Re: sysctl -a panic on VIMAGE kernels

On Sun, Aug 09, 2015 at 08:25:56PM +0200, Kristof Provost wrote:
K> On 2015-08-09 13:36:35 (+0300), Gleb Smirnoff <glebius_at_FreeBSD.org> wrote:
K> > On Sun, Aug 09, 2015 at 12:28:22PM +0200, Kristof Provost wrote:
K> > K> The following fixes it for me:
K> > K> 
K> > K> diff --git a/sys/netinet/tcp_reass.c b/sys/netinet/tcp_reass.c
K> > K> index 77d8940..3913ef3 100644
K> > K> --- a/sys/netinet/tcp_reass.c
K> > K> +++ b/sys/netinet/tcp_reass.c
K> > K> _at__at_ -84,7 +84,7 _at__at_ SYSCTL_INT(_net_inet_tcp_reass, OID_AUTO, maxsegments, CTLFLAG_RDTUN,
K> > K>      "Global maximum number of TCP Segments in Reassembly Queue");
K> > K> 
K> > K>  static uma_zone_t tcp_reass_zone;
K> > K> -SYSCTL_UMA_CUR(_net_inet_tcp_reass, OID_AUTO, cursegments, CTLFLAG_VNET,
K> > K> +SYSCTL_UMA_CUR(_net_inet_tcp_reass, OID_AUTO, cursegments, 0,
K> > K>      &tcp_reass_zone,
K> > K>      "Global number of TCP Segments currently in Reassembly Queue”);
K> > 
K> > Right, if a variable isn't virtualized, the CTLFLAG_VNET must be removed.
K> > 
K> > Patrick, how is your progress wuth improved reassembly?
K> > 
K> Any opposition to me committing the above patch? It'll at least make us
K> stop panic()ing and I don't think it'll make Patrick's life any harder.

Of course no objections, please do.

-- 
Totus tuus, Glebius.