Re: [PATCH 1/2]MSI(X) save/restore for suspend/resume

[Matthew Wilcox <matthew@wil.cx>]

On Thu, Jan 12, 2006 at 10:30:59AM +0800, Shaohua Li wrote:

Did you not understand my comment about using the hlist functions?

> +struct pci_cap_saved_state {
> +	struct list_head next;

struct hlist_node list;

> +	char cap_nr;
> +	u32 data[0];
> +};

>  	u32		saved_config_space[16]; /* config space saved at suspend time */
> +	struct list_head saved_cap_space; /* extend config space saved at suspend */

struct hlist_head saved_cap_space;

BTW, the comment is confusing and wrong.  Extended config space is
0x100-0xfff.  How about simply:

/* Capabilities saved on suspend */

Or even omit it.  I think it's pretty obvious what "saved_cap_space" is.

> +static inline struct pci_cap_saved_state *pci_find_saved_cap(
> +	struct pci_dev *pci_dev,char cap)
> +{
> +	struct pci_cap_saved_state *tmp;
> +
> +	list_for_each_entry(tmp, &pci_dev->saved_cap_space, next) {

hlist_for_each_entry

> +		if (tmp->cap_nr == cap)
> +			return tmp;
> +	}
> +	return NULL;
> +}
> +
> +static inline void pci_add_saved_cap(struct pci_dev *pci_dev,
> +	struct pci_cap_saved_state *new_cap)
> +{
> +	list_add(&new_cap->next, &pci_dev->saved_cap_space);

hlist_add_head

> +}
> +
> +static inline void pci_remove_saved_cap(struct pci_cap_saved_state *cap)
> +{
> +	list_del(&cap->next);

hlist_del

> +}
> +
>  /*
>   *  For PCI devices, the region numbers are assigned this way:
>   *
> diff -puN drivers/pci/probe.c~msi_save_restore drivers/pci/probe.c
> --- linux-2.6.15-rc5/drivers/pci/probe.c~msi_save_restore	2006-01-12 10:12:00.000000000 +0800
> +++ linux-2.6.15-rc5-root/drivers/pci/probe.c	2006-01-12 10:12:55.000000000 +0800
> @@ -788,6 +788,7 @@ void __devinit pci_device_add(struct pci
>  	/* Fix up broken headers */
>  	pci_fixup_device(pci_fixup_header, dev);
>  
> +	INIT_LIST_HEAD(&dev->saved_cap_space);

and while you could use INIT_HLIST_HEAD, you shouldn't need to since an
empty hlist is NULL.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/