Hi, I'm doing a partiy check and have run into the system locking up after 10hours x3 time now.

It always gives a genral protection fault.

 

The system becomes unresponsive and has to be manually reset. No response from IP.

 

Afraid no diagnostic file due to complete lock-up. Manual reset required.

 

Have read some similar thread about C states possibly causing this and fixed once disabled. 

 

Also wondered about memory but can't do a test because I boot in UEFI and the memest is BIOS only.

Unless I can partition the unraid USB key?

 

 

 

Jan  9 08:50:14 Moulin-Rouge kernel: general protection fault, probably for non-canonical address 0xff7e888107bcaaf8: 0000 [#1] PREEMPT SMP NOPTI
Jan  9 08:50:14 Moulin-Rouge kernel: CPU: 6 PID: 102 Comm: khugepaged Tainted: G           O      5.19.17-Unraid #2
Jan  9 08:50:14 Moulin-Rouge kernel: Hardware name: Micro-Star International Co., Ltd. MS-7D17/MAG B560M MORTAR WIFI (MS-7D17), BIOS 1.90 07/15/2022
Jan  9 08:50:14 Moulin-Rouge kernel: RIP: 0010:down_read_trylock+0xf/0x3b
Jan  9 08:50:14 Moulin-Rouge kernel: Code: 48 8b 14 25 c0 bb 01 00 83 e0 02 48 09 d0 48 83 c8 01 48 89 47 08 c3 cc cc cc cc 0f 1f 44 00 00 48 b9 07 00 00 00 00 00 00 80 <48> 8b 07 48 85 c8 75 1d 48 8d 90 00 01 00 00 f0 48 0f b1 17 75 ed
Jan  9 08:50:14 Moulin-Rouge kernel: RSP: 0018:ffffc90000487dc0 EFLAGS: 00010246
Jan  9 08:50:14 Moulin-Rouge kernel: RAX: ff7e888107bcaa80 RBX: ff7e888107bcaa80 RCX: 8000000000000007
Jan  9 08:50:14 Moulin-Rouge kernel: RDX: 0000000000000001 RSI: 0000000000000000 RDI: ff7e888107bcaaf8
Jan  9 08:50:14 Moulin-Rouge kernel: RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000
Jan  9 08:50:14 Moulin-Rouge kernel: R10: 0000000000000000 R11: ffff8881004000e8 R12: 0000000000000001
Jan  9 08:50:14 Moulin-Rouge kernel: R13: ffff88827ff1e000 R14: 0000000000000000 R15: 0000000000000002
Jan  9 08:50:14 Moulin-Rouge kernel: FS:  0000000000000000(0000) GS:ffff88885f580000(0000) knlGS:0000000000000000
Jan  9 08:50:14 Moulin-Rouge kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Jan  9 08:50:14 Moulin-Rouge kernel: CR2: 000014d0686b3a60 CR3: 000000017a8c8004 CR4: 0000000000770ee0
Jan  9 08:50:14 Moulin-Rouge kernel: PKRU: 55555554
Jan  9 08:50:14 Moulin-Rouge kernel: Call Trace:
Jan  9 08:50:14 Moulin-Rouge kernel: <TASK>
Jan  9 08:50:14 Moulin-Rouge kernel: khugepaged+0x2ad/0x179c
Jan  9 08:50:14 Moulin-Rouge kernel: ? _raw_spin_rq_lock_irqsave+0x20/0x20
Jan  9 08:50:14 Moulin-Rouge kernel: ? preempt_latency_start+0x2b/0x46
Jan  9 08:50:14 Moulin-Rouge kernel: ? collapse_pte_mapped_thp+0x275/0x275
Jan  9 08:50:14 Moulin-Rouge kernel: kthread+0xe4/0xef
Jan  9 08:50:14 Moulin-Rouge kernel: ? kthread_complete_and_exit+0x1b/0x1b
Jan  9 08:50:14 Moulin-Rouge kernel: ret_from_fork+0x1f/0x30
Jan  9 08:50:14 Moulin-Rouge kernel: </TASK>
Jan  9 08:50:14 Moulin-Rouge kernel: Modules linked in: xpad xt_connmark xt_comment iptable_raw xt_mark wireguard curve25519_x86_64 libcurve25519_generic libchacha20poly1305 chacha_x86_64 poly1305_x86_64 ip6_udp_tunnel udp_tunnel libchacha xt_CHECKSUM ipt_REJECT nf_reject_ipv4 ip6table_mangle ip6table_nat iptable_mangle vhost_net tun vhost vhost_iotlb tap veth xt_nat xt_tcpudp xt_conntrack xt_MASQUERADE nf_conntrack_netlink nfnetlink xfrm_user xfrm_algo xt_addrtype iptable_nat nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 br_netfilter xfs md_mod nct6687(O) hwmon_vid nct6683 efivarfs ip6table_filter ip6_tables iptable_filter ip_tables x_tables af_packet 8021q garp mrp bridge stp llc i915 iosf_mbi drm_buddy i2c_algo_bit ttm drm_display_helper drm_kms_helper wmi_bmof mxm_wmi x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm btusb crct10dif_pclmul crc32_pclmul crc32c_intel btrtl btbcm ghash_clmulni_intel aesni_intel drm btintel crypto_simd mpt3sas cryptd intel_cstate r8169 i2c_i801 intel_gtt
Jan  9 08:50:14 Moulin-Rouge kernel: raid_class intel_uncore i2c_smbus agpgart ahci realtek bluetooth scsi_transport_sas libahci i2c_core nvme syscopyarea sysfillrect input_leds ecdh_generic sysimgblt nvme_core led_class joydev ecc thermal fb_sys_fops fan video wmi backlight acpi_pad acpi_tad button unix
Jan  9 08:50:14 Moulin-Rouge kernel: ---[ end trace 0000000000000000 ]---
Jan  9 08:50:14 Moulin-Rouge kernel: RIP: 0010:down_read_trylock+0xf/0x3b
Jan  9 08:50:14 Moulin-Rouge kernel: Code: 48 8b 14 25 c0 bb 01 00 83 e0 02 48 09 d0 48 83 c8 01 48 89 47 08 c3 cc cc cc cc 0f 1f 44 00 00 48 b9 07 00 00 00 00 00 00 80 <48> 8b 07 48 85 c8 75 1d 48 8d 90 00 01 00 00 f0 48 0f b1 17 75 ed
Jan  9 08:50:14 Moulin-Rouge kernel: RSP: 0018:ffffc90000487dc0 EFLAGS: 00010246
Jan  9 08:50:14 Moulin-Rouge kernel: RAX: ff7e888107bcaa80 RBX: ff7e888107bcaa80 RCX: 8000000000000007
Jan  9 08:50:14 Moulin-Rouge kernel: RDX: 0000000000000001 RSI: 0000000000000000 RDI: ff7e888107bcaaf8
Jan  9 08:50:14 Moulin-Rouge kernel: RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000
Jan  9 08:50:14 Moulin-Rouge kernel: R10: 0000000000000000 R11: ffff8881004000e8 R12: 0000000000000001
Jan  9 08:50:14 Moulin-Rouge kernel: R13: ffff88827ff1e000 R14: 0000000000000000 R15: 0000000000000002
Jan  9 08:50:14 Moulin-Rouge kernel: FS:  0000000000000000(0000) GS:ffff88885f580000(0000) knlGS:0000000000000000
Jan  9 08:50:14 Moulin-Rouge kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Jan  9 08:50:14 Moulin-Rouge kernel: CR2: 000014d0686b3a60 CR3: 000000017a8c8004 CR4: 0000000000770ee0
Jan  9 08:50:14 Moulin-Rouge kernel: PKRU: 55555554

 

3 minutes ago, dopeytree said:

but can't do a test because I boot in UEFI and the memest is BIOS only.

Use the free Passmark memtest.

Still not quite sure what caused this.

 

Did notice possibly ram slightly undervolted while CPU overvolted or by default allowed to boost too much.

Ram is now overclocked to 3200Mhz too using automatic settings.

CPU is now limited in total power. Its sold as a 35w chip but is easily capable of 115w overclock.

Which is what MSI the motherboard has suggested by default.

This is great for gamers but not a server!!

 

Server has maintained itself overnight although with dockers turned off.

 

I can't update the motherboard's bios because fans seem to full full wack and black screen.

 

As a side question how do I tell the unraid linux kernal which CPU I have as I think this may be causing some issues.

 

I have an i9 11-900t but its an ES test version so isn't in the database so to speak. It loads all CPU INFO it's numbers correctly but I dont know what information Intels Pstate is expecting. and in any case the kernal has put it in ACPI CPUfreq driver instead of PSTATE Driver

Edited January 11, 20233 yr by dopeytree

I'm a dipstick.

 

Ok some progress. The bios updater seems to default to using the nvidia card rather than the build in igpu. No idea why as the main bios uses the igpu.

Any was able to update the bios hooray. This should help things.

 

Also changed a few more things in bios and unlocked c1e which might be useful. before it was showing as N/A for cpu features.

Again have limited the over chip power to 65w on short boosts with overall to 35w.

By default they give it 115w and although it's an i9 I want it to be efficient too.

 

Have also used unbalance to clean up my drives to a single drive (with parity backup) ready for zfs.

Possibly I should make this a cache pool drive & mirror it for speedier backup this would avoid the need for unraid to correct parity when I remove the other empty drives.

Edited January 11, 20233 yr by dopeytree

Best if you don't attempt overclock or other tweaks with your server.

Any advice on writting a memtest86+ to disk with mac. 

The readme.txt is terrible.

 

Followed a video but something has changed since his tutorial.

Might try on the windows VM.

 

As soon as I find a nice linux laptop I'm taking it over this mac. 10hours battery is impressive but not at the cost of not being able to do basic tasks.

Probably go AMD as they are working directly on better battery efficiency in linux because of their work on the steam-deck & its custom APU chip.

 

Normally I'd agree but tweaking has enabled the idle to work properly. See before / after

 

 

 

 

After CPU now idles properly.

 

 

 

 

 

Edited January 11, 20233 yr by dopeytree