Re: powernv_rng_read: Oops: Kernel access of bad area, sig: 11 [#1]

Dear Madhavan, dear Jason,


Am 11.05.26 um 09:00 schrieb Paul Menzel:

Am 07.05.26 um 04:40 schrieb Madhavan Srinivasan:

quoted

On 5/6/26 7:31 PM, Paul Menzel wrote:

quoted

quoted

After a long while, on the 8335-GCA POWER8 (raw) 0x4d0200 
opal:skiboot-5.4.8-5787ad3 PowerNV, I built Linux from Linus’ master 
branch and rebooted via kexec.

[    0.000000] Linux version 7.1.0-rc2+ (pmenzel@flughafenberlinbrandenburgwillybrandt.molgen.mpg.de) (gcc (Ubuntu 11.2.0-7ubuntu2) 11.2.0, GNU ld (GNU Binutils for Ubuntu) 2.37) #3 SMP PREEMPT Wed May  6 08:50:58 CEST 2026
[…]
[   17.901992] Kernel attempted to read user page (0) - exploit attempt? (uid: 0)
[   17.902011] BUG: Kernel NULL pointer dereference on read at 0x00000000
[   17.902018] Faulting instruction address: 0xc0000000000e7138
[   17.902027] Oops: Kernel access of bad area, sig: 11 [#1]
[   17.902034] LE PAGE_SIZE=64K MMU=Hash  SMP NR_CPUS=2048 NUMA PowerNV
[   17.902045] Modules linked in: powernv_rng(+) bnx2x ofpart ibmpowernv bfq mdio cmdlinepart powernv_flash ipmi_powernv ipmi_devintf mtd ipmi_msghandler at24(+) vmx_crypto opal_prd sch_fq_codel nfsd parport_pc ppdev auth_rpcgss nfs_acl lp lockd grace parport sunrpc autofs4 btrfs xor libblake2b raid6_pq ast drm_shmem_helper drm_client_lib i2c_algo_bit drm_kms_helper drm ahci drm_panel_orientation_quirks libahci
[   17.902185] CPU: 147 UID: 0 PID: 2626 Comm: hwrng Not tainted 7.1.0-rc2+ #3 PREEMPTLAZY
[   17.902197] Hardware name: 8335-GCA POWER8 (raw) 0x4d0200 opal:skiboot-5.4.8-5787ad3 PowerNV
[   17.902204] NIP:  c0000000000e7138 LR: c00800001ec8013c CTR: c0000000000e70fc
[   17.902212] REGS: c000000092913c50 TRAP: 0300   Not tainted (7.1.0-rc2+)
[   17.902222] MSR:  900000000280b033 <SF,HV,VEC,VSX,EE,FP,ME,IR,DR,RI,LE>  CR: 44420220  XER: 20000000
[   17.902269] CFAR: c00800001ec8026c DAR: 0000000000000000 DSISR: 40000000 IRQMASK: 0
               GPR00: c00800001ec8013c c000000092913ef0 c000000001c18100 c00000002222d900
               GPR04: c00000002222d900 0000000000000080 0000000000000001 0000000000000000
               GPR08: 0000000000000000 c000000002212000 c0000000951e1780 c00800001ec80258
               GPR12: c0000000000e70fc c00000ffff6fd700 c0000000001d11c0 c00000001b99b9c0
               GPR16: 0000000000000000 0000000000000000 0000000000000000 0000000000000000
               GPR20: 0000000000000000 0000000000000000 0000000000000000 0000000000000000
               GPR24: 0000000000000000 c000000002fe6a58 0000000000000000 0000000000000000
               GPR28: c000000002fe6a20 0000000000000010 000000000000000f c00000002222d900
[   17.902406] NIP [c0000000000e7138] pnv_get_random_long+0x3c/0x114
[   17.902426] LR [c00800001ec8013c] powernv_rng_read+0x78/0xc4 [powernv_rng]
[   17.902444] Call Trace:
[   17.902448] [c000000092913ef0] [c000000092913f30] 0xc000000092913f30 (unreliable)
[   17.902463] [c000000092913f30] [c000000000decd58] hwrng_fillfn+0xd4/0x3dc
[   17.902484] [c000000092913f90] [c0000000001d1328] kthread+0x170/0x1a4
[   17.902498] [c000000092913fe0] [c00000000000d030] start_kernel_thread+0x14/0x18
[   17.902513] Code: 60000000 7d2000a6 71290010 418200bc e94d0908 812a0000 39290001 912a0000 e90d0030 3d220060 39299f00 7d08482a <e9280000> 7c0004ac e8e90000 0c070000
[   17.902569] ---[ end trace 0000000000000000 ]---
[   18.008801] pstore: backend (nvram) writing error (-1)

[   18.015458] note: hwrng[2626] exited with irqs disabled
[   18.015483] note: hwrng[2626] exited with preempt_count 1

Please find the output of `dmesg` attached.

This is from my yesterday's boot test log in my P8, did not see this 
fail.

root@ltcppm1:~# uname -a
Linux ltcppm1.ltc.tadn.ibm.com 7.1.0-rc2-00021-gf583bd5f64d4 #1 SMP 
PREEMPT Wed May  6 00:55:45 EDT 2026 ppc64le GNU/Linux
root@ltcppm1:~# dmesg
[    0.000000] [      T0] random: crng init done
[    0.000000] [      T0] hash-mmu: Page sizes from device-tree:
[    0.000000] [      T0] hash-mmu: base_shift=12: shift=12, sllp=0x0000, avpnm=0x00000000, tlbiel=1, penc=0
[    0.000000] [      T0] hash-mmu: base_shift=12: shift=16, sllp=0x0000, avpnm=0x00000000, tlbiel=1, penc=7
[    0.000000] [      T0] hash-mmu: base_shift=12: shift=24, sllp=0x0000, avpnm=0x00000000, tlbiel=1, penc=56
[    0.000000] [      T0] hash-mmu: base_shift=16: shift=16, sllp=0x0110, avpnm=0x00000000, tlbiel=1, penc=1
[    0.000000] [      T0] hash-mmu: base_shift=16: shift=24, sllp=0x0110, avpnm=0x00000000, tlbiel=1, penc=8
[    0.000000] [      T0] hash-mmu: base_shift=24: shift=24, sllp=0x0100, avpnm=0x00000001, tlbiel=0, penc=0
[    0.000000] [      T0] hash-mmu: base_shift=34: shift=34, sllp=0x0120, avpnm=0x000007ff, tlbiel=0, penc=3
[    0.000000] [      T0] Enabling pkeys with max key count 32
[    0.000000] [      T0] Activating Kernel Userspace Access Prevention
[    0.000000] [      T0] Activating Kernel Userspace Execution Prevention
[    0.000000] [      T0] hash-mmu: Page orders: linear mapping = 24, virtual = 16, io = 16, vmemmap = 24
[    0.000000] [      T0] hash-mmu: Using 1TB segments
[    0.000000] [      T0] hash-mmu: Initializing hash mmu with SLB
[    0.000000] [      T0] Linux version 7.1.0-rc2-00021-gf583bd5f64d4 (root@ltcppm1.ltc.tadn.ibm.com) (gcc (GCC) 16.1.1 20260501 (Red Hat 16.1.1-1), GNU ld version 2.46-1.fc44) #1 SMP PREEMPT Wed May  6 
00:55:45 EDT 2026
[    0.000000] [      T0] OF: reserved mem: 0x0000000039c00000..0x000000003b6801ff (27136 KiB) map non-reusable ibm,firmware-allocs-memory@39c00000
[    0.000000] [      T0] OF: reserved mem: 0x0000000800000000..0x0000000800e801ff (14848 KiB) map non-reusable ibm,firmware-allocs-memory@800000000
[    0.000000] [      T0] OF: reserved mem: 0x0000001000000000..0x0000001000dc01ff (14080 KiB) map non-reusable ibm,firmware-allocs-memory@1000000000
[    0.000000] [      T0] OF: reserved mem: 0x0000001800000000..0x0000001800e801ff (14848 KiB) map non-reusable ibm,firmware-allocs-memory@1800000000
[    0.000000] [      T0] OF: reserved mem: 0x0000000030000000..0x00000000302fffff (3072 KiB) map non-reusable ibm,firmware-code@30000000
[    0.000000] [      T0] OF: reserved mem: 0x0000000031000000..0x0000000031bfffff (12288 KiB) map non-reusable ibm,firmware-data@31000000
[    0.000000] [      T0] OF: reserved mem: 0x0000000030300000..0x0000000030ffffff (13312 KiB) map non-reusable ibm,firmware-heap@30300000
[    0.000000] [      T0] OF: reserved mem: 0x0000000031c00000..0x0000000033fdffff (36736 KiB) map non-reusable ibm,firmware-stacks@31c00000
[    0.000000] [      T0] OF: reserved mem: 0x0000001ffd510000..0x0000001ffd69ffff (1600 KiB) map non-reusable ibm,hbrt-code-image@1ffd510000
[    0.000000] [      T0] OF: reserved mem: 0x0000001ffd6a0000..0x0000001ffd6fffff (384 KiB) map non-reusable ibm,hbrt-target-image@1ffd6a0000
[    0.000000] [      T0] OF: reserved mem: 0x0000001ffd700000..0x0000001ffd7fffff (1024 KiB) map non-reusable ibm,hbrt-vpd-image@1ffd700000
[    0.000000] [      T0] OF: reserved mem: 0x0000001ffda00000..0x0000001ffdafffff (1024 KiB) map non-reusable ibm,slw-image@1ffda00000
[    0.000000] [      T0] OF: reserved mem: 0x0000001ffde00000..0x0000001ffdefffff (1024 KiB) map non-reusable ibm,slw-image@1ffde00000
[    0.000000] [      T0] OF: reserved mem: 0x0000001ffe200000..0x0000001ffe2fffff (1024 KiB) map non-reusable ibm,slw-image@1ffe200000
[    0.000000] [      T0] OF: reserved mem: 0x0000001ffe600000..0x0000001ffe6fffff (1024 KiB) map non-reusable ibm,slw-image@1ffe600000
[    0.000000] [      T0] Found initrd at 0xc000000006a40000:0xc00000000815ae9e
[    0.000000] [      T0] Hardware name: 8247-22L POWER8E (raw) 0x4b0201 opal:skiboot-v5.4.12 PowerNV
[    0.000000] [      T0] printk: legacy bootconsole [udbg0] enabled
[    0.000000] [      T0] CPU maps initialized for 8 threads per core
[    0.000000] [      T0]  (thread shift is 3)

quoted

quoted

But I my opal version 5.4.12.

Thanks for reporting the issue, will have an look at it.

I bisected it to a change between 5.19-rc3 and 5.19-rc4, and merge 
commit 8100775d59a6 (Merge tag 'powerpc-5.19-3' of git://git.kernel.org/ 
pub/scm/linux/kernel/git/powerpc/linux) [1] indeed has rng related changes

quoted

 - Three fixes to wire up our various RNGs earlier in boot so they're
   available for use in the initial seeding in random_init().

I confirmed, that commit f3eac426657d (powerpc/powernv: wire up rng 
during setup_arch) [2] introduced the Oops.

quoted

[    0.000000] [      T0] Allocated 4608 bytes for 160 pacas
[    0.000000] [      T0] 
-----------------------------------------------------

.......

[   37.407674] [    T900] audit: type=1130 audit(1778043621.931:10): pid=1 uid=0 auid=4294967295 ses=4294967295 msg='unit=lvm2-monitor comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? 
terminal=? res=success'
[   37.413015] [    T900] audit: type=1130 audit(1778043621.937:11): pid=1 uid=0 auid=4294967295 ses=4294967295 msg='unit=systemd-sysctl comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
[   38.448156] [   T2286] powernv_rng: Registered powernv hwrng.
[   38.575227] [   T2264] tg3 0005:09:00.1 enP5p9s0f1: renamed from eth1
[   38.582176] [   T2223] tg3 0005:09:00.2 enP5p9s0f2: renamed from eth2
........

////cpuinfo output

processor    : 159

cpu        : POWER8E (raw), altivec supported
clock        : 2061.000000MHz
revision    : 2.1 (pvr 004b 0201)

timebase    : 512000000
platform    : PowerNV
model        : 8247-22L
machine        : PowerNV 8247-22L
firmware    : OPAL
MMU        : Hash


But my system opal version 5.4.12.
Thanks for reporting the issue, will have an look at it.

Thank you.

Kind regards,

Paul

[1]: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=8100775d59a6789c3c6c309de26fac52f129cba8

[2]: 
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=f3eac426657d985b97c92fa5f7ae1d43f04721f3