lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [day] [month] [year] [list]
Message-ID: <CA+Lpn7bkfcQu+i=cHW8Yo6PsJUJzFVbhbMyTosH0dzttf+bNHQ@mail.gmail.com>
Date:   Thu, 19 Oct 2017 00:26:14 +0530
From:   Kartikeyan Sadasivuni <s.kartikeyan@...il.com>
To:     linux-kernel@...r.kernel.org
Subject: driver mmap()'ed vma in process coredump

I have a process which does mmap into a kernel driver to map DMA
packet memory into user space.
When the process crashes want this memory to be part of coredump for
post crash analysis.

As well documented VM_IO flag for this vma prevents it.
Is always_dump_vma() calling vma->vm_ops "name" function pointer
return non-NULL a workaround/solution  ?

When trying this by setting vma->vm_ops name function pointer
returning non-NULL,
VMA is filled with all zeroes in coredump because get_dump_page() is
returning NULL for VM_IO flag set vmas.

Previously saw a thread w.r.t framebuffer device having similar requirement,
but I see VM_ALWAYSDUMP added and removed from linux kernel.
What is the current way to achieve this functionality if at all its possible ?


Kartikeyan

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ