Re: [Libunwind-devel] mincore_validate fails sometimes

[Dave Watson]

On 01/15/18 12:00 PM, Daniel Vassdal wrote:
> We have an issue where on Yocto Morty 2.2 mincore_validate() sometimes 
> unexpectedly returns -1.
> This seems to happen most often directly after the system has been booted.

Do you know what the errno is when it returns -1?  Is there some condition we 
should be checking for?

> The check that fails is
> > if (!(mvec[i] & 1)) return -1;
> 
> The reason I suspect this is a libunwind problem is that when HAVE_MINCORE 
> isn't defined, it works well.
> This workaround is sufficient for us, but in case this indeed is a bug I 
> figured I'd let you guys know.