2022-11-15 16:32:25

by George Kennedy

[permalink] [raw]
Subject: [PATCH] ubi: ensure that VID header offset + VID header size <= alloc, size

Ensure that the VID header offset + VID header size does not exceed
the allocated area to avoid slab OOB.

BUG: KASAN: slab-out-of-bounds in crc32_body lib/crc32.c:111 [inline]
BUG: KASAN: slab-out-of-bounds in crc32_le_generic lib/crc32.c:179 [inline]
BUG: KASAN: slab-out-of-bounds in crc32_le_base+0x58c/0x626 lib/crc32.c:197
Read of size 4 at addr ffff88802bb36f00 by task syz-executor136/1555

CPU: 2 PID: 1555 Comm: syz-executor136 Tainted: G W
6.0.0-1868 #1
Hardware name: Red Hat KVM, BIOS 1.13.0-2.module+el8.3.0+7860+a7792d29
04/01/2014
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0x85/0xad lib/dump_stack.c:106
print_address_description mm/kasan/report.c:317 [inline]
print_report.cold.13+0xb6/0x6bb mm/kasan/report.c:433
kasan_report+0xa7/0x11b mm/kasan/report.c:495
crc32_body lib/crc32.c:111 [inline]
crc32_le_generic lib/crc32.c:179 [inline]
crc32_le_base+0x58c/0x626 lib/crc32.c:197
ubi_io_write_vid_hdr+0x1b7/0x472 drivers/mtd/ubi/io.c:1067
create_vtbl+0x4d5/0x9c4 drivers/mtd/ubi/vtbl.c:317
create_empty_lvol drivers/mtd/ubi/vtbl.c:500 [inline]
ubi_read_volume_table+0x67b/0x288a drivers/mtd/ubi/vtbl.c:812
ubi_attach+0xf34/0x1603 drivers/mtd/ubi/attach.c:1601
ubi_attach_mtd_dev+0x6f3/0x185e drivers/mtd/ubi/build.c:965
ctrl_cdev_ioctl+0x2db/0x347 drivers/mtd/ubi/cdev.c:1043
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:870 [inline]
__se_sys_ioctl fs/ioctl.c:856 [inline]
__x64_sys_ioctl+0x193/0x213 fs/ioctl.c:856
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x3e/0x86 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0x0
RIP: 0033:0x7f96d5cf753d
Code:
RSP: 002b:00007fffd72206f8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f96d5cf753d
RDX: 0000000020000080 RSI: 0000000040186f40 RDI: 0000000000000003
RBP: 0000000000400cd0 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000400be0
R13: 00007fffd72207e0 R14: 0000000000000000 R15: 0000000000000000
</TASK>

Allocated by task 1555:
kasan_save_stack+0x20/0x3d mm/kasan/common.c:38
kasan_set_track mm/kasan/common.c:45 [inline]
set_alloc_info mm/kasan/common.c:437 [inline]
____kasan_kmalloc mm/kasan/common.c:516 [inline]
__kasan_kmalloc+0x88/0xa3 mm/kasan/common.c:525
kasan_kmalloc include/linux/kasan.h:234 [inline]
__kmalloc+0x138/0x257 mm/slub.c:4429
kmalloc include/linux/slab.h:605 [inline]
ubi_alloc_vid_buf drivers/mtd/ubi/ubi.h:1093 [inline]
create_vtbl+0xcc/0x9c4 drivers/mtd/ubi/vtbl.c:295
create_empty_lvol drivers/mtd/ubi/vtbl.c:500 [inline]
ubi_read_volume_table+0x67b/0x288a drivers/mtd/ubi/vtbl.c:812
ubi_attach+0xf34/0x1603 drivers/mtd/ubi/attach.c:1601
ubi_attach_mtd_dev+0x6f3/0x185e drivers/mtd/ubi/build.c:965
ctrl_cdev_ioctl+0x2db/0x347 drivers/mtd/ubi/cdev.c:1043
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:870 [inline]
__se_sys_ioctl fs/ioctl.c:856 [inline]
__x64_sys_ioctl+0x193/0x213 fs/ioctl.c:856
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x3e/0x86 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0x0

The buggy address belongs to the object at ffff88802bb36e00
which belongs to the cache kmalloc-256 of size 256
The buggy address is located 0 bytes to the right of
256-byte region [ffff88802bb36e00, ffff88802bb36f00)

The buggy address belongs to the physical page:
page:00000000ea4d1263 refcount:1 mapcount:0 mapping:0000000000000000
index:0x0 pfn:0x2bb36
head:00000000ea4d1263 order:1 compound_mapcount:0 compound_pincount:0
flags: 0xfffffc0010200(slab|head|node=0|zone=1|lastcpupid=0x1fffff)
raw: 000fffffc0010200 ffffea000066c300 dead000000000003 ffff888100042b40
raw: 0000000000000000 0000000000100010 00000001ffffffff 0000000000000000
page dumped because: kasan: bad access detected

Memory state around the buggy address:
ffff88802bb36e00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
ffff88802bb36e80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>ffff88802bb36f00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
^
ffff88802bb36f80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
ffff88802bb37000: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
==================================================================

Fixes: 801c135ce73d ("UBI: Unsorted Block Images")
Reported-by: syzkaller <[email protected]>
Signed-off-by: George Kennedy <[email protected]>
---
drivers/mtd/ubi/build.c | 6 ++++++
1 file changed, 6 insertions(+)

diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c
index a32050fecabf..53aa4de6b963 100644
--- a/drivers/mtd/ubi/build.c
+++ b/drivers/mtd/ubi/build.c
@@ -663,6 +663,12 @@ static int io_init(struct ubi_device *ubi, int max_beb_per1024)
ubi->ec_hdr_alsize = ALIGN(UBI_EC_HDR_SIZE, ubi->hdrs_min_io_size);
ubi->vid_hdr_alsize = ALIGN(UBI_VID_HDR_SIZE, ubi->hdrs_min_io_size);

+ if (ubi->vid_hdr_offset && ((ubi->vid_hdr_offset + UBI_VID_HDR_SIZE) >
+ ubi->vid_hdr_alsize)) {
+ ubi_err(ubi, "VID header offset %d too large.", ubi->vid_hdr_offset);
+ return -EINVAL;
+ }
+
dbg_gen("min_io_size %d", ubi->min_io_size);
dbg_gen("max_write_size %d", ubi->max_write_size);
dbg_gen("hdrs_min_io_size %d", ubi->hdrs_min_io_size);
--
2.31.1




2022-11-19 21:39:32

by Richard Weinberger

[permalink] [raw]
Subject: Re: [PATCH] ubi: ensure that VID header offset + VID header size <= alloc, size

----- Ursprüngliche Mail -----
> Fixes: 801c135ce73d ("UBI: Unsorted Block Images")
> Reported-by: syzkaller <[email protected]>
> Signed-off-by: George Kennedy <[email protected]>
> ---
> drivers/mtd/ubi/build.c | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c
> index a32050fecabf..53aa4de6b963 100644
> --- a/drivers/mtd/ubi/build.c
> +++ b/drivers/mtd/ubi/build.c
> @@ -663,6 +663,12 @@ static int io_init(struct ubi_device *ubi, int
> max_beb_per1024)
> ubi->ec_hdr_alsize = ALIGN(UBI_EC_HDR_SIZE, ubi->hdrs_min_io_size);
> ubi->vid_hdr_alsize = ALIGN(UBI_VID_HDR_SIZE, ubi->hdrs_min_io_size);
>
> + if (ubi->vid_hdr_offset && ((ubi->vid_hdr_offset + UBI_VID_HDR_SIZE) >
> + ubi->vid_hdr_alsize)) {
> + ubi_err(ubi, "VID header offset %d too large.", ubi->vid_hdr_offset);
> + return -EINVAL;
> + }
> +

Thanks for fixing this, applied to the fixes branch.
That said, something broke whitespaces in your patch, I applied it manually.
Context lines before and after contain an extra space and caused both
git am and patch to give up.

Thanks,
//richard

2023-04-17 16:02:51

by Uwe Kleine-König

[permalink] [raw]
Subject: [Regression] Cannot overwrite VID header offset any more [Was: [PATCH] ubi: ensure that VID header offset + VID header size <= alloc, size]

Hello,

On Tue, Nov 15, 2022 at 10:14:44AM -0500, George Kennedy wrote:
> Ensure that the VID header offset + VID header size does not exceed
> the allocated area to avoid slab OOB.
>
> BUG: KASAN: slab-out-of-bounds in crc32_body lib/crc32.c:111 [inline]
> BUG: KASAN: slab-out-of-bounds in crc32_le_generic lib/crc32.c:179 [inline]
> BUG: KASAN: slab-out-of-bounds in crc32_le_base+0x58c/0x626 lib/crc32.c:197
> Read of size 4 at addr ffff88802bb36f00 by task syz-executor136/1555
>
> CPU: 2 PID: 1555 Comm: syz-executor136 Tainted: G W
> 6.0.0-1868 #1
> Hardware name: Red Hat KVM, BIOS 1.13.0-2.module+el8.3.0+7860+a7792d29
> 04/01/2014
> Call Trace:
> <TASK>
> __dump_stack lib/dump_stack.c:88 [inline]
> dump_stack_lvl+0x85/0xad lib/dump_stack.c:106
> print_address_description mm/kasan/report.c:317 [inline]
> print_report.cold.13+0xb6/0x6bb mm/kasan/report.c:433
> kasan_report+0xa7/0x11b mm/kasan/report.c:495
> crc32_body lib/crc32.c:111 [inline]
> crc32_le_generic lib/crc32.c:179 [inline]
> crc32_le_base+0x58c/0x626 lib/crc32.c:197
> ubi_io_write_vid_hdr+0x1b7/0x472 drivers/mtd/ubi/io.c:1067
> create_vtbl+0x4d5/0x9c4 drivers/mtd/ubi/vtbl.c:317
> create_empty_lvol drivers/mtd/ubi/vtbl.c:500 [inline]
> ubi_read_volume_table+0x67b/0x288a drivers/mtd/ubi/vtbl.c:812
> ubi_attach+0xf34/0x1603 drivers/mtd/ubi/attach.c:1601
> ubi_attach_mtd_dev+0x6f3/0x185e drivers/mtd/ubi/build.c:965
> ctrl_cdev_ioctl+0x2db/0x347 drivers/mtd/ubi/cdev.c:1043
> vfs_ioctl fs/ioctl.c:51 [inline]
> __do_sys_ioctl fs/ioctl.c:870 [inline]
> __se_sys_ioctl fs/ioctl.c:856 [inline]
> __x64_sys_ioctl+0x193/0x213 fs/ioctl.c:856
> do_syscall_x64 arch/x86/entry/common.c:50 [inline]
> do_syscall_64+0x3e/0x86 arch/x86/entry/common.c:80
> entry_SYSCALL_64_after_hwframe+0x63/0x0
> RIP: 0033:0x7f96d5cf753d
> Code:
> RSP: 002b:00007fffd72206f8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
> RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f96d5cf753d
> RDX: 0000000020000080 RSI: 0000000040186f40 RDI: 0000000000000003
> RBP: 0000000000400cd0 R08: 0000000000000000 R09: 0000000000000000
> R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000400be0
> R13: 00007fffd72207e0 R14: 0000000000000000 R15: 0000000000000000
> </TASK>
>
> Allocated by task 1555:
> kasan_save_stack+0x20/0x3d mm/kasan/common.c:38
> kasan_set_track mm/kasan/common.c:45 [inline]
> set_alloc_info mm/kasan/common.c:437 [inline]
> ____kasan_kmalloc mm/kasan/common.c:516 [inline]
> __kasan_kmalloc+0x88/0xa3 mm/kasan/common.c:525
> kasan_kmalloc include/linux/kasan.h:234 [inline]
> __kmalloc+0x138/0x257 mm/slub.c:4429
> kmalloc include/linux/slab.h:605 [inline]
> ubi_alloc_vid_buf drivers/mtd/ubi/ubi.h:1093 [inline]
> create_vtbl+0xcc/0x9c4 drivers/mtd/ubi/vtbl.c:295
> create_empty_lvol drivers/mtd/ubi/vtbl.c:500 [inline]
> ubi_read_volume_table+0x67b/0x288a drivers/mtd/ubi/vtbl.c:812
> ubi_attach+0xf34/0x1603 drivers/mtd/ubi/attach.c:1601
> ubi_attach_mtd_dev+0x6f3/0x185e drivers/mtd/ubi/build.c:965
> ctrl_cdev_ioctl+0x2db/0x347 drivers/mtd/ubi/cdev.c:1043
> vfs_ioctl fs/ioctl.c:51 [inline]
> __do_sys_ioctl fs/ioctl.c:870 [inline]
> __se_sys_ioctl fs/ioctl.c:856 [inline]
> __x64_sys_ioctl+0x193/0x213 fs/ioctl.c:856
> do_syscall_x64 arch/x86/entry/common.c:50 [inline]
> do_syscall_64+0x3e/0x86 arch/x86/entry/common.c:80
> entry_SYSCALL_64_after_hwframe+0x63/0x0
>
> The buggy address belongs to the object at ffff88802bb36e00
> which belongs to the cache kmalloc-256 of size 256
> The buggy address is located 0 bytes to the right of
> 256-byte region [ffff88802bb36e00, ffff88802bb36f00)
>
> The buggy address belongs to the physical page:
> page:00000000ea4d1263 refcount:1 mapcount:0 mapping:0000000000000000
> index:0x0 pfn:0x2bb36
> head:00000000ea4d1263 order:1 compound_mapcount:0 compound_pincount:0
> flags: 0xfffffc0010200(slab|head|node=0|zone=1|lastcpupid=0x1fffff)
> raw: 000fffffc0010200 ffffea000066c300 dead000000000003 ffff888100042b40
> raw: 0000000000000000 0000000000100010 00000001ffffffff 0000000000000000
> page dumped because: kasan: bad access detected
>
> Memory state around the buggy address:
> ffff88802bb36e00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> ffff88802bb36e80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> > ffff88802bb36f00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> ^
> ffff88802bb36f80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> ffff88802bb37000: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> ==================================================================
>
> Fixes: 801c135ce73d ("UBI: Unsorted Block Images")
> Reported-by: syzkaller <[email protected]>
> Signed-off-by: George Kennedy <[email protected]>
> ---
> drivers/mtd/ubi/build.c | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c
> index a32050fecabf..53aa4de6b963 100644
> --- a/drivers/mtd/ubi/build.c
> +++ b/drivers/mtd/ubi/build.c
> @@ -663,6 +663,12 @@ static int io_init(struct ubi_device *ubi, int max_beb_per1024)
> ubi->ec_hdr_alsize = ALIGN(UBI_EC_HDR_SIZE, ubi->hdrs_min_io_size);
> ubi->vid_hdr_alsize = ALIGN(UBI_VID_HDR_SIZE, ubi->hdrs_min_io_size);
> + if (ubi->vid_hdr_offset && ((ubi->vid_hdr_offset + UBI_VID_HDR_SIZE) >
> + ubi->vid_hdr_alsize)) {
> + ubi_err(ubi, "VID header offset %d too large.", ubi->vid_hdr_offset);
> + return -EINVAL;
> + }
> +

This patch is in mainline as 1b42b1a36fc946f0d7088425b90d491b4257ca3e,
and backported to various stable releases.

For me this breaks

ubiattach -m 0 -O 2048

I think the check

ubi->vid_hdr_offset + UBI_VID_HDR_SIZE > ubi->vid_hdr_alsize

is wrong. Without -O passed to ubiattach (and dynamic debug enabled) I
get:

[ 5294.936762] UBI DBG gen (pid 9619): sizeof(struct ubi_ainf_peb) 56
[ 5294.936769] UBI DBG gen (pid 9619): sizeof(struct ubi_wl_entry) 32
[ 5294.936774] UBI DBG gen (pid 9619): min_io_size 2048
[ 5294.936779] UBI DBG gen (pid 9619): max_write_size 2048
[ 5294.936783] UBI DBG gen (pid 9619): hdrs_min_io_size 512
[ 5294.936787] UBI DBG gen (pid 9619): ec_hdr_alsize 512
[ 5294.936791] UBI DBG gen (pid 9619): vid_hdr_alsize 512
[ 5294.936796] UBI DBG gen (pid 9619): vid_hdr_offset 512
[ 5294.936800] UBI DBG gen (pid 9619): vid_hdr_aloffset 512
[ 5294.936804] UBI DBG gen (pid 9619): vid_hdr_shift 0
[ 5294.936808] UBI DBG gen (pid 9619): leb_start 2048
[ 5294.936812] UBI DBG gen (pid 9619): max_erroneous 409

So the check would only pass for vid_hdr_offset <= 512 -
UBI_VID_HDR_SIZE; note that even specifying the default value 512 (i.e.

ubiattach -m 0 -O 512

) fails the check.

A less strong check would be:

diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c
index 0904eb40c95f..69c28a862430 100644
--- a/drivers/mtd/ubi/build.c
+++ b/drivers/mtd/ubi/build.c
@@ -666,8 +666,8 @@ static int io_init(struct ubi_device *ubi, int max_beb_per1024)
ubi->ec_hdr_alsize = ALIGN(UBI_EC_HDR_SIZE, ubi->hdrs_min_io_size);
ubi->vid_hdr_alsize = ALIGN(UBI_VID_HDR_SIZE, ubi->hdrs_min_io_size);

- if (ubi->vid_hdr_offset && ((ubi->vid_hdr_offset + UBI_VID_HDR_SIZE) >
- ubi->vid_hdr_alsize)) {
+ if (ubi->vid_hdr_offset &&
+ ubi->vid_hdr_offset + UBI_VID_HDR_SIZE > ubi->peb_size) {
ubi_err(ubi, "VID header offset %d too large.", ubi->vid_hdr_offset);
return -EINVAL;
}

But I'm unsure if this would be too lax?!

Best regards
Uwe

--
Pengutronix e.K. | Uwe Kleine-K?nig |
Industrial Linux Solutions | https://www.pengutronix.de/ |


Attachments:
(No filename) (7.63 kB)
signature.asc (499.00 B)
Download all attachments

2023-04-17 16:06:35

by Uwe Kleine-König

[permalink] [raw]
Subject: Re: [Regression] Cannot overwrite VID header offset any more [Was: [PATCH] ubi: ensure that VID header offset + VID header size <= alloc, size]

Hello,

On Mon, Apr 17, 2023 at 06:01:02PM +0200, Uwe Kleine-K?nig wrote:
> On Tue, Nov 15, 2022 at 10:14:44AM -0500, George Kennedy wrote:
> > Ensure that the VID header offset + VID header size does not exceed
> > the allocated area to avoid slab OOB.
> >
> > BUG: KASAN: slab-out-of-bounds in crc32_body lib/crc32.c:111 [inline]
> > BUG: KASAN: slab-out-of-bounds in crc32_le_generic lib/crc32.c:179 [inline]
> > BUG: KASAN: slab-out-of-bounds in crc32_le_base+0x58c/0x626 lib/crc32.c:197
> > Read of size 4 at addr ffff88802bb36f00 by task syz-executor136/1555
> >
> > CPU: 2 PID: 1555 Comm: syz-executor136 Tainted: G W
> > 6.0.0-1868 #1
> > Hardware name: Red Hat KVM, BIOS 1.13.0-2.module+el8.3.0+7860+a7792d29
> > 04/01/2014
> > Call Trace:
> > <TASK>
> > __dump_stack lib/dump_stack.c:88 [inline]
> > dump_stack_lvl+0x85/0xad lib/dump_stack.c:106
> > print_address_description mm/kasan/report.c:317 [inline]
> > print_report.cold.13+0xb6/0x6bb mm/kasan/report.c:433
> > kasan_report+0xa7/0x11b mm/kasan/report.c:495
> > crc32_body lib/crc32.c:111 [inline]
> > crc32_le_generic lib/crc32.c:179 [inline]
> > crc32_le_base+0x58c/0x626 lib/crc32.c:197
> > ubi_io_write_vid_hdr+0x1b7/0x472 drivers/mtd/ubi/io.c:1067
> > create_vtbl+0x4d5/0x9c4 drivers/mtd/ubi/vtbl.c:317
> > create_empty_lvol drivers/mtd/ubi/vtbl.c:500 [inline]
> > ubi_read_volume_table+0x67b/0x288a drivers/mtd/ubi/vtbl.c:812
> > ubi_attach+0xf34/0x1603 drivers/mtd/ubi/attach.c:1601
> > ubi_attach_mtd_dev+0x6f3/0x185e drivers/mtd/ubi/build.c:965
> > ctrl_cdev_ioctl+0x2db/0x347 drivers/mtd/ubi/cdev.c:1043
> > vfs_ioctl fs/ioctl.c:51 [inline]
> > __do_sys_ioctl fs/ioctl.c:870 [inline]
> > __se_sys_ioctl fs/ioctl.c:856 [inline]
> > __x64_sys_ioctl+0x193/0x213 fs/ioctl.c:856
> > do_syscall_x64 arch/x86/entry/common.c:50 [inline]
> > do_syscall_64+0x3e/0x86 arch/x86/entry/common.c:80
> > entry_SYSCALL_64_after_hwframe+0x63/0x0
> > RIP: 0033:0x7f96d5cf753d
> > Code:
> > RSP: 002b:00007fffd72206f8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
> > RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f96d5cf753d
> > RDX: 0000000020000080 RSI: 0000000040186f40 RDI: 0000000000000003
> > RBP: 0000000000400cd0 R08: 0000000000000000 R09: 0000000000000000
> > R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000400be0
> > R13: 00007fffd72207e0 R14: 0000000000000000 R15: 0000000000000000
> > </TASK>
> >
> > Allocated by task 1555:
> > kasan_save_stack+0x20/0x3d mm/kasan/common.c:38
> > kasan_set_track mm/kasan/common.c:45 [inline]
> > set_alloc_info mm/kasan/common.c:437 [inline]
> > ____kasan_kmalloc mm/kasan/common.c:516 [inline]
> > __kasan_kmalloc+0x88/0xa3 mm/kasan/common.c:525
> > kasan_kmalloc include/linux/kasan.h:234 [inline]
> > __kmalloc+0x138/0x257 mm/slub.c:4429
> > kmalloc include/linux/slab.h:605 [inline]
> > ubi_alloc_vid_buf drivers/mtd/ubi/ubi.h:1093 [inline]
> > create_vtbl+0xcc/0x9c4 drivers/mtd/ubi/vtbl.c:295
> > create_empty_lvol drivers/mtd/ubi/vtbl.c:500 [inline]
> > ubi_read_volume_table+0x67b/0x288a drivers/mtd/ubi/vtbl.c:812
> > ubi_attach+0xf34/0x1603 drivers/mtd/ubi/attach.c:1601
> > ubi_attach_mtd_dev+0x6f3/0x185e drivers/mtd/ubi/build.c:965
> > ctrl_cdev_ioctl+0x2db/0x347 drivers/mtd/ubi/cdev.c:1043
> > vfs_ioctl fs/ioctl.c:51 [inline]
> > __do_sys_ioctl fs/ioctl.c:870 [inline]
> > __se_sys_ioctl fs/ioctl.c:856 [inline]
> > __x64_sys_ioctl+0x193/0x213 fs/ioctl.c:856
> > do_syscall_x64 arch/x86/entry/common.c:50 [inline]
> > do_syscall_64+0x3e/0x86 arch/x86/entry/common.c:80
> > entry_SYSCALL_64_after_hwframe+0x63/0x0
> >
> > The buggy address belongs to the object at ffff88802bb36e00
> > which belongs to the cache kmalloc-256 of size 256
> > The buggy address is located 0 bytes to the right of
> > 256-byte region [ffff88802bb36e00, ffff88802bb36f00)
> >
> > The buggy address belongs to the physical page:
> > page:00000000ea4d1263 refcount:1 mapcount:0 mapping:0000000000000000
> > index:0x0 pfn:0x2bb36
> > head:00000000ea4d1263 order:1 compound_mapcount:0 compound_pincount:0
> > flags: 0xfffffc0010200(slab|head|node=0|zone=1|lastcpupid=0x1fffff)
> > raw: 000fffffc0010200 ffffea000066c300 dead000000000003 ffff888100042b40
> > raw: 0000000000000000 0000000000100010 00000001ffffffff 0000000000000000
> > page dumped because: kasan: bad access detected
> >
> > Memory state around the buggy address:
> > ffff88802bb36e00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> > ffff88802bb36e80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
> > > ffff88802bb36f00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> > ^
> > ffff88802bb36f80: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
> > ffff88802bb37000: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
> > ==================================================================
> >
> > Fixes: 801c135ce73d ("UBI: Unsorted Block Images")
> > Reported-by: syzkaller <[email protected]>
> > Signed-off-by: George Kennedy <[email protected]>
> > ---
> > drivers/mtd/ubi/build.c | 6 ++++++
> > 1 file changed, 6 insertions(+)
> >
> > diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c
> > index a32050fecabf..53aa4de6b963 100644
> > --- a/drivers/mtd/ubi/build.c
> > +++ b/drivers/mtd/ubi/build.c
> > @@ -663,6 +663,12 @@ static int io_init(struct ubi_device *ubi, int max_beb_per1024)
> > ubi->ec_hdr_alsize = ALIGN(UBI_EC_HDR_SIZE, ubi->hdrs_min_io_size);
> > ubi->vid_hdr_alsize = ALIGN(UBI_VID_HDR_SIZE, ubi->hdrs_min_io_size);
> > + if (ubi->vid_hdr_offset && ((ubi->vid_hdr_offset + UBI_VID_HDR_SIZE) >
> > + ubi->vid_hdr_alsize)) {
> > + ubi_err(ubi, "VID header offset %d too large.", ubi->vid_hdr_offset);
> > + return -EINVAL;
> > + }
> > +
>
> This patch is in mainline as 1b42b1a36fc946f0d7088425b90d491b4257ca3e,
> and backported to various stable releases.
>
> For me this breaks
>
> ubiattach -m 0 -O 2048
>
> I think the check
>
> ubi->vid_hdr_offset + UBI_VID_HDR_SIZE > ubi->vid_hdr_alsize
>
> is wrong. Without -O passed to ubiattach (and dynamic debug enabled) I
> get:
>
> [ 5294.936762] UBI DBG gen (pid 9619): sizeof(struct ubi_ainf_peb) 56
> [ 5294.936769] UBI DBG gen (pid 9619): sizeof(struct ubi_wl_entry) 32
> [ 5294.936774] UBI DBG gen (pid 9619): min_io_size 2048
> [ 5294.936779] UBI DBG gen (pid 9619): max_write_size 2048
> [ 5294.936783] UBI DBG gen (pid 9619): hdrs_min_io_size 512
> [ 5294.936787] UBI DBG gen (pid 9619): ec_hdr_alsize 512
> [ 5294.936791] UBI DBG gen (pid 9619): vid_hdr_alsize 512
> [ 5294.936796] UBI DBG gen (pid 9619): vid_hdr_offset 512
> [ 5294.936800] UBI DBG gen (pid 9619): vid_hdr_aloffset 512
> [ 5294.936804] UBI DBG gen (pid 9619): vid_hdr_shift 0
> [ 5294.936808] UBI DBG gen (pid 9619): leb_start 2048
> [ 5294.936812] UBI DBG gen (pid 9619): max_erroneous 409
>
> So the check would only pass for vid_hdr_offset <= 512 -
> UBI_VID_HDR_SIZE; note that even specifying the default value 512 (i.e.
>
> ubiattach -m 0 -O 512
>
> ) fails the check.
>
> A less strong check would be:
>
> diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c
> index 0904eb40c95f..69c28a862430 100644
> --- a/drivers/mtd/ubi/build.c
> +++ b/drivers/mtd/ubi/build.c
> @@ -666,8 +666,8 @@ static int io_init(struct ubi_device *ubi, int max_beb_per1024)
> ubi->ec_hdr_alsize = ALIGN(UBI_EC_HDR_SIZE, ubi->hdrs_min_io_size);
> ubi->vid_hdr_alsize = ALIGN(UBI_VID_HDR_SIZE, ubi->hdrs_min_io_size);
>
> - if (ubi->vid_hdr_offset && ((ubi->vid_hdr_offset + UBI_VID_HDR_SIZE) >
> - ubi->vid_hdr_alsize)) {
> + if (ubi->vid_hdr_offset &&
> + ubi->vid_hdr_offset + UBI_VID_HDR_SIZE > ubi->peb_size) {
> ubi_err(ubi, "VID header offset %d too large.", ubi->vid_hdr_offset);
> return -EINVAL;
> }
>
> But I'm unsure if this would be too lax?!

I just learned in irc that

1e020e1b96af ("ubi: Fix failure attaching when vid_hdr offset equals to (sub)page size")

is supposed to fix that.

Best regards
Uwe

--
Pengutronix e.K. | Uwe Kleine-K?nig |
Industrial Linux Solutions | https://www.pengutronix.de/ |


Attachments:
(No filename) (8.17 kB)
signature.asc (499.00 B)
Download all attachments

2023-04-17 16:08:25

by Richard Weinberger

[permalink] [raw]
Subject: Re: [Regression] Cannot overwrite VID header offset any more [Was: [PATCH] ubi: ensure that VID header offset + VID header size <= alloc, size]

Uwe,

----- Ursprüngliche Mail -----
> Von: "Uwe Kleine-König" <[email protected]>
> This patch is in mainline as 1b42b1a36fc946f0d7088425b90d491b4257ca3e,
> and backported to various stable releases.
>
> For me this breaks
>
> ubiattach -m 0 -O 2048
>
> I think the check
>
> ubi->vid_hdr_offset + UBI_VID_HDR_SIZE > ubi->vid_hdr_alsize
>
> is wrong. Without -O passed to ubiattach (and dynamic debug enabled) I
> get:
>
> [ 5294.936762] UBI DBG gen (pid 9619): sizeof(struct ubi_ainf_peb) 56
> [ 5294.936769] UBI DBG gen (pid 9619): sizeof(struct ubi_wl_entry) 32
> [ 5294.936774] UBI DBG gen (pid 9619): min_io_size 2048
> [ 5294.936779] UBI DBG gen (pid 9619): max_write_size 2048
> [ 5294.936783] UBI DBG gen (pid 9619): hdrs_min_io_size 512
> [ 5294.936787] UBI DBG gen (pid 9619): ec_hdr_alsize 512
> [ 5294.936791] UBI DBG gen (pid 9619): vid_hdr_alsize 512
> [ 5294.936796] UBI DBG gen (pid 9619): vid_hdr_offset 512
> [ 5294.936800] UBI DBG gen (pid 9619): vid_hdr_aloffset 512
> [ 5294.936804] UBI DBG gen (pid 9619): vid_hdr_shift 0
> [ 5294.936808] UBI DBG gen (pid 9619): leb_start 2048
> [ 5294.936812] UBI DBG gen (pid 9619): max_erroneous 409
>
> So the check would only pass for vid_hdr_offset <= 512 -
> UBI_VID_HDR_SIZE; note that even specifying the default value 512 (i.e.
>
> ubiattach -m 0 -O 512
>
> ) fails the check.
>
> A less strong check would be:
>
> diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c
> index 0904eb40c95f..69c28a862430 100644
> --- a/drivers/mtd/ubi/build.c
> +++ b/drivers/mtd/ubi/build.c
> @@ -666,8 +666,8 @@ static int io_init(struct ubi_device *ubi, int
> max_beb_per1024)
> ubi->ec_hdr_alsize = ALIGN(UBI_EC_HDR_SIZE, ubi->hdrs_min_io_size);
> ubi->vid_hdr_alsize = ALIGN(UBI_VID_HDR_SIZE, ubi->hdrs_min_io_size);
>
> - if (ubi->vid_hdr_offset && ((ubi->vid_hdr_offset + UBI_VID_HDR_SIZE) >
> - ubi->vid_hdr_alsize)) {
> + if (ubi->vid_hdr_offset &&
> + ubi->vid_hdr_offset + UBI_VID_HDR_SIZE > ubi->peb_size) {
> ubi_err(ubi, "VID header offset %d too large.", ubi->vid_hdr_offset);
> return -EINVAL;
> }
>
> But I'm unsure if this would be too lax?!

As written on IRC, 1e020e1b96af ("ubi: Fix failure attaching when vid_hdr offset equals to (sub)page size") is supposed to fix that
and on it's way into stable.

Thanks,
//richard