linux-kernel - Re: [PATCH] slub: fix a crash with SLUB_DEBUG + KASAN_SW

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
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAAeHK+wp=UtzoNzcnmnef_uMpEMSeXBgKYOJz7V_tZXs-RK+kQ@mail.gmail.com>
Date:   Wed, 20 Feb 2019 13:51:35 +0100
From:   Andrey Konovalov <andreyknvl@...gle.com>
To:     Qian Cai <cai@....pw>
Cc:     Andrew Morton <akpm@...ux-foundation.org>,
        Christoph Lameter <cl@...ux.com>,
        Pekka Enberg <penberg@...nel.org>,
        David Rientjes <rientjes@...gle.com>,
        Joonsoo Kim <iamjoonsoo.kim@....com>,
        Linux Memory Management List <linux-mm@...ck.org>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] slub: fix a crash with SLUB_DEBUG + KASAN_SW_TAGS

On Wed, Feb 20, 2019 at 3:03 AM Qian Cai <cai@....pw> wrote:
>
> In process_slab(), "p = get_freepointer()" could return a tagged
> pointer, but "addr = page_address()" always return a native pointer. As
> the result, slab_index() is messed up here,
>
> return (p - addr) / s->size;
>
> All other callers of slab_index() have the same situation where "addr"
> is from page_address(), so just need to untag "p".
>
>  # cat /sys/kernel/slab/hugetlbfs_inode_cache/alloc_calls
>
> [18868.759419] Unable to handle kernel paging request at virtual address 2bff808aa4856d48
> [18868.767341] Mem abort info:
> [18868.770133]   ESR = 0x96000007
> [18868.773187]   Exception class = DABT (current EL), IL = 32 bits
> [18868.779103]   SET = 0, FnV = 0
> [18868.782155]   EA = 0, S1PTW = 0
> [18868.785292] Data abort info:
> [18868.788170]   ISV = 0, ISS = 0x00000007
> [18868.792003]   CM = 0, WnR = 0
> [18868.794973] swapper pgtable: 64k pages, 48-bit VAs, pgdp = 0000000002498338
> [18868.801932] [2bff808aa4856d48] pgd=00000097fcfd0003, pud=00000097fcfd0003, pmd=00000097fca30003, pte=00e8008b24850712
> [18868.812597] Internal error: Oops: 96000007 [#1] SMP
> [18868.835088] CPU: 3 PID: 79210 Comm: read_all Tainted: G             L    5.0.0-rc7+ #84
> [18868.843087] Hardware name: HPE Apollo 70             /C01_APACHE_MB         , BIOS L50_5.13_1.0.6 07/10/2018
> [18868.852915] pstate: 00400089 (nzcv daIf +PAN -UAO)
> [18868.857710] pc : get_map+0x78/0xec
> [18868.861109] lr : get_map+0xa0/0xec
> [18868.864505] sp : aeff808989e3f8e0
> [18868.867816] x29: aeff808989e3f940 x28: ffff800826200000
> [18868.873128] x27: ffff100012d47000 x26: 9700000000002500
> [18868.878440] x25: 0000000000000001 x24: 52ff8008200131f8
> [18868.883753] x23: 52ff8008200130a0 x22: 52ff800820013098
> [18868.889065] x21: ffff800826200000 x20: ffff100013172ba0
> [18868.894377] x19: 2bff808a8971bc00 x18: ffff1000148f5538
> [18868.899690] x17: 000000000000001b x16: 00000000000000ff
> [18868.905002] x15: ffff1000148f5000 x14: 00000000000000d2
> [18868.910314] x13: 0000000000000001 x12: 0000000000000000
> [18868.915626] x11: 0000000020000002 x10: 2bff808aa4856d48
> [18868.920937] x9 : 0000020000000000 x8 : 68ff80082620ebb0
> [18868.926249] x7 : 0000000000000000 x6 : ffff1000105da1dc
> [18868.931561] x5 : 0000000000000000 x4 : 0000000000000000
> [18868.936872] x3 : 0000000000000010 x2 : 2bff808a8971bc00
> [18868.942184] x1 : ffff7fe002098800 x0 : ffff80082620ceb0
> [18868.947499] Process read_all (pid: 79210, stack limit = 0x00000000f65b9361)
> [18868.954454] Call trace:
> [18868.956899]  get_map+0x78/0xec
> [18868.959952]  process_slab+0x7c/0x47c
> [18868.963526]  list_locations+0xb0/0x3c8
> [18868.967273]  alloc_calls_show+0x34/0x40
> [18868.971107]  slab_attr_show+0x34/0x48
> [18868.974768]  sysfs_kf_seq_show+0x2e4/0x570
> [18868.978864]  kernfs_seq_show+0x12c/0x1a0
> [18868.982786]  seq_read+0x48c/0xf84
> [18868.986099]  kernfs_fop_read+0xd4/0x448
> [18868.989935]  __vfs_read+0x94/0x5d4
> [18868.993334]  vfs_read+0xcc/0x194
> [18868.996560]  ksys_read+0x6c/0xe8
> [18868.999786]  __arm64_sys_read+0x68/0xb0
> [18869.003622]  el0_svc_handler+0x230/0x3bc
> [18869.007544]  el0_svc+0x8/0xc
> [18869.010428] Code: d3467d2a 9ac92329 8b0a0e6a f9800151 (c85f7d4b)
> [18869.016742] ---[ end trace a383a9a44ff13176 ]---
> [18869.021356] Kernel panic - not syncing: Fatal exception
> [18869.026705] SMP: stopping secondary CPUs
> [18870.254279] SMP: failed to stop secondary CPUs 1-7,32,40,127
> [18870.259942] Kernel Offset: disabled
> [18870.263434] CPU features: 0x002,20000c18
> [18870.267358] Memory Limit: none
> [18870.270725] ---[ end Kernel panic - not syncing: Fatal exception ]---
>
> Signed-off-by: Qian Cai <cai@....pw>

Reviewed-by: Andrey Konovalov <andreyknvl@...gle.com>

> ---
>  mm/slub.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/mm/slub.c b/mm/slub.c
> index 4a61959e1887..289c22f1b0c4 100644
> --- a/mm/slub.c
> +++ b/mm/slub.c
> @@ -311,7 +311,7 @@ static inline void set_freepointer(struct kmem_cache *s, void *object, void *fp)
>  /* Determine object index from a given position */
>  static inline unsigned int slab_index(void *p, struct kmem_cache *s, void *addr)
>  {
> -       return (p - addr) / s->size;
> +       return (kasan_reset_tag(p) - addr) / s->size;
>  }
>
>  static inline unsigned int order_objects(unsigned int order, unsigned int size)
> --
> 2.17.2 (Apple Git-113)
>