|
From: Chris E. <chr...@gm...> - 2007-01-08 23:20:15
|
Hello, When I run my java application using lackey, the program runs at normal speed and then finishes with no follow-up data that lackey normally prints out. I've tried looking at the verbose information during execution: dizzy% valgrind --smc-check=all -v --tool=lackey ../../java/bin/java binarytrees 5 ==30242== Lackey, an example Valgrind tool. ==30242== Copyright (C) 2002-2007, and GNU GPL'd, by Nicholas Nethercote. ==30242== Using LibVEX rev 1720, a library for dynamic binary translation. ==30242== Copyright (C) 2004-2007, and GNU GPL'd, by OpenWorks LLP. ==30242== Using valgrind-3.3.0.SVN, a dynamic binary instrumentation framework. ==30242== Copyright (C) 2000-2007, and GNU GPL'd, by Julian Seward et al. ==30242== --30242-- Command line --30242-- ../../java/bin/java --30242-- binarytrees --30242-- 5 --30242-- Startup, with flags: --30242-- --smc-check=all --30242-- -v --30242-- --tool=lackey --30242-- Contents of /proc/version: --30242-- Linux version 2.6.18-1.2869.fc6 ( bre...@hs...) (gcc version 4.1.1 20061011 (Red Hat 4.1.1-30)) #1 SMP Wed Dec 20 14:51:46 EST 2006 --30242-- Arch and hwcaps: X86, x86-sse1-sse2 --30242-- Page sizes: currently 4096, max supported 4096 --30242-- Valgrind library directory: /cs/student/ceberz/arch/valgrind/lib/valgrind --30242-- Reading syms from /lib/ld-2.5.so (0x31A000) --30242-- Reading syms from /cs/student/ceberz/arch/j2re1.4.2_13/bin/java (0x8048000) --30242-- object doesn't have a symbol table --30242-- Reading syms from /cs/student/ceberz/arch/valgrind/lib/valgrind/x86-linux/lackey (0x38000000) --30242-- object doesn't have a dynamic symbol table --30242-- Reading syms from /cs/student/ceberz/arch/valgrind/lib/valgrind/x86-linux/vgpreload_core.so (0x4001000) --30242-- Reading syms from /lib/libpthread-2.5.so (0xD18000) --30242-- Reading syms from /lib/libdl-2.5.so (0xD12000) --30242-- Reading syms from /lib/libc-2.5.so (0x101000) stretch tree of depth 7 check: -1 128 trees of depth 4 check: -128 32 trees of depth 6 check: -32 long lived tree of depth 6 check: -1 dizzy% But there is nothing there that I understand as an error. Java applications had been running just fine, but recently all the computers in my lab had been upgraded with a new OS, moving from fedora core 4 to fedora core 6, and now I have this problem. In both cases I had been using sun JDK 1.5. I'm hoping someone can point me in the right direction as to what might cause a problem like this, as right now I have very few details other than an almost useless "one day it stopped working." Thanks for any help, - Chris Eberz |
|
From: Julian S. <js...@ac...> - 2007-01-08 23:27:05
|
> When I run my java application using lackey, the program runs at normal > speed and then finishes with no follow-up data that lackey normally prints > out. My guess is your java command is just a script or whatever that starts another process to do the real work, so by default V doesn't trace into that. Try again with --trace-children=yes. J |
|
From: Nicholas N. <nj...@cs...> - 2007-01-08 23:50:48
|
On Mon, 8 Jan 2007, Julian Seward wrote: >> When I run my java application using lackey, the program runs at normal >> speed and then finishes with no follow-up data that lackey normally prints >> out. > > My guess is your java command is just a script or whatever that starts > another process to do the real work, so by default V doesn't trace into > that. Try again with --trace-children=yes. This is an FAQ, although it's in the "Memcheck doesn't find my bug" section, so the connection might not have been obvious in this case. Should we change the default to --trace-children=yes? It seems to be what people want. (The last survey showed 10% of people always use it, and 6.5% sometimes use it, and the rest probably don't have multi-process programs...) N |
|
From: Chris E. <chr...@gm...> - 2007-01-08 23:56:53
|
Thanks for the very quick reply. Tracing the children seems to work for me. Searching through the mailing list archives (something I should have done more thuroughly before asking) I see the same answer is given. Thanks again, - Chris On 1/8/07, Nicholas Nethercote <nj...@cs...> wrote: > > On Mon, 8 Jan 2007, Julian Seward wrote: > > >> When I run my java application using lackey, the program runs at normal > >> speed and then finishes with no follow-up data that lackey normally > prints > >> out. > > > > My guess is your java command is just a script or whatever that starts > > another process to do the real work, so by default V doesn't trace into > > that. Try again with --trace-children=yes. > > This is an FAQ, although it's in the "Memcheck doesn't find my bug" > section, > so the connection might not have been obvious in this case. > > Should we change the default to --trace-children=yes? It seems to be what > people want. (The last survey showed 10% of people always use it, and > 6.5% > sometimes use it, and the rest probably don't have multi-process > programs...) > > N > > ------------------------------------------------------------------------- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to share > your > opinions on IT & business topics through brief surveys - and earn cash > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > _______________________________________________ > Valgrind-users mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-users > |