Hi all,
I've just checked in a some changes to BRAT:
* Removed serious memory leaks in BrDigitizeMultSi. Here is some
memory stats for "TestGeantDig", run on a Redhat 6.1 Linux, 733 MHz
Pentium III, 256 MB RAM, with 100000 single Lambda events output
from gbrahms
CPU Time Before fix After fix After Fix with -g
---------+------------+-----------+------------------
0:05 7 MB 7 MB 7 MB
0:25 49 14 14
1:00 86 15 16
2:30 139 18 19
5:00 205 22 25
7:00 233 26 28
7:30 238 26 29
8:15 241 27 30
11:30 <killed> <killed> 35
16:00 42
19:00 47
So apperantly there is still some memory leak somewhere (though
probably not in "BrDigitizeMultSi")
* "BrDigitizeMultSi" now uses "BrDigSi" for output, similar to
"BrDigitizeMultTile"
* "BrDigitizeMultTile" uses name for object from "BrTableNames.h"
* Some minor changes to "BrDigMultSi" and "BrDigSi"
And finally two small pleas:
When you commit change to the CVS, could you please remove as many as
possible of the compiler warnings. The output from the compiler/linker
is scrambled with this info, which makes it harder to detect
errors. Also, in general, compiler shouldn't be ignored. They contain
usefull information of weaknesses in the code, and is meant to help
you write better and safer code. Thanks for helping out keeping the
code clean.
As Flemming pointed out in an earlier mail, could you please send a
small note to brahms-dev-l@bnl.gov. There were at least two
un-announced updates the last two-days. Again, this is something that
makes life a bit harder for everyone. Thanks for helping everyone
out.
Cheers,
Christian -----------------------------------------------------------
Holm Christensen Phone: (+45) 35 35 96 91
Sankt Hansgade 23, 1. th. Office: (+45) 353 25 305
DK-2200 Copenhagen N Web: www.nbi.dk/~cholm
Denmark Email: cholm@nbi.dk
This archive was generated by hypermail 2b29 : Wed May 24 2000 - 10:40:06 EDT