[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-ha-dev
Subject:    Re: [Linux-ha-dev] PRE-2.0: crash
From:       Alan Robertson <alanr () unix ! sh>
Date:       2005-07-25 21:46:15
Message-ID: 42E55DA7.6040209 () unix ! sh
[Download RAW message or body]

Guochun Shi wrote:
> At 03:04 PM 7/25/2005 +0100, you wrote:
> > On Mon, 25 Jul 2005, Lars Marowsky-Bree wrote:
> > 
> > > On 2005-07-25T11:04:18, David Lee <t.d.lee@durham.ac.uk> wrote:
> > > 
> > > > I could probably quite easily fudge around this with a "memcpy(...)".
> > > > But how many other instances of such coding might there be?  Might it be
> > > > cleaner overall to abstract buffer processing and transmission/reception
> > > > via some sort of packing/unpacking interface?
> > > Yes, that sounds like a great idea, together with a small unit test to
> > > make sure it all works.
> > > 
> > > Do you have a patch? ;-)
> > > 
> > > Surely 2.0.1 material though I assume.
> > Thanks, Lars.  Your reply is appreciated!
> > 
> > I wondered whether this coding was an isolated problem, or perhaps the tip of a \
> > much larger coding iceberg.  So I did a quick patch to investigate. Excellent \
> > news!  No iceberg. 
> > Without the patch, the heartbeat process (1.99.6) were crashing almost \
> > immediately on sparc architecture.  With the patch they run! 
> > The patch is attached.  It is tightly contained within two files.  It replaces \
> > the "longclock_t" assignments in non-aligned bytestream buffers with "memcpy()".  \
> > (It also localises some things from a general ".h" into the only ".c" that \
> > references them.) 
> > Please, please, Alan, could this be considered for 2.0??  Pretty please?
> > 
> > Without the patch, heartbeat on sparc (e.g. Solaris) fails immediately (core \
> > dump, signal 10).  It is a "blocker" for such architectures.  But with the patch, \
> > the heartbeat processes start, and stay running! 
> > At last: heartbeat (PRE-)2.0 processes run on Solaris.  (There may yet be further \
> > runtime problems, but at least now sparc and Solaris users would have a better \
> > chance of finding and debugging them.)
> 
> Thanks, Davlid, the patch looks good.
> 
> Alan, how about apply the patch and disable the debug code (assume u don't want it \
> removed )?

Disabling it and applying the patch seems perfect!


-- 
     Alan Robertson <alanr@unix.sh>

"Openness is the foundation and preservative of friendship...  Let me 
claim from you at all times your undisguised opinions." - William 
Wilberforce
_______________________________________________________
Linux-HA-Dev: Linux-HA-Dev@lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/


[prev in list] [next in list] [prev in thread] [next in thread]