From: Thomas Gleixner <tglx@linutronix.de>
To: "Charles (Chas) Williams" <ciwillia@brocade.com>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
x86@kernel.org, linux-kernel@vger.kernel.org,
"M. Vefa Bicakci" <m.v.b@runbox.com>
Subject: Re: [RFC PATCH] perf/x86/intel/rapl: avoid access unallocate memory
Date: Mon, 7 Nov 2016 21:20:53 +0100 (CET) [thread overview]
Message-ID: <alpine.DEB.2.20.1611072117040.3501@nanos> (raw)
In-Reply-To: <58e229e2-91f4-a97f-1b9f-089f48ef994a@brocade.com>
On Mon, 7 Nov 2016, Charles (Chas) Williams wrote:
> On 11/07/2016 11:19 AM, Thomas Gleixner wrote:
> > On Wed, 2 Nov 2016, Charles (Chas) Williams wrote:
> > > I don't know why the CPU's phys_proc_id is 2.
> >
> > max_physical_pkg_id gets initialized via:
> >
> > cpus = boot_cpu_data.x86_max_cores;
> > max_physical_pkg_id = DIV_ROUND_UP(MAX_LOCAL_APIC, ncpus);
> >
> > What's the value of boot_cpu_data.x86_max_cores and MAX_LOCAL_APIC?
>
> I have discovered that that is not the problem. smp_init_package_map()
> is calculating the physical core id using the following:
>
> for_each_present_cpu(cpu) {
> unsigned int apicid = apic->cpu_present_to_apicid(cpu);
>
> ...
> if (!topology_update_package_map(apicid, cpu))
> continue;
>
> ...
> int topology_update_package_map(unsigned int apicid, unsigned int cpu)
> {
> unsigned int new, pkg = apicid >>
> boot_cpu_data.x86_coreid_bits;
>
> But later when the secondary CPU's are identified they use a different
> calculation using the local APIC ID from the CPU's registers:
>
> static void generic_identify(struct cpuinfo_x86 *c)
> ...
> if (c->cpuid_level >= 0x00000001) {
> c->initial_apicid = (cpuid_ebx(1) >> 24) & 0xFF;
> ...
> c->phys_proc_id = c->initial_apicid;
>
> So at the end of identify_cpu() when the boot/hotplug assignment is
> put back:
>
> c->logical_proc_id = topology_phys_to_logical_pkg(c->phys_proc_id);
>
> topology_phys_to_logical_pkg() is returning an invalid logical processor
> since one isn't configured.
>
> It's not clear to me what the right thing to do is or which is right.
Nice detective work! So the issue is that the package mapping code honours
boot_cpu_data.x86_coreid_bit, while generic_identify does
not. boot_cpu_data.x86_coreid_bit is obviously 1 in your case. Tentative
fix below. I still need to gow through that maze and figure out what could
go wrong with that :(
Thanks,
tglx
8<------------------------
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -905,6 +905,8 @@ static void detect_null_seg_behavior(str
static void generic_identify(struct cpuinfo_x86 *c)
{
+ unsigned int pkg;
+
c->extended_cpuid_level = 0;
if (!have_cpuid_p())
@@ -929,7 +931,8 @@ static void generic_identify(struct cpui
c->apicid = c->initial_apicid;
# endif
#endif
- c->phys_proc_id = c->initial_apicid;
+ pkg = c->initial_apicid >> boot_cpu_data.x86_coreid_bits;
+ c->phys_proc_id = pkg;
}
get_model_name(c); /* Default name */
next prev parent reply other threads:[~2016-11-07 20:23 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-11-02 12:25 [RFC PATCH] perf/x86/intel/rapl: avoid access unallocate memory Sebastian Andrzej Siewior
2016-11-02 22:47 ` Charles (Chas) Williams
2016-11-03 17:47 ` Sebastian Andrzej Siewior
2016-11-04 12:20 ` Charles (Chas) Williams
2016-11-04 18:03 ` Sebastian Andrzej Siewior
2016-11-04 20:42 ` Charles (Chas) Williams
2016-11-04 20:57 ` Sebastian Andrzej Siewior
2016-11-07 16:19 ` Thomas Gleixner
2016-11-07 16:59 ` Charles (Chas) Williams
2016-11-07 20:20 ` Thomas Gleixner [this message]
2016-11-08 14:20 ` Charles (Chas) Williams
2016-11-08 14:31 ` Thomas Gleixner
2016-11-08 14:57 ` Charles (Chas) Williams
2016-11-08 16:22 ` Thomas Gleixner
2016-11-09 15:35 ` [PATCH] x86/cpuid: Deal with broken firmware once more Thomas Gleixner
2016-11-09 15:37 ` Thomas Gleixner
2016-11-09 16:03 ` Peter Zijlstra
2016-11-09 16:34 ` Charles (Chas) Williams
2016-11-09 18:37 ` Thomas Gleixner
2016-11-09 18:15 ` Charles (Chas) Williams
2016-11-09 20:27 ` [tip:x86/urgent] x86/cpu: Deal with broken firmware (VMWare/XEN) tip-bot for Thomas Gleixner
2016-11-11 5:49 ` Alok Kataria
2016-11-10 3:57 ` [PATCH] x86/cpuid: Deal with broken firmware once more M. Vefa Bicakci
2016-11-10 10:50 ` Charles (Chas) Williams
2016-11-10 11:14 ` Thomas Gleixner
2016-11-12 22:05 ` M. Vefa Bicakci
2016-11-10 11:13 ` Thomas Gleixner
2016-11-10 11:39 ` Peter Zijlstra
2016-11-10 14:02 ` Boris Ostrovsky
2016-11-10 15:05 ` Charles (Chas) Williams
2016-11-10 15:31 ` Boris Ostrovsky
2016-11-10 15:54 ` Sebastian Andrzej Siewior
2016-11-10 17:15 ` Thomas Gleixner
2016-11-12 22:05 ` M. Vefa Bicakci
2016-11-13 18:04 ` Boris Ostrovsky
2016-11-13 23:42 ` M. Vefa Bicakci
2016-11-15 1:21 ` Boris Ostrovsky
2016-11-18 11:16 ` Thomas Gleixner
2016-11-18 14:22 ` Boris Ostrovsky
2016-11-10 15:12 ` Thomas Gleixner
2016-11-10 15:38 ` Boris Ostrovsky
2016-11-10 17:13 ` Thomas Gleixner
2016-11-10 18:01 ` Boris Ostrovsky
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=alpine.DEB.2.20.1611072117040.3501@nanos \
--to=tglx@linutronix.de \
--cc=bigeasy@linutronix.de \
--cc=ciwillia@brocade.com \
--cc=linux-kernel@vger.kernel.org \
--cc=m.v.b@runbox.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).