<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Hello,<br>
I've tried to submit the new test for WM_SETTEXT button handler
(user32/tests/msg.c), corresponding to the patch that I proposed for
the bug #25790.<br>
But Marvin reports that this test (user32 msg) fails on W7PROX64
(64-bit app).<br>
I've tried to run Marvin's user32_test64.exe file on a Windows 2008
R2 and got the same. Apparently, it stucks in a loop (somewhere
after my code). Then I compiled 64-bit user32_test.exe (using
amd64-mingw32msvc-gcc (GCC) 4.4.4
<style type="text/css">p { margin-bottom: 0.21cm; }code.western { font-family: "DejaVu Sans Mono",monospace; }code.cjk { font-family: "WenQuanYi Zen Hei",monospace; }code.ctl { font-family: "DejaVu Sans Mono",monospace; }</style><font
face="DejaVu Sans Mono, monospace"><code class="western"></code></font>)
and ran it on the same machine. And it passed without any errors.
I've setup another MinGW-w64 package, compiled another version of
user32_test.exe (using x86_64-w64-mingw32-gcc (GCC) 4.4.5 20101001)
and it worked too.<br>
Meanwhile, Marvin uses GCC 4.4.1-1a (as I found from his
user32_test64.exe).<br>
I think that user32_test64 fail related to GCC version (and the test
would fail with or without my additions).<br>
<br>
So, what can I do to get my initial patch for the bug commited? To
setup GCC 4.4.1 and get a user32_test.exe, that would not work on
Windows x64, then to find what causes the app to stuck and propose a
patch for msg.c to workaround the issue? Or maybe just wait till
Marvin gets GCC updated? Or maybe the initial patch can be commited
without the corresponding test?<br>
<br>
WBR,<br>
Alexander<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
</body>
</html>