You are not logged in.

#1 2017-05-20 07:02:46

albinoloverats
Member
From: London
Registered: 2007-12-18
Posts: 29
Website

Radeon crash causing X/lightdm/Xfce restart

Sometime last night my Radeon GPU started playing up; everything seemed to continue to work but the GPU load was at 100% (as indicated by radeontop). This was causing X to restart every now and then. My machine had been up for about a week or so before becoming unstable, (and is running fine again now following a reboot) so unless I wait another week to see if it happens again, so I guess it's likely some resource leak that may or may not happen again.

Details:

  1. CPU: AMD FX(tm)-9590 Eight-Core Processor

  2. GPU: AMD/ATI Cape Verde XT [Radeon HD 7770/8760 / R7 250X

  3. MB: ASUS SABERTOOTH 990FX R2.0 (DDR3, USB3.0, 6Gb/s, CrossFireX/Sli)

  4. Kernel: Linux insomniac 4.10.13-1-ARCH #1 SMP PREEMPT Thu Apr 27 12:15:09 CEST 2017 x86_64 GNU/Linux

  5. ATI driver: xf86-video-ati 1:7.9.0-1

  6. Log file of the event

Anything else necessary/useful to know?

So if anybody can shed any light on this that would be great, thank you.

Offline

#2 2017-05-21 20:48:33

R00KIE
Forum Fellow
From: Between a computer and a chair
Registered: 2008-09-14
Posts: 4,734

Re: Radeon crash causing X/lightdm/Xfce restart

You say it's working fine now, so unless you find a way to reproduce the problem we can't help.


R00KIE
Tm90aGluZyB0byBzZWUgaGVyZSwgbW92ZSBhbG9uZy4K

Offline

#3 2017-05-21 21:54:38

loqs
Member
Registered: 2014-03-06
Posts: 17,196

Re: Radeon crash causing X/lightdm/Xfce restart

Are the kernel oops occuring after the reboot?

May 19 23:31:40 insomniac kernel: ------------[ cut here ]------------
May 19 23:31:40 insomniac kernel: WARNING: CPU: 3 PID: 12742 at drivers/gpu/drm/radeon/radeon_object.c:84 radeon_ttm_bo_destroy+0xea/0xf0 [radeon]
May 19 23:31:40 insomniac kernel: Modules linked in: joydev hid_sony blowfish_generic blowfish_x86_64 blowfish_common des3_ede_x86_64 des_generic cast5_avx_x86_64 cast5_generic cast_common cbc twofish_generic twofish_avx_x86_64 twofish_x86_64_3way twofish_x86_64 twofish_common msr nls_iso8859_1 nls_cp437 vfat fat cpufreq_conservative cpufreq_powersave tun amdgpu fuse xt_helper xt_tcpudp xt_multiport xt_set iptable_filter iptable_nat nf_nat_ipv4 nf_nat nf_conntrack_ipv4 nf_defrag_ipv4 xt_connmark xt_conntrack nf_conntrack iptable_mangle ip_set_hash_ip ip_set nfnetlink xfs usblp serpent_avx_x86_64 serpent_sse2_x86_64 serpent_generic lrw ablk_helper snd_usb_audio snd_usbmidi_lib snd_rawmidi snd_seq_device eeepc_wmi asus_wmi sparse_keymap mxm_wmi video edac_mce_amd edac_core kvm_amd kvm irqbypass evdev amdkfd input_leds mac_hid
May 19 23:31:40 insomniac kernel:  gspca_stv06xx gspca_main v4l2_common videodev amd_iommu_v2 snd_hda_codec_realtek snd_hda_codec_hdmi media snd_hda_codec_generic radeon fam15h_power k10temp sp5100_tco snd_hda_intel ttm r8169 snd_hda_codec drm_kms_helper mii snd_hda_core i2c_piix4 snd_hwdep drm syscopyarea snd_pcm sysfillrect sysimgblt snd_timer fb_sys_fops snd soundcore i2c_algo_bit tpm_infineon shpchp wmi button fjes acpi_cpufreq tpm_tis tpm_tis_core sch_fq_codel b43 bcma mac80211 cfg80211 ssb mmc_core rfkill led_class pcmcia pcmcia_core pci_stub vboxpci(O) vboxnetflt(O) vboxnetadp(O) vboxdrv(O) tpm_rng rng_core tpm it87 hwmon_vid sg ip_tables x_tables algif_skcipher af_alg dm_crypt sr_mod cdrom sd_mod hid_logitech ff_memless usbhid hid uas usb_storage ohci_pci crct10dif_pclmul crc32_pclmul ghash_clmulni_intel pcbc
May 19 23:31:40 insomniac kernel:  aesni_intel aes_x86_64 crypto_simd glue_helper cryptd xhci_pci ahci sata_sil24 libahci ehci_pci xhci_hcd ehci_hcd ohci_hcd libata usbcore usb_common scsi_mod serio dm_mod btrfs ext4 crc16 jbd2 fscrypto mbcache reiserfs xts gf128mul tgr192 raid456 libcrc32c crc32c_generic crc32c_intel async_raid6_recov async_memcpy async_pq async_xor async_tx xor raid6_pq raid1 md_mod
May 19 23:31:40 insomniac kernel: CPU: 3 PID: 12742 Comm: TaskSchedulerSe Tainted: G        W  O    4.10.13-1-ARCH #1
May 19 23:31:40 insomniac kernel: Hardware name: To be filled by O.E.M. To be filled by O.E.M./SABERTOOTH 990FX R2.0, BIOS 2501 04/08/2014
May 19 23:31:40 insomniac kernel: Call Trace:
May 19 23:31:40 insomniac kernel:  dump_stack+0x63/0x83
May 19 23:31:40 insomniac kernel:  __warn+0xcb/0xf0
May 19 23:31:40 insomniac kernel:  warn_slowpath_null+0x1d/0x20
May 19 23:31:40 insomniac kernel:  radeon_ttm_bo_destroy+0xea/0xf0 [radeon]
May 19 23:31:40 insomniac kernel:  ttm_bo_release_list+0xcb/0x220 [ttm]
May 19 23:31:40 insomniac kernel:  ? ttm_bo_man_put_node+0x40/0x50 [ttm]
May 19 23:31:40 insomniac kernel:  ttm_bo_unref+0x1ae/0x220 [ttm]
May 19 23:31:40 insomniac kernel:  radeon_bo_unref+0x39/0x70 [radeon]
May 19 23:31:40 insomniac kernel:  radeon_gem_object_free+0x57/0x70 [radeon]
May 19 23:31:40 insomniac kernel:  drm_gem_object_free+0x29/0x70 [drm]
May 19 23:31:40 insomniac kernel:  drm_gem_object_unreference_unlocked+0x3a/0x90 [drm]
May 19 23:31:40 insomniac kernel:  drm_gem_object_handle_unreference_unlocked+0x65/0xb0 [drm]
May 19 23:31:40 insomniac kernel:  drm_gem_object_release_handle+0x53/0x90 [drm]
May 19 23:31:40 insomniac kernel:  idr_for_each+0xb0/0x110
May 19 23:31:40 insomniac kernel:  ? drm_gem_object_init+0x40/0x40 [drm]
May 19 23:31:40 insomniac kernel:  drm_gem_release+0x20/0x30 [drm]
May 19 23:31:40 insomniac kernel:  drm_release+0x316/0x370 [drm]
May 19 23:31:40 insomniac kernel:  __fput+0xa2/0x1f0
May 19 23:31:40 insomniac kernel:  ____fput+0xe/0x10
May 19 23:31:40 insomniac kernel:  task_work_run+0x80/0xa0
May 19 23:31:40 insomniac kernel:  do_exit+0x2b9/0xb40
May 19 23:31:40 insomniac kernel:  ? pick_next_task_fair+0x451/0x4c0
May 19 23:31:40 insomniac kernel:  do_group_exit+0x3b/0xb0
May 19 23:31:40 insomniac kernel:  get_signal+0x268/0x640
May 19 23:31:40 insomniac kernel:  do_signal+0x37/0x690
May 19 23:31:40 insomniac kernel:  ? kmem_cache_alloc+0xdb/0x1b0
May 19 23:31:40 insomniac kernel:  ? kmem_cache_alloc+0xdb/0x1b0
May 19 23:31:40 insomniac kernel:  ? add_wait_queue+0x44/0x50
May 19 23:31:40 insomniac kernel:  ? ep_poll+0x346/0x350
May 19 23:31:40 insomniac kernel:  ? wake_up_q+0x80/0x80
May 19 23:31:40 insomniac kernel:  exit_to_usermode_loop+0x8c/0xb0
May 19 23:31:40 insomniac kernel:  syscall_return_slowpath+0x59/0x60
May 19 23:31:40 insomniac kernel:  entry_SYSCALL_64_fastpath+0xa7/0xa9
May 19 23:31:40 insomniac kernel: RIP: 0033:0x7f5d9da97aa3
May 19 23:31:40 insomniac kernel: RSP: 002b:00007f5d8ca97a00 EFLAGS: 00000293 ORIG_RAX: 00000000000000e8
May 19 23:31:40 insomniac kernel: RAX: fffffffffffffffc RBX: 00000e35889ff000 RCX: 00007f5d9da97aa3
May 19 23:31:40 insomniac kernel: RDX: 0000000000000020 RSI: 00000e35889ff000 RDI: 0000000000000017
May 19 23:31:40 insomniac kernel: RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000
May 19 23:31:40 insomniac kernel: R10: 00000000ffffffff R11: 0000000000000293 R12: 00000e35885f15f8
May 19 23:31:40 insomniac kernel: R13: 00000e35885f1618 R14: 00000e3588b061e0 R15: 00000e35885f1000
May 19 23:31:40 insomniac kernel: ---[ end trace 7f022e51da968477 ]---
May 19 23:31:40 insomniac kernel: ------------[ cut here ]------------

Also a significant number of kernel messages were lost ( lots of entries similar to )

May 19 23:31:39 insomniac systemd-journald[401]: Missed 23 kernel messages

Are you aware that 4.10.13 as well as the 4.10 series itself is classed as end of life by upstream so if you you report the issue upstream you may be required to recreate the issue on a kernel upstream still supports?

Offline

#4 2017-05-22 07:28:09

albinoloverats
Member
From: London
Registered: 2007-12-18
Posts: 29
Website

Re: Radeon crash causing X/lightdm/Xfce restart

R00KIE wrote:

You say it's working fine now, so unless you find a way to reproduce the problem we can't help.

Yes, unfortunately I'm not exactly sure how to recreate the problem as it happened while I was away from my computer, and hadn't left anything running that hadn't already been running for several days. This is why I wondered whether it was some sort of resource leak.

loqs wrote:

Also a significant number of kernel messages were lost ( lots of entries similar to )

I saw that too - is that likely to be because the log was flooded with entries from the kernel and radeon driver?

I was rather hoping that someone else might have had a similar problem and found a solution (as is often the case, which is nice). And hopefully if it does happen again I might actually be sat at my desk and able to identify a particular task that caused the problem.

Thanks

Offline

Board footer

Powered by FluxBB