From: Ingo Molnar <mingo@kernel.org>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: LKML <linux-kernel@vger.kernel.org>,
x86@kernel.org, Peter Zijlstra <peterz@infradead.org>,
Andy Lutomirski <luto@kernel.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Jiri Kosina <jkosina@suse.cz>,
Tom Lendacky <thomas.lendacky@amd.com>,
Josh Poimboeuf <jpoimboe@redhat.com>,
Andrea Arcangeli <aarcange@redhat.com>,
David Woodhouse <dwmw@amazon.co.uk>,
Andi Kleen <ak@linux.intel.com>,
Dave Hansen <dave.hansen@intel.com>,
Casey Schaufler <casey.schaufler@intel.com>,
Asit Mallick <asit.k.mallick@intel.com>,
Arjan van de Ven <arjan@linux.intel.com>,
Jon Masters <jcm@redhat.com>, Waiman Long <longman9394@gmail.com>,
Greg KH <gregkh@linuxfoundation.org>,
Dave Stewart <david.c.stewart@intel.com>,
Kees Cook <keescook@chromium.org>
Subject: Re: [patch 17/24] x86/speculation: Move IBPB control out of switch_mm()
Date: Thu, 22 Nov 2018 08:52:06 +0100 [thread overview]
Message-ID: <20181122075206.GG41788@gmail.com> (raw)
In-Reply-To: <20181121201723.948990148@linutronix.de>
* Thomas Gleixner <tglx@linutronix.de> wrote:
> IBPB control is currently in switch_mm() to avoid issuing IBPB when
> switching between tasks of the same process.
>
> But that's not covering the case of sandboxed tasks which get the
> TIF_SPEC_IB flag set via seccomp. There the barrier is required when the
> potentially malicious task is switched out because the task which is
> switched in might have it not set and would still be attackable.
>
> For tasks which mark themself with TIF_SPEC_IB via the prctl, the barrier
> needs to be when the tasks switches in because the previous one might be an
> attacker.
s/themself
/themselves
>
> Move the code out of switch_mm() and evaluate the TIF bit in
> switch_to(). Make it an inline function so it can be used both in 32bit and
> 64bit code.
s/32bit
/32-bit
s/64bit
/64-bit
>
> This loses the optimization of switching back to the same process, but
> that's wrong in the context of seccomp anyway as it does not protect tasks
> of the same process against each other.
>
> This could be optimized by keeping track of the last user task per cpu and
> avoiding the barrier when the task is immediately scheduled back and the
> thread inbetween was a kernel thread. It's dubious whether that'd be worth
> the extra load/store and conditional operations. Keep it optimized for the
> common case where the TIF bit is not set.
s/cpu/CPU
>
> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
> ---
> arch/x86/include/asm/nospec-branch.h | 2 +
> arch/x86/include/asm/spec-ctrl.h | 46 +++++++++++++++++++++++++++++++++++
> arch/x86/include/asm/tlbflush.h | 2 -
> arch/x86/kernel/cpu/bugs.c | 16 +++++++++++-
> arch/x86/kernel/process_32.c | 11 ++++++--
> arch/x86/kernel/process_64.c | 11 ++++++--
> arch/x86/mm/tlb.c | 39 -----------------------------
> 7 files changed, 81 insertions(+), 46 deletions(-)
>
> --- a/arch/x86/include/asm/nospec-branch.h
> +++ b/arch/x86/include/asm/nospec-branch.h
> @@ -312,6 +312,8 @@ do { \
> } while (0)
>
> DECLARE_STATIC_KEY_FALSE(switch_to_cond_stibp);
> +DECLARE_STATIC_KEY_FALSE(switch_to_cond_ibpb);
> +DECLARE_STATIC_KEY_FALSE(switch_to_always_ibpb);
>
> #endif /* __ASSEMBLY__ */
>
> --- a/arch/x86/include/asm/spec-ctrl.h
> +++ b/arch/x86/include/asm/spec-ctrl.h
> @@ -76,6 +76,52 @@ static inline u64 ssbd_tif_to_amd_ls_cfg
> return (tifn & _TIF_SSBD) ? x86_amd_ls_cfg_ssbd_mask : 0ULL;
> }
>
> +/**
> + * switch_to_ibpb - Issue IBPB on task switch
> + * @next: Pointer to the next task
> + * @prev_tif: Threadinfo flags of the previous task
> + * @next_tif: Threadinfo flags of the next task
> + *
> + * IBPB flushes the branch predictor, which stops Spectre-v2 attacks
> + * between user space tasks. Depending on the mode the flush is made
> + * conditional.
> + */
> +static inline void switch_to_ibpb(struct task_struct *next,
> + unsigned long prev_tif,
> + unsigned long next_tif)
> +{
> + if (static_branch_unlikely(&switch_to_always_ibpb)) {
> + /* Only flush when switching to a user task. */
> + if (next->mm)
> + indirect_branch_prediction_barrier();
> + }
> +
> + if (static_branch_unlikely(&switch_to_cond_ibpb)) {
> + /*
> + * Both tasks' threadinfo flags are checked for TIF_SPEC_IB.
> + *
> + * For an outgoing sandboxed task which has TIF_SPEC_IB set
> + * via seccomp this is needed because it might be malicious
> + * and the next user task switching in might not have it
> + * set.
> + *
> + * For an incoming task which has set TIF_SPEC_IB itself
> + * via prctl() this is needed because the previous user
> + * task might be malicious and have the flag unset.
> + *
> + * This could be optimized by keeping track of the last
> + * user task per cpu and avoiding the barrier when the task
> + * is immediately scheduled back and the thread inbetween
> + * was a kernel thread. It's dubious whether that'd be
> + * worth the extra load/store and conditional operations.
> + * Keep it optimized for the common case where the TIF bit
> + * is not set.
> + */
> + if ((prev_tif | next_tif) & _TIF_SPEC_IB)
> + indirect_branch_prediction_barrier();
s/cpu/CPU
> +
> + switch (mode) {
> + case SPECTRE_V2_APP2APP_STRICT:
> + static_branch_enable(&switch_to_always_ibpb);
> + break;
> + default:
> + break;
> + }
> +
> + pr_info("mitigation: Enabling %s Indirect Branch Prediction Barrier\n",
> + mode == SPECTRE_V2_APP2APP_STRICT ? "forced" : "conditional");
Maybe s/forced/always-on, to better match the code?
> @@ -617,11 +619,16 @@ void compat_start_thread(struct pt_regs
> /* Reload sp0. */
> update_task_stack(next_p);
>
> + prev_tif = task_thread_info(prev_p)->flags;
> + next_tif = task_thread_info(next_p)->flags;
> + /* Indirect branch prediction barrier control */
> + switch_to_ibpb(next_p, prev_tif, next_tif);
> +
> /*
> * Now maybe reload the debug registers and handle I/O bitmaps
> */
> - if (unlikely(task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT ||
> - task_thread_info(prev_p)->flags & _TIF_WORK_CTXSW_PREV))
> + if (unlikely(next_tif & _TIF_WORK_CTXSW_NEXT ||
> + prev_tif & _TIF_WORK_CTXSW_PREV))
> __switch_to_xtra(prev_p, next_p, tss);
Hm, the repetition between process_32.c and process_64.c is getting
stronger - could some of this be unified into process.c? (in later
patches)
Thanks,
Ingo
next prev parent reply other threads:[~2018-11-22 7:52 UTC|newest]
Thread overview: 95+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-11-21 20:14 [patch 00/24] x86/speculation: Remedy the STIBP/IBPB overhead Thomas Gleixner
2018-11-21 20:14 ` [patch 01/24] x86/speculation: Update the TIF_SSBD comment Thomas Gleixner
2018-11-21 20:28 ` Linus Torvalds
2018-11-21 20:30 ` Thomas Gleixner
2018-11-21 20:33 ` Linus Torvalds
2018-11-21 22:48 ` Thomas Gleixner
2018-11-21 22:53 ` Borislav Petkov
2018-11-21 22:55 ` Thomas Gleixner
2018-11-21 22:55 ` Arjan van de Ven
2018-11-21 22:56 ` Borislav Petkov
2018-11-21 23:07 ` Borislav Petkov
2018-11-21 23:04 ` Josh Poimboeuf
2018-11-21 23:08 ` Borislav Petkov
2018-11-22 17:30 ` Josh Poimboeuf
2018-11-22 17:52 ` Borislav Petkov
2018-11-22 21:17 ` Thomas Gleixner
2018-11-21 20:14 ` [patch 02/24] x86/speculation: Clean up spectre_v2_parse_cmdline() Thomas Gleixner
2018-11-21 20:14 ` [patch 03/24] x86/speculation: Remove unnecessary ret variable in cpu_show_common() Thomas Gleixner
2018-11-21 20:14 ` [patch 04/24] x86/speculation: Reorganize cpu_show_common() Thomas Gleixner
2018-11-21 20:14 ` [patch 05/24] x86/speculation: Disable STIBP when enhanced IBRS is in use Thomas Gleixner
2018-11-21 20:33 ` Borislav Petkov
2018-11-21 20:36 ` Thomas Gleixner
2018-11-21 22:01 ` Thomas Gleixner
2018-11-21 20:14 ` [patch 06/24] x86/speculation: Rename SSBD update functions Thomas Gleixner
2018-11-21 20:14 ` [patch 07/24] x86/speculation: Reorganize speculation control MSRs update Thomas Gleixner
2018-11-21 20:14 ` [patch 08/24] sched/smt: Make sched_smt_present track topology Thomas Gleixner
2018-11-21 20:14 ` [patch 09/24] x86/Kconfig: Select SCHED_SMT if SMP enabled Thomas Gleixner
2018-11-21 20:14 ` [patch 10/24] sched/smt: Expose sched_smt_present static key Thomas Gleixner
2018-11-21 20:41 ` Thomas Gleixner
2018-11-21 20:14 ` [patch 11/24] x86/speculation: Rework SMT state change Thomas Gleixner
2018-11-21 20:14 ` [patch 12/24] x86/l1tf: Show actual SMT state Thomas Gleixner
2018-11-21 20:14 ` [patch 13/24] x86/speculation: Reorder the spec_v2 code Thomas Gleixner
2018-11-21 20:14 ` [patch 14/24] x86/speculation: Unify conditional spectre v2 print functions Thomas Gleixner
2018-11-22 7:59 ` Ingo Molnar
2018-11-21 20:14 ` [patch 15/24] x86/speculation: Add command line control for indirect branch speculation Thomas Gleixner
2018-11-21 23:43 ` Borislav Petkov
2018-11-22 8:14 ` Thomas Gleixner
2018-11-22 9:07 ` Thomas Gleixner
2018-11-22 9:18 ` Peter Zijlstra
2018-11-22 10:10 ` Borislav Petkov
2018-11-22 10:48 ` Thomas Gleixner
2018-11-21 20:14 ` [patch 16/24] x86/speculation: Prepare for per task indirect branch speculation control Thomas Gleixner
2018-11-22 7:57 ` Ingo Molnar
2018-11-21 20:14 ` [patch 17/24] x86/speculation: Move IBPB control out of switch_mm() Thomas Gleixner
2018-11-22 0:01 ` Andi Kleen
2018-11-22 7:42 ` Jiri Kosina
2018-11-22 9:18 ` Thomas Gleixner
2018-11-22 1:40 ` Tim Chen
2018-11-22 7:52 ` Ingo Molnar [this message]
2018-11-22 22:29 ` Thomas Gleixner
2018-11-21 20:14 ` [patch 18/24] x86/speculation: Avoid __switch_to_xtra() calls Thomas Gleixner
2018-11-22 1:23 ` Tim Chen
2018-11-22 7:44 ` Ingo Molnar
2018-11-21 20:14 ` [patch 19/24] ptrace: Remove unused ptrace_may_access_sched() and MODE_IBRS Thomas Gleixner
2018-11-21 20:14 ` [patch 20/24] x86/speculation: Split out TIF update Thomas Gleixner
2018-11-22 2:13 ` Tim Chen
2018-11-22 23:00 ` Thomas Gleixner
2018-11-23 7:37 ` Ingo Molnar
2018-11-26 18:35 ` Tim Chen
2018-11-26 21:55 ` Thomas Gleixner
2018-11-27 7:05 ` Jiri Kosina
2018-11-27 7:13 ` Thomas Gleixner
2018-11-27 7:30 ` Jiri Kosina
2018-11-27 12:52 ` Jiri Kosina
2018-11-27 13:18 ` Jiri Kosina
2018-11-27 21:57 ` Thomas Gleixner
2018-11-27 22:07 ` Jiri Kosina
2018-11-27 22:20 ` Jiri Kosina
2018-11-27 22:36 ` Thomas Gleixner
2018-11-28 1:50 ` Tim Chen
2018-11-28 10:43 ` Thomas Gleixner
2018-11-28 6:05 ` Jiri Kosina
2018-11-28 14:33 ` [tip:x86/pti] x86/speculation: Prevent stale SPEC_CTRL msr content tip-bot for Thomas Gleixner
2018-11-22 7:43 ` [patch 20/24] x86/speculation: Split out TIF update Ingo Molnar
2018-11-22 23:04 ` Thomas Gleixner
2018-11-23 7:37 ` Ingo Molnar
2018-11-21 20:14 ` [patch 21/24] x86/speculation: Prepare arch_smt_update() for PRCTL mode Thomas Gleixner
2018-11-22 7:34 ` Ingo Molnar
2018-11-22 23:17 ` Thomas Gleixner
2018-11-22 23:28 ` Jiri Kosina
2018-11-21 20:14 ` [patch 22/24] x86/speculation: Create PRCTL interface to restrict indirect branch speculation Thomas Gleixner
2018-11-22 7:10 ` Ingo Molnar
2018-11-22 9:03 ` Peter Zijlstra
2018-11-22 9:08 ` Thomas Gleixner
2018-11-22 12:26 ` Borislav Petkov
2018-11-22 12:33 ` Peter Zijlstra
2018-11-21 20:14 ` [patch 23/24] x86/speculation: Enable PRCTL mode for spectre_v2_app2app Thomas Gleixner
2018-11-22 7:17 ` Ingo Molnar
2018-11-21 20:14 ` [patch 24/24] x86/speculation: Add seccomp Spectre v2 app to app protection mode Thomas Gleixner
2018-11-22 2:24 ` Tim Chen
2018-11-22 7:26 ` Ingo Molnar
2018-11-22 23:45 ` Thomas Gleixner
2018-11-21 23:48 ` [patch 00/24] x86/speculation: Remedy the STIBP/IBPB overhead Tim Chen
2018-11-22 9:55 ` Thomas Gleixner
2018-11-22 9:45 ` Peter Zijlstra
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=20181122075206.GG41788@gmail.com \
--to=mingo@kernel.org \
--cc=aarcange@redhat.com \
--cc=ak@linux.intel.com \
--cc=arjan@linux.intel.com \
--cc=asit.k.mallick@intel.com \
--cc=casey.schaufler@intel.com \
--cc=dave.hansen@intel.com \
--cc=david.c.stewart@intel.com \
--cc=dwmw@amazon.co.uk \
--cc=gregkh@linuxfoundation.org \
--cc=jcm@redhat.com \
--cc=jkosina@suse.cz \
--cc=jpoimboe@redhat.com \
--cc=keescook@chromium.org \
--cc=linux-kernel@vger.kernel.org \
--cc=longman9394@gmail.com \
--cc=luto@kernel.org \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=thomas.lendacky@amd.com \
--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;
as well as URLs for NNTP newsgroup(s).