2022-12-08 12:42:15

by Wei Chen

[permalink] [raw]
Subject: divide error in mb_update_avg_fragment_size

Dear Linux Developers,

Recently, when using our tool to fuzz kernel, the following crash was triggered.

HEAD commit: 147307c69ba
git tree: linux-next
compiler: clang 12.0.0
console output:
https://drive.google.com/file/d/1tComytzKYQKzKfiS1PwrKBLQJwZn4usl/view?usp=share_link
kernel config: https://drive.google.com/file/d/1NAf4S43d9VOKD52xbrqw-PUP1Mbj8z-S/view?usp=share_link

Unfortunately, I didn't have a reproducer for this crash. When calling
ioctl with command FITRIM, the variable grp->bb_fragments is zero,
thus leading to divide error when executing grp->bb_free /
grp->bb_fragments. I'm wondering if a sanity check is missing when
handling malicious disk image.

IMPORTANT: if you fix the bug, please add the following tag to the commit:
Reported-by: Wei Chen <[email protected]>

divide error: 0000 [#1] PREEMPT SMP
CPU: 0 PID: 8874 Comm: syz-executor.0 Not tainted 6.1.0-rc5-next-20221118 #2
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS
rel-1.13.0-48-gd9c812dda519-prebuilt.qemu.org 04/01/2014
RIP: 0010:mb_update_avg_fragment_size+0x84/0x2e0 fs/ext4/mballoc.c:835
Code: 41 5f 5d c3 49 8d 7e 14 e8 a9 38 ca ff 41 8b 5e 14 31 ff 89 de
e8 ac d3 b6 ff 85 db 74 61 49 8d 7e 18 e8 8f 38 ca ff 89 d8 99 <41> f7
7e 18 89 44 24 14 bb ff ff ff ff 0f bd 5c 24 14 8d 73 ff 45
RSP: 0018:ffffc90001bf3a60 EFLAGS: 00010246
RAX: 0000000000000456 RBX: 0000000000000456 RCX: ffffffff8170e251
RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff88810204e018
RBP: ffff8881030de800 R08: 000188810204e01b R09: 0000000000000000
R10: 0001ffffffffffff R11: 000188810204e020 R12: 0000000000008000
R13: ffff88810315b000 R14: ffff88810204e000 R15: ffffc90001bf3cc0
FS: 00007f4a1db97700(0000) GS:ffff88813bc00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 000055ebcf207bd0 CR3: 0000000131681000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000600
Call Trace:
<TASK>
mb_mark_used+0xb5c/0xcc0 fs/ext4/mballoc.c:1958
ext4_trim_extent+0xd4/0x340 fs/ext4/mballoc.c:6352
ext4_try_to_trim_range+0x194/0x460 fs/ext4/mballoc.c:6382
ext4_trim_all_free+0x26e/0x3d0 fs/ext4/mballoc.c:6444
ext4_trim_fs+0x56b/0x6d0 fs/ext4/mballoc.c:6540
__ext4_ioctl+0x13b6/0x2990 fs/ext4/ioctl.c:1483
ext4_ioctl+0x1e/0x30 fs/ext4/ioctl.c:1606
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:870 [inline]
__se_sys_ioctl+0xcb/0x140 fs/ioctl.c:856
__x64_sys_ioctl+0x3f/0x50 fs/ioctl.c:856
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x2b/0x70 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
RIP: 0033:0x4697f9
Code: f7 d8 64 89 02 b8 ff ff ff ff c3 66 0f 1f 44 00 00 48 89 f8 48
89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d
01 f0 ff ff 73 01 c3 48 c7 c1 bc ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f4a1db96c48 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
RAX: ffffffffffffffda RBX: 000000000077bf80 RCX: 00000000004697f9
RDX: 0000000020000040 RSI: 00000000c0185879 RDI: 0000000000000004
RBP: 00000000004d29e9 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 000000000077bf80
R13: 0000000000000000 R14: 000000000077bf80 R15: 00007fffbabb2d30
</TASK>
Modules linked in:
---[ end trace 0000000000000000 ]---
RIP: 0010:mb_update_avg_fragment_size+0x84/0x2e0 fs/ext4/mballoc.c:835
Code: 41 5f 5d c3 49 8d 7e 14 e8 a9 38 ca ff 41 8b 5e 14 31 ff 89 de
e8 ac d3 b6 ff 85 db 74 61 49 8d 7e 18 e8 8f 38 ca ff 89 d8 99 <41> f7
7e 18 89 44 24 14 bb ff ff ff ff 0f bd 5c 24 14 8d 73 ff 45
RSP: 0018:ffffc90001bf3a60 EFLAGS: 00010246
RAX: 0000000000000456 RBX: 0000000000000456 RCX: ffffffff8170e251
RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff88810204e018
RBP: ffff8881030de800 R08: 000188810204e01b R09: 0000000000000000
R10: 0001ffffffffffff R11: 000188810204e020 R12: 0000000000008000
R13: ffff88810315b000 R14: ffff88810204e000 R15: ffffc90001bf3cc0
FS: 00007f4a1db97700(0000) GS:ffff88813bc00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 000055ebcf207bd0 CR3: 0000000131681000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000600
----------------
Code disassembly (best guess):
0: 41 5f pop %r15
2: 5d pop %rbp
3: c3 retq
4: 49 8d 7e 14 lea 0x14(%r14),%rdi
8: e8 a9 38 ca ff callq 0xffca38b6
d: 41 8b 5e 14 mov 0x14(%r14),%ebx
11: 31 ff xor %edi,%edi
13: 89 de mov %ebx,%esi
15: e8 ac d3 b6 ff callq 0xffb6d3c6
1a: 85 db test %ebx,%ebx
1c: 74 61 je 0x7f
1e: 49 8d 7e 18 lea 0x18(%r14),%rdi
22: e8 8f 38 ca ff callq 0xffca38b6
27: 89 d8 mov %ebx,%eax
29: 99 cltd
* 2a: 41 f7 7e 18 idivl 0x18(%r14) <-- trapping instruction
2e: 89 44 24 14 mov %eax,0x14(%rsp)
32: bb ff ff ff ff mov $0xffffffff,%ebx
37: 0f bd 5c 24 14 bsr 0x14(%rsp),%ebx
3c: 8d 73 ff lea -0x1(%rbx),%esi
3f: 45 rex.RB

Best,
Wei


2023-12-08 14:26:08

by lee bruce

[permalink] [raw]
Subject: Re: divide error in mb_update_avg_fragment_size

Hello I saw you can't reproduce this bug and I reproduce it with
repro.c and repro.txt
I test the repro.c in the lastest HEAD: 5e3f5b81de80c98338bcb47c233aebefee5a4801
kernel config: https://syzkaller.appspot.com/text?tag=KernelConfig&x=6ae1a4ee971a7305
and the bug also existed.

=* repro.c =*
// autogenerated by syzkaller (https://github.com/google/syzkaller)

#define _GNU_SOURCE

#include <dirent.h>
#include <endian.h>
#include <errno.h>
#include <fcntl.h>
#include <signal.h>
#include <stdarg.h>
#include <stdbool.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/prctl.h>
#include <sys/stat.h>
#include <sys/syscall.h>
#include <sys/types.h>
#include <sys/wait.h>
#include <time.h>
#include <unistd.h>

static unsigned long long procid;

static void sleep_ms(uint64_t ms) { usleep(ms * 1000); }

static uint64_t current_time_ms(void) {
struct timespec ts;
if (clock_gettime(CLOCK_MONOTONIC, &ts)) exit(1);
return (uint64_t)ts.tv_sec * 1000 + (uint64_t)ts.tv_nsec / 1000000;
}

static bool write_file(const char* file, const char* what, ...) {
char buf[1024];
va_list args;
va_start(args, what);
vsnprintf(buf, sizeof(buf), what, args);
va_end(args);
buf[sizeof(buf) - 1] = 0;
int len = strlen(buf);
int fd = open(file, O_WRONLY | O_CLOEXEC);
if (fd == -1) return false;
if (write(fd, buf, len) != len) {
int err = errno;
close(fd);
errno = err;
return false;
}
close(fd);
return true;
}

static void kill_and_wait(int pid, int* status) {
kill(-pid, SIGKILL);
kill(pid, SIGKILL);
for (int i = 0; i < 100; i++) {
if (waitpid(-1, status, WNOHANG | __WALL) == pid) return;
usleep(1000);
}
DIR* dir = opendir("/sys/fs/fuse/connections");
if (dir) {
for (;;) {
struct dirent* ent = readdir(dir);
if (!ent) break;
if (strcmp(ent->d_name, ".") == 0 || strcmp(ent->d_name, "..") == 0)
continue;
char abort[300];
snprintf(abort, sizeof(abort), "/sys/fs/fuse/connections/%s/abort",
ent->d_name);
int fd = open(abort, O_WRONLY);
if (fd == -1) {
continue;
}
if (write(fd, abort, 1) < 0) {
}
close(fd);
}
closedir(dir);
} else {
}
while (waitpid(-1, status, __WALL) != pid) {
}
}

static void setup_test() {
prctl(PR_SET_PDEATHSIG, SIGKILL, 0, 0, 0);
setpgrp();
write_file("/proc/self/oom_score_adj", "1000");
}

static void execute_one(void);

#define WAIT_FLAGS __WALL

static void loop(void) {
int iter = 0;
for (;; iter++) {
int pid = fork();
if (pid < 0) exit(1);
if (pid == 0) {
setup_test();
execute_one();
exit(0);
}
int status = 0;
uint64_t start = current_time_ms();
for (;;) {
if (waitpid(-1, &status, WNOHANG | WAIT_FLAGS) == pid) break;
sleep_ms(1);
if (current_time_ms() - start < 5000) continue;
kill_and_wait(pid, &status);
break;
}
}
}

uint64_t r[5] = {0xffffffffffffffff, 0xffffffffffffffff, 0xffffffffffffffff,
0xffffffffffffffff, 0xffffffffffffffff};

void execute_one(void) {
intptr_t res = 0;
memcpy((void*)0x20000280, "cgroup.controllers\000", 19);
res = syscall(__NR_openat, /*fd=*/0xffffff9c, /*file=*/0x20000280ul,
/*flags=*/0x275aul, /*mode=*/0ul);
if (res != -1) r[0] = res;
memcpy((void*)0x20000180, "cgroup.controllers\000", 19);
res = syscall(__NR_openat, /*fd=*/0xffffff9c, /*file=*/0x20000180ul,
/*flags=*/0x275aul, /*mode=*/0ul);
if (res != -1) r[1] = res;
res = syscall(__NR_dup3, /*oldfd=*/r[1], /*newfd=*/r[0], /*flags=*/0ul);
if (res != -1) r[2] = res;
*(uint32_t*)0x20000140 = 0x20;
*(uint32_t*)0x20000144 = 0x8c8c;
*(uint32_t*)0x20000148 = 0;
*(uint32_t*)0x2000014c = 0;
*(uint32_t*)0x20000150 = 0;
memset((void*)0x20000154, 0, 8);
syscall(__NR_ioctl, /*fd=*/r[0], /*cmd=*/0x401c5820, /*arg=*/0x20000140ul);
sprintf((char*)0x20000040, "0x%016llx", (long long)0);
syscall(__NR_write, /*fd=*/r[0], /*buf=*/0x20000040ul, /*len=*/0xfea0ul);
memcpy((void*)0x200001c0, "cpuset.effective_cpus\000", 22);
res = syscall(__NR_openat, /*fd=*/0xffffff9c, /*file=*/0x200001c0ul,
/*flags=*/0x275aul, /*mode=*/0ul);
if (res != -1) r[3] = res;
sprintf((char*)0x20000380, "0x%016llx", (long long)0);
syscall(__NR_write, /*fd=*/r[3], /*buf=*/0x20000380ul, /*len=*/0x101bful);
syscall(__NR_ioctl, /*fd=*/r[3], /*cmd=*/0x660c, 0);
*(uint32_t*)0x200000c0 = 0;
*(uint32_t*)0x200000c4 = r[3];
*(uint64_t*)0x200000c8 = 7;
*(uint64_t*)0x200000d0 = 0;
*(uint64_t*)0x200000d8 = 0;
*(uint64_t*)0x200000e0 = 0;
syscall(__NR_ioctl, /*fd=*/r[2], /*cmd=*/0xc028660f, /*arg=*/0x200000c0ul);
syscall(__NR_writev, /*fd=*/-1, /*vec=*/0ul, /*vlen=*/0ul);
syscall(__NR_ioctl, /*fd=*/-1, /*cmd=*/0x40045569, /*arg=*/9ul);
syscall(__NR_openat, /*fd=*/0xffffff9c, /*file=*/0ul, /*flags=*/0x275aul,
/*mode=*/0ul);
memcpy((void*)0x20000180, "cgroup.controllers\000", 19);
res = syscall(__NR_openat, /*fd=*/0xffffff9c, /*file=*/0x20000180ul,
/*flags=*/0x275aul, /*mode=*/0ul);
if (res != -1) r[4] = res;
*(uint64_t*)0x200000c0 = 4;
*(uint64_t*)0x200000c8 = 0x16000000000000;
*(uint64_t*)0x200000d0 = 0x20;
syscall(__NR_ioctl, /*fd=*/r[4], /*cmd=*/0xc0185879, /*arg=*/0x200000c0ul);
}
int main(void) {
syscall(__NR_mmap, /*addr=*/0x1ffff000ul, /*len=*/0x1000ul, /*prot=*/0ul,
/*flags=*/0x32ul, /*fd=*/-1, /*offset=*/0ul);
syscall(__NR_mmap, /*addr=*/0x20000000ul, /*len=*/0x1000000ul, /*prot=*/7ul,
/*flags=*/0x32ul, /*fd=*/-1, /*offset=*/0ul);
syscall(__NR_mmap, /*addr=*/0x21000000ul, /*len=*/0x1000ul, /*prot=*/0ul,
/*flags=*/0x32ul, /*fd=*/-1, /*offset=*/0ul);
for (procid = 0; procid < 4; procid++) {
if (fork() == 0) {
loop();
}
}
sleep(1000000);
return 0;
}

=* repro.txt =*
r0 = openat$cgroup_ro(0xffffffffffffff9c,
&(0x7f0000000280)='cgroup.controllers\x00', 0x275a, 0x0)
r1 = openat$cgroup_ro(0xffffffffffffff9c,
&(0x7f0000000180)='cgroup.controllers\x00', 0x275a, 0x0)
r2 = dup3(r1, r0, 0x0)
ioctl$FS_IOC_FSSETXATTR(r0, 0x401c5820, &(0x7f0000000140)={0x20, 0x8c8c})
write$cgroup_int(r0, &(0x7f0000000040), 0xfea0)
r3 = openat$cgroup_ro(0xffffffffffffff9c,
&(0x7f00000001c0)='cpuset.effective_cpus\x00', 0x275a, 0x0)
write$cgroup_int(r3, &(0x7f0000000380), 0x101bf)
ioctl$EXT4_IOC_ALLOC_DA_BLKS(r3, 0x660c)
ioctl$EXT4_IOC_MOVE_EXT(r2, 0xc028660f, &(0x7f00000000c0)={0x0, r3, 0x7})
writev(0xffffffffffffffff, 0x0, 0x0)
ioctl$UI_SET_LEDBIT(0xffffffffffffffff, 0x40045569, 0x9)
openat$cgroup_ro(0xffffffffffffff9c, 0x0, 0x275a, 0x0)
r4 = openat$cgroup_ro(0xffffffffffffff9c,
&(0x7f0000000180)='cgroup.controllers\x00', 0x275a, 0x0)
ioctl$FITRIM(r4, 0xc0185879, &(0x7f00000000c0)={0x4, 0x16000000000000, 0x20})

and also https://gist.github.com/xrivendell7/bad992c2b716ed14310efa2c6f878b7c