Re: [PATCH 1/2] driver: core: add dedicated workqueue for devlink removal

["Nuno Sá" <noname.nuno@gmail.com>]

On Fri, 2024-02-02 at 16:59 +0100, Rafael J. Wysocki wrote:
> On Fri, Feb 2, 2024 at 1:18 PM Nuno Sa via B4 Relay
> <devnull+nuno.sa.analog.com@kernel.org> wrote:
> > 
> > From: Nuno Sa <nuno.sa@analog.com>
> > 
> > Let's use a dedicated queue for devlinks since releasing a link happens
> > asynchronously but some code paths, like DT overlays, have some
> > expectations regarding the of_node when being removed (the refcount must
> > be 1). Given how devlinks are released that cannot be assured. Hence, add a
> > dedicated queue so that it's easy to sync against devlinks removal.
> 
> Thanks for following my suggestion!
> 
> > While at it, make sure to explicitly include <linux/workqueue.h>.
> > 
> > Signed-off-by: Nuno Sa <nuno.sa@analog.com>
> > ---
> >  drivers/base/core.c    | 33 +++++++++++++++++++++++++++++----
> >  include/linux/fwnode.h |  1 +
> >  2 files changed, 30 insertions(+), 4 deletions(-)
> > 
> > diff --git a/drivers/base/core.c b/drivers/base/core.c
> > index 14d46af40f9a..06e7766b5227 100644
> > --- a/drivers/base/core.c
> > +++ b/drivers/base/core.c
> > @@ -31,6 +31,7 @@
> >  #include <linux/swiotlb.h>
> >  #include <linux/sysfs.h>
> >  #include <linux/dma-map-ops.h> /* for dma_default_coherent */
> > +#include <linux/workqueue.h>
> > 
> >  #include "base.h"
> >  #include "physical_location.h"
> > @@ -44,6 +45,7 @@ static bool fw_devlink_is_permissive(void);
> >  static void __fw_devlink_link_to_consumers(struct device *dev);
> >  static bool fw_devlink_drv_reg_done;
> >  static bool fw_devlink_best_effort;
> > +static struct workqueue_struct *devlink_release_queue __ro_after_init;
> > 
> >  /**
> >   * __fwnode_link_add - Create a link between two fwnode_handles.
> > @@ -235,6 +237,11 @@ static void __fw_devlink_pickup_dangling_consumers(struct
> > fwnode_handle *fwnode,
> >                 __fw_devlink_pickup_dangling_consumers(child, new_sup);
> >  }
> > 
> > +void fwnode_links_flush_queue(void)
> > +{
> > +       flush_workqueue(devlink_release_queue);
> > +}
> > +
> >  static DEFINE_MUTEX(device_links_lock);
> >  DEFINE_STATIC_SRCU(device_links_srcu);
> > 
> > @@ -531,9 +538,10 @@ static void devlink_dev_release(struct device *dev)
> >          * It may take a while to complete this work because of the SRCU
> >          * synchronization in device_link_release_fn() and if the consumer or
> >          * supplier devices get deleted when it runs, so put it into the "long"
> > -        * workqueue.
> > +        * devlink workqueue.
> > +        *
> >          */
> > -       queue_work(system_long_wq, &link->rm_work);
> > +       queue_work(devlink_release_queue, &link->rm_work);
> >  }
> > 
> >  static struct class devlink_class = {
> > @@ -636,10 +644,27 @@ static int __init devlink_class_init(void)
> >                 return ret;
> > 
> >         ret = class_interface_register(&devlink_class_intf);
> > -       if (ret)
> > +       if (ret) {
> > +               class_unregister(&devlink_class);
> > +               return ret;
> > +       }
> > +
> > +       /*
> > +        * Using a dedicated queue for devlinks since releasing a link happens
> > +        * asynchronously but some code paths, like DT overlays, have some
> > +        * expectations regarding the of_node when being removed (the refcount
> > +        * must be 1). Given how devlinks are released that cannot be assured.
> > +        * Hence, add a dedicated queue so that it's easy to sync against
> > +        * devlinks removal.
> > +        */
> > +       devlink_release_queue = alloc_workqueue("devlink_release", 0, 0);
> > +       if (!devlink_release_queue) {
> > +               class_interface_unregister(&devlink_class_intf);
> >                 class_unregister(&devlink_class);
> 
> This is a bit drastic.
> 
> I think that device links can still work if devlink_release_queue is
> NULL, just devlink_dev_release() needs to check it and release
> synchronously if it is NULL.
> 

Agreed, I'll do that way. It will always synchronously remove the links (which is
different than before) but I guess that failing in allocating the queue is rather
unlikely.

- Nuno Sá