| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <m4nmhm2zmgrelxtyhftmg6kkyxexmznyo7khfmjxphkzr37ea6@44ga7yswli3c>
Date: Fri, 16 Jan 2026 14:46:44 -0500
From: "Liam R. Howlett" <Liam.Howlett@...cle.com>
To: kernel test robot <lkp@...el.com>
Cc: Andrew Morton <akpm@...ux-foundation.org>, oe-kbuild-all@...ts.linux.dev,
Linux Memory Management List <linux-mm@...ck.org>,
maple-tree@...ts.infradead.org, linux-kernel@...r.kernel.org,
Suren Baghdasaryan <surenb@...gle.com>,
Matthew Wilcox <willy@...radead.org>,
Sidhartha Kumar <sidhartha.kumar@...cle.com>,
Vlastimil Babka <vbabka@...e.cz>, Alice Ryhl <aliceryhl@...gle.com>,
Kuninori Morimoto <kuninori.morimoto.gx@...esas.com>,
Geert Uytterhoeven <geert@...ux-m68k.org>,
Arnd Bergmann <arnd@...db.de>, Christian Kujau <lists@...dbynature.de>
Subject: Re: [PATCH 10/28] maple_tree: Introduce maple_copy node and use it
in mas_spanning_rebalance()
* kernel test robot <lkp@...el.com> [260116 02:45]:
> Hi Liam,
>
> kernel test robot noticed the following build warnings:
>
> [auto build test WARNING on akpm-mm/mm-everything]
> [also build test WARNING on next-20260115]
> [cannot apply to akpm-mm/mm-nonmm-unstable soc/for-next linus/master v6.19-rc5]
> [If your patch is applied to the wrong git tree, kindly drop us a note.
> And when submitting patch, we suggest to use '--base' as documented in
> https://git-scm.com/docs/git-format-patch#_base_tree_information]
>
> url: https://github.com/intel-lab-lkp/linux/commits/Liam-R-Howlett/maple_tree-Move-mas_spanning_rebalance-loop-to-function/20260116-034135
> base: https://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git mm-everything
> patch link: https://lore.kernel.org/r/20260115193647.1695937-11-Liam.Howlett%40oracle.com
> patch subject: [PATCH 10/28] maple_tree: Introduce maple_copy node and use it in mas_spanning_rebalance()
> config: x86_64-randconfig-121-20260116 (https://download.01.org/0day-ci/archive/20260116/202601161501.NZ3GNDCG-lkp@intel.com/config)
> compiler: gcc-13 (Debian 13.3.0-16) 13.3.0
> reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20260116/202601161501.NZ3GNDCG-lkp@intel.com/reproduce)
>
> If you fix the issue in a separate patch/commit (i.e. not just a new version of
> the same patch/commit), kindly add following tags
> | Reported-by: kernel test robot <lkp@...el.com>
> | Closes: https://lore.kernel.org/oe-kbuild-all/202601161501.NZ3GNDCG-lkp@intel.com/
>
> sparse warnings: (new ones prefixed by >>)
> >> lib/maple_tree.c:2621:29: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected void [noderef] __rcu * @@ got void *content @@
> lib/maple_tree.c:2621:29: sparse: expected void [noderef] __rcu *
> lib/maple_tree.c:2621:29: sparse: got void *content
> >> lib/maple_tree.c:2624:23: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected void [noderef] __rcu * @@ got void *entry @@
> lib/maple_tree.c:2624:23: sparse: expected void [noderef] __rcu *
> lib/maple_tree.c:2624:23: sparse: got void *entry
> lib/maple_tree.c:2680:17: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected void [noderef] __rcu **b_slots @@ got void [noderef] __rcu * @@
> lib/maple_tree.c:2680:17: sparse: expected void [noderef] __rcu **b_slots
> lib/maple_tree.c:2680:17: sparse: got void [noderef] __rcu *
> lib/maple_tree.c:6412:30: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected void [noderef] __rcu * @@ got struct maple_node * @@
> lib/maple_tree.c:6412:30: sparse: expected void [noderef] __rcu *
> lib/maple_tree.c:6412:30: sparse: got struct maple_node *
> lib/maple_tree.c:6412:30: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected void [noderef] __rcu * @@ got struct maple_node * @@
> lib/maple_tree.c:6412:30: sparse: expected void [noderef] __rcu *
> lib/maple_tree.c:6412:30: sparse: got struct maple_node *
>
> vim +2621 lib/maple_tree.c
>
> 2603
> 2604 /*
> 2605 * cp_leaf_init() - Initialize a maple_copy node for the leaf level of a
> 2606 * spanning store
> 2607 * @cp: The maple copy node
> 2608 * @mas: The maple state
> 2609 * @l_wr_mas: The left write state of the spanning store
> 2610 * @r_wr_mas: The right write state of the spanning store
> 2611 */
> 2612 static inline void cp_leaf_init(struct maple_copy *cp,
> 2613 struct ma_state *mas, struct ma_wr_state *l_wr_mas,
> 2614 struct ma_wr_state *r_wr_mas)
> 2615 {
> 2616 unsigned char end = 0;
> 2617
> 2618 /* Create entries to insert including split entries to left and right */
> 2619 if (l_wr_mas->r_min < mas->index) {
> 2620 end++;
> > 2621 cp->slot[0] = l_wr_mas->content;
> 2622 cp->pivot[0] = mas->index - 1;
> 2623 }
> > 2624 cp->slot[end] = l_wr_mas->entry;
> 2625 cp->pivot[end] = mas->last;
> 2626
> 2627 if (r_wr_mas->end_piv > mas->last) {
> 2628 end++;
> 2629 cp->slot[end] = r_wr_mas->slots[r_wr_mas->offset_end];
> 2630 cp->pivot[end] = r_wr_mas->end_piv;
> 2631 }
> 2632
> 2633 cp->min = l_wr_mas->r_min;
> 2634 cp->max = cp->pivot[end];
> 2635 cp->end = end;
> 2636 }
> 2637
>
#syz test: https://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git mm-everything
--- a/lib/maple_tree.c
+++ b/lib/maple_tree.c
@@ -2092,19 +2092,28 @@ static inline void cp_leaf_init(struct maple_copy *cp,
{
unsigned char end = 0;
+ /*
+ * It is very important that the maple_copy node is never seen before a
+ * write side barrier happens. Otherwise, there could be a reordering
+ * in the RCU code below that will be extremely difficult to debug.
+ *
+ * In this case, it is fine because the maple_copy is never exposed to
+ * readers until it is put into a final node and a wmb() occurs.
+ */
cp->height = 1;
/* Create entries to insert including split entries to left and right */
if (l_wr_mas->r_min < mas->index) {
end++;
- cp->slot[0] = l_wr_mas->content;
+ RCU_INIT_POINTER(cp->slot[0], l_wr_mas->content);
cp->pivot[0] = mas->index - 1;
}
- cp->slot[end] = l_wr_mas->entry;
+ RCU_INIT_POINTER(cp->slot[end], l_wr_mas->entry);
cp->pivot[end] = mas->last;
if (r_wr_mas->end_piv > mas->last) {
end++;
- cp->slot[end] = r_wr_mas->slots[r_wr_mas->offset_end];
+ RCU_INIT_POINTER(cp->slot[end],
+ r_wr_mas->slots[r_wr_mas->offset_end]);
cp->pivot[end] = r_wr_mas->end_piv;
}
Powered by blists - more mailing lists