linux-kernel - Re: [tip:perf/core] perf/x86: Add an MSR PMU driver

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CALCETrXP0-jRiq_qz=hwPxjMxGrbke9rtembh4_XBq-vq+cPNg@mail.gmail.com>
Date:	Tue, 4 Aug 2015 11:13:25 -0700
From:	Andy Lutomirski <luto@...capital.net>
To:	"Liang, Kan" <kan.liang@...el.com>
Cc:	Thomas Gleixner <tglx@...utronix.de>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	"H. Peter Anvin" <hpa@...or.com>,
	Andrew Lutomirski <luto@...nel.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	Peter Zijlstra <peterz@...radead.org>,
	Ingo Molnar <mingo@...nel.org>,
	"linux-tip-commits@...r.kernel.org" 
	<linux-tip-commits@...r.kernel.org>
Subject: Re: [tip:perf/core] perf/x86: Add an MSR PMU driver

On Tue, Aug 4, 2015 at 11:11 AM, Liang, Kan <kan.liang@...el.com> wrote:
>
>
>>
>> On Tue, Aug 4, 2015 at 8:03 AM, Liang, Kan <kan.liang@...el.com> wrote:
>> >
>> >> > +
>> >> > +enum perf_msr_id {
>> >> > +       PERF_MSR_TSC                    = 0,
>> >> > +       PERF_MSR_APERF                  = 1,
>> >> > +       PERF_MSR_MPERF                  = 2,
>> >> > +       PERF_MSR_PPERF                  = 3,
>> >> > +       PERF_MSR_SMI                    = 4,
>> >> > +
>> >> > +       PERF_MSR_EVENT_MAX,
>> >> > +};
>> >> > +
>> >> > +struct perf_msr {
>> >> > +       int     id;
>> >> > +       u64     msr;
>> >> > +};
>> >> > +
>> >> > +static struct perf_msr msr[] = {
>> >> > +       { PERF_MSR_TSC, 0 },
>> >> > +       { PERF_MSR_APERF, MSR_IA32_APERF },
>> >> > +       { PERF_MSR_MPERF, MSR_IA32_MPERF },
>> >> > +       { PERF_MSR_PPERF, MSR_PPERF },
>> >> > +       { PERF_MSR_SMI, MSR_SMI_COUNT }, };
>> >>
>> >> I think this could be easier to work with if it were [PERF_MSR_TSC] =
>> >> {...}, etc.  No big deal, though, until the list gets long.  However,
>> >> it might make fixing the apparent issue below easier...
>> >>
>> >> > +static int msr_event_init(struct perf_event *event) {
>> >> > +       u64 cfg = event->attr.config;
>> >>
>> >> ...
>> >>
>> >> > +       event->hw.event_base = msr[cfg].msr;
>> >>
>> >> Shouldn't this verify that the fancy enumeration code actually
>> >> believes that msr[cfg] exists on this system? Otherwise we might have
>> >> a very short wait until the perf fuzzer oopses this thing :)
>> >>
>> >
>> > I think we already did the check before using msr[cfg].
>>
>> Where?  All I see is:
>>
>> +       if (cfg >= PERF_MSR_EVENT_MAX)
>> +               return -EINVAL;
>
> Yes, we check cfg here. So msr[cfg] should be always available.
>

PERF_MSR_EVENT_MAX is a constant.  If I run this thing on an AMD CPU
that supports TSC, APERF, MPERF, and nothing else, and someone asks
for PPERF, then the check will succeed and we'll oops, right?

--Andy
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/