|
From: Karim B. <kar...@gm...> - 2006-02-08 16:36:58
|
Hi I have a (big) code we run on a cluster and which runs ok almost everywhere. The cluster is made of 2 types of computers with the same operating system (linux). For one type, no problems but for the other it fails with : ToolSvc.myBTagTool inserting: SV1Tag to tools list. TH1.Print Name = N2TEffSV2, Entries= 195322, Total sum= 195286 TH1.Print Name = N2TNormSV2, Entries= 291102, Total sum= 287128 TH1.Print Name = N2TEffSV2, Entries= 8513, Total sum= 8512 TH1.Print Name = N2TNormSV2, Entries= 417022, Total sum= 354695 TH1.Print Name = TridimMEN2T, Entries= 195322, Total sum= 195247 ==4122== ==4122== Invalid write of size 1 ==4122== at 0x3E5B22FB: Analysis::HistoHelperRoot::smoothASH3D(TH3*, int, int, int, bool) (in /home/atlassgm/releases/rel_11/dist/11.0.4/PhysicsAnalysis/JetTagging/JetTagTools/JetTagTools-00-02-11/i686-slc3-gcc323-opt/libJetTagToolsLib.so) ==4122== Address 0x9C1A0DFF is on thread 1's stack ==4122== Stack overflow in thread 1: can't grow stack to 0x9C1A0DFF ==4122== Can't extend stack to 0x9C1A09D0 during signal delivery for thread 1: ==4122== no stack segment ==4122== ==4122== Process terminating with default action of signal 11 (SIGSEGV) ==4122== Access not within mapped region at address 0x9C1A09D0 ==4122== at 0x3E5B22FB: Analysis::HistoHelperRoot::smoothASH3D(TH3*, int, int, int, bool) (in /home/atlassgm/releases/rel_11/dist/11.0.4/PhysicsAnalysis/JetTagging/JetTagTools/JetTagTools-00-02-11/i686-slc3-gcc323-opt/libJetTagToolsLib.so) ==4122== Stack overflow in thread 1: can't grow stack to 0x9C1A0CDC ==4122== ==4122== Process terminating with default action of signal 11 (SIGSEGV) ==4122== Access not within mapped region at address 0x9C1A0CDC ==4122== at 0x34145998: _vgw(float, long double,...)(...)(long double,...)(short) (vg_intercept.c:51) ==4122== Thanks for any advice to understand what is wrong ... Karim. |
|
From: Julian S. <js...@ac...> - 2006-02-14 16:12:25
|
Looks like you might have a stack overflow. Check out the latest sources from svn and try (easy; see http://www.valgrind.org/downloads/repository.html) I recently increased the allowable stack size to 16M (it was 8). Make sure also your ulimit -s value is set correctly. J On Wednesday 08 February 2006 16:35, Karim Bernardet wrote: > Hi > > I have a (big) code we run on a cluster and which runs ok almost > everywhere. The cluster is made of 2 types of computers with the same > operating system (linux). For one type, no problems but for the other it > fails with : > > ToolSvc.myBTagTool inserting: SV1Tag to tools list. > TH1.Print Name = N2TEffSV2, Entries= 195322, Total sum= 195286 > TH1.Print Name = N2TNormSV2, Entries= 291102, Total sum= 287128 > TH1.Print Name = N2TEffSV2, Entries= 8513, Total sum= 8512 > TH1.Print Name = N2TNormSV2, Entries= 417022, Total sum= 354695 > TH1.Print Name = TridimMEN2T, Entries= 195322, Total sum= 195247 > ==4122== > ==4122== Invalid write of size 1 > ==4122== at 0x3E5B22FB: Analysis::HistoHelperRoot::smoothASH3D(TH3*, > int, int, int, bool) (in > /home/atlassgm/releases/rel_11/dist/11.0.4/PhysicsAnalysis/JetTagging/JetTa >gTools/JetTagTools-00-02-11/i686-slc3-gcc323-opt/libJetTagToolsLib.so) > ==4122== Address 0x9C1A0DFF is on thread 1's stack > ==4122== Stack overflow in thread 1: can't grow stack to 0x9C1A0DFF > ==4122== Can't extend stack to 0x9C1A09D0 during signal delivery for > thread 1: > ==4122== no stack segment > ==4122== > ==4122== Process terminating with default action of signal 11 (SIGSEGV) > ==4122== Access not within mapped region at address 0x9C1A09D0 > ==4122== at 0x3E5B22FB: Analysis::HistoHelperRoot::smoothASH3D(TH3*, > int, int, int, bool) (in > /home/atlassgm/releases/rel_11/dist/11.0.4/PhysicsAnalysis/JetTagging/JetTa >gTools/JetTagTools-00-02-11/i686-slc3-gcc323-opt/libJetTagToolsLib.so) > ==4122== Stack overflow in thread 1: can't grow stack to 0x9C1A0CDC > ==4122== > ==4122== Process terminating with default action of signal 11 (SIGSEGV) > ==4122== Access not within mapped region at address 0x9C1A0CDC > ==4122== at 0x34145998: _vgw(float, long double,...)(...)(long > double,...)(short) (vg_intercept.c:51) > ==4122== > > Thanks for any advice to understand what is wrong ... > > Karim. > > > > ------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. Do you grep through log > files for problems? Stop! Download the new AJAX search engine that makes > searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642 > _______________________________________________ > Valgrind-users mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-users |
|
From: Karim B. <kar...@gm...> - 2006-02-14 16:26:31
|
Julian Seward wrote: >Looks like you might have a stack overflow. Check out the latest >sources from svn and try (easy; see >http://www.valgrind.org/downloads/repository.html) > >I recently increased the allowable stack size to 16M (it was 8). >Make sure also your ulimit -s value is set correctly. > > > Hi ! You are right ! we found that the problem was due to an array float tab[100][100][100] allocate this array on the heap (new/delete) solves the problem. I try all the ulimit options and it didn t work. Cheers Karim >J > >On Wednesday 08 February 2006 16:35, Karim Bernardet wrote: > > >>Hi >> >>I have a (big) code we run on a cluster and which runs ok almost >>everywhere. The cluster is made of 2 types of computers with the same >>operating system (linux). For one type, no problems but for the other it >>fails with : >> >>ToolSvc.myBTagTool inserting: SV1Tag to tools list. >>TH1.Print Name = N2TEffSV2, Entries= 195322, Total sum= 195286 >>TH1.Print Name = N2TNormSV2, Entries= 291102, Total sum= 287128 >>TH1.Print Name = N2TEffSV2, Entries= 8513, Total sum= 8512 >>TH1.Print Name = N2TNormSV2, Entries= 417022, Total sum= 354695 >>TH1.Print Name = TridimMEN2T, Entries= 195322, Total sum= 195247 >>==4122== >>==4122== Invalid write of size 1 >>==4122== at 0x3E5B22FB: Analysis::HistoHelperRoot::smoothASH3D(TH3*, >>int, int, int, bool) (in >>/home/atlassgm/releases/rel_11/dist/11.0.4/PhysicsAnalysis/JetTagging/JetTa >>gTools/JetTagTools-00-02-11/i686-slc3-gcc323-opt/libJetTagToolsLib.so) >>==4122== Address 0x9C1A0DFF is on thread 1's stack >>==4122== Stack overflow in thread 1: can't grow stack to 0x9C1A0DFF >>==4122== Can't extend stack to 0x9C1A09D0 during signal delivery for >>thread 1: >>==4122== no stack segment >>==4122== >>==4122== Process terminating with default action of signal 11 (SIGSEGV) >>==4122== Access not within mapped region at address 0x9C1A09D0 >>==4122== at 0x3E5B22FB: Analysis::HistoHelperRoot::smoothASH3D(TH3*, >>int, int, int, bool) (in >>/home/atlassgm/releases/rel_11/dist/11.0.4/PhysicsAnalysis/JetTagging/JetTa >>gTools/JetTagTools-00-02-11/i686-slc3-gcc323-opt/libJetTagToolsLib.so) >>==4122== Stack overflow in thread 1: can't grow stack to 0x9C1A0CDC >>==4122== >>==4122== Process terminating with default action of signal 11 (SIGSEGV) >>==4122== Access not within mapped region at address 0x9C1A0CDC >>==4122== at 0x34145998: _vgw(float, long double,...)(...)(long >>double,...)(short) (vg_intercept.c:51) >>==4122== >> >>Thanks for any advice to understand what is wrong ... >> >>Karim. >> >> >> >>------------------------------------------------------- >>This SF.net email is sponsored by: Splunk Inc. Do you grep through log >>files for problems? Stop! Download the new AJAX search engine that makes >>searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! >>http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642 >>_______________________________________________ >>Valgrind-users mailing list >>Val...@li... >>https://lists.sourceforge.net/lists/listinfo/valgrind-users >> >> > > > |