Re: [PATCH v1 01/25] net: core: device_rename: Use rwsem instead of a seqcount

[Stephen Hemminger]

On Wed, 20 May 2020 00:23:48 +0200
Thomas Gleixner <tglx@xxxxxxxxxxxxx> wrote:

> Stephen Hemminger <stephen@xxxxxxxxxxxxxxxxxx> writes:
> > On Tue, 19 May 2020 23:45:23 +0200
> > "Ahmed S. Darwish" <a.darwish@xxxxxxxxxxxxx> wrote:
> >  
> >> Sequence counters write paths are critical sections that must never be
> >> preempted, and blocking, even for CONFIG_PREEMPTION=n, is not allowed.
> >> 
> >> Commit 5dbe7c178d3f ("net: fix kernel deadlock with interface rename and
> >> netdev name retrieval.") handled a deadlock, observed with
> >> CONFIG_PREEMPTION=n, where the devnet_rename seqcount read side was
> >> infinitely spinning: it got scheduled after the seqcount write side
> >> blocked inside its own critical section.
> >> 
> >> To fix that deadlock, among other issues, the commit added a
> >> cond_resched() inside the read side section. While this will get the
> >> non-preemptible kernel eventually unstuck, the seqcount reader is fully
> >> exhausting its slice just spinning -- until TIF_NEED_RESCHED is set.
> >> 
> >> The fix is also still broken: if the seqcount reader belongs to a
> >> real-time scheduling policy, it can spin forever and the kernel will
> >> livelock.
> >> 
> >> Disabling preemption over the seqcount write side critical section will
> >> not work: inside it are a number of GFP_KERNEL allocations and mutex
> >> locking through the drivers/base/ :: device_rename() call chain.
> >> 
> >> From all the above, replace the seqcount with a rwsem.
> >> 
> >> Fixes: 5dbe7c178d3f (net: fix kernel deadlock with interface rename and netdev name retrieval.)
> >> Fixes: 30e6c9fa93cf (net: devnet_rename_seq should be a seqcount)
> >> Fixes: c91f6df2db49 (sockopt: Change getsockopt() of SO_BINDTODEVICE to return an interface name)
> >> Cc: <stable@xxxxxxxxxxxxxxx>
> >> Signed-off-by: Ahmed S. Darwish <a.darwish@xxxxxxxxxxxxx>
> >> Reviewed-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>  
> >
> > Have your performance tested this with 1000's of network devices?  
> 
> No. We did not. -ENOTESTCASE

Please try, it isn't that hard..

# time for ((i=0;i<1000;i++)); do ip li add dev dummy$i type dummy; done

real	0m17.002s
user	0m1.064s
sys	0m0.375s