Messages in this thread | | | Date | Wed, 16 Jan 2013 10:53:26 +0800 | From | Fengguang Wu <> | Subject | Re: [PATCH v2] writeback: mark sysctl vm.block_dump for removal |
| |
On Wed, Jan 16, 2013 at 03:07:53AM +0100, Jan Kara wrote: > On Tue 15-01-13 08:28:13, Tejun Heo wrote: > > vm.block_dump is nasty in that it dumps IO traces directly to printk. > > It isn't scalable and can easily lead to looping behavior - IO > > generating kernel message which in turn genreates log IO, ad > > infinitum. > > > > Now that proper tracepoints are in place, vm.block_dump isn't > > necessary. Generate a warning if used so that it can be removed down > > the road.
> > + if (!warned && block_dump) { > > + pr_warning("sysctl: vm.block_dump is scheduled for removal. Use tracepoints instead.\n"); > > + pr_warning("sysctl: Please contact Jens Axboe <axboe@kernel.dk> if you require this.\n"); > > + warned = true;
Looks good to me. A warning in dmesg is pertinent because the block_dump users will be looking at dmesg for block dump results.
Hmm, perhaps a more practical tip on the usage of replacement tracepoints or tool will help eliminate some user emails to Jens.
> > diff --git a/patches/series b/patches/series > > index 9fb9bfe..b2582cb 100644 > > --- a/patches/series > > +++ b/patches/series > > @@ -18,3 +18,4 @@ kill-pool-gcwq > > remove-gcwq > > rename-nr_running > > cleanup-leftovers > > +dbg
An extra chunk.
Reviewed-by: Fengguang Wu <fengguang.wu@intel.com>
Thanks!
| |