Re: [Question] panic when write file cpuset.cpus

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



+Cc linux-kernel@xxxxxxxxxxxxxxx

On 2019/6/6 11:04, Chen Zhou wrote:
> Hi all,
> 
> I hit the following issue in linux 4.4 which is hard to reproduce.
> 
> [20190527221106][bsp_pci_device_get_bar]--- pBasePhyAddr :3a008000000, len:4000000  ---
> [20190527221106]Unable to handle kernel paging request at virtual address 100000010
> [20190527221107]pgd = ffffffd3c696b000
> [20190527221107][100000010] *pgd=0000000000000000, *pud=0000000000000000
> [20190527221107]Internal error: Oops: 96000005 [#1] PREEMPT SMP
> [20190527221107]Modules linked in: linux_user_bde(O) linux_kernel_bde(O) cmac(O) nse(O) pp(O) tm(O) lfe(O) tipc(O) pcie_aer_hisi(O) brd_drv_lpu(O) hi161x_glf(O) hi161x_glc(O) chip_sdk_adpt(O) bonding(O) memenv(O) iof_sal(O) iof_dlog(O) iof_devent(O) iof_iomm(O) drv_bsp_pic(O) bsp_common(O) pramdisk(O) bsp_proc(O) kdc_uio_log(O) vrp_env_log_area(O) drv_bsp_fmea(O) Drv_LastWds_K(O) Drv_CpuDfxInfo_K(O) Drv_CpuDfx_K(O) v8_dfx_cpu(O) Drv_Dfx_K(O) Drv_CpuRegInject_K(O) Drv_ResetCause_K(O) Drv_Kbox_K(O) env_core(O) iof_data(O) Drv_L2flush_K(O) arm64_cache_dfx(O) mmapdev(O) drv_extern_int(O) irq_monitor(O) drv_bsp_avs(O) Drv_Pmbus_K(O) Drv_Smc_K(O) bsp_sal(O) Drv_Ipsec_K(O) Drv_Tsensor_K(O) pci_hisi(O) serdes(O) Drv_CheckBoot_K(O) Drv_Djtag_K(O) addr_win(O) iof_cbb(O) Drv_I2c_K(O) hns_uio_enet(O) hns_enet_drv(O) hns_dsaf(O) hnae(O) hns_mdio(O) mdio(O) Drv_FlowCtrl_K(O) Drv_Gpio_K(O) Drv_SysClk_K(O) physmap_of(O) map_rom(O) cfi_cmdset_0002(O) cfi_probe(O) cfi_util(O) gen_probe(O) chipreg(O) rsm(O) rtos_snapshot(O) rtos_kbox_panic(O) bsp_wdt(O) drv_bsp_ddr(O) bsp_reg(O) Drv_Dts_K(O) Drv_SysCtl_K(O) arm_sal_issu(O) ksecurec(PO) ext4 jbd2 ext2 mbcache ofpart i2c_dev i2c_core uio nand nand_ecc nand_ids cmdlinepart mtdblock mtd_blkdevs mtd
> [20190527221107]CPU: 2 PID: 2656 Comm: monitor Tainted: P        W  O    4.4.171 #1
> [20190527221107]Hardware name: Hisilicon chip6_16 Product Board (DT)
> [20190527221107]task: ffffffd3bf81c250 task.stack: ffffffd3bff0c000
> [20190527221107]PC is at rb_erase+0x14/0x320
> [20190527221107]LR is at erase_header+0x50/0x54
> [20190527221107]pc : [<ffffff800846e528>] lr : [<ffffff80083d2468>] pstate: 00000145
> [20190527221107]sp : ffffffd3bff0f9a0
> [20190527221107]x29: ffffffd3bff0f9a0 x28: ffffff69fe8b1980
> [20190527221107]x27: 0000000000000001 x26: ffffff8008e7e340
> [20190527221107]x25: ffffff8008e05000 x24: ffffff8008e32e28
> [20190527221107]x23: ffffffd3c0542500 x22: ffffff8008e32000
> [20190527221107]x21: ffffff8008e05000 x20: ffffffd3c0542f00
> [20190527221107]x19: ffffffd3c0542fb8 x18: 000000000000000f
> [20190527221107]x17: 0000007f9bd20e10 x16: ffffff8008367108
> [20190527221107]x15: 0000000000001fee x14: 0000000000000000
> [20190527221107]x13: 0000000000000000 x12: 0000000000000000
> [20190527221107]x11: 0000000000000001 x10: 0000000000000001
> [20190527221107]x9 : 0000000000000001 x8 : ffffff800894622d
> [20190527221107]x7 : ffffffd3c65dd7d0 x6 : 0000000000000000
> [20190527221107]x5 : ffffffd3bf81a740 x4 : 0000000000000000
> [20190527221107]x3 : 0000000100000001 x2 : 0000000100000000
> [20190527221107]x1 : ffffffd3c0542550 x0 : ffffffd3c0542f58
> [20190527221107]Process monitor (pid: 2656, stack limit = 0xffffffd3bff0c000)
> [20190527221107]
> [20190527221107][<ffffff800846e528>] rb_erase+0x14/0x320
> [20190527221107][<ffffff80083d2f5c>] drop_sysctl_table+0x17c/0x1d4
> [20190527221107][<ffffff80083d2f84>] drop_sysctl_table+0x1a4/0x1d4
> [20190527221107][<ffffff80083d3050>] unregister_sysctl_table+0x9c/0xa8
> [20190527221107][<ffffff80083d3014>] unregister_sysctl_table+0x60/0xa8
> [20190527221107][<ffffff800825b880>] partition_sched_domains+0x64/0x338
> [20190527221107][<ffffff80082bd37c>] rebuild_sched_domains_locked+0xe0/0x3c0
> [20190527221107][<ffffff80082be590>] cpuset_write_resmask+0x288/0x8cc	
> [20190527221107][<ffffff80082b5600>] cgroup_file_write+0x64/0x128
> [20190527221107][<ffffff80083daa50>] kernfs_fop_write+0x15c/0x1ac
> [20190527221107][<ffffff8008365c9c>] __vfs_write+0x60/0x124
> [20190527221107][<ffffff800836666c>] vfs_write+0xb0/0x184
> [20190527221107][<ffffff8008367174>] SyS_write+0x6c/0xcc
> [20190527221107][<ffffff8008202cb8>] __sys_trace_return+0x0/0x4
> 
> 
> The disassembler and the source code about the backtrace are as below:
> 
> rb_erase()->__rb_erase_augmented()->__rb_change_child()
> __rb_erase_augmented():
> ffffff800846e514:       a9409006        ldp     x6, x4, [x0, #8]
> ffffff800846e518:       b5000244        cbnz    x4, ffffff800846e560 <rb_erase+0x4c>
> ffffff800846e51c:       f9400003        ldr     x3, [x0]
> __rb_change_child():
> ffffff800846e520:       f27ef462        ands    x2, x3, #0xfffffffffffffffc
> ffffff800846e524:       54000140        b.eq    ffffff800846e54c <rb_erase+0x38>  // b.none
> ffffff800846e528:       f9400844        ldr     x4, [x2, #16]
> ffffff800846e52c:       eb04001f        cmp     x0, x4
> 
> ffffff800846e530:       540000a1        b.ne    ffffff800846e544 <rb_erase+0x30>  // b.any
> __write_once_size():
> ffffff800846e534:       f9000846        str     x6, [x2, #16]
> 
> rb_erase()->__rb_erase_augmented()->__rb_change_child()
> static __always_inline struct rb_node *
> __rb_erase_augmented(struct rb_node *node, struct rb_root *root,
> 		     const struct rb_augment_callbacks *augment)
> {
> 	...
> 	if (!tmp) {
> 		...
> 		pc = node->__rb_parent_color;
> 		parent = __rb_parent(pc);
> 		__rb_change_child(node, child, parent, root);
> 	...
> }
> static inline void
> __rb_change_child(struct rb_node *old, struct rb_node *new,
> 		  struct rb_node *parent, struct rb_root *root)
> {
> 	if (parent) {
> 		if (parent->rb_left == old)
> 			WRITE_ONCE(parent->rb_left, new);
> 		else
> 			WRITE_ONCE(parent->rb_right, new);
> 	} else
> 		WRITE_ONCE(root->rb_node, new);
> }
> 
> 
> When panic, the x0 is ffffffd3c0542f58 which indicates the first parameter of function __rb_erase_augmented.
> 
> 2f38  c0542500 ffffffd3 c0542f58 ffffffd3 00000000 00000000 00000000 00000000
> 2f58  00000001 00000001 00000000 00000000 00000000 00000000 c0542f00 ffffffd3
> 2f78  c0542ff8 ffffffd3 00000000 00000000 00000000 0000416d 00000000 00000000
> 
> x0 is the "struct rb_node *node", that is, the content of the node is:
> struct rb_node {
> 	unsigned long  __rb_parent_color;             0000000100000001
> 	struct rb_node *rb_right;                     ffffffd3c0542558
> 	struct rb_node *rb_left;                      0000000000000000
> } __attribute__((aligned(sizeof(long))));
> 
> The value of __rb_parent_color is 0000000100000001 and the parent address is 0000000100000000. Generally, the parent
> address should be NULL or a valid address.
> 
> 
> Is there any idea about this issue?
> 
> Thanks,
> Chen Zhou
> 
> 
> 
> 
> 




[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Security]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]     [Monitors]

  Powered by Linux