From mboxrd@z Thu Jan 1 00:00:00 1970
From: bugzilla-daemon@freedesktop.org
Subject: [Bug 93301] ns2_linux32: radeon VM fault /
VM_CONTEXT1_PROTECTION_FAULT_ADDR
Date: Wed, 09 Dec 2015 04:10:26 +0000
Message-ID:
Bug ID
93301
Summary
ns2_linux32: radeon VM fault / VM_CONTEXT1_PROTECTION_FAULT_ADDR
Product
Mesa
Version
git
Hardware
Other
OS
All
Status
NEW
Severity
normal
Priority
medium
Component
Drivers/Gallium/radeonsi
Assignee
dri-devel@lists.freedesktop.org
Reporter
dev@illwieckz.net
QA Contact
dri-devel@lists.freedesktop.org
Created attachment 120427 [details]
syslog (Linux 4.3, mesa git, radeon hawaii r9 390X)
Hi I (re) tried Natural Selection 2 yesterday, and I it crashed (as expected),
and this time the syslog said:
```
Dec 8 07:20:25 gollum kernel: [16588.241364] radeon 0000:01:00.0: GPU fault
detected: 146 0x0e65e014
Dec 8 07:20:25 gollum kernel: [16588.241368] radeon 0000:01:00.0:
VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x00007673
Dec 8 07:20:25 gollum kernel: [16588.241369] radeon 0000:01:00.0:
VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x050E0014
Dec 8 07:20:25 gollum kernel: [16588.241370] VM fault (0x04, vmid 2) at page
30323, write from 'CB1' (0x43423100) (224)
---8<--- 120 000+ lines ---------------------
Dec 8 07:22:58 gollum kernel: [16741.758495] radeon 0000:01:00.0: GPU fault
detected: 146 0x06a59014
Dec 8 07:22:58 gollum kernel: [16741.758506] radeon 0000:01:00.0:
VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x00007673
Dec 8 07:22:58 gollum kernel: [16741.758512] radeon 0000:01:00.0:
VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x05150014
Dec 8 07:22:58 gollum kernel: [16741.758518] VM fault (0x04, vmid 2) at page
30323, write from 'CB6' (0x43423600) (336)
Dec 8 07:23:03 gollum kernel: [16746.694729] ns2_linux32[16192]: segfault at 1
ip 00000000f3bf8d12 sp 00000000f0b9adf8 error 4 in
libc-2.21.so[f3acc000+1b4000]
```
Well, I'm not expecting to see this game working one day, I own this game since
2013-12-15 (it's almost two years) and it never worked, so these guys proved
their product is a joke and I don't expect nothing from them. But perhaps this
bug can help you to track things on your side, that's why I report it.
When the error happened, NS2 was loading assets since 20min to load a tutorial
map.
I don't have more logs currently but if you have some ideas of things I can
record, I can try to reproduce it. I agree the current log is a bit quiet.
I've already seen radeon errors like that before in a dpm issue (see #93288 )
but the current bug is unrelated to dpm (dpm was inactive in the current case).
I use a Radeon R9 390X with Mesa git on Linux 4.3 here.