<html>
<head>
<base href="https://bugs.freedesktop.org/">
</head>
<body><span class="vcard"><a class="email" href="mailto:fourdan@xfce.org" title="Olivier Fourdan <fourdan@xfce.org>"> <span class="fn">Olivier Fourdan</span></a>
</span> changed
<a class="bz_bug_link
bz_status_NEW "
title="NEW - Crash on resuming laptop with screen connected"
href="https://bugs.freedesktop.org/show_bug.cgi?id=107314">bug 107314</a>
<br>
<table border="1" cellspacing="0" cellpadding="8">
<tr>
<th>What</th>
<th>Removed</th>
<th>Added</th>
</tr>
<tr>
<td style="text-align:right;">CC</td>
<td>
</td>
<td>fourdan@xfce.org
</td>
</tr></table>
<p>
<div>
<b><a class="bz_bug_link
bz_status_NEW "
title="NEW - Crash on resuming laptop with screen connected"
href="https://bugs.freedesktop.org/show_bug.cgi?id=107314#c7">Comment # 7</a>
on <a class="bz_bug_link
bz_status_NEW "
title="NEW - Crash on resuming laptop with screen connected"
href="https://bugs.freedesktop.org/show_bug.cgi?id=107314">bug 107314</a>
from <span class="vcard"><a class="email" href="mailto:fourdan@xfce.org" title="Olivier Fourdan <fourdan@xfce.org>"> <span class="fn">Olivier Fourdan</span></a>
</span></b>
<pre>Created <span class=""><a href="attachment.cgi?id=141463" name="attach_141463" title="[RFC PATCH xserver] xwayland: ignore sync callback if window is destroyed">attachment 141463</a> <a href="attachment.cgi?id=141463&action=edit" title="[RFC PATCH xserver] xwayland: ignore sync callback if window is destroyed">[details]</a></span> <a href='page.cgi?id=splinter.html&bug=107314&attachment=141463'>[review]</a>
[RFC PATCH xserver] xwayland: ignore sync callback if window is destroyed
(In reply to Daniel Stone from <a href="show_bug.cgi?id=107314#c6">comment #6</a>)
<span class="quote">> Seems safe enough. When we free the vblanks (which, when we're aborting
> here, can only happen because the window has been destroyed), [...]</span >
Roman pointed out on the ML a crash possibly caused by that patch.
What about this patch then?
On destroy, `xwl_present_cleanup()` would free the `xwl_present_window` but
leave its reference in the window's privates, so that other functions could
still find it and use freed memory.
That patch there <a href="https://patchwork.freedesktop.org/patch/247271/">https://patchwork.freedesktop.org/patch/247271/</a> fixes that, so
with this if we just check for `xwl_present_window` being nun-null in
`xwl_present_sync_callback()` we should avoid the crash, cleanly, no?
If the idea makes sense (and if someone cwould be willing to try it) and if it
works, I would send that to the ML for further review.</pre>
</div>
</p>
<hr>
<span>You are receiving this mail because:</span>
<ul>
<li>You are the assignee for the bug.</li>
</ul>
</body>
</html>