Re: [PATCH V2] block: brd: associate with queue until adding disk
On 11/1/18 6:50 PM, Ming Lei wrote: > brd_free() may be called in failure path on one brd instance which > disk isn't added yet, so release handler of gendisk may free the > associated request_queue early and causes the following use-after-free[1]. > > This patch fixes this issue by associating gendisk with request_queue > just before adding disk. > > [1] KASAN: use-after-free Read in del_timer_syncNon-volatile memory driver > v1.3 > Linux agpgart interface v0.103 > [drm] Initialized vgem 1.0.0 20120112 for virtual device on minor 0 > usbcore: registered new interface driver udl > == > BUG: KASAN: use-after-free in __lock_acquire+0x36d9/0x4c20 > kernel/locking/lockdep.c:3218 > Read of size 8 at addr 8801d1b6b540 by task swapper/0/1 > > CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.19.0+ #88 > Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS > Google 01/01/2011 > Call Trace: > __dump_stack lib/dump_stack.c:77 [inline] > dump_stack+0x244/0x39d lib/dump_stack.c:113 > print_address_description.cold.7+0x9/0x1ff mm/kasan/report.c:256 > kasan_report_error mm/kasan/report.c:354 [inline] > kasan_report.cold.8+0x242/0x309 mm/kasan/report.c:412 > __asan_report_load8_noabort+0x14/0x20 mm/kasan/report.c:433 > __lock_acquire+0x36d9/0x4c20 kernel/locking/lockdep.c:3218 > lock_acquire+0x1ed/0x520 kernel/locking/lockdep.c:3844 > del_timer_sync+0xb7/0x270 kernel/time/timer.c:1283 > blk_cleanup_queue+0x413/0x710 block/blk-core.c:809 > brd_free+0x5d/0x71 drivers/block/brd.c:422 > brd_init+0x2eb/0x393 drivers/block/brd.c:518 > do_one_initcall+0x145/0x957 init/main.c:890 > do_initcall_level init/main.c:958 [inline] > do_initcalls init/main.c:966 [inline] > do_basic_setup init/main.c:984 [inline] > kernel_init_freeable+0x5c6/0x6b9 init/main.c:1148 > kernel_init+0x11/0x1ae init/main.c:1068 > ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:350 > > Reported-by: syzbot+3701447012fe951da...@syzkaller.appspotmail.com > Signed-off-by: Ming Lei Replaced the other patch with this one. -- Jens Axboe
[PATCH V2] block: brd: associate with queue until adding disk
brd_free() may be called in failure path on one brd instance which
disk isn't added yet, so release handler of gendisk may free the
associated request_queue early and causes the following use-after-free[1].
This patch fixes this issue by associating gendisk with request_queue
just before adding disk.
[1] KASAN: use-after-free Read in del_timer_syncNon-volatile memory driver v1.3
Linux agpgart interface v0.103
[drm] Initialized vgem 1.0.0 20120112 for virtual device on minor 0
usbcore: registered new interface driver udl
==
BUG: KASAN: use-after-free in __lock_acquire+0x36d9/0x4c20
kernel/locking/lockdep.c:3218
Read of size 8 at addr 8801d1b6b540 by task swapper/0/1
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.19.0+ #88
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS
Google 01/01/2011
Call Trace:
__dump_stack lib/dump_stack.c:77 [inline]
dump_stack+0x244/0x39d lib/dump_stack.c:113
print_address_description.cold.7+0x9/0x1ff mm/kasan/report.c:256
kasan_report_error mm/kasan/report.c:354 [inline]
kasan_report.cold.8+0x242/0x309 mm/kasan/report.c:412
__asan_report_load8_noabort+0x14/0x20 mm/kasan/report.c:433
__lock_acquire+0x36d9/0x4c20 kernel/locking/lockdep.c:3218
lock_acquire+0x1ed/0x520 kernel/locking/lockdep.c:3844
del_timer_sync+0xb7/0x270 kernel/time/timer.c:1283
blk_cleanup_queue+0x413/0x710 block/blk-core.c:809
brd_free+0x5d/0x71 drivers/block/brd.c:422
brd_init+0x2eb/0x393 drivers/block/brd.c:518
do_one_initcall+0x145/0x957 init/main.c:890
do_initcall_level init/main.c:958 [inline]
do_initcalls init/main.c:966 [inline]
do_basic_setup init/main.c:984 [inline]
kernel_init_freeable+0x5c6/0x6b9 init/main.c:1148
kernel_init+0x11/0x1ae init/main.c:1068
ret_from_fork+0x3a/0x50 arch/x86/entry/entry_64.S:350
Reported-by: syzbot+3701447012fe951da...@syzkaller.appspotmail.com
Signed-off-by: Ming Lei
---
V2:
- don't reference queue via disk->queue in brd_alloc()
drivers/block/brd.c | 16 +++-
1 file changed, 11 insertions(+), 5 deletions(-)
diff --git a/drivers/block/brd.c b/drivers/block/brd.c
index df8103dd40ac..c18586fccb6f 100644
--- a/drivers/block/brd.c
+++ b/drivers/block/brd.c
@@ -396,15 +396,14 @@ static struct brd_device *brd_alloc(int i)
disk->first_minor = i * max_part;
disk->fops = _fops;
disk->private_data = brd;
- disk->queue = brd->brd_queue;
disk->flags = GENHD_FL_EXT_DEVT;
sprintf(disk->disk_name, "ram%d", i);
set_capacity(disk, rd_size * 2);
- disk->queue->backing_dev_info->capabilities |= BDI_CAP_SYNCHRONOUS_IO;
+ brd->brd_queue->backing_dev_info->capabilities |=
BDI_CAP_SYNCHRONOUS_IO;
/* Tell the block layer that this is not a rotational device */
- blk_queue_flag_set(QUEUE_FLAG_NONROT, disk->queue);
- blk_queue_flag_clear(QUEUE_FLAG_ADD_RANDOM, disk->queue);
+ blk_queue_flag_set(QUEUE_FLAG_NONROT, brd->brd_queue);
+ blk_queue_flag_clear(QUEUE_FLAG_ADD_RANDOM, brd->brd_queue);
return brd;
@@ -436,6 +435,7 @@ static struct brd_device *brd_init_one(int i, bool *new)
brd = brd_alloc(i);
if (brd) {
+ brd->brd_disk->queue = brd->brd_queue;
add_disk(brd->brd_disk);
list_add_tail(>brd_list, _devices);
}
@@ -503,8 +503,14 @@ static int __init brd_init(void)
/* point of no return */
- list_for_each_entry(brd, _devices, brd_list)
+ list_for_each_entry(brd, _devices, brd_list) {
+ /*
+* associate with queue just before adding disk for
+* avoiding to mess up failure path
+*/
+ brd->brd_disk->queue = brd->brd_queue;
add_disk(brd->brd_disk);
+ }
blk_register_region(MKDEV(RAMDISK_MAJOR, 0), 1UL << MINORBITS,
THIS_MODULE, brd_probe, NULL, NULL);
--
2.9.5
loopback block device will not free
Jens, I am running the latest 4.19 kernel. When I mount a disk image using option 'loop' in /etc/fstab, it mounts just fine. When I un-mount the loopback device, lsof verifies that nothing else is using it. However, losetup -d will not free the device. Also, the reference counts in lsmod show 2 references for the one loopback device. Where should I start debugging? /usr/src/linux.stable # losetup NAME SIZELIMIT OFFSET AUTOCLEAR RO BACK-FILE DIO LOG-SEC /dev/loop0 0 0 1 0 /mnt/images0/i486.img 0 512 /usr/src/linux.stable # lsmod | grep loop loop 24576 2 - Matthew Whitehead
Re: [PATCH blktests] fix discontiguous-io compile error on 32 bit systems
On Thu, 2018-11-01 at 14:35 +0800, Yufen Yu wrote:
> When make discontiguous-io.cpp with -m32, g++ compiler reports
> error for std::min(long unsigned int, size_t) has diffent
> arguments type.
>
> fixes: fd21728886e7 ("Add the discontiguous-io test program")
> Signed-off-by: Yufen Yu
> ---
> src/discontiguous-io.cpp | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/src/discontiguous-io.cpp b/src/discontiguous-io.cpp
> index 5e0ee0f..855aba9 100644
> --- a/src/discontiguous-io.cpp
> +++ b/src/discontiguous-io.cpp
> @@ -291,7 +291,7 @@ int main(int argc, char **argv)
> unsigned char *p = &*buf.begin();
> for (int i = 0; i < len / 4; i++)
> iov.append(p + 4 + i * 8,
> -std::min(4ul, len - i * 4));
> +std::min((size_t)4, len - i * 4));
> } else {
> iov.append(&*buf.begin(), buf.size());
> }
Are you reading the messages posted on linux-block? An alternative that I like
better has been discussed in this e-mail thread:
https://www.spinics.net/lists/linux-block/msg32181.html
Bart.
[PATCH blktests] fix discontiguous-io compile error on 32 bit systems
When make discontiguous-io.cpp with -m32, g++ compiler reports
error for std::min(long unsigned int, size_t) has diffent
arguments type.
fixes: fd21728886e7 ("Add the discontiguous-io test program")
Signed-off-by: Yufen Yu
---
src/discontiguous-io.cpp | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/discontiguous-io.cpp b/src/discontiguous-io.cpp
index 5e0ee0f..855aba9 100644
--- a/src/discontiguous-io.cpp
+++ b/src/discontiguous-io.cpp
@@ -291,7 +291,7 @@ int main(int argc, char **argv)
unsigned char *p = &*buf.begin();
for (int i = 0; i < len / 4; i++)
iov.append(p + 4 + i * 8,
- std::min(4ul, len - i * 4));
+ std::min((size_t)4, len - i * 4));
} else {
iov.append(&*buf.begin(), buf.size());
}
--
2.7.4
