Mirror of the gdb mailing list
 help / color / mirror / Atom feed
* RE: Reconciling PowerPC simulator traces with performance monitor results
@ 2007-03-20 18:35 Wealand, Barry
  0 siblings, 0 replies; 5+ messages in thread
From: Wealand, Barry @ 2007-03-20 18:35 UTC (permalink / raw)
  To: gdb


Some additional information:  The application was run on the PowerPC
750FX with all interrupts disabled, and all memory mapping was performed
via BATs.  To the best of my knowledge, no exceptions should have been
generated (none were generated when running the application in the
PowerPC simulator).

Barry Wealand

-----Original Message-----
From: Wealand, Barry 
Sent: Tuesday, March 20, 2007 10:51 AM
To: 'gdb@sourceware.org'
Subject: Reconciling PowerPC simulator traces with performance monitor
results


Hello -

I'm using the PowerPC simulator in GDB to collect instruction (-t
semantics) and load-store (-t load-store) traces for a pretty-good-sized
program (several hundred million trace lines).  I then apply various
filters to the traces to collect measurements of interest.  All of this
seems to work very well.

Then I run the exact same code on a PowerPC 750FX processor, which I set
up to collect performance monitor statistics.  For example, I can
collect "completed load and store instructions", and "number of
instructions completed from the FPU".  What puzzles me is that the
results that I collect from the simulator for these two counts in
particular are 10 to 15 percent lower than the results reported by the
performance monitor.

I'm guessing that this discrepancy might have something to do with the
details of how the performance monitor works, but the descriptions above
are typical of the level of detail provided in the 750 FX manual.  I
wonder if anyone else has grappled with this, or might have any
suggestions.

Thanks!

Barry Wealand
Lockheed Martin Space Systems Company


^ permalink raw reply	[flat|nested] 5+ messages in thread
* Reconciling PowerPC simulator traces with performance monitor results
@ 2007-03-20 17:52 Wealand, Barry
  2007-03-20 19:34 ` Mark Kettenis
  0 siblings, 1 reply; 5+ messages in thread
From: Wealand, Barry @ 2007-03-20 17:52 UTC (permalink / raw)
  To: gdb


Hello -

I'm using the PowerPC simulator in GDB to collect instruction (-t
semantics) and load-store (-t load-store) traces for a pretty-good-sized
program (several hundred million trace lines).  I then apply various
filters to the traces to collect measurements of interest.  All of this
seems to work very well.

Then I run the exact same code on a PowerPC 750FX processor, which I set
up to collect performance monitor statistics.  For example, I can
collect "completed load and store instructions", and "number of
instructions completed from the FPU".  What puzzles me is that the
results that I collect from the simulator for these two counts in
particular are 10 to 15 percent lower than the results reported by the
performance monitor.

I'm guessing that this discrepancy might have something to do with the
details of how the performance monitor works, but the descriptions above
are typical of the level of detail provided in the 750 FX manual.  I
wonder if anyone else has grappled with this, or might have any
suggestions.

Thanks!

Barry Wealand
Lockheed Martin Space Systems Company


^ permalink raw reply	[flat|nested] 5+ messages in thread
end of thread, other threads:[~2007-03-20 19:55 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2007-03-20 18:35 Reconciling PowerPC simulator traces with performance monitor results Wealand, Barry
  -- strict thread matches above, loose matches on Subject: below --
2007-03-20 17:52 Wealand, Barry
2007-03-20 19:34 ` Mark Kettenis
2007-03-20 19:49   ` Paul Koning
2007-03-20 19:55   ` Stan Shebs

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox