From: Josh Poimboeuf <jpoimboe@kernel.org>
To: David Kaplan <david.kaplan@amd.com>
Cc: Thomas Gleixner <tglx@linutronix.de>,
Borislav Petkov <bp@alien8.de>,
Peter Zijlstra <peterz@infradead.org>,
Pawan Gupta <pawan.kumar.gupta@linux.intel.com>,
Ingo Molnar <mingo@redhat.com>,
Dave Hansen <dave.hansen@linux.intel.com>,
x86@kernel.org, "H . Peter Anvin" <hpa@zytor.com>,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH v3 21/35] x86/bugs: Determine relevant vulnerabilities based on attack vector controls.
Date: Tue, 11 Feb 2025 10:41:30 -0800 [thread overview]
Message-ID: <20250211184130.2fl4jvdwd4e5y32a@jpoimboe> (raw)
In-Reply-To: <20250108202515.385902-22-david.kaplan@amd.com>
On Wed, Jan 08, 2025 at 02:25:01PM -0600, David Kaplan wrote:
> diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c
> index 88eba8e4c7fb..175dbbf9b06e 100644
> --- a/arch/x86/kernel/cpu/bugs.c
> +++ b/arch/x86/kernel/cpu/bugs.c
> @@ -347,6 +347,75 @@ static void x86_amd_ssb_disable(void)
> wrmsrl(MSR_AMD64_LS_CFG, msrval);
> }
>
> +/*
> + * Returns true if vulnerability should be mitigated based on the
> + * selected attack vector controls
This needs a period.
> + *
> + * See Documentation/admin-guide/hw-vuln/attack_vector_controls.rst
> + */
> +static bool __init should_mitigate_vuln(unsigned int bug)
> +{
> + switch (bug) {
> + /*
> + * The only spectre_v1 mitigations in the kernel are related to
> + * SWAPGS protection on kernel entry. Therefore, protection is
> + * only required for the user->kernel attack vector.
> + */
This comment isn't quite correct, there are things like
array_index_nospec() and barrier_nospec() being used, but those aren't
being controlled by bugs.c. They should at least be mentioned here.
> + case X86_BUG_SPECTRE_V1:
> + return cpu_mitigate_attack_vector(CPU_MITIGATE_USER_KERNEL);
> +
> + /*
> + * Both spectre_v2 and srso may allow user->kernel or
> + * guest->host attacks through branch predictor manipulation.
> + */
I don't think this comment adds anything, the code already makes this
obvious.
> + case X86_BUG_SPECTRE_V2:
> + case X86_BUG_SRSO:
> + return cpu_mitigate_attack_vector(CPU_MITIGATE_USER_KERNEL) ||
> + cpu_mitigate_attack_vector(CPU_MITIGATE_GUEST_HOST);
This needs aligned:
return cpu_mitigate_attack_vector(CPU_MITIGATE_USER_KERNEL) ||
cpu_mitigate_attack_vector(CPU_MITIGATE_GUEST_HOST);
Also, aren't cross-thread attacks possible here, thus the need for
STIBP? More questions about the cross-thread "vector" below, at the
bottom.
> + /*
> + * spectre_v2_user refers to user->user or guest->guest branch
> + * predictor attacks only. Other indirect branch predictor attacks
> + * are covered by the spectre_v2 vulnerability.
> + */
The code is already self-evident IMO, I don't think the comment adds
anything.
> + case X86_BUG_SPECTRE_V2_USER:
> + return cpu_mitigate_attack_vector(CPU_MITIGATE_USER_USER) ||
> + cpu_mitigate_attack_vector(CPU_MITIGATE_GUEST_GUEST);
Another alignment issue.
> +
> + /* L1TF is only possible as a guest->host attack */
That's not quite correct, PTE inversion is also done to protect against
the user->kernel vector.
Also, IIRC the full l1tf mitigation requires disabling SMT, does that
not qualify as CPU_MITIGATE_CROSS_THREAD?
> + case X86_BUG_L1TF:
> + return cpu_mitigate_attack_vector(CPU_MITIGATE_GUEST_HOST);
> +
> + /*
> + * All the vulnerabilities below allow potentially leaking data
> + * across address spaces. Therefore, mitigation is required for
> + * any of these 4 attack vectors.
> + */
> + case X86_BUG_MDS:
> + case X86_BUG_TAA:
> + case X86_BUG_MMIO_STALE_DATA:
> + case X86_BUG_RFDS:
> + case X86_BUG_SRBDS:
> + return cpu_mitigate_attack_vector(CPU_MITIGATE_USER_KERNEL) ||
> + cpu_mitigate_attack_vector(CPU_MITIGATE_GUEST_HOST) ||
> + cpu_mitigate_attack_vector(CPU_MITIGATE_USER_USER) ||
> + cpu_mitigate_attack_vector(CPU_MITIGATE_GUEST_GUEST);
Some of these also require disabling SMT for their complete mitigations?
> + /*
> + * GDS can potentially leak data across address spaces and
> + * threads. Mitigation is required under all attack vectors.
> + */
> + case X86_BUG_GDS:
> + return cpu_mitigate_attack_vector(CPU_MITIGATE_USER_KERNEL) ||
> + cpu_mitigate_attack_vector(CPU_MITIGATE_GUEST_HOST) ||
> + cpu_mitigate_attack_vector(CPU_MITIGATE_USER_USER) ||
> + cpu_mitigate_attack_vector(CPU_MITIGATE_GUEST_GUEST) ||
> + cpu_mitigate_attack_vector(CPU_MITIGATE_CROSS_THREAD);
I'm confused by CPU_MITIGATE_CROSS_THREAD here, as the GDS mitigation
doesn't seem to disable SMT?
Am I just completely misunderstanding the meaning of
CPU_MITIGATE_CROSS_THREAD?
I assumed it's not a vector per se, but rather it means to force nosmt
if one of the other enabled mitigations requires doing so for its "full"
mitigation. But the implementation doesn't seem to match that.
On the other hand if it really is considered to be its own vector, that
doesn't make sense either, as "cross-thread attack" is really a subset
of each of the other vectors. For example, a user->kernel attack can
often be done either via syscall/irq or via cross-thread.
So I'm really confused. Am I missing something?
--
Josh
next prev parent reply other threads:[~2025-02-11 18:41 UTC|newest]
Thread overview: 138+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-01-08 20:24 [PATCH v3 00/35] x86/bugs: Attack " David Kaplan
2025-01-08 20:24 ` [PATCH v3 01/35] x86/bugs: Add X86_BUG_SPECTRE_V2_USER David Kaplan
2025-02-28 11:53 ` [tip: x86/bugs] " tip-bot2 for David Kaplan
2025-01-08 20:24 ` [PATCH v3 02/35] x86/bugs: Relocate mds/taa/mmio/rfds defines David Kaplan
2025-02-28 11:53 ` [tip: x86/bugs] " tip-bot2 for David Kaplan
2025-01-08 20:24 ` [PATCH v3 03/35] x86/bugs: Add AUTO mitigations for mds/taa/mmio/rfds David Kaplan
2025-02-28 11:53 ` [tip: x86/bugs] " tip-bot2 for David Kaplan
2025-01-08 20:24 ` [PATCH v3 04/35] x86/bugs: Restructure mds mitigation David Kaplan
2025-02-10 16:13 ` Brendan Jackman
2025-02-10 17:17 ` Kaplan, David
2025-02-10 17:28 ` Brendan Jackman
2025-02-10 22:25 ` Josh Poimboeuf
2025-02-10 22:33 ` Kaplan, David
2025-01-08 20:24 ` [PATCH v3 05/35] x86/bugs: Restructure taa mitigation David Kaplan
2025-02-10 16:24 ` Brendan Jackman
2025-02-10 17:19 ` Kaplan, David
2025-02-10 22:50 ` Josh Poimboeuf
2025-02-11 17:17 ` Kaplan, David
2025-02-11 19:17 ` Josh Poimboeuf
2025-01-08 20:24 ` [PATCH v3 06/35] x86/bugs: Restructure mmio mitigation David Kaplan
2025-02-10 16:42 ` Brendan Jackman
2025-02-10 17:22 ` Kaplan, David
2025-02-10 17:35 ` Brendan Jackman
2025-02-10 23:29 ` Josh Poimboeuf
2025-02-11 20:35 ` Kaplan, David
2025-02-11 23:18 ` Josh Poimboeuf
2025-02-12 17:28 ` Kaplan, David
2025-02-12 23:16 ` Josh Poimboeuf
2025-02-19 18:20 ` Borislav Petkov
2025-02-21 21:48 ` Kaplan, David
2025-01-08 20:24 ` [PATCH v3 07/35] x86/bugs: Restructure rfds mitigation David Kaplan
2025-02-10 23:36 ` Josh Poimboeuf
2025-02-11 22:49 ` Kaplan, David
2025-01-08 20:24 ` [PATCH v3 08/35] x86/bugs: Remove md_clear_*_mitigation() David Kaplan
2025-01-08 20:24 ` [PATCH v3 09/35] x86/bugs: Restructure srbds mitigation David Kaplan
2025-02-10 23:44 ` Josh Poimboeuf
2025-02-11 22:59 ` Kaplan, David
2025-01-08 20:24 ` [PATCH v3 10/35] x86/bugs: Restructure gds mitigation David Kaplan
2025-02-10 17:06 ` Brendan Jackman
2025-02-10 17:27 ` Kaplan, David
2025-02-10 17:40 ` Brendan Jackman
2025-02-10 23:52 ` Josh Poimboeuf
2025-02-12 15:36 ` Kaplan, David
2025-01-08 20:24 ` [PATCH v3 11/35] x86/bugs: Restructure spectre_v1 mitigation David Kaplan
2025-01-08 20:24 ` [PATCH v3 12/35] x86/bugs: Restructure retbleed mitigation David Kaplan
2025-01-09 5:22 ` Pawan Gupta
2025-01-09 15:26 ` Kaplan, David
2025-01-09 16:40 ` Pawan Gupta
2025-01-09 16:42 ` Kaplan, David
2025-01-10 18:45 ` David Laight
2025-01-10 20:30 ` Pawan Gupta
2025-01-10 20:35 ` Borislav Petkov
2025-02-10 18:35 ` Brendan Jackman
2025-02-10 20:50 ` Kaplan, David
2025-02-11 0:10 ` Josh Poimboeuf
2025-02-24 15:45 ` Borislav Petkov
2025-02-24 15:59 ` Kaplan, David
2025-01-08 20:24 ` [PATCH v3 13/35] x86/bugs: Restructure spectre_v2_user mitigation David Kaplan
2025-02-11 0:53 ` Josh Poimboeuf
2025-02-12 15:59 ` Kaplan, David
2025-02-12 21:35 ` Josh Poimboeuf
2025-01-08 20:24 ` [PATCH v3 14/35] x86/bugs: Restructure bhi mitigation David Kaplan
2025-01-08 20:24 ` [PATCH v3 15/35] x86/bugs: Restructure spectre_v2 mitigation David Kaplan
2025-02-11 1:07 ` Josh Poimboeuf
2025-02-12 16:40 ` Kaplan, David
2025-01-08 20:24 ` [PATCH v3 16/35] x86/bugs: Restructure ssb mitigation David Kaplan
2025-02-11 1:10 ` Josh Poimboeuf
2025-02-12 16:45 ` Kaplan, David
2025-01-08 20:24 ` [PATCH v3 17/35] x86/bugs: Restructure l1tf mitigation David Kaplan
2025-02-11 1:21 ` Josh Poimboeuf
2025-02-12 16:47 ` Kaplan, David
2025-01-08 20:24 ` [PATCH v3 18/35] x86/bugs: Restructure srso mitigation David Kaplan
2025-02-11 16:39 ` Josh Poimboeuf
2025-02-12 17:01 ` Kaplan, David
2025-01-08 20:24 ` [PATCH v3 19/35] Documentation/x86: Document the new attack vector controls David Kaplan
2025-02-11 16:43 ` Josh Poimboeuf
2025-02-11 16:57 ` Kaplan, David
2025-01-08 20:25 ` [PATCH v3 20/35] x86/bugs: Define attack vectors David Kaplan
2025-02-11 18:07 ` Josh Poimboeuf
2025-02-12 17:20 ` Kaplan, David
2025-02-17 17:33 ` Kaplan, David
2025-02-17 20:19 ` Josh Poimboeuf
2025-02-17 20:38 ` Kaplan, David
2025-02-17 23:39 ` Josh Poimboeuf
2025-02-18 2:24 ` Kaplan, David
2025-02-18 7:05 ` Josh Poimboeuf
2025-02-18 8:52 ` Borislav Petkov
2025-02-20 22:04 ` Josh Poimboeuf
2025-02-26 18:57 ` Kaplan, David
2025-02-26 20:14 ` Pawan Gupta
2025-02-26 21:01 ` Borislav Petkov
2025-02-26 21:51 ` Pawan Gupta
2025-02-27 13:39 ` Borislav Petkov
2025-02-26 21:03 ` Kaplan, David
2025-02-26 22:13 ` Pawan Gupta
2025-02-26 22:18 ` Kaplan, David
2025-02-26 22:34 ` Pawan Gupta
2025-02-26 23:44 ` Josh Poimboeuf
2025-02-27 0:35 ` Pawan Gupta
2025-02-27 1:23 ` Josh Poimboeuf
2025-02-27 3:50 ` Pawan Gupta
2025-02-27 14:08 ` Borislav Petkov
2025-02-27 14:36 ` Kaplan, David
2025-02-27 15:01 ` Borislav Petkov
2025-02-27 15:22 ` Kaplan, David
2025-02-27 15:37 ` Borislav Petkov
2025-02-27 16:05 ` Kaplan, David
2025-02-27 17:07 ` Borislav Petkov
2025-01-08 20:25 ` [PATCH v3 21/35] x86/bugs: Determine relevant vulnerabilities based on attack vector controls David Kaplan
2025-01-09 3:43 ` Pawan Gupta
2025-01-09 15:08 ` Kaplan, David
2025-02-11 18:41 ` Josh Poimboeuf [this message]
2025-02-11 18:54 ` Josh Poimboeuf
2025-02-11 19:04 ` Kaplan, David
2025-02-11 20:34 ` Josh Poimboeuf
2025-02-11 20:53 ` Kaplan, David
2025-02-11 22:38 ` Josh Poimboeuf
2025-02-11 18:55 ` Kaplan, David
2025-01-08 20:25 ` [PATCH v3 22/35] x86/bugs: Add attack vector controls for mds David Kaplan
2025-01-08 20:25 ` [PATCH v3 23/35] x86/bugs: Add attack vector controls for taa David Kaplan
2025-02-11 19:01 ` Josh Poimboeuf
2025-01-08 20:25 ` [PATCH v3 24/35] x86/bugs: Add attack vector controls for mmio David Kaplan
2025-01-08 20:25 ` [PATCH v3 25/35] x86/bugs: Add attack vector controls for rfds David Kaplan
2025-01-08 20:25 ` [PATCH v3 26/35] x86/bugs: Add attack vector controls for srbds David Kaplan
2025-01-08 20:25 ` [PATCH v3 27/35] x86/bugs: Add attack vector controls for gds David Kaplan
2025-01-08 20:25 ` [PATCH v3 28/35] x86/bugs: Add attack vector controls for spectre_v1 David Kaplan
2025-01-08 20:25 ` [PATCH v3 29/35] x86/bugs: Add attack vector controls for retbleed David Kaplan
2025-01-08 20:25 ` [PATCH v3 30/35] x86/bugs: Add attack vector controls for spectre_v2_user David Kaplan
2025-02-11 19:03 ` Josh Poimboeuf
2025-02-12 17:22 ` Kaplan, David
2025-01-08 20:25 ` [PATCH v3 31/35] x86/bugs: Add attack vector controls for bhi David Kaplan
2025-01-08 20:25 ` [PATCH v3 32/35] x86/bugs: Add attack vector controls for spectre_v2 David Kaplan
2025-01-08 20:25 ` [PATCH v3 33/35] x86/bugs: Add attack vector controls for l1tf David Kaplan
2025-01-08 20:25 ` [PATCH v3 34/35] x86/bugs: Add attack vector controls for srso David Kaplan
2025-01-08 20:25 ` [PATCH v3 35/35] x86/pti: Add attack vector controls for pti David Kaplan
[not found] ` <20250110083627.xankiqhczr7ksldv@desk>
2025-01-10 15:39 ` [PATCH v3 00/35] x86/bugs: Attack vector controls Borislav Petkov
[not found] ` <20250110171410.ttbt7cohzdjwi4hk@desk>
2025-01-12 11:38 ` Borislav Petkov
2025-01-13 17:41 ` Pawan Gupta
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=20250211184130.2fl4jvdwd4e5y32a@jpoimboe \
--to=jpoimboe@kernel.org \
--cc=bp@alien8.de \
--cc=dave.hansen@linux.intel.com \
--cc=david.kaplan@amd.com \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=pawan.kumar.gupta@linux.intel.com \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--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®