After enabling alignment checks in UBSan I've noticed a lot of
reports like this:
UBSan: Undefined behaviour in ../kernel/irq/chip.c:195:14
member access within misaligned address ffff88003e80d6f8
for type 'struct irq_desc' which requires 64 byte alignment
struct irq_desc declared with ____cacheline_internodealigned_in_smp
attribute. However in some cases it allocated dynamically via kmalloc().
In general case kmalloc() guaranties only sizeof(void *) alignment.
We should use a separate slab cache to make struct irq_desc
properly aligned on SMP configuration.
This also could slightly reduce memory usage on some configurations.
E.g. in my setup sizeof(struct irq_desc) == 320. Which means that
kmalloc-512 will be used for allocating irg_desc via kmalloc().
In that case using separate slab cache will save us 192 bytes per
each irq_desc.
Signed-off-by: Andrey Ryabinin <[email protected]>
Acked-by: David Rientjes <[email protected]>
---
Changes since v1:
- Drop kmem_cache_zalloc_node() and use kmem_cache_alloc_node with __GFP_ZERO flag.
kernel/irq/irqdesc.c | 10 +++++++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/kernel/irq/irqdesc.c b/kernel/irq/irqdesc.c
index a1782f8..c7a812c 100644
--- a/kernel/irq/irqdesc.c
+++ b/kernel/irq/irqdesc.c
@@ -23,6 +23,8 @@
*/
static struct lock_class_key irq_desc_lock_class;
+static struct kmem_cache *irq_desc_cachep;
+
#if defined(CONFIG_SMP)
static void __init init_irq_default_affinity(void)
{
@@ -137,7 +139,7 @@ static struct irq_desc *alloc_desc(int irq, int node, struct module *owner)
struct irq_desc *desc;
gfp_t gfp = GFP_KERNEL;
- desc = kzalloc_node(sizeof(*desc), gfp, node);
+ desc = kmem_cache_alloc_node(irq_desc_cachep, gfp | __GFP_ZERO, node);
if (!desc)
return NULL;
/* allocate based on nr_cpu_ids */
@@ -158,7 +160,7 @@ static struct irq_desc *alloc_desc(int irq, int node, struct module *owner)
err_kstat:
free_percpu(desc->kstat_irqs);
err_desc:
- kfree(desc);
+ kmem_cache_free(irq_desc_cachep, desc);
return NULL;
}
@@ -174,7 +176,7 @@ static void free_desc(unsigned int irq)
free_masks(desc);
free_percpu(desc->kstat_irqs);
- kfree(desc);
+ kmem_cache_free(irq_desc_cachep, desc);
}
static int alloc_descs(unsigned int start, unsigned int cnt, int node,
@@ -218,6 +220,8 @@ int __init early_irq_init(void)
init_irq_default_affinity();
+ irq_desc_cachep = KMEM_CACHE(irq_desc, SLAB_PANIC);
+
/* Let arch update nr_irqs and return the nr of preallocated irqs */
initcnt = arch_probe_nr_irqs();
printk(KERN_INFO "NR_IRQS:%d nr_irqs:%d %d\n", NR_IRQS, nr_irqs, initcnt);
--
2.1.3
On Fri, 21 Nov 2014, Andrey Ryabinin wrote:
> Signed-off-by: Andrey Ryabinin <[email protected]>
> Acked-by: David Rientjes <[email protected]>
Acked-by: Christoph Lameter <[email protected]>