2.6.18-rc4-mm1 Run-time of Locking API testsuite

2.6.18-rc4-mm1 Run-time of Locking API testsuite

[Jim Cromie]

Note the non-trivial execution time difference:

soekris:~/pinlab# egrep -e 'Locking|Good' dmesg-2.6.18-rc4-*
dmesg-2.6.18-rc4-mm1-sk:[   16.044699] | Locking API testsuite:
dmesg-2.6.18-rc4-mm1-sk:[   96.083576] Good, all 218 testcases passed! |
dmesg-2.6.18-rc4-sk:[   18.563808] | Locking API testsuite:
dmesg-2.6.18-rc4-sk:[   19.693692] Good, all 218 testcases passed! |



I diffed the respective configs, saw nothing obvious in the following,
some of them look like theyd favor -mm1, if anything.

<trimming the obvious irrelevants>

[jimc@harpo lxbuild]$ diff linux-2.6.18-rc4-sk/.config 
linux-2.6.18-rc4-mm1-sk/.config
3,4c3,4
< # Linux kernel version: 2.6.18-rc4
< # Mon Aug  7 17:53:01 2006
---
 > # Linux kernel version: 2.6.18-rc4-mm1
 > # Sun Aug 13 09:19:25 2006

1451a1491
 > CONFIG_ENABLE_MUST_CHECK=y
1454a1495
 > CONFIG_DEBUG_SHIRQ=y
1482,1483c1525,1528
< CONFIG_FORCED_INLINING=y
< CONFIG_RCU_TORTURE_TEST=y
---
 > # CONFIG_PROFILE_LIKELY is not set
 > # CONFIG_FORCED_INLINING is not set
 > # CONFIG_DEBUG_SYNCHRO_TEST is not set
 > # CONFIG_RCU_TORTURE_TEST is not set
1486c1531
< CONFIG_DEBUG_STACK_USAGE=y
---
 > # CONFIG_DEBUG_STACK_USAGE is not set


will send full configs if useful.

Re: 2.6.18-rc4-mm1 Run-time of Locking API testsuite

[Jesper Juhl]

On 17/08/06, Jim Cromie <jim.cromie@gmail.com> wrote:
>
> Note the non-trivial execution time difference:
>
> soekris:~/pinlab# egrep -e 'Locking|Good' dmesg-2.6.18-rc4-*
> dmesg-2.6.18-rc4-mm1-sk:[   16.044699] | Locking API testsuite:
> dmesg-2.6.18-rc4-mm1-sk:[   96.083576] Good, all 218 testcases passed! |
> dmesg-2.6.18-rc4-sk:[   18.563808] | Locking API testsuite:
> dmesg-2.6.18-rc4-sk:[   19.693692] Good, all 218 testcases passed! |
>
Interresting. On my box it takes at most a few seconds (don't have
printk times enabled, so I can't give exact numbers). My best estimate
is 2-3 seconds to run the self tests.

I wonder what's so different about our machines. Mine is a Athlon64 X2
4400+ w/ 2GB RAM.

relevant config options look identical to yours... Strange..

-- 
Jesper Juhl <jesper.juhl@gmail.com>
Don't top-post  http://www.catb.org/~esr/jargon/html/T/top-post.html
Plain text mails only, please      http://www.expita.com/nomime.html

Re: 2.6.18-rc4-mm1 Run-time of Locking API testsuite

[Jim Cromie]

Jesper Juhl wrote:
> On 17/08/06, Jim Cromie <jim.cromie@gmail.com> wrote:
>>
>> Note the non-trivial execution time difference:
>>
>> soekris:~/pinlab# egrep -e 'Locking|Good' dmesg-2.6.18-rc4-*
>> dmesg-2.6.18-rc4-mm1-sk:[   16.044699] | Locking API testsuite:
>> dmesg-2.6.18-rc4-mm1-sk:[   96.083576] Good, all 218 testcases passed! |
>> dmesg-2.6.18-rc4-sk:[   18.563808] | Locking API testsuite:
>> dmesg-2.6.18-rc4-sk:[   19.693692] Good, all 218 testcases passed! |
>>
> Interresting. On my box it takes at most a few seconds (don't have
> printk times enabled, so I can't give exact numbers). My best estimate
> is 2-3 seconds to run the self tests.
>
> I wonder what's so different about our machines. Mine is a Athlon64 X2
> 4400+ w/ 2GB RAM.
>
> relevant config options look identical to yours... Strange..
>

Mines a toy in comparison :-( but has its uses )

# cat /proc/cpuinfo
processor       : 0
vendor_id       : Geode by NSC
cpu family      : 5
model           : 9
model name      : Unknown
stepping        : 1
cpu MHz         : 266.670
fdiv_bug        : no
hlt_bug         : no
f00f_bug        : no
coma_bug        : no
fpu             : yes
fpu_exception   : yes
cpuid level     : 2
wp              : yes
flags           : fpu tsc msr cx8 cmov mmx cxmmx up
bogomips        : 535.83

these are my closest-fit cpu choices

# CONFIG_M586MMX is not set
CONFIG_MGEODEGX1=y
# CONFIG_MGEODE_LX is not set

Re: 2.6.18-rc4-mm1 Run-time of Locking API testsuite

[Nish Aravamudan]

On 8/17/06, Jim Cromie <jim.cromie@gmail.com> wrote:
>
> Note the non-trivial execution time difference:
>
> soekris:~/pinlab# egrep -e 'Locking|Good' dmesg-2.6.18-rc4-*
> dmesg-2.6.18-rc4-mm1-sk:[   16.044699] | Locking API testsuite:
> dmesg-2.6.18-rc4-mm1-sk:[   96.083576] Good, all 218 testcases passed! |
> dmesg-2.6.18-rc4-sk:[   18.563808] | Locking API testsuite:
> dmesg-2.6.18-rc4-sk:[   19.693692] Good, all 218 testcases passed! |

This is more than just a dmesg difference, I assume? As in, you can
actually tell the difference in time it takes?

Thanks,
Nish

Re: 2.6.18-rc4-mm1 Run-time of Locking API testsuite

[Jim Cromie]

Nish Aravamudan wrote:
> On 8/17/06, Jim Cromie <jim.cromie@gmail.com> wrote:
>>
>> Note the non-trivial execution time difference:
>>
>> soekris:~/pinlab# egrep -e 'Locking|Good' dmesg-2.6.18-rc4-*
>> dmesg-2.6.18-rc4-mm1-sk:[   16.044699] | Locking API testsuite:
>> dmesg-2.6.18-rc4-mm1-sk:[   96.083576] Good, all 218 testcases passed! |
>> dmesg-2.6.18-rc4-sk:[   18.563808] | Locking API testsuite:
>> dmesg-2.6.18-rc4-sk:[   19.693692] Good, all 218 testcases passed! |
>
> This is more than just a dmesg difference, I assume? As in, you can
> actually tell the difference in time it takes?
>

yes - I can actually see individual cells (the 'ok's) in the table being 
rendered separately.
This is across a serial line, at 115 kbaud, if that matters..

[   22.762505]              A-B-C-A-B-C deadlock:  ok  |  ok  |  ok  |  
ok  |  ok  |  ok  |
[   26.589133]          A-B-B-C-C-D-D-A deadlock:  ok  |  ok  |  ok  |  
ok  |  ok  |  ok  |
[   31.854688]          A-B-C-D-B-D-D-A deadlock:  ok  |  ok  |  ok  |  
ok  |  ok  |  ok  |

> Thanks,
> Nish
>