[PATCHv8 1/5] powerpc/setup : Enable boot_cpu_hwid for PPC32

Tue Oct 10 19:24:48 AEDT 2023

Hello Pingfan,

>
> With this patch series applied, the kdump kernel fails to boot on 
> powerpc with nr_cpus=1.
>
> Console logs:
> -------------------
> [root]# echo c > /proc/sysrq-trigger
> [   74.783235] sysrq: Trigger a crash
> [   74.783244] Kernel panic - not syncing: sysrq triggered crash
> [   74.783252] CPU: 58 PID: 3838 Comm: bash Kdump: loaded Not tainted 
> 6.6.0-rc5pf-nr-cpus+ #3
> [   74.783259] Hardware name: POWER10 (raw) phyp pSeries
> [   74.783275] Call Trace:
> [   74.783280] [c00000020f4ebac0] [c000000000ed9f38] 
> dump_stack_lvl+0x6c/0x9c (unreliable)
> [   74.783291] [c00000020f4ebaf0] [c000000000150300] panic+0x178/0x438
> [   74.783298] [c00000020f4ebb90] [c000000000936d48] 
> sysrq_handle_crash+0x28/0x30
> [   74.783304] [c00000020f4ebbf0] [c00000000093773c] 
> __handle_sysrq+0x10c/0x250
> [   74.783309] [c00000020f4ebc90] [c000000000937fa8] 
> write_sysrq_trigger+0xc8/0x168
> [   74.783314] [c00000020f4ebcd0] [c000000000665d8c] 
> proc_reg_write+0x10c/0x1b0
> [   74.783321] [c00000020f4ebd00] [c00000000058da54] 
> vfs_write+0x104/0x4b0
> [   74.783326] [c00000020f4ebdc0] [c00000000058dfdc] 
> ksys_write+0x7c/0x140
> [   74.783331] [c00000020f4ebe10] [c000000000033a64] 
> system_call_exception+0x144/0x3a0
> [   74.783337] [c00000020f4ebe50] [c00000000000c554] 
> system_call_common+0xf4/0x258
> [   74.783343] --- interrupt: c00 at 0x7fffa0721594
> [   74.783352] NIP:  00007fffa0721594 LR: 00007fffa0697bf4 CTR: 
> 0000000000000000
> [   74.783364] REGS: c00000020f4ebe80 TRAP: 0c00   Not tainted 
> (6.6.0-rc5pf-nr-cpus+)
> [   74.783376] MSR:  800000000280f033 
> <SF,VEC,VSX,EE,PR,FP,ME,IR,DR,RI,LE>  CR: 28222202  XER: 00000000
> [   74.783394] IRQMASK: 0
> [   74.783394] GPR00: 0000000000000004 00007ffffc4b6800 
> 00007fffa0807300 0000000000000001
> [   74.783394] GPR04: 000000013549ea60 0000000000000002 
> 0000000000000010 0000000000000000
> [   74.783394] GPR08: 0000000000000000 0000000000000000 
> 0000000000000000 0000000000000000
> [   74.783394] GPR12: 0000000000000000 00007fffa0abaf70 
> 0000000040000000 000000011a0f9798
> [   74.783394] GPR16: 000000011a0f9724 000000011a097688 
> 000000011a02ff70 000000011a0fd568
> [   74.783394] GPR20: 0000000135554bf0 0000000000000001 
> 000000011a0aa478 00007ffffc4b6a24
> [   74.783394] GPR24: 00007ffffc4b6a20 000000011a0faf94 
> 0000000000000002 000000013549ea60
> [   74.783394] GPR28: 0000000000000002 00007fffa08017a0 
> 000000013549ea60 0000000000000002
> [   74.783440] NIP [00007fffa0721594] 0x7fffa0721594
> [   74.783443] LR [00007fffa0697bf4] 0x7fffa0697bf4
> [   74.783447] --- interrupt: c00
> I'm in purgatory
> [    0.000000] radix-mmu: Page sizes from device-tree:
> [    0.000000] radix-mmu: Page size shift = 12 AP=0x0
> [    0.000000] radix-mmu: Page size shift = 16 AP=0x5
> [    0.000000] radix-mmu: Page size shift = 21 AP=0x1
> [    0.000000] radix-mmu: Page size shift = 30 AP=0x2
> [    0.000000] Activating Kernel Userspace Access Prevention
> [    0.000000] Activating Kernel Userspace Execution Prevention
> [    0.000000] radix-mmu: Mapped 0x0000000000000000-0x0000000000010000 
> with 64.0 KiB pages (exec)
> [    0.000000] radix-mmu: Mapped 0x0000000000010000-0x0000000000200000 
> with 64.0 KiB pages
> [    0.000000] radix-mmu: Mapped 0x0000000000200000-0x0000000020000000 
> with 2.00 MiB pages
> [    0.000000] radix-mmu: Mapped 0x0000000020000000-0x0000000022600000 
> with 2.00 MiB pages (exec)
> [    0.000000] radix-mmu: Mapped 0x0000000022600000-0x0000000040000000 
> with 2.00 MiB pages
> [    0.000000] radix-mmu: Mapped 0x0000000040000000-0x0000000180000000 
> with 1.00 GiB pages
> [    0.000000] radix-mmu: Mapped 0x0000000180000000-0x00000001a0000000 
> with 2.00 MiB pages
> [    0.000000] lpar: Using radix MMU under hypervisor
> [    0.000000] Linux version 6.6.0-rc5pf-nr-cpus+ 
> (root at ltcever7x0-lp1.aus.stglabs.ibm.com) (gcc (GCC) 8.5.0 20210514 
> (Red Hat 8.5.0-20), GNU ld version 2.30-123.el8) #3 SMP Mon Oct  9 11:07:
> 41 CDT 2023
> [    0.000000] Found initrd at 0xc000000022e60000:0xc0000000248f08d8
> [    0.000000] Hardware name: IBM,9043-MRX POWER10 (raw) 0x800200 
> 0xf000006 of:IBM,FW1060.00 (NM1060_016) hv:phyp pSeries
> [    0.000000] printk: bootconsole [udbg0] enabled
> [    0.000000] the round shift between dt seq and the cpu logic 
> number: 56
> [    0.000000] BUG: Unable to handle kernel data access on write at 
> 0xc0000001a0000000
> [    0.000000] Faulting instruction address: 0xc000000022009c64
> [    0.000000] Oops: Kernel access of bad area, sig: 11 [#1]
> [    0.000000] LE PAGE_SIZE=64K MMU=Radix SMP NR_CPUS=2048 NUMA pSeries
> [    0.000000] Modules linked in:
> [    0.000000] CPU: 2 PID: 0 Comm: swapper Not tainted 
> 6.6.0-rc5pf-nr-cpus+ #3
> [    0.000000] Hardware name:  POWER10 (raw)  hv:phyp pSeries
> [    0.000000] NIP:  c000000022009c64 LR: c000000022009c54 CTR: 
> c0000000201ff348
> [    0.000000] REGS: c000000022aebb00 TRAP: 0300   Not tainted 
> (6.6.0-rc5pf-nr-cpus+)
> [    0.000000] MSR:  8000000000001033 <SF,ME,IR,DR,RI,LE> CR: 
> 28222824  XER: 00000001
> [    0.000000] CFAR: c000000020031574 DAR: c0000001a0000000 DSISR: 
> 42000000 IRQMASK: 1
> [    0.000000] GPR00: c000000022009ba0 c000000022aebda0 
> c0000000213d1300 0000000000000004
> [    0.000000] GPR04: 0000000000000001 c000000022aebbc0 
> c000000022aebbb8 0000000000000000
> [    0.000000] GPR08: 0000000000000001 c00000019ffffff8 
> 000000000000003a c0000000229c8a78
> [    0.000000] GPR12: 0000000000002000 c000000022e4a800 
> c0000000211d34b8 c0000000211d3aa8
> [    0.000000] GPR16: c0000000211d75a0 c0000000211d75b0 
> c0000000225f3b98 0000000000000000
> [    0.000000] GPR20: 0000000000000001 0000000000000001 
> 0000000000000001 0000000000000001
> [    0.000000] GPR24: 0000000000000008 0000000000000000 
> 0000000000000001 c00000019ffffdc0
> [    0.000000] GPR28: 0000000000000002 c000000022b368e0 
> c000000022aebe08 0000000000000008
> [    0.000000] NIP [c000000022009c64] smp_setup_cpu_maps+0x420/0x724
> [    0.000000] LR [c000000022009c54] smp_setup_cpu_maps+0x410/0x724
> [    0.000000] Call Trace:
> [    0.000000] [c000000022aebda0] [c000000022009ba0] 
> smp_setup_cpu_maps+0x35c/0x724 (unreliable)
> [    0.000000] [c000000022aebeb0] [c00000002200a19c] 
> setup_arch+0x1b8/0x54c
> [    0.000000] [c000000022aebf30] [c000000022003f88] 
> start_kernel+0xb0/0x768
> [    0.000000] [c000000022aebfe0] [c00000002000d888] 
> start_here_common+0x1c/0x20
> [    0.000000] Code: 3929ffff 7f89e040 409c002c 7ec4b378 7f83e378 
> 4a027939 7f83e378 4a0278e5 e95b0018 3d22017d e929f028 7d4ac42c 
> <7d49c12e> eb7b0000 7e99a378 4bffff3c
> [    0.000000] ---[ end trace 0000000000000000 ]---
> [    0.000000]
> [    0.000000] Kernel panic - not syncing: Fatal exception
> [    0.000000] Rebooting in 180 seconds..
>
> However, the kdump kernel boots fine if the kernel crashes on CPU 0.

Found a pattern in kdump kernel failure with nr_cpus=1.

On CPU 0, 8, 16, 24, 32, 40, it boots fine.
On CPUs 1-7, 9-15, 17-23, 25-31, 33-39, it fails to boot.

Hope this helps.

Thanks,
Sourabh