Re: 5.?? regression: strace testsuite OOpses kernel on ia64

From: Yonghong Song
Date: Sat Apr 10 2021 - 13:37:02 EST

Next message: Shakeel Butt: "Re: [PATCH net-next v3 2/5] mm: add a signature in struct page"
Previous message: Ilya Lipnitskiy: "[PATCH] staging: mt7621-pci: stop using of_pci_range_to_resource"
In reply to: Sergei Trofimovich: "Re: 5.?? regression: strace testsuite OOpses kernel on ia64"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On 4/9/21 2:20 PM, Sergei Trofimovich wrote:

On Tue, 23 Feb 2021 18:53:21 +0000
Sergei Trofimovich <slyich@xxxxxxxxx> wrote:

The crash seems to be related to sock_filter-v test from strace:
https://github.com/strace/strace/blob/master/tests/seccomp-filter-v.c

Here is an OOps:

[ 818.089904] BUG: Bad page map in process sock_filter-v pte:00000001 pmd:118580001
[ 818.089904] page:00000000e6a429c8 refcount:1 mapcount:-1 mapping:0000000000000000 index:0x0 pfn:0x0
[ 818.089904] flags: 0x1000(reserved)
[ 818.089904] raw: 0000000000001000 a000400000000008 a000400000000008 0000000000000000
[ 818.089904] raw: 0000000000000000 0000000000000000 00000001fffffffe
[ 818.089904] page dumped because: bad pte
[ 818.089904] addr:0000000000000000 vm_flags:04044011 anon_vma:0000000000000000 mapping:0000000000000000 index:0
[ 818.095483] file:(null) fault:0x0 mmap:0x0 readpage:0x0
[ 818.095483] CPU: 0 PID: 5990 Comm: sock_filter-v Not tainted 5.11.0-00003-gbfa5a4929c90 #57
[ 818.095483] Hardware name: hp server rx3600 , BIOS 04.03 04/08/2008
[ 818.095483]
[ 818.095483] Call Trace:
[ 818.095483] [<a000000100014a70>] show_stack+0x90/0xc0
[ 818.095483] sp=e000000118707bb0 bsp=e0000001187013c0
[ 818.095483] [<a00000010100c2b0>] dump_stack+0x120/0x160
[ 818.095483] sp=e000000118707d80 bsp=e000000118701348
[ 818.095483] [<a0000001002e7b40>] print_bad_pte+0x300/0x3a0
[ 818.095483] sp=e000000118707d80 bsp=e0000001187012e0
[ 818.099483] [<a0000001002ec190>] unmap_page_range+0xa90/0x11a0
[ 818.099483] sp=e000000118707d80 bsp=e000000118701140
[ 818.099483] [<a0000001002ecba0>] unmap_vmas+0xc0/0x100
[ 818.099483] sp=e000000118707da0 bsp=e000000118701108
[ 818.099483] [<a000000100302db0>] exit_mmap+0x150/0x320
[ 818.099483] sp=e000000118707da0 bsp=e0000001187010d8
[ 818.099483] [<a00000010005dba0>] mmput+0x60/0x200
[ 818.099483] sp=e000000118707e20 bsp=e0000001187010b0
[ 818.103482] [<a00000010006e570>] do_exit+0x6f0/0x18a0
[ 818.103482] sp=e000000118707e20 bsp=e000000118701038
[ 818.103482] [<a000000100072070>] do_group_exit+0x90/0x2a0
[ 818.103482] sp=e000000118707e30 bsp=e000000118700ff0
[ 818.103482] [<a0000001000722a0>] sys_exit_group+0x20/0x40
[ 818.103482] sp=e000000118707e30 bsp=e000000118700f98
[ 818.107482] [<a00000010000c5f0>] ia64_trace_syscall+0xf0/0x130
[ 818.107482] sp=e000000118707e30 bsp=e000000118700f98
[ 818.107482] [<a000000000040720>] ia64_ivt+0xffffffff00040720/0x400
[ 818.107482] sp=e000000118708000 bsp=e000000118700f98
[ 818.115482] Disabling lock debugging due to kernel taint
[ 818.115482] BUG: Bad rss-counter state mm:000000002eec6412 type:MM_FILEPAGES val:-1
[ 818.132256] Unable to handle kernel NULL pointer dereference (address 0000000000000068)
[ 818.133904] sock_filter-v-X[5999]: Oops 11012296146944 [1]
[ 818.133904] Modules linked in: acpi_ipmi ipmi_si usb_storage e1000 ipmi_devintf ipmi_msghandler rtc_efi
[ 818.133904]
[ 818.133904] CPU: 0 PID: 5999 Comm: sock_filter-v-X Tainted: G B 5.11.0-00003-gbfa5a4929c90 #57
[ 818.133904] Hardware name: hp server rx3600 , BIOS 04.03 04/08/2008
[ 818.133904] psr : 0000121008026010 ifs : 8000000000000288 ip : [<a0000001001eaa61>] Tainted: G B (5.11.0-00003-gbfa5a4929c90)
[ 818.133904] ip is at bpf_prog_free+0x21/0xe0
[ 818.133904] unat: 0000000000000000 pfs : 0000000000000307 rsc : 0000000000000003
[ 818.133904] rnat: 0000000000000000 bsps: 0000000000000000 pr : 00106a5a51665965
[ 818.133904] ldrs: 0000000000000000 ccv : 0000000012088904 fpsr: 0009804c8a70033f
[ 818.133904] csd : 0000000000000000 ssd : 0000000000000000
[ 818.133904] b0 : a000000100d54080 b6 : a000000100d53fe0 b7 : a00000010000cef0
[ 818.133904] f6 : 0ffefb0c50daa1b67f89a f7 : 0ffed8b3e4fdb08000000
[ 818.133904] f8 : 10017fbd1bc0000000000 f9 : 1000eb95f000000000000
[ 818.133904] f10 : 10008ade20716a6c83cc1 f11 : 1003e00000000000002b7
[ 818.133904] r1 : a00000010176b300 r2 : a000000200008004 r3 : 0000000000000000
[ 818.133904] r8 : 0000000000000008 r9 : e00000011873f800 r10 : e000000102c18600
[ 818.133904] r11 : e000000102c19600 r12 : e00000011873f7f0 r13 : e000000118738000
[ 818.133904] r14 : 0000000000000068 r15 : a000000200008028 r16 : e000000005606a70
[ 818.133904] r17 : e000000102c18600 r18 : e000000104370748 r19 : e000000102c18600
[ 818.133904] r20 : e000000102c18600 r21 : e000000005606a78 r22 : a00000010156bd28
[ 818.133904] r23 : a00000010147fdf4 r24 : 0000000000004000 r25 : e000000104370750
[ 818.133904] r26 : a0000001012f7088 r27 : a000000100d53fe0 r28 : 0000000000000001
[ 818.133904] r29 : e00000011873f800 r30 : e00000011873f810 r31 : e00000011873f808
[ 818.133904]
[ 818.133904] Call Trace:
[ 818.133904] [<a000000100014a70>] show_stack+0x90/0xc0
[ 818.133904] sp=e00000011873f420 bsp=e0000001187396d0
[ 818.133904] [<a000000100015170>] show_regs+0x6d0/0xa40
[ 818.133904] sp=e00000011873f5f0 bsp=e000000118739660
[ 818.133904] [<a000000100026e90>] die+0x1b0/0x4a0
[ 818.133904] sp=e00000011873f610 bsp=e000000118739620
[ 818.133904] [<a000000100059220>] ia64_do_page_fault+0x820/0xb60
[ 818.133904] sp=e00000011873f610 bsp=e000000118739580
[ 818.133904] [<a00000010000c8e0>] ia64_leave_kernel+0x0/0x270
[ 818.133904] sp=e00000011873f620 bsp=e000000118739580
[ 818.133904] [<a0000001001eaa60>] bpf_prog_free+0x20/0xe0
[ 818.133904] sp=e00000011873f7f0 bsp=e000000118739540
[ 818.133904] [<a000000100d54080>] sk_filter_release_rcu+0xa0/0x120
[ 818.133904] sp=e00000011873f7f0 bsp=e000000118739510
[ 818.133904] [<a00000010016a7f0>] rcu_core+0x530/0xf20
[ 818.133904] sp=e00000011873f7f0 bsp=e0000001187394a8
[ 818.133904] [<a00000010016b200>] rcu_core_si+0x20/0x40
[ 818.133904] sp=e00000011873f810 bsp=e000000118739490
[ 818.133904] [<a0000001010279f0>] __do_softirq+0x230/0x640
[ 818.133904] sp=e00000011873f810 bsp=e0000001187393a0
[ 818.133904] [<a000000100074770>] irq_exit+0x170/0x200
[ 818.133904] sp=e00000011873f810 bsp=e000000118739388
[ 818.133904] [<a000000100013310>] ia64_handle_irq+0x1b0/0x360
[ 818.133904] sp=e00000011873f810 bsp=e000000118739308
[ 818.133904] [<a00000010000c8e0>] ia64_leave_kernel+0x0/0x270
[ 818.133904] sp=e00000011873f820 bsp=e000000118739308
[ 818.133904] [<a000000101027700>] flush_icache_range+0x80/0xa0
[ 818.133904] sp=e00000011873f9f0 bsp=e0000001187392f8
[ 818.133904] [<a0000001002f7ae0>] __access_remote_vm+0x1e0/0x320
[ 818.133904] sp=e00000011873f9f0 bsp=e000000118739258
[ 818.133904] [<a0000001002f7c80>] access_process_vm+0x60/0xa0
[ 818.133904] sp=e00000011873fa00 bsp=e000000118739210
[ 818.133904] [<a000000100018210>] ia64_sync_user_rbs+0x70/0xe0
[ 818.133904] sp=e00000011873fa00 bsp=e0000001187391d0
[ 818.133904] [<a000000100018b00>] do_sync_rbs+0xc0/0x100
[ 818.133904] sp=e00000011873fa10 bsp=e000000118739198
[ 818.133904] [<a00000010000cf30>] unw_init_running+0x70/0xa0
[ 818.133904] sp=e00000011873fa10 bsp=e000000118739170
[ 818.133904] [<a00000010001ae90>] ia64_ptrace_stop+0x130/0x160
[ 818.133904] sp=e00000011873fdf0 bsp=e000000118739158
[ 818.133904] [<a00000010008af60>] ptrace_stop+0xc0/0x880
[ 818.133904] sp=e00000011873fdf0 bsp=e000000118739118
[ 818.133904] [<a00000010008b820>] ptrace_do_notify+0x100/0x120
[ 818.133904] sp=e00000011873fdf0 bsp=e0000001187390e8
[ 818.133904] [<a00000010008b8d0>] ptrace_notify+0x90/0x260
[ 818.133904] sp=e00000011873fe30 bsp=e0000001187390c8
[ 818.133904] [<a00000010001e3d0>] syscall_trace_enter+0xf0/0x2c0
[ 818.133904] sp=e00000011873fe30 bsp=e000000118739070
[ 818.133904] [<a00000010000c540>] ia64_trace_syscall+0x40/0x130
[ 818.133904] sp=e00000011873fe30 bsp=e000000118739020
[ 818.186114] Kernel panic - not syncing: Aiee, killing interrupt handler!
[ 818.186114] ---[ end Kernel panic - not syncing: Aiee, killing interrupt handler! ]---

I'm not sure how to interpret it. It looks like 'bpf_prog_free'
frees the memory that is not there anymore, but previous crash
hints at already broken page tables. Maybe VM is already corrupted
by previous strace tests?

I wonder if I can enable a bit more kernel VM debugging to catch the corruption earlier.

I threw a bunch of CONFIG_DEBUG_* options at the kernel.
None of them exposed any extra details or changed initial crash report.

I found out a few workarounds:

Workaround 1: switching vmalloc() to kmalloc() in kernel/bpf/core.c makes
the crash disappear (largeish patch: https://bugs.gentoo.org/769614#c26 ).
That might hint at the bug in ia64's virtual memory implementation
(like missing tlb flush?) but nothing else seems to exhibit any problems.

Workaround 2: switching sk_filter_release() from RCU free to inline free:

--- a/net/core/filter.c
+++ b/net/core/filter.c
static void sk_filter_release(struct sk_filter *fp)
{
if (refcount_dec_and_test(&fp->refcnt))
- call_rcu(&fp->rcu, sk_filter_release_rcu);
+ //call_rcu(&fp->rcu, sk_filter_release_rcu);
+ __sk_filter_release(fp);
}

+netdev@, +bpf@ for possible debugging help.

Now I can easily trigger the failure with the following test program:

$ cat bug.c
#include <unistd.h>
#include <netinet/in.h>
#include <sys/socket.h>
#include <linux/filter.h>

int
main(void)
{
struct sock_filter bpf_filter[] = {
BPF_STMT(BPF_RET|BPF_K, 0)
};
struct sock_fprog prog = {
.len = 1,
.filter = bpf_filter,
};

int fd = socket(AF_INET, SOCK_DGRAM, 0);
setsockopt(fd, SOL_SOCKET, SO_ATTACH_FILTER, &prog, sizeof(prog));

return 0;
}

$ gcc bug.c -o bug; while ./bug; do echo again; done

After 8 iterations machine crashes.

If I interpret backtrace correctly:

[ 818.133904] [<a0000001001eaa60>] bpf_prog_free+0x20/0xe0
[ 818.133904] sp=e00000011873f7f0 bsp=e000000118739540

; void bpf_prog_free(struct bpf_prog *fp):
; fp is in r32
Dump of assembler code for function bpf_prog_free:
0xa00000010023bea0 <+0>: [MMI] alloc r35=ar.pfs,8,5,0
0xa00000010023bea1 <+1>: adds r32=56,r32 ; r32 = fp + 56 (&fp->aux)
0xa00000010023bea2 <+2>: mov r34=b0;;
0xa00000010023beb0 <+16>: [MMI] ld8 r33=[r32];; ; r33 = fp->aux
0xa00000010023beb1 <+17>: adds r14=112,r33 ; r14 = &(fp->aux->dst_prog)
0xa00000010023beb2 <+18>: nop.i 0x0;;
0xa00000010023bec0 <+32>: [MMI] nop.m 0x0
0xa00000010023bec1 <+33>: ld8 r37=[r14] ; r37 = fp->aux->dst_prog
0xa00000010023bec2 <+34>: nop.i 0x0;;

kernel crashes on aux NULL-ish dereference of fp->aux->dst_prog at:

[ 818.133904] r14 : 0000000000000068 r15 : a000000200008028 r16 : e000000005606a70

/* Free internal BPF program */
void bpf_prog_free(struct bpf_prog *fp)
{
struct bpf_prog_aux *aux = fp->aux;

if (aux->dst_prog)
bpf_prog_put(aux->dst_prog);
INIT_WORK(&aux->work, bpf_prog_free_deferred);
schedule_work(&aux->work);
}

Could it be that fp->aux is sometimes free'd earlier than it should?
It's a non-preemptible kernel.

I cannot reproduce the bug on x86_64. Maybe checking the jit difference
between x86 and ia64, e.g., regarding to freeing memory/prog?

Next message: Shakeel Butt: "Re: [PATCH net-next v3 2/5] mm: add a signature in struct page"
Previous message: Ilya Lipnitskiy: "[PATCH] staging: mt7621-pci: stop using of_pci_range_to_resource"
In reply to: Sergei Trofimovich: "Re: 5.?? regression: strace testsuite OOpses kernel on ia64"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]