stack out-of-bounds write in mce-inject.c

stack out-of-bounds write in mce-inject.c

[Dmitry Vyukov]

Hi,

Opening /sys/kernel/debug/mce-inject/flags overwrites stack:

==================================================================
BUG: KASAN: stack-out-of-bounds in vsnprintf+0x1b23/0x1b40 lib/vsprintf.c:2365
Write of size 1 at addr ffff8800627b7abb by task egrep/4309

CPU: 3 PID: 4309 Comm: egrep Not tainted 4.17.0-rc2+ #179
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/01/2014
Call Trace:
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x1b9/0x294 lib/dump_stack.c:113
 print_address_description+0x6c/0x20b mm/kasan/report.c:256
 kasan_report_error mm/kasan/report.c:366 [inline]
 kasan_report.cold.7+0x242/0x2fe mm/kasan/report.c:424
 __asan_report_store1_noabort+0x17/0x20 mm/kasan/report.c:447
 vsnprintf+0x1b23/0x1b40 lib/vsprintf.c:2365
 sprintf+0xa7/0xd0 lib/vsprintf.c:2494
 flags_read+0xa3/0xf0 arch/x86/kernel/cpu/mcheck/mce-inject.c:345
 full_proxy_read+0x125/0x190 fs/debugfs/file.c:195
 __vfs_read+0x10f/0xa50 fs/read_write.c:416
 vfs_read+0x17f/0x3d0 fs/read_write.c:452
 ksys_read+0xf9/0x250 fs/read_write.c:578
 __do_sys_read fs/read_write.c:588 [inline]
 __se_sys_read fs/read_write.c:586 [inline]
 __x64_sys_read+0x73/0xb0 fs/read_write.c:586
 do_syscall_64+0x1b1/0x800 arch/x86/entry/common.c:287
 entry_SYSCALL_64_after_hwframe+0x49/0xbe
RIP: 0033:0x7fb17c6c3310
RSP: 002b:00007ffca98af8e8 EFLAGS: 00000246 ORIG_RAX: 0000000000000000
RAX: ffffffffffffffda RBX: 0000000001d72000 RCX: 00007fb17c6c3310
RDX: 0000000000008000 RSI: 0000000001d72000 RDI: 0000000000000005
RBP: 0000000000001000 R08: 0000000001d72000 R09: 00007fb17c979e90
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 0000000000008000 R14: 0000000000000000 R15: 0000000000000003

The buggy address belongs to the page:
page:ffffea000189edc0 count:0 mapcount:0 mapping:0000000000000000 index:0x0
flags: 0x1fffc0000000000()
raw: 01fffc0000000000 0000000000000000 0000000000000000 00000000ffffffff
raw: 0000000000000000 ffffea0001890101 0000000000000000 0000000000000000
page dumped because: kasan: bad access detected

Memory state around the buggy address:
 ffff8800627b7980: f3 f3 f3 f3 00 00 00 00 00 00 00 f1 f1 f1 f1 00
 ffff8800627b7a00: 00 00 f2 f3 f3 f3 f3 00 00 00 00 00 00 00 00 00
>ffff8800627b7a80: 00 00 00 f1 f1 f1 f1 03 f2 f2 f2 f3 f3 f3 f3 00
                                        ^
 ffff8800627b7b00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 ffff8800627b7b80: 00 00 00 00 00 f1 f1 f1 f1 00 00 f2 f2 f2 f2 f2
==================================================================

On upstream commit 0644f186fc9d77bb5bd198369e59fb28927a3692.

Re: stack out-of-bounds write in mce-inject.c

[Borislav Petkov]

On Fri, Apr 27, 2018 at 05:24:24PM +0200, Dmitry Vyukov wrote:
> Hi,
> 
> Opening /sys/kernel/debug/mce-inject/flags overwrites stack:
> 
> ==================================================================
> BUG: KASAN: stack-out-of-bounds in vsnprintf+0x1b23/0x1b40 lib/vsprintf.c:2365
> Write of size 1 at addr ffff8800627b7abb by task egrep/4309

How do you trigger this exactly?

You grep it for something apparently...

Thx.

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.

Re: stack out-of-bounds write in mce-inject.c

[Dmitry Vyukov]

On Fri, Apr 27, 2018 at 5:41 PM, Borislav Petkov <bp@alien8.de> wrote:
> On Fri, Apr 27, 2018 at 05:24:24PM +0200, Dmitry Vyukov wrote:
>> Hi,
>>
>> Opening /sys/kernel/debug/mce-inject/flags overwrites stack:
>>
>> ==================================================================
>> BUG: KASAN: stack-out-of-bounds in vsnprintf+0x1b23/0x1b40 lib/vsprintf.c:2365
>> Write of size 1 at addr ffff8800627b7abb by task egrep/4309
>
> How do you trigger this exactly?
>
> You grep it for something apparently...


find /sys -exec grep "64" {} \; -print

[PATCH] x86, mce: Fix stack out-of-bounds write in mce-inject.c:flags_read()

[Luck, Tony]

Each of the strings that we want to put into the buf[MAX_FLAG_OPT_SIZE]
in flags_read() is two characters. But the sprintf() adds a trailing newline
and will add a terminating NUL byte. So MAX_FLAG_OPT_SIZE needs to be 4.

Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Tony Luck <tony.luck@intel.com>

---

diff --git a/arch/x86/kernel/cpu/mcheck/mce-inject.c b/arch/x86/kernel/cpu/mcheck/mce-inject.c
index 475cb4f5f14f..c805a06e14c3 100644
--- a/arch/x86/kernel/cpu/mcheck/mce-inject.c
+++ b/arch/x86/kernel/cpu/mcheck/mce-inject.c
@@ -48,7 +48,7 @@ static struct dentry *dfs_inj;
 
 static u8 n_banks;
 
-#define MAX_FLAG_OPT_SIZE	3
+#define MAX_FLAG_OPT_SIZE	4
 #define NBCFG			0x44
 
 enum injection_type {

Re: [PATCH] x86, mce: Fix stack out-of-bounds write in mce-inject.c:flags_read()

[Borislav Petkov]

On Fri, Apr 27, 2018 at 09:37:08AM -0700, Luck, Tony wrote:
> Each of the strings that we want to put into the buf[MAX_FLAG_OPT_SIZE]
> in flags_read() is two characters. But the sprintf() adds a trailing newline
> and will add a terminating NUL byte. So MAX_FLAG_OPT_SIZE needs to be 4.

... and I dumped the n here:

	n = sprintf(buf, "%s\n", flags_options[inj_type])

and it was 3 but sprintf() calls vsnprintf() and *that* does return:

" * The return value is the number of characters which would
 * be generated for the given input, excluding the trailing
 * '\0', as per ISO C99."

So I'll extend the commit message with that and apply it.

Thanks.

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.

[tip:ras/core] x86/MCE: Fix stack out-of-bounds write in mce-inject.c: Flags_read()

[tip-bot for Luck, Tony]

Commit-ID:  985c78d3ff8e9c74450fa2bb08eb55e680d999ca
Gitweb:     https://git.kernel.org/tip/985c78d3ff8e9c74450fa2bb08eb55e680d999ca
Author:     Luck, Tony <tony.luck@intel.com>
AuthorDate: Fri, 27 Apr 2018 09:37:08 -0700
Committer:  Thomas Gleixner <tglx@linutronix.de>
CommitDate: Sun, 6 May 2018 12:46:39 +0200

x86/MCE: Fix stack out-of-bounds write in mce-inject.c: Flags_read()

Each of the strings that we want to put into the buf[MAX_FLAG_OPT_SIZE]
in flags_read() is two characters long. But the sprintf() adds
a trailing newline and will add a terminating NUL byte. So
MAX_FLAG_OPT_SIZE needs to be 4.

sprintf() calls vsnprintf() and *that* does return:

" * The return value is the number of characters which would
 * be generated for the given input, excluding the trailing
 * '\0', as per ISO C99."

Note the "excluding".

Reported-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Cc: linux-edac <linux-edac@vger.kernel.org>
Link: http://lkml.kernel.org/r/20180427163707.ktaiysvbk3yhk4wm@agluck-desk
---
 arch/x86/kernel/cpu/mcheck/mce-inject.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/x86/kernel/cpu/mcheck/mce-inject.c b/arch/x86/kernel/cpu/mcheck/mce-inject.c
index 475cb4f5f14f..c805a06e14c3 100644
--- a/arch/x86/kernel/cpu/mcheck/mce-inject.c
+++ b/arch/x86/kernel/cpu/mcheck/mce-inject.c
@@ -48,7 +48,7 @@ static struct dentry *dfs_inj;
 
 static u8 n_banks;
 
-#define MAX_FLAG_OPT_SIZE	3
+#define MAX_FLAG_OPT_SIZE	4
 #define NBCFG			0x44
 
 enum injection_type {