From: Borislav Petkov <bp@suse.de>
To: Kyle Huey <me@kylehuey.com>
Cc: "Robert O'Callahan" <robert@ocallahan.org>,
linux-api@vger.kernel.org, Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@redhat.com>, "H. Peter Anvin" <hpa@zytor.com>,
"maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)"
<x86@kernel.org>, "Peter Zijlstra (Intel)" <peterz@infradead.org>,
"Rafael J. Wysocki" <rafael.j.wysocki@intel.com>,
Len Brown <len.brown@intel.com>,
Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>,
Huang Rui <ray.huang@amd.com>,
Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com>,
Alexander Shishkin <alexander.shishkin@linux.intel.com>,
Vladimir Zapolskiy <vladimir_zapolskiy@mentor.com>,
Andy Lutomirski <luto@kernel.org>,
Juergen Gross <jgross@suse.com>,
Fenghua Yu <fenghua.yu@intel.com>,
"Luis R. Rodriguez" <mcgrof@kernel.org>,
Denys Vlasenko <dvlasenk@redhat.com>,
Andrew Morton <akpm@linux-foundation.org>,
Kees Cook <keescook@chromium.org>,
Dmitry Vyukov <dvyukov@google.com>,
Paul Gortmaker <paul.gortmaker@windriver.com>,
"Michael S. Tsirkin" <mst@redhat.com>,
Andrey Ryabinin <aryabinin@virtuozzo.com>,
Jiri Slaby <jslaby@suse.cz>, Michal Hocko <mhocko@suse.com>,
Alex Thorlton <athorlton@sgi.com>,
Vlastimil Babka <vbabka@suse.cz>,
Mateusz Guzik <mguzik@redhat.com>,
Ben Segall <bsegall@google.com>,
John Stultz <john.stultz@linaro.org>,
"open list:X86 ARCHITECTURE (32-BIT AND 64-BIT)"
<linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] prctl,x86 Add PR_[GET|SET]_CPUID for controlling the CPUID instruction.
Date: Mon, 12 Sep 2016 11:07:22 +0200 [thread overview]
Message-ID: <20160912090722.2yal3mucf2x6j7pi@pd.tnic> (raw)
In-Reply-To: <1473640169-24145-1-git-send-email-khuey@kylehuey.com>
On Sun, Sep 11, 2016 at 05:29:23PM -0700, Kyle Huey wrote:
> rr (http://rr-project.org/), a userspace record-and-replay reverse-
> execution debugger, would like to trap and emulate the CPUID instruction.
> This would allow us to a) mask away certain hardware features that rr does
> not support (e.g. RDRAND) and b) enable trace portability across machines
> by providing constant results.
>
> Intel supports faulting on the CPUID instruction in newer processors. Bit
> 31 of MSR_PLATFORM_INFO advertises support for this feature. It is
> documented in detail in Section 2.3.2 of
> http://www.intel.com/content/dam/www/public/us/en/documents/application-notes/virtualization-technology-flexmigration-application-note.pdf.
>
> I would like to thank Trevor Saunders <tbsaunde@tbsaunde.org> for drafting
> an earlier version of this patch.
>
> Signed-off-by Kyle Huey <khuey@kylehuey.com>
> ---
> arch/x86/include/asm/msr-index.h | 1 +
> arch/x86/include/asm/processor.h | 7 ++++
> arch/x86/include/asm/thread_info.h | 4 +-
> arch/x86/kernel/process.c | 79 ++++++++++++++++++++++++++++++++++++++
> include/uapi/linux/prctl.h | 6 +++
> kernel/sys.c | 12 ++++++
> 6 files changed, 108 insertions(+), 1 deletion(-)
...
> diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c
> index 62c0b0e..a189516 100644
> --- a/arch/x86/kernel/process.c
> +++ b/arch/x86/kernel/process.c
> @@ -191,6 +191,76 @@ int set_tsc_mode(unsigned int val)
> return 0;
> }
>
> +static void hard_disable_CPUID(void)
Why hard_disable? I don't see any soft_disable.
Also, I can't say that I like all that screaming "CPUID" :-)
disable_cpuid() looks just fine to me too.
> +{
> + msr_set_bit(MSR_MISC_FEATURES_ENABLES, 0);
> +}
> +
> +static void disable_CPUID(void)
> +{
> + preempt_disable();
> + if (!test_and_set_thread_flag(TIF_NOCPUID))
> + /*
> + * Must flip the CPU state synchronously with
> + * TIF_NOCPUID in the current running context.
> + */
> + hard_disable_CPUID();
> + preempt_enable();
> +}
> +
> +static void hard_enable_CPUID(void)
> +{
> + msr_clear_bit(MSR_MISC_FEATURES_ENABLES, 0);
> +}
> +
> +static void enable_CPUID(void)
> +{
> + preempt_disable();
> + if (test_and_clear_thread_flag(TIF_NOCPUID))
> + /*
> + * Must flip the CPU state synchronously with
> + * TIF_NOCPUID in the current running context.
> + */
> + hard_enable_CPUID();
> + preempt_enable();
> +}
> +
> +static int supports_CPUID_faulting(void)
> +{
> + unsigned int lo, hi;
> +
> + rdmsr(MSR_PLATFORM_INFO, lo, hi);
rdmsr_safe()
> + if ((lo & (1 << 31)))
> + return 1;
> + else
> + return 0;
> +}
>
> +int get_cpuid_mode(unsigned long adr)
> +{
> + unsigned int val;
> +
> + if (test_thread_flag(TIF_NOCPUID))
> + val = PR_CPUID_SIGSEGV;
> + else
> + val = PR_CPUID_ENABLE;
> +
> + return put_user(val, (unsigned int __user *)adr);
> +}
> +
> +int set_cpuid_mode(unsigned int val)
> +{
> + // Only disable/enable_CPUID() if it is supported on this hardware.
Use /* ... */ for comments in the kernel.
> + if (val == PR_CPUID_SIGSEGV && supports_CPUID_faulting())
> + disable_CPUID();
> + else if (val == PR_CPUID_ENABLE && supports_CPUID_faulting())
> + enable_CPUID();
> + else
> + return -EINVAL;
> +
> + return 0;
> +}
> +
> void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p,
> struct tss_struct *tss)
> {
> @@ -210,6 +280,15 @@ void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p,
> update_debugctlmsr(debugctl);
> }
>
> + if (test_tsk_thread_flag(prev_p, TIF_NOCPUID) ^
> + test_tsk_thread_flag(next_p, TIF_NOCPUID)) {
> + /* prev and next are different */
> + if (test_tsk_thread_flag(next_p, TIF_NOCPUID))
> + hard_disable_CPUID();
> + else
> + hard_enable_CPUID();
> + }
> +
Frankly, I can't say that I'm thrilled by this: if this is a niche
feature which has only a very narrow usage for debugging, I'd much
prefer if this whole thing were implemented with a static_key which was
false on the majority of the systems so that __switch_to() tests it much
cheaply.
Then and only then if your debugger runs arch_prctl(), it would enable
the key and then set_cpuid_mode() can query the MSR directly instead of
using another flag in the thread_info flags.
This would keep this niche feature out of the way of the hot paths.
> if (test_tsk_thread_flag(prev_p, TIF_NOTSC) ^
> test_tsk_thread_flag(next_p, TIF_NOTSC)) {
> /* prev and next are different */
> diff --git a/include/uapi/linux/prctl.h b/include/uapi/linux/prctl.h
> index a8d0759..641d12b 100644
> --- a/include/uapi/linux/prctl.h
> +++ b/include/uapi/linux/prctl.h
> @@ -197,4 +197,10 @@ struct prctl_mm_map {
> # define PR_CAP_AMBIENT_LOWER 3
> # define PR_CAP_AMBIENT_CLEAR_ALL 4
>
> +/* Get/set the process' ability to use the CPUID instruction */
> +#define PR_GET_CPUID 48
> +#define PR_SET_CPUID 49
> +# define PR_CPUID_ENABLE 1 /* allow the use of the CPUID instruction */
> +# define PR_CPUID_SIGSEGV 2 /* throw a SIGSEGV instead of reading the CPUID */
> +
> #endif /* _LINUX_PRCTL_H */
> diff --git a/kernel/sys.c b/kernel/sys.c
> index 89d5be4..997c6519 100644
> --- a/kernel/sys.c
> +++ b/kernel/sys.c
> @@ -91,6 +91,12 @@
> #ifndef SET_TSC_CTL
> # define SET_TSC_CTL(a) (-EINVAL)
> #endif
> +#ifndef GET_CPUID_CTL
> +# define GET_CPUID_CTL(a) (-EINVAL)
> +#endif
> +#ifndef SET_CPUID_CTL
> +# define SET_CPUID_CTL(a) (-EINVAL)
> +#endif
> #ifndef MPX_ENABLE_MANAGEMENT
> # define MPX_ENABLE_MANAGEMENT() (-EINVAL)
> #endif
> @@ -2162,6 +2168,12 @@ SYSCALL_DEFINE5(prctl, int, option, unsigned long, arg2, unsigned long, arg3,
> case PR_SET_TSC:
> error = SET_TSC_CTL(arg2);
> break;
> + case PR_GET_CPUID:
> + error = GET_CPUID_CTL(arg2);
> + break;
> + case PR_SET_CPUID:
> + error = SET_CPUID_CTL(arg2);
> + break;
> case PR_TASK_PERF_EVENTS_DISABLE:
> error = perf_event_task_disable();
> break;
This whole fun should be in arch_prctl() as it is arch-specific.
And wherever it ends, it needs documenting in the man page.
--
Regards/Gruss,
Boris.
SUSE Linux GmbH, GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg)
--
next prev parent reply other threads:[~2016-09-12 9:07 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-09-12 0:29 [PATCH] prctl,x86 Add PR_[GET|SET]_CPUID for controlling the CPUID instruction Kyle Huey
2016-09-12 9:07 ` Borislav Petkov [this message]
2016-09-12 14:15 ` Kyle Huey
2016-09-12 14:34 ` Borislav Petkov
2016-09-13 18:42 ` Kyle Huey
2016-09-12 16:56 ` Andy Lutomirski
2016-09-12 17:18 ` Borislav Petkov
2016-09-12 17:56 ` Jann Horn
2016-09-12 21:07 ` Andy Lutomirski
2016-09-14 6:13 ` Kyle Huey
2016-09-14 18:52 ` Andy Lutomirski
2016-09-14 19:22 ` Andrew Cooper
2016-09-14 19:23 ` Boris Ostrovsky
2016-09-14 19:28 ` Andrew Cooper
2016-09-14 19:36 ` Andy Lutomirski
2016-09-14 19:42 ` Andrew Cooper
2016-09-12 17:37 ` Andi Kleen
2016-09-12 18:25 ` Henrique de Moraes Holschuh
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=20160912090722.2yal3mucf2x6j7pi@pd.tnic \
--to=bp@suse.de \
--cc=Aravind.Gopalakrishnan@amd.com \
--cc=akpm@linux-foundation.org \
--cc=alexander.shishkin@linux.intel.com \
--cc=aryabinin@virtuozzo.com \
--cc=athorlton@sgi.com \
--cc=bsegall@google.com \
--cc=dvlasenk@redhat.com \
--cc=dvyukov@google.com \
--cc=fenghua.yu@intel.com \
--cc=hpa@zytor.com \
--cc=jgross@suse.com \
--cc=john.stultz@linaro.org \
--cc=jslaby@suse.cz \
--cc=keescook@chromium.org \
--cc=len.brown@intel.com \
--cc=linux-api@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@kernel.org \
--cc=mcgrof@kernel.org \
--cc=me@kylehuey.com \
--cc=mguzik@redhat.com \
--cc=mhocko@suse.com \
--cc=mingo@redhat.com \
--cc=mst@redhat.com \
--cc=paul.gortmaker@windriver.com \
--cc=peterz@infradead.org \
--cc=rafael.j.wysocki@intel.com \
--cc=ray.huang@amd.com \
--cc=robert@ocallahan.org \
--cc=srinivas.pandruvada@linux.intel.com \
--cc=tglx@linutronix.de \
--cc=vbabka@suse.cz \
--cc=vladimir_zapolskiy@mentor.com \
--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).