On Mon, 21 Jun 2010, Christoph Lameter wrote:
> > > if (slab_state == DOWN) {
> > > - early_kmem_cache_node_alloc(gfpflags, node);
> > > + early_kmem_cache_node_alloc(node);
> > > continue;
> > > }
> > > n = kmem_cache_alloc_node(kmalloc_caches,
> > > - gfpflags, node);
> > > + GFP_KERNEL, node);
> > >
> > > if (!n) {
> > > free_kmem_cache_nodes(s);
> >
> > Same here, this can still lead to GFP_KERNEL allocations from
> > kmem_cache_init() because slab_state is PARTIAL or UP.
>
> You cannot do that here because this function is also used later when the
> slab is up. There is more in the percpu allocator which we are also trying
> to use to avoid having static kmem_cache_cpu declarations. GFP_KERNEL
> needs to be usable during early boot otherwise functions will have to add
> special casing for boot situations.
>
The gfp_allowed_mask only changes once irqs are enabled, so either the
gfpflags need to be passed into init_kmem_cache_nodes again or we need to
do something like
gfp_t gfpflags = irqs_disabled() ? GFP_NOWAIT : GFP_KERNEL;
locally.
The cleanest solution would probably be to extend slab_state to be set in
kmem_cache_init_late() to determine when we're fully initialized, though.
> > > + for (i = 0; i < SLUB_PAGE_SHIFT; i++) {
> > > + struct kmem_cache *s = &kmalloc_caches[i];
> > > +
> > > + if (s && s->size) {
> > > + char *name = kasprintf(GFP_KERNEL,
> > > + "dma-kmalloc-%d", s->objsize);
> >
> > kasprintf() can return NULL which isn't caught by kmem_cache_open().
>
> Then we will have a nameless cache. We could catch this with a WARN_ON()
> but does this work that early?
>
It works, but this seems to be a forced
if (WARN_ON(!name))
continue;
because although it appears that s->name can be NULL within the slub
layer, the sysfs layer would result in a NULL pointer dereference for
things like strcmp() when looking up the cache's dirent.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxxx For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>