[Bug 36249] valgrind shows two potential leaks in d3d10/tests/effect.c

https://bugs.winehq.org/show_bug.cgi?id=36249

--- Comment #2 from Austin English  ---
==32334== 12,068 bytes in 1 blocks are possibly lost in loss record 865 of 892
==32334==    at 0x7BC50A9B: RtlAllocateHeap (heap.c:255)
==32334==    by 0x5058901: wined3d_create (wined3d_main.c:95)
==32334==    by 0x4F55F50: dxgi_factory_create (factory.c:332)
==32334==    by 0x4F54689: CreateDXGIFactory (dxgi_main.c:80)
==32334==    by 0x4AD134F: D3D10CreateDevice (d3d10_main.c:63)
==32334==    by 0x4A76475: create_device (effect.c:30)
==32334==    by 0x4A861EE: func_effect (effect.c:2723)
==32334==    by 0x4A8FBAC: run_test (test.h:584)
==32334==    by 0x4A8FF9B: main (test.h:654)
==32334==

==364== 12,068 bytes in 1 blocks are definitely lost in loss record 1,794 of
1,854
==364==    at 0x7BC50A9B: RtlAllocateHeap (heap.c:255)
==364==    by 0x506F901: wined3d_create (wined3d_main.c:95)
==364==    by 0x4F70F50: dxgi_factory_create (factory.c:332)
==364==    by 0x4F6F689: CreateDXGIFactory (dxgi_main.c:80)
==364==    by 0x4AB134F: D3D10CreateDevice (d3d10_main.c:63)
==364==    by 0x4A56475: create_device (effect.c:30)
==364==    by 0x4A5651B: test_effect_constant_buffer_type (effect.c:95)
==364==    by 0x4A65D4B: func_effect (effect.c:4278)
==364==    by 0x4A6FBAC: run_test (test.h:584)
==364==    by 0x4A6FF9B: main (test.h:654)
==364==