The Linux Kernel Mailing List
 help / color / mirror / Atom feed
* [BENCHMARK] rmap15, rmap14c and rc1aa1 with contest
@ 2002-11-24 13:17 Nero
  2002-11-24 13:23 ` Nero
                   ` (3 more replies)
  0 siblings, 4 replies; 5+ messages in thread
From: Nero @ 2002-11-24 13:17 UTC (permalink / raw)
  To: lkml; +Cc: Rik van Riel, Con Kolivas

noload:
Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
2.4.19-rmap15 [1]       84.6    95      0       0       1.00
2.4.19-rmap14c [1]      84.2    96      0       0       1.00
2.4.20-rc1aa1 [2]       41.0    49      0       0       inf

process_load:
Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
2.4.19-rmap15 [1]       140.8   55      133     45      1.66
2.4.19-rmap14c [1]      139.4   56      126     45      1.65
2.4.20-rc1aa1 [2]       133.9   14      211     84      inf

io_load:
Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
2.4.19-rmap15 [1]       164.8   51      33      19      1.95
2.4.19-rmap14c [1]      160.1   53      33      21      1.90
2.4.20-rc1aa1 [1]       166.2   49      44      23      inf

read_load:
Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
2.4.19-rmap15 [1]       0.0     0       0       7       0.00
2.4.19-rmap14c [1]      121.0   72      20      7       1.44
2.4.20-rc1aa1 [1]       111.2   79      34      14      inf

rmap15 OOM'd the cc1 process twice and fscked this run up.

list_load:
Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
2.4.19-rmap15 [1]       99.2    84      0       7       1.17
2.4.19-rmap14c [1]      100.7   83      0       7       1.20
2.4.20-rc1aa1 [1]       96.7    85      0       8       inf

mem_load:
Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
2.4.19-rmap15 [1]       111.2   80      96      8       1.31
2.4.19-rmap14c [1]      106.1   82      96      9       1.26
2.4.20-rc1aa1 [1]       126.3   66      77      3       inf






^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [BENCHMARK] rmap15, rmap14c and rc1aa1 with contest
  2002-11-24 13:17 [BENCHMARK] rmap15, rmap14c and rc1aa1 with contest Nero
@ 2002-11-24 13:23 ` Nero
  2002-11-24 13:32 ` Con Kolivas
                   ` (2 subsequent siblings)
  3 siblings, 0 replies; 5+ messages in thread
From: Nero @ 2002-11-24 13:23 UTC (permalink / raw)
  To: lkml

Nero wrote:

> noload:
> Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> 2.4.19-rmap15 [1]       84.6    95      0       0       1.00
> 2.4.19-rmap14c [1]      84.2    96      0       0       1.00
> 2.4.20-rc1aa1 [2]       41.0    49      0       0       inf

The time for rc1aa1 here is obviously wrong - this is a bug in contest, 
according to Dr. Kolivas :-)

>
> process_load:
> Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> 2.4.19-rmap15 [1]       140.8   55      133     45      1.66
> 2.4.19-rmap14c [1]      139.4   56      126     45      1.65
> 2.4.20-rc1aa1 [2]       133.9   14      211     84      inf
>
> io_load:
> Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> 2.4.19-rmap15 [1]       164.8   51      33      19      1.95
> 2.4.19-rmap14c [1]      160.1   53      33      21      1.90
> 2.4.20-rc1aa1 [1]       166.2   49      44      23      inf
>
> read_load:
> Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> 2.4.19-rmap15 [1]       0.0     0       0       7       0.00
> 2.4.19-rmap14c [1]      121.0   72      20      7       1.44
> 2.4.20-rc1aa1 [1]       111.2   79      34      14      inf
>
> rmap15 OOM'd the cc1 process twice and fscked this run up.
>
> list_load:
> Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> 2.4.19-rmap15 [1]       99.2    84      0       7       1.17
> 2.4.19-rmap14c [1]      100.7   83      0       7       1.20
> 2.4.20-rc1aa1 [1]       96.7    85      0       8       inf
>
> mem_load:
> Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> 2.4.19-rmap15 [1]       111.2   80      96      8       1.31
> 2.4.19-rmap14c [1]      106.1   82      96      9       1.26
> 2.4.20-rc1aa1 [1]       126.3   66      77      3       inf
>



^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [BENCHMARK] rmap15, rmap14c and rc1aa1 with contest
  2002-11-24 13:17 [BENCHMARK] rmap15, rmap14c and rc1aa1 with contest Nero
  2002-11-24 13:23 ` Nero
@ 2002-11-24 13:32 ` Con Kolivas
  2002-11-24 13:34 ` Nero
  2002-11-24 17:39 ` Rik van Riel
  3 siblings, 0 replies; 5+ messages in thread
From: Con Kolivas @ 2002-11-24 13:32 UTC (permalink / raw)
  To: Nero, lkml; +Cc: Rik van Riel

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Thanks for this interesting comparison

>noload:
>Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
>2.4.19-rmap15 [1]       84.6    95      0       0       1.00
>2.4.19-rmap14c [1]      84.2    96      0       0       1.00
>2.4.20-rc1aa1 [2]       41.0    49      0       0       inf

Clearly my braindead parser failed here. The time of aa1 should probably be 82 
seconds (41.0 x 2) Can you check the full log of 2.4.20-rc1aa1.log ? It has 
raw results.

>process_load:
>Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
>2.4.19-rmap15 [1]       140.8   55      133     45      1.66
>2.4.19-rmap14c [1]      139.4   56      126     45      1.65
>2.4.20-rc1aa1 [2]       133.9   14      211     84      inf

If the error continued here (2 runs of process_load) then the time of aa1 
should be 133.9 x 2. Can you check the full log of 2.4.20-rc1aa1.log here 
too?

>io_load:
>Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
>2.4.19-rmap15 [1]       164.8   51      33      19      1.95
>2.4.19-rmap14c [1]      160.1   53      33      21      1.90
>2.4.20-rc1aa1 [1]       166.2   49      44      23      inf
>
>read_load:
>Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
>2.4.19-rmap15 [1]       0.0     0       0       7       0.00
>2.4.19-rmap14c [1]      121.0   72      20      7       1.44
>2.4.20-rc1aa1 [1]       111.2   79      34      14      inf
>
>rmap15 OOM'd the cc1 process twice and fscked this run up.
>
>list_load:
>Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
>2.4.19-rmap15 [1]       99.2    84      0       7       1.17
>2.4.19-rmap14c [1]      100.7   83      0       7       1.20
>2.4.20-rc1aa1 [1]       96.7    85      0       8       inf
>
>mem_load:
>Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
>2.4.19-rmap15 [1]       111.2   80      96      8       1.31
>2.4.19-rmap14c [1]      106.1   82      96      9       1.26
>2.4.20-rc1aa1 [1]       126.3   66      77      3       inf

With only one run these numbers appear not to be greatly statistically 
significant, except for the mem_load results.

Con
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.0 (GNU/Linux)

iD8DBQE94NTSF6dfvkL3i1gRAkwXAJ9YJmnxvxdiYyESNYLeQa9sbwR+2wCcCZal
bFKWReK3EP8oayP/GwiIK7c=
=gYaS
-----END PGP SIGNATURE-----

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [BENCHMARK] rmap15, rmap14c and rc1aa1 with contest
  2002-11-24 13:17 [BENCHMARK] rmap15, rmap14c and rc1aa1 with contest Nero
  2002-11-24 13:23 ` Nero
  2002-11-24 13:32 ` Con Kolivas
@ 2002-11-24 13:34 ` Nero
  2002-11-24 17:39 ` Rik van Riel
  3 siblings, 0 replies; 5+ messages in thread
From: Nero @ 2002-11-24 13:34 UTC (permalink / raw)
  To: Con Kolivas; +Cc: lkml

Con Kolivas wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Thanks for this interesting comparison
>
>
> >noload:
> >Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> >2.4.19-rmap15 [1]       84.6    95      0       0       1.00
> >2.4.19-rmap14c [1]      84.2    96      0       0       1.00
> >2.4.20-rc1aa1 [2]       41.0    49      0       0       inf
>
>
> Clearly my braindead parser failed here. The time of aa1 should 
> probably be 82
> seconds (41.0 x 2) Can you check the full log of 2.4.20-rc1aa1.log ? 
> It has
> raw results.
>
Yep.
noload Time: 81.95  CPU: 98%  LoadRuns: 0  LoadCPU%: 0

>
> >process_load:
> >Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> >2.4.19-rmap15 [1]       140.8   55      133     45      1.66
> >2.4.19-rmap14c [1]      139.4   56      126     45      1.65
> >2.4.20-rc1aa1 [2]       133.9   14      211     84      inf
>
>
> If the error continued here (2 runs of process_load) then the time of aa1
> should be 133.9 x 2. Can you check the full log of 2.4.20-rc1aa1.log here
> too?

process_load Time: 267.85  CPU: 29%  LoadRuns: 422  LoadCPU%: 70%

>
>
> >io_load:
> >Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> >2.4.19-rmap15 [1]       164.8   51      33      19      1.95
> >2.4.19-rmap14c [1]      160.1   53      33      21      1.90
> >2.4.20-rc1aa1 [1]       166.2   49      44      23      inf
> >
> >read_load:
> >Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> >2.4.19-rmap15 [1]       0.0     0       0       7       0.00
> >2.4.19-rmap14c [1]      121.0   72      20      7       1.44
> >2.4.20-rc1aa1 [1]       111.2   79      34      14      inf
> >
> >rmap15 OOM'd the cc1 process twice and fscked this run up.
> >
> >list_load:
> >Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> >2.4.19-rmap15 [1]       99.2    84      0       7       1.17
> >2.4.19-rmap14c [1]      100.7   83      0       7       1.20
> >2.4.20-rc1aa1 [1]       96.7    85      0       8       inf
> >
> >mem_load:
> >Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> >2.4.19-rmap15 [1]       111.2   80      96      8       1.31
> >2.4.19-rmap14c [1]      106.1   82      96      9       1.26
> >2.4.20-rc1aa1 [1]       126.3   66      77      3       inf
>
>
> With only one run these numbers appear not to be greatly statistically
> significant, except for the mem_load results.
>
> Con
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.2.0 (GNU/Linux)
>
> iD8DBQE94NTSF6dfvkL3i1gRAkwXAJ9YJmnxvxdiYyESNYLeQa9sbwR+2wCcCZal
> bFKWReK3EP8oayP/GwiIK7c=
> =gYaS
> -----END PGP SIGNATURE-----
>
>



^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [BENCHMARK] rmap15, rmap14c and rc1aa1 with contest
  2002-11-24 13:17 [BENCHMARK] rmap15, rmap14c and rc1aa1 with contest Nero
                   ` (2 preceding siblings ...)
  2002-11-24 13:34 ` Nero
@ 2002-11-24 17:39 ` Rik van Riel
  3 siblings, 0 replies; 5+ messages in thread
From: Rik van Riel @ 2002-11-24 17:39 UTC (permalink / raw)
  To: Nero; +Cc: lkml, Con Kolivas

On Mon, 25 Nov 2002, Nero wrote:

> read_load:
> Kernel [runs]           Time    CPU%    Loads   LCPU%   Ratio
> 2.4.19-rmap15 [1]       0.0     0       0       7       0.00
> 2.4.19-rmap14c [1]      121.0   72      20      7       1.44
> 2.4.20-rc1aa1 [1]       111.2   79      34      14      inf
>
> rmap15 OOM'd the cc1 process twice and fscked this run up.

Known bug, should be fixed with the 2.4.19-rmap15-splitactive
patch I posted on friday:

	http://surriel.com/patches/2.4/2.4.19-rmap15-splitactive

cheers,

Rik
-- 
Bravely reimplemented by the knights who say "NIH".
http://www.surriel.com/		http://guru.conectiva.com/
Current spamtrap:  <a href=mailto:"october@surriel.com">october@surriel.com</a>


^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2002-11-24 17:32 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-11-24 13:17 [BENCHMARK] rmap15, rmap14c and rc1aa1 with contest Nero
2002-11-24 13:23 ` Nero
2002-11-24 13:32 ` Con Kolivas
2002-11-24 13:34 ` Nero
2002-11-24 17:39 ` Rik van Riel

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