[缺陷描述]： 6.6.102-5.2.1_rc1.an23.riscv64内核，xfstests ext4文件系统，generic/211用例执行fail，用例执行过程中触发soft lockup。 测试日志： # ./check generic/211 mount: (hint) your fstab has been modified, but systemd still uses the old version; use 'systemctl daemon-reload' to reload. FSTYP -- ext4 PLATFORM -- Linux/riscv64 localhost 6.6.102-5.2.1_rc1.an23.riscv64 #1 SMP PREEMPT_DYNAMIC Tue Dec 16 04:26:06 EST 2025 mount: (hint) your fstab has been modified, but systemd still uses the old version; use 'systemctl daemon-reload' to reload. generic/211 mount: (hint) your fstab has been modified, but systemd still uses the old version; use 'systemctl daemon-reload' to reload. _check_dmesg: something found in dmesg (see /tmp/tone/run/xfstests/results//generic/211.dmesg) Ran: generic/211 Failures: generic/211 Failed 1 of 1 tests [重现概率]： 必现，测试2次，fail2次 [重现环境]： 环境信息：riscv64物理机 # uname -r 6.6.102-5.2.1_rc1.an23.riscv64 # cat /etc/os-release NAME="Anolis OS" VERSION="23.3" ID="anolis" VERSION_ID="23.3" PLATFORM_ID="platform:an23" PRETTY_NAME="Anolis OS 23.3" ANSI_COLOR="0;31" HOME_URL="https://openanolis.cn/" BUG_REPORT_URL="https://bugzilla.openanolis.cn/" # lscpu Architecture: riscv64 Byte Order: Little Endian CPU(s): 64 On-line CPU(s) list: 0-63 NUMA: NUMA node(s): 4 NUMA node0 CPU(s): 0-7,16-23 NUMA node1 CPU(s): 8-15,24-31 NUMA node2 CPU(s): 32-39,48-55 NUMA node3 CPU(s): 40-47,56-63 # lsblk -f NAME FSTYPE FSVER LABEL UUID FSAVAIL FSUSE% MOUNTPOINTS nvme0n1 ├─nvme0n1p1 vfat FAT16 EFI AF19-A668 74.8M 38% /boot/efi ├─nvme0n1p2 ext4 1.0 BOOT 31f4f3c4-0587-4801-9584-10b131177c0d 221.8M 44% /boot ├─nvme0n1p3 ext4 1.0 ROOT e297bfa3-2c55-44f8-90cb-ae8eb642d188 84.5G 35% / ├─nvme0n1p4 ├─nvme0n1p5 ext4 1.0 ef197a8a-047d-480c-8fe9-902ff347fe30 └─nvme0n1p6 # free -h total used free shared buff/cache available Mem: 61Gi 1.6Gi 51Gi 27Mi 9.2Gi 60Gi Swap: 0B 0B 0B # cat /proc/cmdline root=UUID=e297bfa3-2c55-44f8-90cb-ae8eb642d188 ro console=ttyS0,115200 root=LABEL=ROOT rootfstype=ext4 rootwait rw earlycon selinux=0 LANG=en_US.UTF-8 nvme.use_threaded_interrupts=1 nvme_core.io_timeout=3000 audit=0 per_numa_node_futex=disable [重现步骤]： 1. 准备环境 mkfs.xfs /dev/nvme0n1p5 mkfs.xfs /dev/nvme0n1p6 export FSTYP=ext4 export TEST_DEV=/dev/nvme0n1p5 export SCRATCH_DEV=/dev/nvme0n1p6 export TEST_DIR=/fs/nvme0n1p5 export SCRATCH_MNT=/fs/nvme0n1p6 2.执行测试用例 ./check generic/211 [期望结果]： 用例pass [实际结果]： 用例fail [原因定位]： 用例执行过程中触发soft lockup # cat /var/tmp/tone/run/xfstests/results//generic/211.dmesg [60219.261886] run fstests generic/211 at 2025-12-26 11:27:26 [60976.884874] watchdog: BUG: soft lockup - CPU#18 stuck for 26s! [aio-free-ring-w:3701969] [60976.901094] CPU#18 Utilization every 4s during lockup: [60976.911375] #1: 99% system, 0% softirq, 1% hardirq, 0% idle [60976.923914] #2: 100% system, 0% softirq, 2% hardirq, 0% idle [60976.936451] #3: 100% system, 0% softirq, 2% hardirq, 0% idle [60976.948987] #4: 99% system, 0% softirq, 2% hardirq, 0% idle [60976.961523] #5: 100% system, 0% softirq, 2% hardirq, 0% idle [60976.974059] Modules linked in: pktgen(E) cls_bpf(E) dummy(E) sch_ingress(E) ip6_tunnel(E) tunnel6(E) ipip(E) tunnel4(E) ip_tunnel(E) veth(E) nf_tables(E) overlay(E) dm_thin_pool(E) dm_persistent_data(E) dm_bio_prison(E) sd_mod(E) xfs(E) dm_flakey(E) dm_snapshot(E) dm_bufio(E) dm_zero(E) binfmt_misc(E) rfkill(E) vfat(E) fat(E) r8169(E) realtek(E) uio_pdrv_genirq(E) uio(E) loop(E) fuse(E) dm_mod(E) nfnetli nk(E) amdgpu(E) amdxcp(E) drm_exec(E) gpu_sched(E) drm_buddy(E) radeon(E) video(E) drm_suballoc_helper(E) i2c_algo_bit(E) drm_ttm_helper(E) ttm(E) drm_display_helper(E) sg2042_mcu(E) ahci(E) nvme(E) drm_km s_helper(E) libahci(E) libcrc32c(E) drm(E) sdhci_of_dwcmshc(E) libata(E) nvme_core(E) sdhci_pltfm(E) t10_pi(E) sdhci(E) cqhci(E) cec(E) mmc_core(E) i2c_designware_platform(E) pwm_sophgo_sg2042(E) i2c_desig nware_core(E) pwm_fan(E) [last unloaded: scsi_debug(E)] [60977.126683] CPU: 18 PID: 3701969 Comm: aio-free-ring-w Tainted: G EL 6.6.102-5.2.1_rc1.an23.riscv64 #1 [60977.147910] Hardware name: Milk-V Pioneer (DT) [60977.156797] epc : down_write+0x2/0x76 [60977.164124] ra : free_pgtables+0x166/0x17e [60977.172502] epc : ffffffff80a1d9ce ra : ffffffff80233bc4 sp : ffffffc617d8bc10 [60977.186949] gp : ffffffff81b103f0 tp : ffffffde07104ec0 t0 : 00000026e999bfff [60977.201396] t1 : 0000000000000028 t2 : 000000008040001e s0 : ffffffc617d8bcb0 [60977.215841] s1 : ffffffdefb254190 a0 : ffffffdf14c0b0c8 a1 : 0000000000000068 [60977.230287] a2 : ffffffdef9dbf108 a3 : ffffffdeedda7f00 a4 : 00000000018c825b [60977.244734] a5 : ffffffde1611e1c0 a6 : ffffffdeedda7f40 a7 : 0000000000000001 [60977.259180] s2 : ffffffdefb2544b0 s3 : ffffffc617d8bcb0 s4 : 0000000000000001 [60977.273626] s5 : 0000000000000406 s6 : ffffffffffffffff s7 : 0000000000200000 [60977.288073] s8 : 0000000131fc7000 s9 : ffffffc617d8bcf0 s10: 0000000000000000 [60977.302519] s11: 0000000003e7fc2b t3 : ffffffd76178c830 t4 : 0000000000000006 [60977.316965] t5 : 00000026e9915000 t6 : ffffffff812b2fa8 [60977.327588] status: 0000000200000120 badaddr: 0000000000000000 cause: 8000000000000005 [60977.343427] [<ffffffff80a1d9ce>] down_write+0x2/0x76 [60977.353359] [<ffffffff8024207a>] exit_mmap+0x11c/0x34a [60977.363642] [<ffffffff8001ed9a>] __mmput+0x42/0x106 [60977.373401] [<ffffffff8001ee94>] mmput+0x36/0x40 [60977.382635] [<ffffffff80027cbe>] exit_mm+0x90/0xe4 [60977.392219] [<ffffffff80027e7e>] do_exit+0x16c/0x3be [60977.402151] [<ffffffff80028212>] do_group_exit+0x28/0x74 [60977.412777] [<ffffffff80028276>] __wake_up_parent+0x0/0x24 [60977.423751] [<ffffffff80a162c2>] do_trap_ecall_u+0x114/0x126 [60977.435072] [<ffffffff80a2243e>] ret_from_exception+0x0/0x6e [61110.638017] EXT4-fs (nvme0n1p5): unmounting filesystem ef197a8a-047d-480c-8fe9-902ff347fe30. [61111.951273] EXT4-fs (nvme0n1p5): mounted filesystem ef197a8a-047d-480c-8fe9-902ff347fe30 r/w with ordered data mode. Quota mode: none.