|
From: Aman A. <ama...@gm...> - 2016-04-04 07:15:36
|
Hi, I'm currently working in an application which uses java/jsp in front end (client side) and C++ as middle tier on a Linux server (server side). Will valgrind still be able to help me identify memory leaks? Thanks, Ishan |
|
From: John R. <jr...@bi...> - 2016-04-04 13:41:56
|
> I'm currently working in an application which uses java/jsp in front end (client side) and C++ as middle tier on a Linux server (server side). Will valgrind still be able to help me identify memory leaks? If the C++ server side is a separate process, and if that process can be invoked under valgrind, and if the process calls exit(), and if the monitoring slowdown of 20X to 50X in CPU time can be tolerated, and if the memory size expansion of about 2X can be handled, and if all allocation uses malloc/free/new/delete (no private allocation arenas), then yes, valgrind can help find memory leaks. Java: no. |
|
From: Aman A. <ama...@gm...> - 2016-04-04 16:41:15
|
Thank you so much for your valuable response. On Mon 4 Apr, 2016, 7:12 PM John Reiser, <jr...@bi...> wrote: > > I'm currently working in an application which uses java/jsp in front end > (client side) and C++ as middle tier on a Linux server (server side). Will > valgrind still be able to help me identify memory leaks? > > If the C++ server side is a separate process, and if that process > can be invoked under valgrind, and if the process calls exit(), > and if the monitoring slowdown of 20X to 50X in CPU time can be > tolerated, and if the memory size expansion of about 2X can be > handled, and if all allocation uses malloc/free/new/delete (no > private allocation arenas), then yes, valgrind can help find memory leaks. > Java: no. > > > ------------------------------------------------------------------------------ > _______________________________________________ > Valgrind-developers mailing list > Val...@li... > https://lists.sourceforge.net/lists/listinfo/valgrind-developers > |
|
From: Aman A. <ama...@gm...> - 2016-08-16 13:32:44
|
Hi, Can anyone tell me the meaning of the below error shown in Valgrind report I ran on my code? Below is the error: 48 bytes in 1 block definitely lost in... at 0x4A0610C: malloc (vg_replace_malloc.c:195) .... (and then some files that don't really matter) .... Note: above errors are pointing to files that none can modify. Here's one more such example where I can do nothing. 56 bytes in 1 block are possibly lost in ... at 0x4A0610C: malloc (vg_replace_malloc.c:195) by 0x13120DDF: ??? (in /apps/ecl/oracle/product/11.2.0/client_1/lib/libclntsh.so.11.1) .... (and then repeating the same line above, it doesn't really point to some specific location or line number in my code) Also, is it possible that valgrind might be incorrectly reporting these errors? Please let me know if more information is needed to understand above problem. It's kind of real urgent. Please do reply to this. On Mon, 4 Apr 2016 at 12:45 PM, Aman Ahuja <ama...@gm...> wrote: > Hi, > I'm currently working in an application which uses java/jsp in front end > (client side) and C++ as middle tier on a Linux server (server side). Will > valgrind still be able to help me identify memory leaks? > > Thanks, > Ishan > |