Re: tcp_bind_bucket is missing from slabinfo

[Stephen Hemminger]

On Sat, 24 Feb 2018 06:35:20 -0800
Matthew Wilcox  wrote:

> On Fri, Feb 23, 2018 at 10:50:30PM -0800, Stephen Hemminger wrote:
> > Somewhere back around 3.17 the kmem cache "tcp_bind_bucket" dropped out
> > of /proc/slabinfo. It turns out the ss command was dumpster diving
> > in slabinfo to determine the number of bound sockets and now it always
> > reports 0.
> > 
> > Not sure why, the cache is still created but it doesn't
> > show in slabinfo. Could it be some part of making slab/slub common code
> > (or network namespaces). The cache is created in tcp_init but not visible.
> > 
> > Any ideas?  
> 
> Try booting with slab_nomerge=1

Yes, thats it. 

Re: tcp_bind_bucket is missing from slabinfo

[Matthew Wilcox]

On Fri, Feb 23, 2018 at 10:50:30PM -0800, Stephen Hemminger wrote:
> Somewhere back around 3.17 the kmem cache "tcp_bind_bucket" dropped out
> of /proc/slabinfo. It turns out the ss command was dumpster diving
> in slabinfo to determine the number of bound sockets and now it always
> reports 0.
> 
> Not sure why, the cache is still created but it doesn't
> show in slabinfo. Could it be some part of making slab/slub common code
> (or network namespaces). The cache is created in tcp_init but not visible.
> 
> Any ideas?

Try booting with slab_nomerge=1

tcp_bind_bucket is missing from slabinfo

[Stephen Hemminger]

Somewhere back around 3.17 the kmem cache "tcp_bind_bucket" dropped out
of /proc/slabinfo. It turns out the ss command was dumpster diving
in slabinfo to determine the number of bound sockets and now it always
reports 0.

Not sure why, the cache is still created but it doesn't
show in slabinfo. Could it be some part of making slab/slub common code
(or network namespaces). The cache is created in tcp_init but not visible.

Any ideas?