16.04 user. Since a few months ago (2017 Q1), not sure when, i'm having inestability issues with my machine.
The system is an old i7 920 on a Gigabyte EX58-UD5 motherboard. Currently with 24GB of ram, and a single GTX980 card. I'm using the latest nvidia drivers from the "graphic-drivers" ppa and have hwe stack installed (4.8 kernel and newer Xorg), but this problem is from the 4.4 era (before the ".2" LTS revision) and a few versions back of nvidia driver. The system was stable since 2009 and the current setup (added current gen nvidia card and an extra SATA disk) was from 2015 Q2; nothing changed in 2017. The power supply is a 1000W thermaltake, and have no heat issues.
The problem is: suddenly the screen freezes, and sometimes (not always) the mouse still moves but slow, like jumping from coordinates. Can't also change tty with the function keys combination. But I've been able to ssh to the system from my netbook while in this state; the system is still up, not really frozen.
I guess it may be useful to note that almost every time (about 98%, but not exactly always), it happens while scrolling on firefox. In fact, at first i was blaming some 5X Firefox version that enabled multiprocessing for tabs. It was not the case, as i eventually find out; about a week ago started to (try to) debug this problem, and i saw this messages:
Mar 27 22:17:39 Buda kernel: [ 1841.918831] irq 16: nobody cared (try booting with the "irqpoll" option)
Mar 27 22:17:39 Buda kernel: [ 1841.918836] CPU: 3 PID: 0 Comm: swapper/3 Tainted: P IOE 4.8.0-41-generic #44~16.04.1-Ubuntu
Mar 27 22:17:39 Buda kernel: [ 1841.918837] Hardware name: Gigabyte Technology Co., Ltd. EX58-UD5/EX58-UD5, BIOS F13 01/10/2012
Mar 27 22:17:39 Buda kernel: [ 1841.918838] 0000000000000086 7b4c75c5fc78db31 ffff8d36d72c3e80 ffffffffaa22e043
Mar 27 22:17:39 Buda kernel: [ 1841.918840] ffff8d36d24e8800 ffff8d36d24e88d4 ffff8d36d72c3ea8 ffffffffa9ee2e23
Mar 27 22:17:39 Buda kernel: [ 1841.918842] ffff8d36d24e8800 0000000000000000 0000000000000010 ffff8d36d72c3ee0
Mar 27 22:17:39 Buda kernel: [ 1841.918844] Call Trace:
Mar 27 22:17:39 Buda kernel: [ 1841.918844] <IRQ> [<ffffffffaa22e043>] dump_stack+0x63/0x90
Mar 27 22:17:39 Buda kernel: [ 1841.918851] [<ffffffffa9ee2e23>] __report_bad_irq+0x33/0xc0
Mar 27 22:17:39 Buda kernel: [ 1841.918852] [<ffffffffa9ee31b7>] note_interrupt+0x247/0x290
Mar 27 22:17:39 Buda kernel: [ 1841.918855] [<ffffffffa9ee0174>] handle_irq_event_percpu+0x54/0x80
Mar 27 22:17:39 Buda kernel: [ 1841.918856] [<ffffffffa9ee01de>] handle_irq_event+0x3e/0x60
Mar 27 22:17:39 Buda kernel: [ 1841.918857] [<ffffffffa9ee38b2>] handle_fasteoi_irq+0xa2/0x160
Mar 27 22:17:39 Buda kernel: [ 1841.918859] [<ffffffffa9e302cd>] handle_irq+0x1d/0x30
Mar 27 22:17:39 Buda kernel: [ 1841.918862] [<ffffffffaa69e05b>] do_IRQ+0x4b/0xd0
Mar 27 22:17:39 Buda kernel: [ 1841.918863] [<ffffffffaa69c142>] common_interrupt+0x82/0x82
Mar 27 22:17:39 Buda kernel: [ 1841.918863] <EOI> [<ffffffffaa5198f8>] ? cpuidle_enter_state+0x128/0x2d0
Mar 27 22:17:39 Buda kernel: [ 1841.918869] [<ffffffffaa519ad7>] cpuidle_enter+0x17/0x20
Mar 27 22:17:39 Buda kernel: [ 1841.918870] [<ffffffffa9ec79fa>] call_cpuidle+0x2a/0x50
Mar 27 22:17:39 Buda kernel: [ 1841.918871] [<ffffffffa9ec7dde>] cpu_startup_entry+0x29e/0x350
Mar 27 22:17:39 Buda kernel: [ 1841.918874] [<ffffffffa9e518b1>] start_secondary+0x151/0x190
Mar 27 22:17:39 Buda kernel: [ 1841.918875] handlers:
Mar 27 22:17:39 Buda kernel: [ 1841.918878] [<ffffffffaa465fd0>] usb_hcd_irq
Mar 27 22:17:39 Buda kernel: [ 1841.918879] Disabling IRQ #16
The thing is about IRQ 16 being disabled. I've cheched what could be using IRQ 16:
:~$ lspci -v | grep IRQ
Flags: fast devsel, IRQ 12
Flags: bus master, fast devsel, latency 0, IRQ 16
Flags: bus master, medium devsel, latency 0, IRQ 16
Flags: bus master, medium devsel, latency 0, IRQ 21
Flags: bus master, medium devsel, latency 0, IRQ 18
Flags: bus master, medium devsel, latency 0, IRQ 18
Flags: bus master, fast devsel, latency 0, IRQ 30
Flags: bus master, fast devsel, latency 0, IRQ 24
Flags: bus master, fast devsel, latency 0, IRQ 25
Flags: bus master, fast devsel, latency 0, IRQ 26
Flags: bus master, fast devsel, latency 0, IRQ 27
Flags: bus master, medium devsel, latency 0, IRQ 23
Flags: bus master, medium devsel, latency 0, IRQ 19
Flags: bus master, medium devsel, latency 0, IRQ 18
Flags: bus master, medium devsel, latency 0, IRQ 23
Flags: bus master, 66MHz, medium devsel, latency 0, IRQ 19
Flags: medium devsel, IRQ 7
Flags: bus master, 66MHz, medium devsel, latency 0, IRQ 19
Flags: bus master, fast devsel, latency 0, IRQ 31
Flags: fast devsel, IRQ 17
Flags: bus master, fast devsel, latency 0, IRQ 17
Flags: bus master, fast devsel, latency 0, IRQ 18
Flags: bus master, fast devsel, latency 0, IRQ 28
Flags: bus master, fast devsel, latency 0, IRQ 29
Flags: bus master, medium devsel, latency 32, IRQ 18
And the detail of the two devices using IRQ 16 is this:
00:03.0 PCI bridge: Intel Corporation 5520/5500/X58 I/O Hub PCI Express Root Port 3 (rev 12) (prog-if 00 [Normal decode])
Flags: bus master, fast devsel, latency 0, IRQ 16
Bus: primary=00, secondary=01, subordinate=01, sec-latency=0
I/O behind bridge: 0000b000-0000bfff
Memory behind bridge: f9000000-faffffff
Prefetchable memory behind bridge: 00000000d0000000-00000000efffffff
Capabilities: <access denied>
Kernel driver in use: pcieport
Kernel modules: shpchp
00:1a.0 USB controller: Intel Corporation 82801JI (ICH10 Family) USB UHCI Controller #4 (prog-if 00 [UHCI])
Subsystem: Gigabyte Technology Co., Ltd Motherboard
Flags: bus master, medium devsel, latency 0, IRQ 16
I/O ports at ff00 [size=32]
Capabilities: <access denied>
Kernel driver in use: uhci_hcd
So... i've tried to solve that IRQ "conflict" by changing BIOS settings (didn't found IRQ assigment options) or blacklisting kernel modules ("shpchp", as i'm using both the USB and PCIE controllers) without success.
Then i started to try kernel boot parameters. The first try was using the "irqpoll" option the error suggested. It did not solve anything, and also started to see this other messages:
Mar 26 14:17:23 Buda kernel: [ 1988.976483] hpet1: lost 9599 rtc interrupts
Mar 26 14:18:51 Buda kernel: [ 2076.378021] hpet1: lost 9600 rtc interrupts
Mar 26 14:20:11 Buda kernel: [ 2156.670873] hpet1: lost 9600 rtc interrupts
Mar 26 14:21:33 Buda kernel: [ 2238.533442] hpet1: lost 9599 rtc interrupts
Mar 26 14:22:17 Buda kernel: [ 2282.293272] hpet1: lost 9600 rtc interrupts
Mar 26 14:23:06 Buda kernel: [ 2331.292559] hpet1: lost 9600 rtc interrupts
Mar 26 14:24:42 Buda kernel: [ 2427.306337] hpet1: lost 9600 rtc interrupts
Googling around, i found and tried several other boot options, all without success. Right now I'm using the system with "acpi=off" and it's the first option that "resolves" the IRQ "conflict", as the only device using IRQ16 is the USB controller; but I've lost 4 CPU cores by doing this (I only see 4 out of 8 right now), and don't know what else I'm also disabling (enabled "acpi=off" just a few hours ago).
I blamed firefox, then the nvidia drivers, then waited for a new kernel+xorg stack... nothing helped. I can work for hours with video processing or big data tasks or can even play AAA games on this system without issues, and then go to see some youtuve vid or reading the news and get a system freeze without any cpu or heat problem. It can also happens (and last weeks is happening a lot) just after booting up, when i start thunderbird and firefox. No app error messages in system log, no crashes, just this sudden IRQ stuff.
I'm here asking for help diagnosing this. I do not fully understand all the data I'm gathering, and where else I need to look for. Is there anyone with experience on problems of this kind? What else can i do to debug this?
Thanks in advance.
