Re: [PATCH v2] libnvdimm: fix clear poison locking with spinlock and mempool

["Verma, Vishal L" <vishal.l.verma@intel.com>]

On Tue, 2017-04-11 at 15:09 -0700, Dave Jiang wrote:
> The following warning results from holding a lane spinlock,
> preempt_disable(), or the btt map spinlock and then trying to take the
> reconfig_mutex to walk the poison list and potentially add new
> entries.
> 
> BUG: sleeping function called from invalid context at
> kernel/locking/mutex.
> c:747
> in_atomic(): 1, irqs_disabled(): 0, pid: 17159, name: dd
> [..]
> Call Trace:
> dump_stack+0x85/0xc8
> ___might_sleep+0x184/0x250
> __might_sleep+0x4a/0x90
> __mutex_lock+0x58/0x9b0
> ? nvdimm_bus_lock+0x21/0x30 [libnvdimm]
> ? __nvdimm_bus_badblocks_clear+0x2f/0x60 [libnvdimm]
> ? acpi_nfit_forget_poison+0x79/0x80 [nfit]
> ? _raw_spin_unlock+0x27/0x40
> mutex_lock_nested+0x1b/0x20
> nvdimm_bus_lock+0x21/0x30 [libnvdimm]
> nvdimm_forget_poison+0x25/0x50 [libnvdimm]
> nvdimm_clear_poison+0x106/0x140 [libnvdimm]
> nsio_rw_bytes+0x164/0x270 [libnvdimm]
> btt_write_pg+0x1de/0x3e0 [nd_btt]
> ? blk_queue_enter+0x30/0x290
> btt_make_request+0x11a/0x310 [nd_btt]
> ? blk_queue_enter+0xb7/0x290
> ? blk_queue_enter+0x30/0x290
> generic_make_request+0x118/0x3b0
> 
> Two things are done to address this issue. First, we introduce a
> spinlock to
> protect the poison list. This allows us to not having to acquire the
> reconfig_mutex for touching the poison list. Second, we allocate the
> poison
> entries using GFP_NOWAIT and avoid the sleep with GFP_KERNEL.
> 
> Signed-off-by: Dave Jiang <dave.jiang@intel.com>
> ---
>  drivers/nvdimm/bus.c      |   12 +++++++-----
>  drivers/nvdimm/core.c     |   25 ++++++++-----------------
>  drivers/nvdimm/nd-core.h  |    1 +
>  include/linux/libnvdimm.h |    2 --
>  4 files changed, 16 insertions(+), 24 deletions(-)
> 
> diff --git a/drivers/nvdimm/bus.c b/drivers/nvdimm/bus.c
> index 5ad2e59..fe41146 100644
> --- a/drivers/nvdimm/bus.c
> +++ b/drivers/nvdimm/bus.c
> @@ -296,6 +296,7 @@ struct nvdimm_bus *nvdimm_bus_register(struct
> device *parent,
>  	init_waitqueue_head(&nvdimm_bus->probe_wait);
>  	nvdimm_bus->id = ida_simple_get(&nd_ida, 0, 0, GFP_KERNEL);
>  	mutex_init(&nvdimm_bus->reconfig_mutex);
> +	spin_lock_init(&nvdimm_bus->poison_lock);
>  	if (nvdimm_bus->id < 0) {
>  		kfree(nvdimm_bus);
>  		return NULL;
> @@ -342,8 +343,9 @@ static int child_unregister(struct device *dev,
> void *data)
>  	return 0;
>  }
>  
> -static void free_poison_list(struct list_head *poison_list)
> +static void free_poison_list(struct nvdimm_bus *nvdimm_bus)
>  {
> +	struct list_head *poison_list = &nvdimm_bus->poison_list;
>  	struct nd_poison *pl, *next;

This change isn't needed anymore in v2 right? (and the caller below)

>  
>  	list_for_each_entry_safe(pl, next, poison_list, list) {
> @@ -364,9 +366,9 @@ static int nd_bus_remove(struct device *dev)
>  	nd_synchronize();
>  	device_for_each_child(&nvdimm_bus->dev, NULL,
> child_unregister);
>  
> -	nvdimm_bus_lock(&nvdimm_bus->dev);
> -	free_poison_list(&nvdimm_bus->poison_list);
> -	nvdimm_bus_unlock(&nvdimm_bus->dev);
> +	spin_lock(&nvdimm_bus->poison_lock);
> +	free_poison_list(nvdimm_bus);
> +	spin_unlock(&nvdimm_bus->poison_lock);
>  
>  	nvdimm_bus_destroy_ndctl(nvdimm_bus);
>  
> @@ -990,7 +992,7 @@ static int __nd_ioctl(struct nvdimm_bus
> *nvdimm_bus, struct nvdimm *nvdimm,
>  
>  		if (clear_err->cleared) {
>  			/* clearing the poison list we keep track of
> */
> -			__nvdimm_forget_poison(nvdimm_bus, clear_err-
> >address,
> +			nvdimm_forget_poison(nvdimm_bus, clear_err-
> >address,
>  					clear_err->cleared);
>  
>  			/* now sync the badblocks lists */
> diff --git a/drivers/nvdimm/core.c b/drivers/nvdimm/core.c
> index 40a3da0..9567b08 100644
> --- a/drivers/nvdimm/core.c
> +++ b/drivers/nvdimm/core.c
> @@ -539,7 +539,7 @@ static int bus_add_poison(struct nvdimm_bus
> *nvdimm_bus, u64 addr, u64 length)
>  	struct nd_poison *pl;
>  
>  	if (list_empty(&nvdimm_bus->poison_list))
> -		return add_poison(nvdimm_bus, addr, length,
> GFP_KERNEL);
> +		return add_poison(nvdimm_bus, addr, length,
> GFP_NOWAIT);
>  
>  	/*
>  	 * There is a chance this is a duplicate, check for those
> first.
> @@ -559,30 +559,29 @@ static int bus_add_poison(struct nvdimm_bus
> *nvdimm_bus, u64 addr, u64 length)
>  	 * as any overlapping ranges will get resolved when the list
> is consumed
>  	 * and converted to badblocks
>  	 */
> -	return add_poison(nvdimm_bus, addr, length, GFP_KERNEL);
> +	return add_poison(nvdimm_bus, addr, length, GFP_NOWAIT);
>  }
>  
>  int nvdimm_bus_add_poison(struct nvdimm_bus *nvdimm_bus, u64 addr,
> u64 length)
>  {
>  	int rc;
>  
> -	nvdimm_bus_lock(&nvdimm_bus->dev);
> +	spin_lock(&nvdimm_bus->poison_lock);
>  	rc = bus_add_poison(nvdimm_bus, addr, length);
> -	nvdimm_bus_unlock(&nvdimm_bus->dev);
> +	spin_unlock(&nvdimm_bus->poison_lock);
>  
>  	return rc;
>  }
>  EXPORT_SYMBOL_GPL(nvdimm_bus_add_poison);
>  
> -void __nvdimm_forget_poison(struct nvdimm_bus *nvdimm_bus,
> phys_addr_t start,
> +void nvdimm_forget_poison(struct nvdimm_bus *nvdimm_bus, phys_addr_t
> start,
>  		unsigned int len)
>  {
>  	struct list_head *poison_list = &nvdimm_bus->poison_list;
>  	u64 clr_end = start + len - 1;
>  	struct nd_poison *pl, *next;
>  
> -	lockdep_assert_held(&nvdimm_bus->reconfig_mutex);
> -
> +	spin_lock(&nvdimm_bus->poison_lock);
>  	WARN_ON_ONCE(list_empty(poison_list));
>  
>  	/*
> @@ -629,21 +628,13 @@ void __nvdimm_forget_poison(struct nvdimm_bus
> *nvdimm_bus, phys_addr_t start,
>  			u64 new_len = pl_end - new_start + 1;
>  
>  			/* Add new entry covering the right half */
> -			add_poison(nvdimm_bus, new_start, new_len,
> GFP_NOIO);
> +			add_poison(nvdimm_bus, new_start, new_len,
> GFP_NOWAIT);
>  			/* Adjust this entry to cover the left half
> */
>  			pl->length = start - pl->start;
>  			continue;
>  		}
>  	}
> -}
> -EXPORT_SYMBOL_GPL(__nvdimm_forget_poison);
> -
> -void nvdimm_forget_poison(struct nvdimm_bus *nvdimm_bus,
> -		phys_addr_t start, unsigned int len)
> -{
> -	nvdimm_bus_lock(&nvdimm_bus->dev);
> -	__nvdimm_forget_poison(nvdimm_bus, start, len);
> -	nvdimm_bus_unlock(&nvdimm_bus->dev);
> +	spin_unlock(&nvdimm_bus->poison_lock);
>  }
>  EXPORT_SYMBOL_GPL(nvdimm_forget_poison);
>  
> diff --git a/drivers/nvdimm/nd-core.h b/drivers/nvdimm/nd-core.h
> index 8623e57..4c4bd20 100644
> --- a/drivers/nvdimm/nd-core.h
> +++ b/drivers/nvdimm/nd-core.h
> @@ -32,6 +32,7 @@ struct nvdimm_bus {
>  	struct list_head poison_list;
>  	struct list_head mapping_list;
>  	struct mutex reconfig_mutex;
> +	spinlock_t poison_lock;
>  };
>  
>  struct nvdimm {
> diff --git a/include/linux/libnvdimm.h b/include/linux/libnvdimm.h
> index 1c609e8..98b2076 100644
> --- a/include/linux/libnvdimm.h
> +++ b/include/linux/libnvdimm.h
> @@ -122,8 +122,6 @@ static inline struct nd_blk_region_desc
> *to_blk_region_desc(
>  int nvdimm_bus_add_poison(struct nvdimm_bus *nvdimm_bus, u64 addr,
> u64 length);
>  void nvdimm_forget_poison(struct nvdimm_bus *nvdimm_bus,
>  		phys_addr_t start, unsigned int len);
> -void __nvdimm_forget_poison(struct nvdimm_bus *nvdimm_bus,
> -		phys_addr_t start, unsigned int len);
>  struct nvdimm_bus *nvdimm_bus_register(struct device *parent,
>  		struct nvdimm_bus_descriptor *nfit_desc);
>  void nvdimm_bus_unregister(struct nvdimm_bus *nvdimm_bus);
> 
_______________________________________________
Linux-nvdimm mailing list
Linux-nvdimm@lists.01.org
https://lists.01.org/mailman/listinfo/linux-nvdimm