[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Bug ld/27954] New: Exception handling info garbled on x86_64-w64-mingw3

From: dominik_strasser at mentor dot com
Subject: [Bug ld/27954] New: Exception handling info garbled on x86_64-w64-mingw32
Date: Fri, 04 Jun 2021 08:40:44 +0000


            Bug ID: 27954
           Summary: Exception handling info garbled on x86_64-w64-mingw32
           Product: binutils
           Version: 2.35.2
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: ld
          Assignee: unassigned at sourceware dot org
          Reporter: dominik_strasser at mentor dot com
  Target Milestone: ---

I am having a long standing issue with exception handling on

The symptom is that I am getting a crash in ntdll!RtlUnwindEx when unwinding
the stack between a thrown assertion and the catch handler.

I am suspecting an issue in binutils(ld) because the crash vanishes when I bind
a different program with the identical objects.

I ran my program through wine which gives some additional information, namely
an illegal EH information:
020c:fixme:seh:RtlVirtualUnwind unknown unwind info version 4 at
020c:fixme:seh:dump_unwind_info **** func 4cc6d40-4cc712f
020c:fixme:seh:dump_unwind_info unwind info at 00000000080E350C flags 0 prolog
0xd0 bytes function 00000000050C6D40-00000000050C712F
020c:fixme:seh:dump_unwind_info     0x1: PUSH_MACHFRAME 0
020c:fixme:seh:dump_unwind_info     0x6: pushq %rax
020c:fixme:seh:dump_unwind_info CHANIN 0

According to all documentation I found, only version 1 and 2 are legal values.
Additionally wine prints:
020c:fixme:seh:PUSH_MACHFRAME is not the last opcode.

I cannot provide an example as a lot of source code is involved, but I can
compile binutils myself, and do any experiment you like.

You are receiving this mail because:
You are on the CC list for the bug.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]