Memory leaks in ScardEstablishContext()

[KkVova]

I use libpcsclite and recently faced with some memory leaks while running ScardEstablishContext() under Valgrind.

I came across this link and tried to find topic about it, but unsuccessfully.

I’ve made this sample to reproduce leaks.

g++ test_leaks.cpp -o test_leaks -ldl -I/usr/include/PCSC

//test_leaks.cpp
 
#include <dlfcn.h>
#include <cstddef>
#include <winscard.h>

int main()
{
     typedef LONG (*SCardEstablishContext_t)(DWORD, LPCVOID,LPCVOID,LPSCARDCONTEXT);
     typedef LONG (*SCardReleaseContext_t)(SCARDCONTEXT);

     void* module = dlopen("libpcsclite.so", RTLD_NOW);

     SCardEstablishContext_t SCardEstablishContext = (SCardEstablishContext_t)dlsym(module, "SCardEstablishContext");
     SCardReleaseContext_t SCardReleaseContext = (SCardReleaseContext_t)dlsym(module, "SCardReleaseContext");

     SCARDCONTEXT ctx;
     SCardEstablishContext(0, NULL, NULL, &ctx);
     SCardReleaseContext(ctx);

     dlclose(module);
     return 0;
}

Valgrind report

valgrind ./test_leaks
==9662== Memcheck, a memory error detector
==9662== Copyright (C) 2002-2017, and GNU GPL'd, by Julian Seward et al.
==9662== Using Valgrind-3.13.0 and LibVEX; rerun with -h for copyright info
==9662== Command: ./test_leaks
==9662== 
==9662== 
==9662== HEAP SUMMARY:
==9662==     in use at exit: 1,860 bytes in 9 blocks
==9662==   total heap usage: 19 allocs, 10 frees, 3,741 bytes allocated
==9662== 
==9662== LEAK SUMMARY:
==9662==    definitely lost: 40 bytes in 1 blocks
==9662==    indirectly lost: 72 bytes in 3 blocks
==9662==      possibly lost: 0 bytes in 0 blocks
==9662==    still reachable: 1,748 bytes in 5 blocks
==9662==         suppressed: 0 bytes in 0 blocks
==9662== Rerun with --leak-check=full to see details of leaked memory
==9662== 
==9662== For counts of detected and suppressed errors, rerun with: -v
==9662== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)

Was it fixed or what can I do to imporve the situation?

Thanks for attention.

[LudovicRousseau]

I think this is this bug https://salsa.debian.org/rousseau/PCSC/issues/1
No, it is not yet fixed.

Do you plan to propose a fix?

[LudovicRousseau]

@ValZapod it is not hard to fix. But the issue is really minor.
Please propose a patch if you want.

[wagnst]

We have the same issue. Is there any update here? This seems not to be fixed.

[LudovicRousseau]

@wagnst do you use dlopen("libpcsclite.so", RTLD_NOW); in your code?

[LudovicRousseau]

Fixed in c94aeb5