Dear Kostja,
psst, don't SHOUT... Just couldn't resist - Christian
"Konstantin Popov" <kost@sics.se> wrote in message
news:<200301131751.h0DHpCQ01781@sinuhe.sics.se>...
>
> Dear hackers,
>
> I'm sorry to deliver the following BAD news: between November 1st,
> 2002 and December 1st, 2002 our system has suffered a massive
> performance loss, as illustrated EVEN by our benchmark suite:
>
>
> "2002/12/01"
>
> kost (306) ./ozbench ; ./ozbench ; ./ozbench
> nrev:
> times (ms): t:787.99 r:0.0 g:0.0 c:0.0 p:0.0 s:124.66*
> tak_normal:
> times (ms): t:95.33 r:0.0 g:0.0 c:0.0 p:0.0 s:7.90*
> tak_thread:
> times (ms): t:298.28 r:0.0 g:0.0 c:0.0 p:0.0 s:21.71*
> fd_simple:
> times (ms): t:44.34 r:0.0 g:0.0 c:0.0 p:0.0 s:0.96
> compiler_simple:
> times (ms): t:125.14 r:0.0 g:0.0 c:0.0 p:0.0 s:8.57*
> differentiate:
> times (ms): t:956.66 r:0.0 g:0.0 c:0.0 p:0.0 s:88.66*
> knights:
> times (ms): t:809.33 r:0.0 g:0.0 c:0.0 p:0.0 s:116.66*
> bridge:
> times (ms): t:201.81 r:0.0 g:0.0 c:0.0 p:0.0 s:19.9*
> port:
> times (ms): t:679.0 r:0.0 g:0.0 c:0.0 p:0.0 s:56.0*
> rec_normal:
> times (ms): t:592.0 r:0.0 g:0.0 c:0.0 p:0.0 s:0.50
> nrev:
> times (ms): t:785.99 r:0.0 g:0.0 c:0.0 p:0.0 s:123.33*
> tak_normal:
> times (ms): t:95.21 r:0.0 g:0.0 c:0.0 p:0.0 s:7.82*
> tak_thread:
> times (ms): t:295.42 r:0.0 g:0.0 c:0.0 p:0.0 s:16.0*
> fd_simple:
> times (ms): t:44.62 r:0.0 g:0.0 c:0.0 p:0.0 s:1.37*
> compiler_simple:
> times (ms): t:127.71 r:0.0 g:0.0 c:0.0 p:0.0 s:7.71*
> differentiate:
> times (ms): t:951.33 r:0.0 g:0.0 c:0.0 p:0.0 s:96.0*
> knights:
> times (ms): t:789.33 r:0.0 g:0.0 c:0.0 p:0.0 s:133.33*
> bridge:
> times (ms): t:198.18 r:0.0 g:0.0 c:0.0 p:0.0 s:16.72*
> port:
> times (ms): t:678.0 r:0.0 g:0.0 c:0.0 p:0.0 s:55.0*
> rec_normal:
> times (ms): t:590.50 r:0.0 g:0.0 c:0.0 p:0.0 s:0.0
> nrev:
> times (ms): t:785.33 r:0.0 g:0.0 c:0.0 p:0.0 s:129.33*
> tak_normal:
> times (ms): t:95.13 r:0.0 g:0.0 c:0.0 p:0.0 s:6.60*
> tak_thread:
> times (ms): t:295.71 r:0.0 g:0.0 c:0.0 p:0.0 s:20.57*
> fd_simple:
> times (ms): t:44.55 r:0.0 g:0.0 c:0.0 p:0.0 s:1.65*
> compiler_simple:
> times (ms): t:126.57 r:0.0 g:0.0 c:0.0 p:0.0 s:7.42*
> differentiate:
> times (ms): t:961.33 r:0.0 g:0.0 c:0.0 p:0.0 s:92.66*
> knights:
> times (ms): t:802.0 r:0.0 g:0.0 c:0.0 p:0.0 s:124.66*
> bridge:
> times (ms): t:200.90 r:0.0 g:0.0 c:0.0 p:0.0 s:17.27*
> port:
> times (ms): t:674.66 r:0.0 g:0.0 c:0.0 p:0.0 s:46.66*
> rec_normal:
> times (ms): t:591.50 r:0.0 g:0.0 c:0.0 p:0.0 s:0.0
>
>
> "2002/11/01"
>
> kost (328) ./ozbench ; ./ozbench ; ./ozbench
> nrev:
> times (ms): t:572.0 r:0.0 g:0.0 c:0.0 p:0.0 s:137.0*
> tak_normal:
> times (ms): t:68.11 r:0.0 g:0.0 c:0.0 p:0.0 s:8.35*
> tak_thread:
> times (ms): t:298.85 r:0.0 g:0.0 c:0.0 p:0.0 s:18.28*
> fd_simple:
> times (ms): t:41.76 r:0.0 g:0.0 c:0.0 p:0.0 s:1.35*
> compiler_simple:
> times (ms): t:117.42 r:0.0 g:0.0 c:0.0 p:0.0 s:8.57*
> differentiate:
> times (ms): t:767.33 r:0.0 g:0.0 c:0.0 p:0.0 s:90.66*
> knights:
> times (ms): t:790.0 r:0.0 g:0.0 c:0.0 p:0.0 s:109.33*
> bridge:
> times (ms): t:198.44 r:0.0 g:0.0 c:0.0 p:0.0 s:17.55*
> port:
> times (ms): t:588.0 r:0.0 g:0.0 c:0.0 p:0.0 s:44.0*
> rec_normal:
> times (ms): t:286.85 r:0.0 g:0.0 c:0.0 p:0.0 s:0.0
> nrev:
> times (ms): t:574.0 r:0.0 g:0.0 c:0.0 p:0.0 s:138.50*
> tak_normal:
> times (ms): t:67.70 r:0.0 g:0.0 c:0.0 p:0.0 s:7.70*
> tak_thread:
> times (ms): t:294.25 r:0.0 g:0.0 c:0.0 p:0.0 s:17.25*
> fd_simple:
> times (ms): t:41.23 r:0.0 g:0.0 c:0.0 p:0.0 s:1.29*
> compiler_simple:
> times (ms): t:118.85 r:0.0 g:0.0 c:0.0 p:0.0 s:7.71*
> differentiate:
> times (ms): t:766.0 r:0.0 g:0.0 c:0.0 p:0.0 s:88.66*
> knights:
> times (ms): t:782.66 r:0.0 g:0.0 c:0.0 p:0.0 s:114.66
> bridge:
> times (ms): t:201.80 r:0.0 g:0.0 c:0.0 p:0.0 s:19.60*
> port:
> times (ms): t:592.50 r:0.0 g:0.0 c:0.0 p:0.0 s:55.50
> rec_normal:
> times (ms): t:286.28 r:0.0 g:0.0 c:0.0 p:0.0 s:0.28
> nrev:
> times (ms): t:571.0 r:0.0 g:0.0 c:0.0 p:0.0 s:154.0
> tak_normal:
> times (ms): t:67.64 r:0.0 g:0.0 c:0.0 p:0.0 s:7.52*
> tak_thread:
> times (ms): t:285.71 r:0.0 g:0.0 c:0.0 p:0.0 s:18.28*
> fd_simple:
> times (ms): t:41.10 r:0.0 g:0.0 c:0.0 p:0.0 s:1.10*
> compiler_simple:
> times (ms): t:118.85 r:0.0 g:0.0 c:0.0 p:0.0 s:6.28*
> differentiate:
> times (ms): t:757.33 r:0.0 g:0.0 c:0.0 p:0.0 s:88.0*
> knights:
> times (ms): t:803.33 r:0.0 g:0.0 c:0.0 p:0.0 s:106.0*
> bridge:
> times (ms): t:201.0 r:0.0 g:0.0 c:0.0 p:0.0 s:16.60*
> port:
> times (ms): t:590.50 r:0.0 g:0.0 c:0.0 p:0.0 s:50.0*
> rec_normal:
> times (ms): t:287.14 r:0.0 g:0.0 c:0.0 p:0.0 s:0.0
>
>
> A "real" application is set back up to 20%.
>
> In the log I've noticed some modifications made by Leif (in the
> emulator) and Denys (in the os.cc). I personally have no time to
> investigate it further.
>
> Since we have some important deadlines ahead, we need a solution ASAP.
>
> And at that I'm not even mentioning how that version could make its
> way into CVS in the first place ;-[
>
> Thanks in advance,
> Cheers,
>
> --- Kostja.
> -
> Please send submissions to hackers@mozart-oz.org
> and administriva mail to hackers-request@mozart-oz.org.
> The Mozart Oz web site is at http://www.mozart-oz.org/.
-
Please send submissions to hackers@mozart-oz.org
and administriva mail to hackers-request@mozart-oz.org.
The Mozart Oz web site is at http://www.mozart-oz.org/.