glibc 2.18 machine status?

[David Miller]

From: David Miller <davem@davemloft.net>
Date: Sat, 20 Jul 2013 16:49:54 -0700 (PDT)

> All that is left are failures of nptl/tst-cancel{4,5}* on 64-bit
> sparc.  They all fail the same way saying "minimum write buffer size
> too large".
> 
> Should be easy to fix and looks like just a minor adjustment is
> needed in these test cases.

The problem is that the minimum socket buffer size has been increased
in recent kernels, so WRITE_BUFFER_SIZE in these test cases needs to
bumped up a bit.

I'll make sure I choose a value that works in all cases.