Re: cli/sti vs local_cmpxchg and local_add_return

[Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>]

* Josh Boyer (jwboyer@linux.vnet.ibm.com) wrote:
> On Mon, Mar 16, 2009 at 09:32:20PM -0400, Mathieu Desnoyers wrote:
> >Hi,
> >
> >I am trying to get access to some non-x86 hardware to run some atomic
> >primitive benchmarks for a paper on LTTng I am preparing. That should be
> >useful to argue about performance benefit of per-cpu atomic operations
> >vs interrupt disabling. I would like to run the following benchmark
> >module on CONFIG_SMP :
> >
> >- PowerPC
> >- MIPS
> >- ia64
> >- alpha
> >
> >usage :
> >make
> >insmod test-cmpxchg-nolock.ko
> >insmod: error inserting 'test-cmpxchg-nolock.ko': -1 Resource temporarily unavailable
> >dmesg (see dmesg output)
> >
> >If some of you would be kind enough to run my test module provided below
> >and provide the results of these tests on a recent kernel (2.6.26~2.6.29
> >should be good) along with their cpuinfo, I would greatly appreciate.
> >
> >Here are the CAS results for various Intel-based architectures :
> >
> >Architecture         | Speedup                      |      CAS     |         Interrupts         |
> >                     | (cli + sti) / local cmpxchg  | local | sync | Enable (sti) | Disable (cli)
> >-------------------------------------------------------------------------------------------------
> >Intel Pentium 4      | 5.24                         |  25   | 81   | 70           | 61          |
> >AMD Athlon(tm)64 X2  | 4.57                         |  7    | 17   | 17           | 15          |
> >Intel Core2          | 6.33                         |  6    | 30   | 20           | 18          |
> >Intel Xeon E5405     | 5.25                         |  8    | 24   | 20           | 22          |
> 
> 
> I know you have results from a POWER6 machine already, but
> here are the results on a dual-G5 running 2.6.29-rc7-git4.
> 
> If you are interested, I could get you results from running
> this on an embedded PowerPC board.
> 

Thanks for the results. Well, those already shows that the tradeoff is
different between POWER6 and POWER5, so I guess further powerpc numbers
won't be required.

Mathieu

> josh
> 
> test init
> test results: time for baseline
> number of loops: 20000
> total time: 1532
> -> baseline takes 0 cycles
> test end
> test results: time for locked cmpxchg
> number of loops: 20000
> total time: 48052
> -> locked cmpxchg takes 2 cycles
> test end
> test results: time for non locked cmpxchg
> number of loops: 20000
> total time: 29141
> -> non locked cmpxchg takes 1 cycles
> test end
> test results: time for locked add return
> number of loops: 20000
> total time: 44985
> -> locked add return takes 2 cycles
> test end
> test results: time for non locked add return
> number of loops: 20000
> total time: 32400
> -> non locked add return takes 1 cycles
> test end
> test results: time for enabling interrupts (STI)
> number of loops: 20000
> total time: 65579
> -> enabling interrupts (STI) takes 3 cycles
> test end
> test results: time for disabling interrupts (CLI)
> number of loops: 20000
> total time: 29135
> -> disabling interrupts (CLI) takes 1 cycles
> test end
> test results: time for disabling/enabling interrupts (STI/CLI)
> number of loops: 20000
> total time: 173594
> -> enabling/disabling interrupts (STI/CLI) takes 8 cycles
> test end
> [jwboyer@localhost ~]$

-- 
Mathieu Desnoyers
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68