linux-kernel - Re: [tip:perfcounters/core] perfcounters: IRQ and NMI support on AMD CPUs

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] [day] [month] [year] [list]

Message-Id: <1236849134.5090.26.camel@laptop>
Date:	Thu, 12 Mar 2009 10:12:14 +0100
From:	Peter Zijlstra <a.p.zijlstra@...llo.nl>
To:	Robert Richter <robert.richter@....com>
Cc:	Ingo Molnar <mingo@...e.hu>, linux-kernel@...r.kernel.org
Subject: Re: [tip:perfcounters/core] perfcounters: IRQ and NMI support on
 AMD CPUs

On Thu, 2009-03-12 at 02:44 +0100, Robert Richter wrote:
> On 05.03.09 17:27:29, Peter Zijlstra wrote:
> > Commit-ID:  b0f3f28e0f14eb335f67bfaae33ce8b8d74fd58b
> > Gitweb:     http://git.kernel.org/tip/b0f3f28e0f14eb335f67bfaae33ce8b8d74fd58b
> > Author:     "Peter Zijlstra" <a.p.zijlstra@...llo.nl>
> > AuthorDate: Thu, 5 Mar 2009 18:08:27 +0100
> > Commit:     Ingo Molnar <mingo@...e.hu>
> > CommitDate: Thu, 5 Mar 2009 18:25:16 +0100
> > 
> > perfcounters: IRQ and NMI support on AMD CPUs
> > 
> > The below completes the K7+ performance counter support:
> > 
> >  - IRQ support
> >  - NMI support
> > 
> > KernelTop output works now as well.
> > 
> > Signed-off-by: Peter Zijlstra <a.p.zijlstra@...llo.nl>
> > Cc: Jaswinder Singh Rajput <jaswinder@...nel.org>
> > Cc: Paul Mackerras <paulus@...ba.org>
> > LKML-Reference: <1236273633.5187.286.camel@...top>
> > Signed-off-by: Ingo Molnar <mingo@...e.hu>
> > 
> > 
> > ---
> >  arch/x86/kernel/cpu/perf_counter.c |  272 ++++++++++++++++++++++++++++++------
> >  1 files changed, 228 insertions(+), 44 deletions(-)
> > 
> > diff --git a/arch/x86/kernel/cpu/perf_counter.c b/arch/x86/kernel/cpu/perf_counter.c
> > index 3b65f19..6ebe9ab 100644
> > --- a/arch/x86/kernel/cpu/perf_counter.c
> > +++ b/arch/x86/kernel/cpu/perf_counter.c
> 
> [...]
> 
> > @@ -205,18 +243,24 @@ static u64 pmc_intel_save_disable_all(void)
> >  
> >  static u64 pmc_amd_save_disable_all(void)
> >  {
> > -	int idx;
> > -	u64 val, ctrl = 0;
> > +	struct cpu_hw_counters *cpuc = &__get_cpu_var(cpu_hw_counters);
> > +	int enabled, idx;
> > +
> > +	enabled = cpuc->enabled;
> > +	cpuc->enabled = 0;
> > +	barrier();
> 
> Peter,
> 
> please add comments to all barrier()s you added to this file. Is it
> sufficient in hw_perf_save_disable() to stop NMI's only on the current
> core?

Yes, counters are fundamentally per cpu. But you're right, I'll go
through the code and add comments and fix up some barriers.

--
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/