[LEAPSECS] happy anniversary pips

[Warner Losh]

On Feb 11, 2014, at 11:22 AM, Tim Shepard wrote:


> 

> 

>>>> People have been working for the past 15 years to make leap seconds

>>>> better, yet in the last leap second all Linux kernels crashed due

>>>> to a subtle bug that is only triggered when there was a leap second.

>>> 

>>> My understanding wasn't that all Linux kernels crashed.

>> 

>> Only the ones which cared enough about time-keeping to run NTPD.

> 

> ... and that were running a particular old-but-not-too-old version of

> the Linux kernel.  And it didn't happen everywhere.  And it didn't

> crash machines, just got them very busy looping blocked by in-kernel

> locks (which is perhaps worse than a crash, depending on what

> matters).

> 

> The patch to fix the bug was published in main-line Linux more than

> three months before the leap second occured:

> 

>  https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=6b43ae8a619d17c4935c3320d2ef9e92bdeed05d

> 

> but the patch didn't get deployed everywhere it needed to be deployed,

> and the wedge up of some web site server farms made news:

> 

>  http://www.wired.com/wiredenterprise/2012/07/leap-second-glitch-explained/all/


Right, but none of that detracts from my original point... Leap seconds caused a problem in a widely deployed, presumably widely tested kernel when they should have been well enough known and tested to not to.

Warner