On Mon, Mar 08, 2021 at 11:26:20PM +0800, kernel test robot wrote:
>
> Greeting,
>
> FYI, we noticed the following commit (built with gcc-9):
>
> commit: 9ddc8abf031750362cda61a9fb8a28be8871eaae ("[PATCH v4] mm: cma: support sysfs")
> url: https://github.com/0day-ci/linux/commits/Minchan-Kim/mm-cma-support-sysfs/20210305-002050
> base: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git f69d02e37a85645aa90d18cacfff36dba370f797
>
> in testcase: trinity
> version: trinity-i386-4d2343bd-1_20200320
> with following parameters:
>
> group: group-01
>
> test-description: Trinity is a linux system call fuzz tester.
> test-url: http://codemonkey.org.uk/projects/trinity/
>
>
> on test machine: qemu-system-x86_64 -enable-kvm -cpu SandyBridge -smp 2 -m 8G
>
> caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
>
>
> +-------------------------------------------------------------------------+------------+------------+
> | | f69d02e37a | 9ddc8abf03 |
> +-------------------------------------------------------------------------+------------+------------+
> | BUG:KASAN:null-ptr-deref_in_lockdep_init_map_type | 0 | 12 |
> | BUG:kernel_NULL_pointer_dereference,address | 0 | 12 |
> | Oops:#[##] | 0 | 12 |
> | RIP:lockdep_init_map_type | 0 | 12 |
> +-------------------------------------------------------------------------+------------+------------+
>
>
> If you fix the issue, kindly add following tag
> Reported-by: kernel test robot <oliver.sang@xxxxxxxxx>
>
>
> [ 16.842917] BUG: KASAN: null-ptr-deref in lockdep_init_map_type (kbuild/src/consumer/kernel/locking/lockdep.c:4654)
> [ 16.844311] Write of size 8 at addr 0000000000000030 by task swapper/0/1
> [ 16.844311]
> [ 16.844311] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.12.0-rc1-00023-g9ddc8abf0317 #1
> [ 16.844311] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-1 04/01/2014
> [ 16.844311] Call Trace:
> [ 16.844311] dump_stack (kbuild/src/consumer/lib/dump_stack.c:122)
> [ 16.844311] ? lockdep_init_map_type (kbuild/src/consumer/kernel/locking/lockdep.c:4654)
> [ 16.844311] kasan_report.cold (kbuild/src/consumer/mm/kasan/report.c:403 kbuild/src/consumer/mm/kasan/report.c:416)
> [ 16.844311] ? lockdep_init_map_type (kbuild/src/consumer/kernel/locking/lockdep.c:4654)
> [ 16.844311] lockdep_init_map_type (kbuild/src/consumer/kernel/locking/lockdep.c:4654)
> [ 16.844311] __raw_spin_lock_init (kbuild/src/consumer/kernel/locking/spinlock_debug.c:26)
> [ 16.844311] cma_sysfs_init (kbuild/src/consumer/mm/cma_sysfs.c:91)
> [ 16.844311] ? cma_debugfs_init (kbuild/src/consumer/mm/cma_sysfs.c:74)
> [ 16.844311] do_one_initcall (kbuild/src/consumer/init/main.c:1226)
> [ 16.844311] ? perf_trace_initcall_level (kbuild/src/consumer/init/main.c:1217)
> [ 16.844311] ? rcu_read_lock_sched_held (kbuild/src/consumer/kernel/rcu/update.c:125)
> [ 16.844311] ? trace_event_raw_event_rcu_torture_read (kbuild/src/consumer/kernel/rcu/update.c:120)
> [ 16.844311] ? write_comp_data (kbuild/src/consumer/kernel/kcov.c:218)
> [ 16.844311] ? __sanitizer_cov_trace_pc (kbuild/src/consumer/kernel/kcov.c:197)
> [ 16.844311] kernel_init_freeable (kbuild/src/consumer/init/main.c:1298 kbuild/src/consumer/init/main.c:1315 kbuild/src/consumer/init/main.c:1335 kbuild/src/consumer/init/main.c:1537)
> [ 16.844311] ? console_on_rootfs (kbuild/src/consumer/init/main.c:1503)
> [ 16.844311] ? tracer_hardirqs_on (kbuild/src/consumer/kernel/trace/trace_irqsoff.c:57 kbuild/src/consumer/kernel/trace/trace_irqsoff.c:610)
> [ 16.844311] ? mark_held_locks (kbuild/src/consumer/kernel/locking/lockdep.c:4067)
> [ 16.844311] ? rest_init (kbuild/src/consumer/init/main.c:1421)
> [ 16.844311] kernel_init (kbuild/src/consumer/init/main.c:1426)
> [ 16.844311] ret_from_fork (kbuild/src/consumer/arch/x86/entry/entry_64.S:300)
> [ 16.844311] ==================================================================
> [ 16.844311] Disabling lock debugging due to kernel taint
> [ 16.844425] BUG: kernel NULL pointer dereference, address: 0000000000000030
> [ 16.845925] #PF: supervisor write access in kernel mode
> [ 16.847149] #PF: error_code(0x0002) - not-present page
> [ 16.848311] PGD 0 P4D 0
> [ 16.848311] Oops: 0002 [#1] SMP KASAN PTI
> [ 16.848311] CPU: 0 PID: 1 Comm: swapper/0 Tainted: G B 5.12.0-rc1-00023-g9ddc8abf0317 #1
>From 758182a763fbc0fbd6b5e143ca64a4eb31d22a1a Mon Sep 17 00:00:00 2001
From: Minchan Kim <minchan@xxxxxxxxxx>
Date: Mon, 8 Mar 2021 11:33:47 -0800
Subject: [PATCH] mm: cma: fix ZERO_SIZE_PTR check
If there is no cma instance, cma_area_count will be zero and
kmalloc_arrary will return ZERO_SITE_PTR instead of NULL.
Use ZERO_OR_NULL_PTR to check both cases.
Link: https://lore.kernel.org/linux-mm/20210308152620.GE4324@xsang-OptiPlex-9020/
Reported-by: kernel test robot <oliver.sang@xxxxxxxxx>
Signed-off-by: Minchan Kim <minchan@xxxxxxxxxx>
---
mm/cma_sysfs.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/mm/cma_sysfs.c b/mm/cma_sysfs.c
index 67b63167eaf5..fdcb952ff13f 100644
--- a/mm/cma_sysfs.c
+++ b/mm/cma_sysfs.c
@@ -81,7 +81,7 @@ static int __init cma_sysfs_init(void)
cma_stats = kmalloc_array(cma_area_count, sizeof(struct cma_stat),
GFP_KERNEL|__GFP_ZERO);
- if (!cma_stats)
+ if (ZERO_OR_NULL_PTR(cma_stats))
goto out;
do {
--
2.30.1.766.gb4fecdf3b7-goog
