[Bug 101325] UE4Editor crash after pressing "play" with radeon southern island card (7850 HD)

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=101325

--- Comment #21 from Matt <nonfatmatt at gmail.com> ---
Can confirm this bug on multiple cards / drivers / kernels / Mesa versions
starting with Unreal 4.15.

Radeon 7870
Radeon R7 360
Radeon R9 380 (Using amdgpu)

with anything greater than Mesa 17. Mesa 13 works fine. 

Here's a recent dmesg from a box running Fedora 26 / 4.11 / Mesa 17.07 and a
Radeon 7870.

[  215.746258] WARNING: CPU: 1 PID: 2239 at
drivers/gpu/drm/radeon/radeon_object.c:84 radeon_ttm_bo_destroy+0xf6/0x100
[radeon]
[  215.746258] Modules linked in: ip6t_rpfilter ip6t_REJECT nf_reject_ipv6
xt_conntrack ip_set nfnetlink ebtable_nat ebtable_broute bridge stp llc
ip6table_nat nf_conntrack_ipv6 nf_defrag_ipv6 nf_nat_ipv6 ip6table_mangle
ip6table_raw ip6table_security iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4
nf_nat_ipv4 nf_nat nf_conntrack libcrc32c iptable_mangle iptable_raw
iptable_security ebtable_filter ebtables ip6table_filter ip6_tables sunrpc
intel_rapl snd_hda_codec_realtek snd_hda_codec_generic x86_pkg_temp_thermal
snd_hda_codec_hdmi intel_powerclamp snd_hda_intel snd_hda_codec coretemp
mei_wdt iTCO_wdt iTCO_vendor_support ppdev kvm snd_hda_core snd_hwdep irqbypass
snd_seq intel_cstate snd_seq_device snd_pcm mei_me intel_uncore snd_timer xpad
intel_rapl_perf joydev snd ff_memless mei lpc_ich i2c_i801 soundcore
[  215.746278]  parport_pc tpm_infineon ie31200_edac shpchp parport edac_core
tpm_tis tpm_tis_core video tpm dm_crypt amdkfd amd_iommu_v2 radeon
crct10dif_pclmul i2c_algo_bit crc32_pclmul drm_kms_helper crc32c_intel ttm
ghash_clmulni_intel drm r8169 ata_generic pata_acpi uas mii usb_storage
[  215.746286] CPU: 1 PID: 2239 Comm: UdpMess-eceiver Tainted: G        W      
4.11.11-300.fc26.x86_64 #1
[  215.746286] Hardware name: Gigabyte Technology Co., Ltd. To be filled by
O.E.M./B75M-D3H, BIOS F14 04/17/2013
[  215.746287] Call Trace:
[  215.746288]  dump_stack+0x63/0x84
[  215.746289]  __warn+0xcb/0xf0
[  215.746290]  warn_slowpath_null+0x1d/0x20
[  215.746302]  radeon_ttm_bo_destroy+0xf6/0x100 [radeon]
[  215.746304]  ttm_bo_release_list+0xdb/0x200 [ttm]
[  215.746306]  ? ttm_bo_man_put_node+0x3f/0x50 [ttm]
[  215.746308]  ttm_bo_release+0x190/0x200 [ttm]
[  215.746310]  ttm_bo_unref+0x2c/0x30 [ttm]
[  215.746317]  radeon_bo_unref+0x39/0x70 [radeon]
[  215.746326]  radeon_gem_object_free+0x57/0x70 [radeon]
[  215.746330]  drm_gem_object_free+0x29/0x70 [drm]
[  215.746336]  drm_gem_object_unreference_unlocked+0x3a/0x70 [drm]
[  215.746341]  drm_gem_dmabuf_release+0x17/0x30 [drm]
[  215.746342]  dma_buf_release+0x58/0x1a0
[  215.746343]  __fput+0xdf/0x1e0
[  215.746345]  ____fput+0xe/0x10
[  215.746346]  task_work_run+0x76/0x90
[  215.746347]  do_exit+0x2e9/0xb90
[  215.746348]  do_group_exit+0x47/0xb0
[  215.746349]  get_signal+0x29f/0x640
[  215.746351]  do_signal+0x37/0x6a0
[  215.746352]  ? __check_object_size+0xbb/0x1b3
[  215.746354]  ? _copy_to_user+0x51/0x60
[  215.746355]  ? poll_select_copy_remaining+0xd6/0x140
[  215.746357]  exit_to_usermode_loop+0x76/0xb0
[  215.746358]  syscall_return_slowpath+0xa6/0xb0
[  215.746359]  entry_SYSCALL_64_fastpath+0xa7/0xa9
[  215.746360] RIP: 0033:0x7f0253686cd3
[  215.746360] RSP: 002b:00007f014c084970 EFLAGS: 00000293 ORIG_RAX:
0000000000000017
[  215.746361] RAX: fffffffffffffdfe RBX: 00007f01942e2d80 RCX:
00007f0253686cd3
[  215.746361] RDX: 0000000000000000 RSI: 00007f014c084998 RDI:
00000000000000dd
[  215.746362] RBP: 0000000000000000 R08: 00007f014c084988 R09:
0000000000000000
[  215.746362] R10: 0000000000000000 R11: 0000000000000293 R12:
00007f0193d8ec00
[  215.746363] R13: 00007ffdc4b94320 R14: 00000000000f4240 R15:
0000000000000000
[  215.746364] ---[ end trace 4a4848a1b1423a31 ]---

I'm unfamiliar with using apitrace, but if someone can give me proper reporting
instructions I can reproduce this bug on a wide range of hardware / software if
it would be helpful.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/dri-devel/attachments/20170808/0cf9500b/attachment.html>