Valgrind Memcheck Output

Question

I'm using valgrind to find memory leak on my system, and I received this output

==9697== Memcheck, a memory error detector
==9697== Copyright (C) 2002-2011, and GNU GPL'd, by Julian Seward et al.
==9697== Using Valgrind-3.7.0 and LibVEX; rerun with -h for copyright info
==9697== Command: bin/vfirewall-monitor
==9697== 
==9697== 
==9697== HEAP SUMMARY:
==9697==     in use at exit: 0 bytes in 0 blocks
==9697==   total heap usage: 1 allocs, 1 frees, 37 bytes allocated
==9697== 
==9697== All heap blocks were freed -- no leaks are possible
==9697== 
==9697== For counts of detected and suppressed errors, rerun with: -v
==9697== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 2 from 2)
==9700== Thread 2:
==9700== Conditional jump or move depends on uninitialised value(s)
==9700==    at 0x56DE3B1: vfprintf (vfprintf.c:1630)
==9700==    by 0x5706441: vsnprintf (vsnprintf.c:120)
==9700==    by 0x56E6971: snprintf (snprintf.c:35)
==9700==    by 0x403A1A: save_interfaces_info (interfaces.c:351)
==9700==    by 0x403DC4: get_all_system_info (kernel.c:135)
==9700==    by 0x547DE99: start_thread (pthread_create.c:308)
==9700==    by 0x57873FC: clone (clone.S:112)
==9700==  Uninitialised value was created by a heap allocation
==9700==    at 0x4C2B6CD: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==9700==    by 0x403D6F: get_all_system_info (kernel.c:118)
==9700==    by 0x547DE99: start_thread (pthread_create.c:308)
==9700==    by 0x57873FC: clone (clone.S:112)
==9700==

It says there's no error but I have a uninitialised value on kernel.c:118.

And this is my 1kernel.c1 at 118:

117    Interface * ifaces;
118    ifaces = malloc(sizeof (Interface));
119    ifaces->next_interface = NULL;

I don't understand what is wrong here. Am I looking in the wrong place? Or I'm reading valgrind logs wrong?

Per Johansson · Accepted Answer · 2014-03-05T22:42:10.700

6

The error is at interfaces.c:351, it's calling snprintf with an uninitialized value.

That value was allocated at kernel.c:118 using malloc. As you probably know, malloc does not initialize the memory it returns, it might contain garbage. That's what it's complaining about.

Most likely your Interface object has a char name[] or some such as the first member, and you pass that member to snprintf without ever setting it.

Edit: The reason I know the problem is with the first member of Interface is because if it's some other memory valgrind would say something like Uninitialised value is 4 bytes into a block created by a heap allocation instead of just Uninitialised value was created by a heap allocation.

edited Mar 05 '14 at 22:42

answered Mar 05 '14 at 22:33

Per Johansson

6,697
27
34

Actually, i have this in my interfaces.c:351 `char sql[300] = ""; snprintf(sql, sizeof (sql), "INSERT INTO interfaces (nome,endereco_ipv4,mac_address, ativa) VALUES ('%s','%s','%s','%s')", iface->interface_name, iface->interface_addr_ipv4, iface->interface_mac_addr, iface->interface_active == INTERFACE_ACTIVE ? "true" : "false");` – Vitor Villar Mar 05 '14 at 22:39
The uninitialized value is the first member of `Interface`. See edit. – Per Johansson Mar 05 '14 at 22:42

Valgrind Memcheck Output

1 Answers1