Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-03-23 Thread Ingo Molnar
* Peter Zijlstra wrote: > On Mon, Mar 23, 2015 at 11:39:00AM +0100, Ingo Molnar wrote: > > > http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/5th-gen-core-family-spec-update.pdf > > > > > > BDM11 and BDM55 (not 57) tell us that the PMU will generate crap output

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-03-23 Thread Peter Zijlstra
On Mon, Mar 23, 2015 at 11:39:00AM +0100, Ingo Molnar wrote: > > http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/5th-gen-core-family-spec-update.pdf > > > > BDM11 and BDM55 (not 57) tell us that the PMU will generate crap output > > if you don't do this.

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-03-23 Thread Ingo Molnar
* Peter Zijlstra wrote: > On Mon, Mar 23, 2015 at 10:38:54AM +0100, Ingo Molnar wrote: > > > > * Andi Kleen wrote: > > > > > From: Andi Kleen > > > > > > On Broadwell INST_RETIRED.ALL cannot be used with any period > > > that doesn't have the lowest 6 bits cleared. And the period > > >

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-03-23 Thread Peter Zijlstra
On Mon, Mar 23, 2015 at 10:38:54AM +0100, Ingo Molnar wrote: > > * Andi Kleen wrote: > > > From: Andi Kleen > > > > On Broadwell INST_RETIRED.ALL cannot be used with any period > > that doesn't have the lowest 6 bits cleared. And the period > > should not be smaller than 128. > > Sloppy

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-03-23 Thread Ingo Molnar
* Andi Kleen wrote: > From: Andi Kleen > > On Broadwell INST_RETIRED.ALL cannot be used with any period > that doesn't have the lowest 6 bits cleared. And the period > should not be smaller than 128. Sloppy changelog: a most basic question is not answered by the changelog: what happens in

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-03-23 Thread Ingo Molnar
* Peter Zijlstra pet...@infradead.org wrote: On Mon, Mar 23, 2015 at 11:39:00AM +0100, Ingo Molnar wrote: http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/5th-gen-core-family-spec-update.pdf BDM11 and BDM55 (not 57) tell us that the PMU will generate

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-03-23 Thread Peter Zijlstra
On Mon, Mar 23, 2015 at 11:39:00AM +0100, Ingo Molnar wrote: http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/5th-gen-core-family-spec-update.pdf BDM11 and BDM55 (not 57) tell us that the PMU will generate crap output if you don't do this. Non-fatal but

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-03-23 Thread Ingo Molnar
* Andi Kleen a...@firstfloor.org wrote: From: Andi Kleen a...@linux.intel.com On Broadwell INST_RETIRED.ALL cannot be used with any period that doesn't have the lowest 6 bits cleared. And the period should not be smaller than 128. Sloppy changelog: a most basic question is not answered by

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-03-23 Thread Peter Zijlstra
On Mon, Mar 23, 2015 at 10:38:54AM +0100, Ingo Molnar wrote: * Andi Kleen a...@firstfloor.org wrote: From: Andi Kleen a...@linux.intel.com On Broadwell INST_RETIRED.ALL cannot be used with any period that doesn't have the lowest 6 bits cleared. And the period should not be smaller

Re: [PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-03-23 Thread Ingo Molnar
* Peter Zijlstra pet...@infradead.org wrote: On Mon, Mar 23, 2015 at 10:38:54AM +0100, Ingo Molnar wrote: * Andi Kleen a...@firstfloor.org wrote: From: Andi Kleen a...@linux.intel.com On Broadwell INST_RETIRED.ALL cannot be used with any period that doesn't have the lowest

[PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-02-17 Thread Andi Kleen
From: Andi Kleen On Broadwell INST_RETIRED.ALL cannot be used with any period that doesn't have the lowest 6 bits cleared. And the period should not be smaller than 128. Add a new callback to enforce this, and set it for Broadwell. This is erratum BDM57 and BDM11. How does this handle the

[PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-02-17 Thread Andi Kleen
From: Andi Kleen a...@linux.intel.com On Broadwell INST_RETIRED.ALL cannot be used with any period that doesn't have the lowest 6 bits cleared. And the period should not be smaller than 128. Add a new callback to enforce this, and set it for Broadwell. This is erratum BDM57 and BDM11. How does

[PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-02-10 Thread Andi Kleen
From: Andi Kleen On Broadwell INST_RETIRED.ALL cannot be used with any period that doesn't have the lowest 6 bits cleared. And the period should not be smaller than 128. Add a new callback to enforce this, and set it for Broadwell. This is erratum BDM57 and BDM11. How does this handle the

[PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-02-10 Thread Andi Kleen
From: Andi Kleen a...@linux.intel.com On Broadwell INST_RETIRED.ALL cannot be used with any period that doesn't have the lowest 6 bits cleared. And the period should not be smaller than 128. Add a new callback to enforce this, and set it for Broadwell. This is erratum BDM57 and BDM11. How does

[PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-02-09 Thread Andi Kleen
From: Andi Kleen On Broadwell INST_RETIRED.ALL cannot be used with any period that doesn't have the lowest 6 bits cleared. And the period should not be smaller than 128. Add a new callback to enforce this, and set it for Broadwell. This is erratum BDM57 and BDM11. How does this handle the

[PATCH 3/3] perf, x86: Add INST_RETIRED.ALL workarounds

2015-02-09 Thread Andi Kleen
From: Andi Kleen a...@linux.intel.com On Broadwell INST_RETIRED.ALL cannot be used with any period that doesn't have the lowest 6 bits cleared. And the period should not be smaller than 128. Add a new callback to enforce this, and set it for Broadwell. This is erratum BDM57 and BDM11. How does