Re: linux-next: build failure after merge of the rcu tree

From: Paul E. McKenney
Date: Fri Dec 04 2020 - 14:21:15 EST


On Fri, Dec 04, 2020 at 07:25:26PM +1100, Stephen Rothwell wrote:
> Hi all,
>
> After merging the rcu tree, today's linux-next build (sparc defconfig)
> failed like this:
>
> mm/slab_common.o: In function `kmem_last_alloc':
> slab_common.c:(.text+0xc4): undefined reference to `kmem_cache_last_alloc'
>
> Caused by commit
>
> f7c3fb4fc476 ("mm: Add kmem_last_alloc() to return last allocation for memory block")
>
> in mm/slab.c, kmem_cache_last_alloc() is only defined when CONFIG_NUMA
> is set - which is not for this build.
>
> I applied the following hack fix patch for today.
>
> From ac5dcf78be1e6da530302966369a3bd63007cf81 Mon Sep 17 00:00:00 2001
> From: Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx>
> Date: Fri, 4 Dec 2020 19:11:01 +1100
> Subject: [PATCH] fixup for "mm: Add kmem_last_alloc() to return last
> allocation for memory block"
>
> Signed-off-by: Stephen Rothwell <sfr@xxxxxxxxxxxxxxxx>

Apologies for the hassle and thank you for the very helpful "hack
fix patch". The kbuild test robot also found this, but I hadn't
quite gotten it through my head that the only slab definition of the
kmem_cache_last_alloc() was under CONFIG_NUMA.

Does the following patch fix things? (Sigh. It won't apply on the
rcu/next that you used. Or even on this moment's dev branch. I will
fold it in with attribution and update. But just to show you what my
thought is.)

Thanx, Paul

------------------------------------------------------------------------

diff --git a/mm/slab.c b/mm/slab.c
index 87aa2c0..ae1a74c 100644
--- a/mm/slab.c
+++ b/mm/slab.c
@@ -3602,27 +3602,6 @@ void *kmem_cache_alloc_node_trace(struct kmem_cache *cachep,
EXPORT_SYMBOL(kmem_cache_alloc_node_trace);
#endif

-void *kmem_cache_last_alloc(struct kmem_cache *cachep, void *object, void **stackp, int nstackp)
-{
-#ifdef DEBUG
- unsigned int objnr;
- void *objp;
- struct page *page;
-
- if (!(cachep->flags & SLAB_STORE_USER))
- return ERR_PTR(-KMEM_LA_NO_DEBUG);
- objp = object - obj_offset(cachep);
- page = virt_to_head_page(objp);
- objnr = obj_to_index(cachep, page, objp);
- objp = index_to_obj(cachep, page, objnr);
- if (stackp && nstackp)
- stackp[0] = NULL;
- return *dbg_userword(cachep, objp);
-#else
- return NULL;
-#endif
-}
-
static __always_inline void *
__do_kmalloc_node(size_t size, gfp_t flags, int node, unsigned long caller)
{
@@ -3654,6 +3633,27 @@ void *__kmalloc_node_track_caller(size_t size, gfp_t flags,
EXPORT_SYMBOL(__kmalloc_node_track_caller);
#endif /* CONFIG_NUMA */

+void *kmem_cache_last_alloc(struct kmem_cache *cachep, void *object, void **stackp, int nstackp)
+{
+#ifdef DEBUG
+ unsigned int objnr;
+ void *objp;
+ struct page *page;
+
+ if (!(cachep->flags & SLAB_STORE_USER))
+ return ERR_PTR(-KMEM_LA_NO_DEBUG);
+ objp = object - obj_offset(cachep);
+ page = virt_to_head_page(objp);
+ objnr = obj_to_index(cachep, page, objp);
+ objp = index_to_obj(cachep, page, objnr);
+ if (stackp && nstackp)
+ stackp[0] = NULL;
+ return *dbg_userword(cachep, objp);
+#else
+ return NULL;
+#endif
+}
+
/**
* __do_kmalloc - allocate memory
* @size: how many bytes of memory are required.