[Top][All Lists]

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

Re: [Libunwind-devel] SIGSEGV in ARM+Alarm handler

From: Harald Servat
Subject: Re: [Libunwind-devel] SIGSEGV in ARM+Alarm handler
Date: Tue, 13 Oct 2015 11:37:47 +0200
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.3.0

On 10/10/2015 08:00 PM, Arun Sharma wrote:
On Thu, Oct 8, 2015 at 3:40 AM, Harald Servat <address@hidden> wrote:

   according to the same core and listing in gdb, the instr value is set by
dwarf_get which does not returns an error.

   So can anyone experienced with ARM give his or her thoughts?

I'm not experienced with ARM. But here's what I'd do:

First of all, thank you for the pointers Arun.

Run with UNW_DEBUG_LEVEL=n (list archives have more info) and get a
trace that shows why dwarf_step() failed.

That generated lots of outputs. You can get it from

Recognizing signal frames using instruction signature matching such
as: 0xE92DD800 is fragile and error prone. If the libc you're using
has the right CFI annotation around the signal handler ('S' character
in the CIE augmentation), you'll probably end up with more reliable
stack unwinding.

After looking at the output that I shared, I'm afraid that CFI annotation is only available on some libraries because I only see some

parse_cie: CIE parsed OK, augmentation = "S", handler=0x0

  should I ask my sysadmins to install any libc package in particular?

Again, thank you!!


Libunwind-devel mailing list

WARNING / LEGAL TEXT: This message is intended only for the use of the
individual or entity to which it is addressed and may contain
information which is privileged, confidential, proprietary, or exempt
from disclosure under applicable law. If you are not the intended
recipient or the person responsible for delivering the message to the
intended recipient, you are strictly prohibited from disclosing,
distributing, copying, or in any way using this message. If you have
received this communication in error, please notify the sender and
destroy and delete any copies you may have received.

reply via email to

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