NPTL Thread lockup issue on ARMv7

Laurent Bercot ska-dietlibc at skarnet.org
Wed Feb 2 20:29:21 UTC 2011


> I think many people would argue that given the stability of given a 
> system as a whole, a random crash once per 5 years would be prefereble to 
> a severe performance penelty which impacts the system on a continuous 
> basis.

 I would argue the exact opposite. We're talking ARM here, i.e.
embedded systems. On those systems, it's much easier to counter
performance issues than stability issues. If a box crashes, there's
nothing we can do - the customer will just return it.

 I'm advocating the use of uClibc in the embedded boxes I'm writing
software for. My arguments are low resource consumption, and stability.
What am I going to tell my management if you purposefully remove the
stability? I will have to switch libcs. Again.

 Make it work, then make it work fast.

-- 
 Laurent


More information about the uClibc mailing list