[PATCH 1/2] kasan, mm: reset tag when access metadata

Kuan-Ying Lee Kuan-Ying.Lee at mediatek.com
Tue Jul 27 01:32:02 PDT 2021


On Tue, 2021-07-27 at 09:10 +0200, Marco Elver wrote:
> +Cc Catalin
> 
> On Tue, 27 Jul 2021 at 06:00, Kuan-Ying Lee <
> Kuan-Ying.Lee at mediatek.com> wrote:
> > 
> > Hardware tag-based KASAN doesn't use compiler instrumentation, we
> > can not use kasan_disable_current() to ignore tag check.
> > 
> > Thus, we need to reset tags when accessing metadata.
> > 
> > Signed-off-by: Kuan-Ying Lee <Kuan-Ying.Lee at mediatek.com>
> 
> This looks reasonable, but the patch title is not saying this is
> kmemleak, nor does the description say what the problem is. What
> problem did you encounter? Was it a false positive?

kmemleak would scan kernel memory to check memory leak.
When it scans on the invalid slab and dereference, the issue
will occur like below.

So I think we should reset the tag before scanning.

# echo scan > /sys/kernel/debug/kmemleak
[  151.905804]
==================================================================
[  151.907120] BUG: KASAN: out-of-bounds in scan_block+0x58/0x170
[  151.908773] Read at addr f7ff0000c0074eb0 by task kmemleak/138
[  151.909656] Pointer tag: [f7], memory tag: [fe]
[  151.910195]
[  151.910876] CPU: 7 PID: 138 Comm: kmemleak Not tainted 5.14.0-rc2-
00001-g8cae8cd89f05-dirty #134
[  151.912085] Hardware name: linux,dummy-virt (DT)
[  151.912868] Call trace:
[  151.913211]  dump_backtrace+0x0/0x1b0
[  151.913796]  show_stack+0x1c/0x30
[  151.914248]  dump_stack_lvl+0x68/0x84
[  151.914778]  print_address_description+0x7c/0x2b4
[  151.915340]  kasan_report+0x138/0x38c
[  151.915804]  __do_kernel_fault+0x190/0x1c4
[  151.916386]  do_tag_check_fault+0x78/0x90
[  151.916856]  do_mem_abort+0x44/0xb4
[  151.917308]  el1_abort+0x40/0x60
[  151.917754]  el1h_64_sync_handler+0xb4/0xd0
[  151.918270]  el1h_64_sync+0x78/0x7c
[  151.918714]  scan_block+0x58/0x170
[  151.919157]  scan_gray_list+0xdc/0x1a0
[  151.919626]  kmemleak_scan+0x2ac/0x560
[  151.920129]  kmemleak_scan_thread+0xb0/0xe0
[  151.920635]  kthread+0x154/0x160
[  151.921115]  ret_from_fork+0x10/0x18
[  151.921717]
[  151.922077] Allocated by task 0:
[  151.922523]  kasan_save_stack+0x2c/0x60
[  151.923099]  __kasan_kmalloc+0xec/0x104
[  151.923502]  __kmalloc+0x224/0x3c4
[  151.924172]  __register_sysctl_paths+0x200/0x290
[  151.924709]  register_sysctl_table+0x2c/0x40
[  151.925175]  sysctl_init+0x20/0x34
[  151.925665]  proc_sys_init+0x3c/0x48
[  151.926136]  proc_root_init+0x80/0x9c
[  151.926547]  start_kernel+0x648/0x6a4
[  151.926987]  __primary_switched+0xc0/0xc8
[  151.927557]
[  151.927994] Freed by task 0:
[  151.928340]  kasan_save_stack+0x2c/0x60
[  151.928766]  kasan_set_track+0x2c/0x40
[  151.929173]  kasan_set_free_info+0x44/0x54
[  151.929568]  ____kasan_slab_free.constprop.0+0x150/0x1b0
[  151.930063]  __kasan_slab_free+0x14/0x20
[  151.930449]  slab_free_freelist_hook+0xa4/0x1fc
[  151.930924]  kfree+0x1e8/0x30c
[  151.931285]  put_fs_context+0x124/0x220
[  151.931731]  vfs_kern_mount.part.0+0x60/0xd4
[  151.932280]  kern_mount+0x24/0x4c
[  151.932686]  bdev_cache_init+0x70/0x9c
[  151.933122]  vfs_caches_init+0xdc/0xf4
[  151.933578]  start_kernel+0x638/0x6a4
[  151.934014]  __primary_switched+0xc0/0xc8
[  151.934478]
[  151.934757] The buggy address belongs to the object at
ffff0000c0074e00
[  151.934757]  which belongs to the cache kmalloc-256 of size 256
[  151.935744] The buggy address is located 176 bytes inside of
[  151.935744]  256-byte region [ffff0000c0074e00, ffff0000c0074f00)
[  151.936702] The buggy address belongs to the page:
[  151.937378] page:(____ptrval____) refcount:1 mapcount:0
mapping:0000000000000000 index:0x0 pfn:0x100074
[  151.938682] head:(____ptrval____) order:2 compound_mapcount:0
compound_pincount:0
[  151.939440] flags:
0xbfffc0000010200(slab|head|node=0|zone=2|lastcpupid=0xffff|kasantag=0x
0)
[  151.940886] raw: 0bfffc0000010200 0000000000000000 dead000000000122
f5ff0000c0002300
[  151.941634] raw: 0000000000000000 0000000000200020 00000001ffffffff
0000000000000000
[  151.942353] page dumped because: kasan: bad access detected
[  151.942923]
[  151.943214] Memory state around the buggy address:
[  151.943896]  ffff0000c0074c00: f0 f0 f0 f0 f0 f0 f0 f0 f0 fe fe fe
fe fe fe fe
[  151.944857]  ffff0000c0074d00: fe fe fe fe fe fe fe fe fe fe fe fe
fe fe fe fe
[  151.945892] >ffff0000c0074e00: f7 f7 f7 f7 f7 f7 f7 f7 f7 f7 f7 fe
fe fe fe fe
[  151.946407]                                                     ^
[  151.946939]  ffff0000c0074f00: fe fe fe fe fe fe fe fe fe fe fe fe
fe fe fe fe
[  151.947445]  ffff0000c0075000: fb fb fb fb fb fb fb fb fb fb fb fb
fb fb fb fb
[  151.947999]
==================================================================
[  151.948524] Disabling lock debugging due to kernel taint
[  156.434569] kmemleak: 181 new suspected memory leaks (see
/sys/kernel/debug/kmemleak)

> 
> Perhaps this should have been "kmemleak, kasan: reset pointer tags to
> avoid false positives" ?

Thanks for the suggestions.
But I think it doesn't belong to false
positive becuase scan block
touched invalid metadata certainly.

Maybe "kmemleak, kasan: reset tags when scanning block"?

> 
> > ---
> >  mm/kmemleak.c | 6 +++---
> >  1 file changed, 3 insertions(+), 3 deletions(-)
> > 
> > diff --git a/mm/kmemleak.c b/mm/kmemleak.c
> > index 228a2fbe0657..73d46d16d575 100644
> > --- a/mm/kmemleak.c
> > +++ b/mm/kmemleak.c
> > @@ -290,7 +290,7 @@ static void hex_dump_object(struct seq_file
> > *seq,
> >         warn_or_seq_printf(seq, "  hex dump (first %zu bytes):\n",
> > len);
> >         kasan_disable_current();
> >         warn_or_seq_hex_dump(seq, DUMP_PREFIX_NONE, HEX_ROW_SIZE,
> > -                            HEX_GROUP_SIZE, ptr, len, HEX_ASCII);
> > +                            HEX_GROUP_SIZE, kasan_reset_tag((void
> > *)ptr), len, HEX_ASCII);
> >         kasan_enable_current();
> >  }
> > 
> > @@ -1171,7 +1171,7 @@ static bool update_checksum(struct
> > kmemleak_object *object)
> > 
> >         kasan_disable_current();
> >         kcsan_disable_current();
> > -       object->checksum = crc32(0, (void *)object->pointer,
> > object->size);
> > +       object->checksum = crc32(0, kasan_reset_tag((void *)object-
> > >pointer), object->size);
> >         kasan_enable_current();
> >         kcsan_enable_current();
> > 
> > @@ -1246,7 +1246,7 @@ static void scan_block(void *_start, void
> > *_end,
> >                         break;
> > 
> >                 kasan_disable_current();
> > -               pointer = *ptr;
> > +               pointer = *(unsigned long *)kasan_reset_tag((void
> > *)ptr);
> >                 kasan_enable_current();
> > 
> >                 untagged_ptr = (unsigned long)kasan_reset_tag((void
> > *)pointer);
> > --
> > 2.18.0
> > 
> > --
> > You received this message because you are subscribed to the Google
> > Groups "kasan-dev" group.
> > To unsubscribe from this group and stop receiving emails from it,
> > send an email to kasan-dev+unsubscribe at googlegroups.com.
> > To view this discussion on the web visit 
> > https://urldefense.com/v3/__https://groups.google.com/d/msgid/kasan-dev/20210727040021.21371-2-Kuan-Ying.Lee*40mediatek.com__;JQ!!CTRNKA9wMg0ARbw!wNP4ZkYDM7Xvs9xfzKwYuG1X2h9zFqST8_Vm2jSvZUl9BiS8SPFMTvMp3VAPKCnuWELL7Q$
> >  .


More information about the Linux-mediatek mailing list