When CPUs are in an offline state, it can lead to mapping errors.
We need to map them to the correct positions one by one.

Before:
n_namesz: 5 ("CPU2")
n_descsz: 392
  n_type: 1 (NT_PRSTATUS)
          si.signo: 0  si.code: 0  si.errno: 0
          cursig: 0  sigpend: 0  sighold: 0
          pid: 3  ppid: 0  pgrp: 0  sid:0
          utime: 0.000000  stime: 0.000000
          cutime: 0.000000  cstime: 0.000000
           X0: ffffffc000fc8818   X1: 0000000000000000   X2: ffffffc000fc84c8
           X3: 0000000000000000   X4: ffffffc0405e37bf   X5: ffffffc00a07372f
           X6: 322e34323320205b   X7: 545b5d3539383334   X8: ffffffc000fc2f0c
           X9: 89fece0a9ef8cb00  X10: c0000001001f75f4  X11: 00000001001f75f4
          X12: 0000000000000003  X13: 00000000000005f4  X14: ffffffc009eb1210
          X15: 0000000000000004  X16: 000000002a4cec24  X17: 000000002a4cec24
          X18: ffffffc009e7d140  X19: ffffffc00a04c670  X20: 0000000000000000
          X21: 0000000000000000  X22: ffffff8027f22280  X23: 0000000000000009
          X24: 0000000000000007  X25: ffffffc009f839c0  X26: ffffffc0090f87f8
          X27: 0000000000000000  X28: ffffff80454f3840  X29: ffffffc0405e3b60
           LR: ffffffc0080e57fc   SP: ffffffc0405e3b60   PC: ffffffc000fc2f84

CPU 0: [OFFLINE]
CPU 1: [OFFLINE]
CPU 2:
    X0: 0000000000000000   X1: 0000000000000000   X2: 0000000000000000
    X3: 000000000003fcbc   X4: 0000000000000001   X5: 0000000000000000
    X6: 0000000000000000   X7: 0000000000000000   X8: 00000000ffffffff
    X9: ffffffc009e6ae48  X10: ffffffc009e6ae20  X11: 0000000000000000
   X12: 0000000000000002  X13: 0000000000000004  X14: 0000000000000000
   X15: 0000000000004000  X16: 00000000f90f05f6  X17: 00000000f90f05f6
   X18: 0000000000000000  X19: 0000000000000002  X20: ffffffc009e3b008
   X21: ffffffc00a01d020  X22: ffffffc009f798f0  X23: 0000000060001000
   X24: 0000000000000000  X25: 0000000000000000  X26: 0000000000000000
   X27: 0000000000000000  X28: ffffff8111eecb00  X29: ffffffc008003f50
    LR: ffffffc00802df88   SP: ffffffc008003f40   PC: ffffffc00802df94
   PSTATE: 024003c5   FPVALID: 00000000

After:
CPU 2:
    X0: ffffffc000fc8818   X1: 0000000000000000   X2: ffffffc000fc84c8
    X3: 0000000000000000   X4: ffffffc0405e37bf   X5: ffffffc00a07372f
    X6: 322e34323320205b   X7: 545b5d3539383334   X8: ffffffc000fc2f0c
    X9: 89fece0a9ef8cb00  X10: c0000001001f75f4  X11: 00000001001f75f4
   X12: 0000000000000003  X13: 00000000000005f4  X14: ffffffc009eb1210
   X15: 0000000000000004  X16: 000000002a4cec24  X17: 000000002a4cec24
   X18: ffffffc009e7d140  X19: ffffffc00a04c670  X20: 0000000000000000
   X21: 0000000000000000  X22: ffffff8027f22280  X23: 0000000000000009
   X24: 0000000000000007  X25: ffffffc009f839c0  X26: ffffffc0090f87f8
   X27: 0000000000000000  X28: ffffff80454f3840  X29: ffffffc0405e3b60
    LR: ffffffc0080e57fc   SP: ffffffc0405e3b60   PC: ffffffc000fc2f84
   PSTATE: 600000c5   FPVALID: 00000000

crash> bt
PID: 15959    TASK: ffffff80454f3840  CPU: 2    COMMAND: "AnrConsumer"
    [ffffffc0405e3b60] ipanic at ffffffc000fc2f80 [mrdump]
    [ffffffc0405e3b70] atomic_notifier_call_chain at ffffffc0080e57f8
    [ffffffc0405e3c30] panic at ffffffc008f734d0
    [ffffffc0405e3c80] sysrq_handle_crash at ffffffc0087f3c18
    [ffffffc0405e3c90] __handle_sysrq at ffffffc0087f3798
    [ffffffc0405e3ce0] write_sysrq_trigger at ffffffc0087f49c0
    [ffffffc0405e3d00] proc_reg_write at ffffffc00842e4b8
    [ffffffc0405e3d80] vfs_write at ffffffc008381eb4
    [ffffffc0405e3dd0] ksys_write at ffffffc008382200
    [ffffffc0405e3e10] __arm64_sys_write at ffffffc00838228c
    [ffffffc0405e3e20] invoke_syscall at ffffffc00802efe0
    [ffffffc0405e3e40] el0_svc_common at ffffffc00802eef4
    [ffffffc0405e3e70] do_el0_svc at ffffffc00802ede8
    [ffffffc0405e3e80] el0_svc at ffffffc008f7a7d0
    [ffffffc0405e3ea0] el0t_64_sync_handler at ffffffc008f7a758
    [ffffffc0405e3fe0] el0t_64_sync at ffffffc00801157c
     PC: 00000077c798ca28   LR: 00000077a82e19f4   SP: 000000761c517af0
    X29: 000000761c517b00  X28: 000000761c517db8  X27: 000000761c517c90
    X26: 000000761c517c98  X25: 000000761c517bf9  X24: 000000761c519000
    X23: 000000761c517be1  X22: 0000000000000001  X21: 00000000000003e3
    X20: 000000761c517c11  X19: 000000761c517bf8  X18: 0000007568224000
    X17: 00000077c798ca20  X16: 00000077c79b2ae0  X15: b4000077202cc480
    X14: 0000000000000000  X13: 000000761c517a70  X12: ffffff80ffffffd0
    X11: 000000761c517a40  X10: 0000000000000001   X9: 0000000000000000
     X8: 0000000000000040   X7: 7f7f7f7f7f7f7f7f   X6: 0000000000000010
     X5: 000000761c517c0c   X4: ffffffffffffffff   X3: ffffffffffffffff
     X2: 0000000000000001   X1: 000000761c517c11   X0: 00000000000003e3
    ORIG_X0: 00000000000003e3  SYSCALLNO: 40  PSTATE: 00001000

Signed-off-by: Guanyou.Chen <chenguanyou@xiaomi.com>
---
 netdump.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/netdump.c b/netdump.c
index b4e2a5c..8ea5159 100644
--- a/netdump.c
+++ b/netdump.c
@@ -75,7 +75,7 @@ void
 map_cpus_to_prstatus(void)
 {
    void **nt_ptr;
-   int online, i, j, nrcpus;
+   int online, i, nrcpus;
    size_t size;

    if (pc->flags2 & QEMU_MEM_DUMP_ELF)  /* notes exist for all cpus */
@@ -100,9 +100,9 @@ map_cpus_to_prstatus(void)
     */
    nrcpus = (kt->kernel_NR_CPUS ? kt->kernel_NR_CPUS : NR_CPUS);

-   for (i = 0, j = 0; i < nrcpus; i++) {
+   for (i = 0; i < nrcpus; i++) {
        if (in_cpu_map(ONLINE_MAP, i) && machdep->is_cpu_prstatus_valid(i)) {
-           nd->nt_prstatus_percpu[i] = nt_ptr[j++];
+           nd->nt_prstatus_percpu[i] = nt_ptr[i];
            nd->num_prstatus_notes =
                MAX(nd->num_prstatus_notes, i+1);
        }
--
2.34.1

Guanyou.

Thanks.