https://bugs.winehq.org/show_bug.cgi?id=56256 --- Comment #3 from Bernhard Übelacker <bernhardu(a)mailbox.org> --- Created attachment 75968 --> https://bugs.winehq.org/attachment.cgi?id=75968 Successful startup, when correcting stackpointer in windbg. The crash happens because of having an invalid memory address in register $esi. I think this register $esi is getting pushed to the stack at 0x4aa605, and getting retrieved at 0x4aa9df. When this part is executed the third and the following times, the stackpointer is 4 bytes too low, when $esi should be retrieved from the stack. Unfortunately I could not find out where this is caused. There are several functions called, but I saw no obvious candidate for a wrong calling convention. If I adjust the stackpointer in debugger before the pop instruction, procexp could startup successfully (session attached). 0x4aa605: push %esi 0x4aa9df: pop %esi -- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.