<div dir="ltr"><div><div><div><div>Yes that is very useful!<br></div>I took a look at the first one I saw on the list. It goes:<br><br><pre class="gmail-bz_comment_text">==3551== 8 bytes in 1 blocks are definitely lost in loss record 63 of 766
==3551==    at 0x7BC51061: notify_alloc (heap.c:254)
==3551==    by 0x7BC5554F: RtlAllocateHeap (heap.c:1716)
==3551==    by 0x5C85281: XAudio2Create (xaudio_dll.c:2159)
==3551==    by 0x4A1B741: func_xaudio2 (xaudio2.c:1150)
==3551==    by 0x4A1C74F: run_test (test.h:603)
==3551==    by 0x4A1CBAD: main (test.h:687)
==3551== </pre>To fix it I looked at the XAudio2Create function and noticed that IClassFactory *cf was assigned in the call <br>make_xaudio2_factory(&IID_IClassFactory, (void**)&cf);<br></div>which contains<br>struct xaudio2_cf *ret = HeapAlloc(GetProcessHeap(), 0, sizeof(struct xaudio2_cf));<br><br></div>Later in the code IClassFactory_Release(cf); is used, but this wouldn't free up the space on the heap because of how that function is implemented.<br><br></div>I'm typing this all out because I don't know how good my thought process is on this. I'm not used to it so trying to chase all these function calls around is kind of difficult. I would appreciate any feedback<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Sat, Mar 17, 2018 at 12:27 PM, Henri Verbeet <span dir="ltr"><<a href="mailto:hverbeet@gmail.com" target="_blank">hverbeet@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On 17 March 2018 at 19:06, Kieran Duggan <<a href="mailto:kieranduggan15@gmail.com">kieranduggan15@gmail.com</a>> wrote:<br>
> So I've gone off to look for some fixmes. I started looking at a game I've<br>
> tried to run in the past with no success, Fortnite.<br>
> Do any of the fixmes in the bug report look doable for a persons first<br>
> patch?<br>
><br>
</span>In case it helps you or anyone else, we also have a number of issues<br>
found by Valgrind [1], most of which should be fairly straightforward<br>
to debug and most of which should only have a limited scope. If you're<br>
feeling a little more ambitious, I also have lists of Direct3D [2] and<br>
D3DX [3] bugs with various levels of analysis I'd love for someone to<br>
take a look at.<br>
<br>
[1] <a href="https://bugs.winehq.org/buglist.cgi?bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&columnlist=bug_id%2Cshort_desc%2Cbug_status%2Cresolution&keywords=valgrind&order=bug_id%20DESC" rel="noreferrer" target="_blank">https://bugs.winehq.org/<wbr>buglist.cgi?bug_status=<wbr>UNCONFIRMED&bug_status=NEW&<wbr>bug_status=ASSIGNED&bug_<wbr>status=REOPENED&columnlist=<wbr>bug_id%2Cshort_desc%2Cbug_<wbr>status%2Cresolution&keywords=<wbr>valgrind&order=bug_id%20DESC</a><br>
[2] <a href="https://bugs.winehq.org/buglist.cgi?bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&columnlist=bug_id%2Cshort_desc%2Cbug_status%2Cresolution&component=directx-d3d&keywords=download&order=bug_id%20DESC" rel="noreferrer" target="_blank">https://bugs.winehq.org/<wbr>buglist.cgi?bug_status=<wbr>UNCONFIRMED&bug_status=NEW&<wbr>bug_status=ASSIGNED&bug_<wbr>status=REOPENED&columnlist=<wbr>bug_id%2Cshort_desc%2Cbug_<wbr>status%2Cresolution&component=<wbr>directx-d3d&keywords=download&<wbr>order=bug_id%20DESC</a><br>
[3] <a href="https://bugs.winehq.org/buglist.cgi?bug_status=UNCONFIRMED&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&columnlist=bug_id%2Cshort_desc%2Cbug_status%2Cresolution&component=directx-d3dx9&component=directx-d3dx10&keywords=download&order=bug_id%20DESC" rel="noreferrer" target="_blank">https://bugs.winehq.org/<wbr>buglist.cgi?bug_status=<wbr>UNCONFIRMED&bug_status=NEW&<wbr>bug_status=ASSIGNED&bug_<wbr>status=REOPENED&columnlist=<wbr>bug_id%2Cshort_desc%2Cbug_<wbr>status%2Cresolution&component=<wbr>directx-d3dx9&component=<wbr>directx-d3dx10&keywords=<wbr>download&order=bug_id%20DESC</a><br>
</blockquote></div><br></div>