|
From: Julian S. <js...@ac...> - 2014-07-15 08:36:49
|
I think we saw recently (last couple of months?) a case where Memcheck didn't do any intercepts with uClibC, and it turned out that the problem was that uClibc had not been built with LD_PRELOAD support. So the intercept .so never got loaded into the process. I wonder if that is the case here? J On 07/09/2014 12:43 PM, Philippe Waroquiers wrote: > On Wed, 2014-07-09 at 15:58 +0530, Jonnavithula Sharma wrote: >> Hi , >> > ... >> But somehow there were no leaks detected and the output is as >> below(attached is the complete valgrind output) >> > ... >> Any suggestions or help is highly appreciated. >> >> >> I am compiling my application with all debug options as well. I was >> struck and not able to proceed ahead. > > The most frequent source of this kind of problems is the fact > that the malloc library is statically linked with your app. > > You must then indicate that to valgrind, using the option > --soname-synonyms=... > (see user manual for details). > > If that is not the case, then the 2nd most frequent source of problem > is that the redirection of the malloc etc symbols are not done > due to the name of the library not being what is expected by valgrind. > Use the options -v -v -v -d -d -d --trace-redir=yes > and try to understand what is going wrong. > > > Philippe > > > > > ------------------------------------------------------------------------------ > Open source business process management suite built on Java and Eclipse > Turn processes into business applications with Bonita BPM Community Edition > Quickly connect people, data, and systems into organized workflows > Winner of BOSSIE, CODIE, OW2 and Gartner awards > http://p.sf.net/sfu/Bonitasoft > _______________________________________________ > Valgrind-users mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-users > |