Re: [PATCH v4] perf, x86: Fix multi-segment problem of perf_event_intel_uncore

[Peter Zijlstra <peterz@infradead.org>]

On Thu, Sep 17, 2015 at 01:17:12AM +0900, Taku Izumi wrote:

> +struct pci2phy_map {
> +	struct list_head list;
> +	int segment;
> +	int pbus_to_physid[256];
> +};

> +struct pci2phy_map *__find_pci2phy_map(int segment)
> +{
> +	struct pci2phy_map *map, *alloc = NULL;
> +
> +	lockdep_assert_held(&pci2phy_map_lock);
> +
> +lookup:
> +	list_for_each_entry(map, &pci2phy_map_head, list) {
> +		if (map->segment == segment)
> +			goto end;
> +	}
> +
> +	if (!alloc) {
> +		raw_spin_unlock(&pci2phy_map_lock);
> +		alloc = kmalloc(sizeof(struct pci2phy_map), GFP_KERNEL);

Do we want kzalloc() ? we unconditionally initialize the list and
segment thingies, but that array is not initialized here. Or is there a
better/safer value to initialize that array with?

Looking at uncore_pcibus_to_physid(), -1 might be a good value to use.

> +		raw_spin_lock(&pci2phy_map_lock);
> +
> +		if (!alloc)
> +			return NULL;
> +
> +		goto lookup;
> +	}
> +
> +	map = alloc;
> +	alloc = NULL;
> +	map->segment = segment;
> +	list_add_tail(&map->list, &pci2phy_map_head);
> +
> +end:
> +	kfree(alloc);
> +	return map;
> +}

Other than that I think the patch looks OK now.