2
+#else
ASM_BARRIER_NOSPEC
+#endif
+ ASM_STAC
UACCESS movzwl (%_ASM_AX),%edx
xor %eax,%eax
ASM_CLAC
@@ -132,8 +140,12 @@ SYM_FUNC_END(__get_user_nocheck_2)
EXPORT_SYMBOL(__get_user_nocheck_2)
SYM_FUNC_START(__get_user_nocheck_4)
- ASM_STAC
+#ifdef CONFIG_X86_64
+ check_range size=4
+#else
ASM_BARRIER_NOSPEC
+#endif
+ ASM_STAC
UACCESS movl (%_ASM_AX),%edx
xor %eax,%eax
ASM_CLAC
@@ -142,8 +154,12 @@ SYM_FUNC_END(__get_user_nocheck_4)
EXPORT_SYMBOL(__get_user_nocheck_4)
SYM_FUNC_START(__get_user_nocheck_8)
- ASM_STAC
+#ifdef CONFIG_X86_64
+ check_range size=8
+#else
ASM_BARRIER_NOSPEC
+#endif
+ ASM_STAC
#ifdef CONFIG_X86_64
UACCESS movq (%_ASM_AX),%rdx
#else
--
2.47.0
next prev parent reply other threads:[~2024-10-29 1:56 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-10-29 1:56 [PATCH v3 0/6] x86/uaccess: avoid barrier_nospec() Josh Poimboeuf
2024-10-29 1:56 ` [PATCH v3 1/6] x86/uaccess: Avoid barrier_nospec() in 64-bit copy_from_user() Josh Poimboeuf
2024-10-29 8:13 ` Kirill A . Shutemov
2024-10-30 2:03 ` Linus Torvalds
2024-10-30 4:59 ` Josh Poimboeuf
2024-10-29 1:56 ` Josh Poimboeuf [this message]
2024-10-29 3:27 ` [PATCH v3 2/6] x86/uaccess: Avoid barrier_nospec() in 64-bit __get_user() Josh Poimboeuf
2024-11-08 17:12 ` David Laight
2024-11-15 23:06 ` 'Josh Poimboeuf'
2024-11-16 1:27 ` Linus Torvalds
2024-11-16 21:38 ` David Laight
2024-11-16 23:08 ` Linus Torvalds
2024-11-21 21:40 ` Josh Poimboeuf
2024-11-21 22:16 ` Linus Torvalds
2024-11-22 0:12 ` Josh Poimboeuf
2024-11-22 1:02 ` Linus Torvalds
2024-11-22 3:11 ` Josh Poimboeuf
2024-11-22 3:57 ` Linus Torvalds
2024-11-22 9:06 ` Christophe Leroy
2024-11-22 18:16 ` Linus Torvalds
2024-11-22 19:13 ` Linus Torvalds
2024-11-22 19:35 ` Linus Torvalds
2024-11-24 16:11 ` David Laight
2024-11-24 18:16 ` Linus Torvalds
2024-11-22 9:38 ` David Laight
2024-10-29 1:56 ` [PATCH v3 3/6] x86/uaccess: Avoid barrier_nospec() in 32-bit copy_from_user() Josh Poimboeuf
2024-10-29 1:56 ` [PATCH v3 4/6] x86/uaccess: Convert 32-bit get_user() to unconditional pointer masking Josh Poimboeuf
2024-10-29 1:56 ` [PATCH v3 5/6] x86/uaccess: Avoid barrier_nospec() in 32-bit __get_user() Josh Poimboeuf
2024-10-29 1:56 ` [PATCH v3 6/6] x86/uaccess: Converge [__]get_user() implementations Josh Poimboeuf
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=82cbb9983fef5ecf6f1cb33661e977172d40a7e6.1730166635.git.jpoimboe@kernel.org \
--to=jpoimboe@kernel.org \
--cc=andrew.cooper3@citrix.com \
--cc=bp@alien8.de \
--cc=dave.hansen@linux.intel.com \
--cc=kirill@shutemov.name \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=longman@redhat.com \
--cc=mark.rutland@arm.com \
--cc=mingo@redhat.com \
--cc=mpe@ellerman.id.au \
--cc=pawan.kumar.gupta@linux.intel.com \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--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
all inboxes | Powered by JetHome®