On 13/08/19 3:25 PM, Keerthy wrote:


On 13/08/19 3:03 PM, Simon Glass wrote:
Hi Keethy,

On Tue, 9 Jul 2019 at 13:38, Eugeniu Rosca <roscaeuge...@gmail.com> wrote:

Hi All,

On Wed, Apr 24, 2019 at 05:19:53PM +0530, Keerthy wrote:
Add ofnode_get_addr_size_index function to fetch the address
and size of the reg space based on index.

Signed-off-by: Keerthy <j-keer...@ti.com>
Reviewed-by: Simon Glass <s...@chromium.org>
---

Previous  discussion can be found here:

   * https://patchwork.ozlabs.org/patch/999659/

  drivers/core/ofnode.c | 13 ++++++++++---
  include/dm/ofnode.h   | 14 ++++++++++++++
  2 files changed, 24 insertions(+), 3 deletions(-)

jFYI/FWIW, this patch, present in U-Boot/master as commit [1], revives
the issue fixed in [2]. More precisely, when compiling the sandbox
U-Boot with -fsanitize=address and linking it against -lasan,
the following backtrace pops up during boot [3].

[1] https://gitlab.denx.de/u-boot/u-boot/commit/e679d03b08fb
     ("core: ofnode: Add ofnode_get_addr_size_index")
[2] https://gitlab.denx.de/u-boot/u-boot/commit/9bfacf249b10
     ("core: ofnode: Fix ASAN-reported stack-buffer-overflow in of_get_address")
[3] The same issue as fixed in [2], now reproduced on v2019.07

==5996==ERROR: AddressSanitizer: stack-buffer-overflow on address 0x7ffc65f869a0 at pc 0x564834c6cd89 bp 0x7ffc65f867b0 sp 0x7ffc65f867a0
WRITE of size 8 at 0x7ffc65f869a0 thread T0
     #0 0x564834c6cd88 in of_get_address drivers/core/of_addr.c:154
     #1 0x564834c74fac in ofnode_get_addr_size_index drivers/core/ofnode.c:262      #2 0x564834c75232 in ofnode_get_addr_index drivers/core/ofnode.c:290      #3 0x564834d35a4f in sb_eth_ofdata_to_platdata drivers/net/sandbox.c:422
     #4 0x564834c5ad22 in device_probe drivers/core/device.c:407
     #5 0x564834f2f1c4 in eth_initialize net/eth-uclass.c:399
     #6 0x564834bf9c34 in initr_net common/board_r.c:565
     #7 0x564834bfa580 in initcall_run_list include/initcall.h:38
     #8 0x564834bfa580 in board_init_r common/board_r.c:867
     #9 0x564834b8156d in main arch/sandbox/cpu/start.c:362
     #10 0x7f2fe7912b6a in __libc_start_main (/lib/x86_64-linux-gnu/libc.so.6+0x26b6a)
     #11 0x564834b85b29 in _start (/srv/R/u-boot-master/u-boot+0x792b29)

Address 0x7ffc65f869a0 is located in stack of thread T0 at offset 32 in frame      #0 0x564834c751c4 in ofnode_get_addr_index drivers/core/ofnode.c:287

   This frame has 1 object(s):
     [32, 36) 'size' (line 288) <== Memory access at offset 32 partially overflows this variable HINT: this may be a false positive if your program uses some custom stack unwind mechanism, swapcontext or vfork
       (longjmp and C++ exceptions *are* supported)
SUMMARY: AddressSanitizer: stack-buffer-overflow drivers/core/of_addr.c:154 in of_get_address
Shadow bytes around the buggy address:
   0x10000cbe8ce0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
   0x10000cbe8cf0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
   0x10000cbe8d00: 00 00 f1 f1 f1 f1 04 f2 04 f2 04 f3 f3 f3 00 00
   0x10000cbe8d10: 00 00 00 00 00 00 00 00 00 00 f1 f1 f1 f1 f1 f1
   0x10000cbe8d20: 04 f2 00 f3 f3 f3 00 00 00 00 00 00 00 00 00 00
=>0x10000cbe8d30: f1 f1 f1 f1[04]f3 f3 f3 00 00 00 00 00 00 00 00
   0x10000cbe8d40: 00 00 00 00 00 00 00 00 00 00 f1 f1 f1 f1 00 00
   0x10000cbe8d50: f3 f3 00 00 00 00 00 00 00 00 00 00 00 00 00 00
   0x10000cbe8d60: f1 f1 f1 f1 00 f3 f3 f3 00 00 00 00 00 00 00 00
   0x10000cbe8d70: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
   0x10000cbe8d80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
Shadow byte legend (one shadow byte represents 8 application bytes):
   Addressable:           00
   Partially addressable: 01 02 03 04 05 06 07
   Heap left redzone:       fa
   Freed heap region:       fd
   Stack left redzone:      f1
   Stack mid redzone:       f2
   Stack right redzone:     f3
   Stack after return:      f5
   Stack use after scope:   f8
   Global redzone:          f9
   Global init order:       f6
   Poisoned by user:        f7
   Container overflow:      fc
   Array cookie:            ac
   Intra object redzone:    bb
   ASan internal:           fe
   Left alloca redzone:     ca
   Right alloca redzone:    cb
   Shadow gap:              cc
==5996==ABORTING


Can you please take a look?

Sure.

Eugeniu,

Could you kindly help me with the steps to reproduce the issue?

I am sending out a fix patch to convert all the size instances to fdt_size_t type.

I have posted: https://patchwork.ozlabs.org/patch/1146830/

I believe the above should fix the size mismatch.

Regards,
Keerthy


Regards,
Simon

_______________________________________________
U-Boot mailing list
U-Boot@lists.denx.de
https://lists.denx.de/listinfo/u-boot
_______________________________________________
U-Boot mailing list
U-Boot@lists.denx.de
https://lists.denx.de/listinfo/u-boot

Reply via email to