[RESOLVED] Anyone got AMD RX 5700 (XT) to work?

sirn · 2019-08-17 08:17:54

Mainline v5.3-rc4 now has amdgpu driver up to drm-next-5.4 branch (plus one commit ahead of drm-next-5.4 as of this writing; newer commits are mostly navi12 support). It also has SMU fan control merged and is compatible with latest 2019-08-08 firmware. You might as well use linux-mainline along with mesa-git instead of linux-drm-next for RX 5700.

ledbettj · 2019-08-20 14:05:15

I'm running linux-mainline 5.3-rc4 along with mesa/llvm from git as of last night and firmware from https://people.freedesktop.org/~agd5f/r … de/navi10/ ; desktop functionality seems to work great, but launching any OpenGL or Vulkan software causes the system to slow down and quickly become unresponsive.

Is anyone else seeing issues like this or have suggestions? Or I just need to wait for the drivers to shake out a bit?

ineedmana · 2019-08-20 19:33:26

I have no issues here, OpenGL and Vulkan work great.

linux-mainline 5.3rc4-1
llvm-git 10.0.0_r324113.2346b92f2bf-1 from mesa-git repo
mesa-git 1:19.2.0_devel.114399.0c563303619-1 from mesa-git repo
firmware last modified at 2019-08-08. https://people.freedesktop.org/~agd5f/r … de/navi10/ has version from 2019-08-19 right now, so check that first I think

ledbettj · 2019-08-20 22:08:44

ineedmana wrote:

I have no issues here, OpenGL and Vulkan work great.
linux-mainline 5.3rc4-1
llvm-git 10.0.0_r324113.2346b92f2bf-1 from mesa-git repo
mesa-git 1:19.2.0_devel.114399.0c563303619-1 from mesa-git repo
firmware last modified at 2019-08-08. https://people.freedesktop.org/~agd5f/r … de/navi10/ has version from 2019-08-19 right now, so check that first I think

Thanks, I wasn't able to find the 8/8 firmware, but I found the previous July 27 firmware and still encounter the same problem:

mesa-git 1:19.2.0_devel.114550.6a09405368a-1
llvm-git 10.0.0_r324692.12cbbab9d93-1
linux-mainline 5.3rc4-1
lib32-vulkan-radeon-git 1:19.2.0_devel.114550.6a09405368a-1
vulkan-radeon-git 1:19.2.0_devel.114550.6a09405368a-1
lib32-mesa-git 1:19.2.0_devel.114550.6a09405368a-1

I also found the following in dmesg:

[  267.068401] ------------[ cut here ]------------
[  267.068545] WARNING: CPU: 8 PID: 395 at drivers/gpu/drm/amd/amdgpu/../display/dc/dcn20/dcn20_hwseq.c:1942 dcn20_setup_gsl_group_as_lock.cold+0x1f/0x4e [amdgpu]
[  267.068546] Modules linked in: fuse ccm xt_MASQUERADE nf_conntrack_netlink nfnetlink xfrm_user xfrm_algo xt_addrtype iptable_filter xt_conntrack br_netfilter bridge stp llc xt_REDIRECT xt_tcpudp iptable_nat nf_nat nf_conntrack nct6775 nf_defrag_ipv6 nf_defrag_ipv4 hwmon_vid nls_iso8859_1 ath9k nls_cp437 ath9k_common vfat fat input_leds ath9k_hw snd_hda_codec_realtek snd_hda_codec_generic ath snd_hda_codec_hdmi ledtrig_audio mac80211 edac_mce_amd snd_usb_audio uvcvideo snd_hda_intel videobuf2_vmalloc videobuf2_memops snd_hda_codec videobuf2_v4l2 ppdev snd_usbmidi_lib videobuf2_common snd_hda_core cfg80211 snd_rawmidi snd_hwdep wmi_bmof snd_seq_device mousedev videodev uas parport_pc r8169 snd_pcm kvm snd_timer joydev rfkill sp5100_tco realtek mc snd i2c_piix4 k10temp pcspkr irqbypass soundcore libphy libarc4 parport gpio_amdpt wmi pinctrl_amd evdev acpi_cpufreq mac_hid sg crypto_user ip_tables x_tables btrfs libcrc32c crc32c_generic xor raid6_pq dm_crypt dm_mod sd_mod hid_generic usbhid hid
[  267.068595]  usb_storage crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel aesni_intel ahci libahci aes_x86_64 crypto_simd cryptd libata xhci_pci glue_helper ccp xhci_hcd rng_core scsi_mod amdgpu amd_iommu_v2 gpu_sched i2c_algo_bit ttm drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops drm agpgart
[  267.068617] CPU: 8 PID: 395 Comm: kworker/u64:7 Not tainted 5.3.0-rc4-mainline #1
[  267.068618] Hardware name: Micro-Star International Co., Ltd. MS-7A34/B350 GAMING PLUS (MS-7A34), BIOS M.HM 07/08/2019
[  267.068630] Workqueue: events_unbound commit_work [drm_kms_helper]
[  267.068707] RIP: 0010:dcn20_setup_gsl_group_as_lock.cold+0x1f/0x4e [amdgpu]
[  267.068710] Code: d8 8b ca e6 0f 0b e9 c5 93 f6 ff 48 c7 c7 f8 1e 71 c0 e8 c5 8b ca e6 0f 0b e9 d2 9e f6 ff 48 c7 c7 f8 1e 71 c0 e8 b2 8b ca e6 <0f> 0b c6 83 60 02 00 00 00 48 c7 c7 f8 1e 71 c0 e8 9d 8b ca e6 0f
[  267.068712] RSP: 0018:ffffac040173fa48 EFLAGS: 00010246
[  267.068715] RAX: 0000000000000024 RBX: ffff9dc610d401b8 RCX: 0000000000000000
[  267.068716] RDX: 0000000000000000 RSI: ffff9dc7ce817708 RDI: 00000000ffffffff
[  267.068718] RBP: 0000000000000001 R08: 000000000000051d R09: 0000000000000004
[  267.068719] R10: 0000000000000000 R11: 0000000000000001 R12: ffff9dc7c21b0000
[  267.068720] R13: ffff9dc610d41d18 R14: ffff9dc610d401b8 R15: 0000000000000006
[  267.068722] FS:  0000000000000000(0000) GS:ffff9dc7ce800000(0000) knlGS:0000000000000000
[  267.068723] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  267.068724] CR2: 00000000ea3b5000 CR3: 0000000250bf6000 CR4: 00000000003406e0
[  267.068725] Call Trace:
[  267.068794]  dcn20_pipe_control_lock.part.0+0xe4/0x1b0 [amdgpu]
[  267.068861]  dc_commit_updates_for_stream+0xec8/0x1390 [amdgpu]
[  267.068867]  ? _raw_spin_lock+0x13/0x30
[  267.068945]  amdgpu_dm_atomic_commit_tail+0x12a6/0x1d00 [amdgpu]
[  267.068961]  ? commit_tail+0x3c/0x70 [drm_kms_helper]
[  267.068969]  commit_tail+0x3c/0x70 [drm_kms_helper]
[  267.068974]  process_one_work+0x1d1/0x3a0
[  267.068976]  worker_thread+0x4a/0x3d0
[  267.068979]  kthread+0xfb/0x130
[  267.068981]  ? process_one_work+0x3a0/0x3a0
[  267.068982]  ? kthread_park+0x80/0x80
[  267.068983]  ret_from_fork+0x1f/0x40
[  267.068987] ---[ end trace 105f1c3037172655 ]---
[  267.068992] ------------[ cut here ]------------
[  267.069068] WARNING: CPU: 8 PID: 395 at drivers/gpu/drm/amd/amdgpu/../display/dc/dcn20/dcn20_hwseq.c:1960 dcn20_setup_gsl_group_as_lock.cold+0x34/0x4e [amdgpu]
[  267.069069] Modules linked in: fuse ccm xt_MASQUERADE nf_conntrack_netlink nfnetlink xfrm_user xfrm_algo xt_addrtype iptable_filter xt_conntrack br_netfilter bridge stp llc xt_REDIRECT xt_tcpudp iptable_nat nf_nat nf_conntrack nct6775 nf_defrag_ipv6 nf_defrag_ipv4 hwmon_vid nls_iso8859_1 ath9k nls_cp437 ath9k_common vfat fat input_leds ath9k_hw snd_hda_codec_realtek snd_hda_codec_generic ath snd_hda_codec_hdmi ledtrig_audio mac80211 edac_mce_amd snd_usb_audio uvcvideo snd_hda_intel videobuf2_vmalloc videobuf2_memops snd_hda_codec videobuf2_v4l2 ppdev snd_usbmidi_lib videobuf2_common snd_hda_core cfg80211 snd_rawmidi snd_hwdep wmi_bmof snd_seq_device mousedev videodev uas parport_pc r8169 snd_pcm kvm snd_timer joydev rfkill sp5100_tco realtek mc snd i2c_piix4 k10temp pcspkr irqbypass soundcore libphy libarc4 parport gpio_amdpt wmi pinctrl_amd evdev acpi_cpufreq mac_hid sg crypto_user ip_tables x_tables btrfs libcrc32c crc32c_generic xor raid6_pq dm_crypt dm_mod sd_mod hid_generic usbhid hid
[  267.069100]  usb_storage crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel aesni_intel ahci libahci aes_x86_64 crypto_simd cryptd libata xhci_pci glue_helper ccp xhci_hcd rng_core scsi_mod amdgpu amd_iommu_v2 gpu_sched i2c_algo_bit ttm drm_kms_helper syscopyarea sysfillrect sysimgblt fb_sys_fops drm agpgart
[  267.069114] CPU: 8 PID: 395 Comm: kworker/u64:7 Tainted: G        W         5.3.0-rc4-mainline #1
[  267.069115] Hardware name: Micro-Star International Co., Ltd. MS-7A34/B350 GAMING PLUS (MS-7A34), BIOS M.HM 07/08/2019
[  267.069123] Workqueue: events_unbound commit_work [drm_kms_helper]
[  267.069182] RIP: 0010:dcn20_setup_gsl_group_as_lock.cold+0x34/0x4e [amdgpu]
[  267.069184] Code: ca e6 0f 0b e9 d2 9e f6 ff 48 c7 c7 f8 1e 71 c0 e8 b2 8b ca e6 0f 0b c6 83 60 02 00 00 00 48 c7 c7 f8 1e 71 c0 e8 9d 8b ca e6 <0f> 0b e9 aa 9e f6 ff 48 c7 c7 f8 1e 71 c0 e8 8a 8b ca e6 0f 0b e9
[  267.069185] RSP: 0018:ffffac040173fa48 EFLAGS: 00010246
[  267.069186] RAX: 0000000000000024 RBX: ffff9dc610d401b8 RCX: 0000000000000000
[  267.069187] RDX: 0000000000000000 RSI: 0000000000000092 RDI: 00000000ffffffff
[  267.069188] RBP: 0000000000000001 R08: 000000000000053d R09: 0000000000000004
[  267.069189] R10: 0000000000000000 R11: 0000000000000001 R12: ffff9dc7c21b0000
[  267.069189] R13: ffff9dc610d41d18 R14: ffff9dc610d401b8 R15: 0000000000000006
[  267.069191] FS:  0000000000000000(0000) GS:ffff9dc7ce800000(0000) knlGS:0000000000000000
[  267.069192] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  267.069193] CR2: 00000000ea3b5000 CR3: 0000000250bf6000 CR4: 00000000003406e0
[  267.069195] Call Trace:
[  267.069260]  dcn20_pipe_control_lock.part.0+0xe4/0x1b0 [amdgpu]
[  267.069314]  dc_commit_updates_for_stream+0xec8/0x1390 [amdgpu]
[  267.069316]  ? _raw_spin_lock+0x13/0x30
[  267.069374]  amdgpu_dm_atomic_commit_tail+0x12a6/0x1d00 [amdgpu]
[  267.069385]  ? commit_tail+0x3c/0x70 [drm_kms_helper]
[  267.069392]  commit_tail+0x3c/0x70 [drm_kms_helper]
[  267.069395]  process_one_work+0x1d1/0x3a0
[  267.069397]  worker_thread+0x4a/0x3d0
[  267.069399]  kthread+0xfb/0x130
[  267.069401]  ? process_one_work+0x3a0/0x3a0
[  267.069402]  ? kthread_park+0x80/0x80
[  267.069405]  ret_from_fork+0x1f/0x40
[  267.069407] ---[ end trace 105f1c3037172656 ]---

Updated my motherboard BIOS as well, with no luck; I guess I'll try rc5 and see if anything changes.

archixxx · 2019-08-20 22:24:04

Haven't tried any Vulkan apps (at least none I'm aware of ;-) ) but KDE with OpenGL 3.1 rendering backend works fine so far. Currently using:

- aur/linux-mainline 5.3rc4-1
- aur/llvm-minimal-git 10.0.0_r324774.c310e5a7ab6-1
- aur/mesa-git 19.2.0_devel.114565.b2839193987-1
- firmware 2019-08-19 from https://people.freedesktop.org/~agd5f/r … de/navi10/
- core/amd-ucode 20190815.07b925b-1

Currently compiling aur/linux-mainline 5.3rc5-1 which was released today.

I've a Sapphire Pulse 5700 XT which is one of the first custom design cards out in the market I guess. My current problem is that power consumption is just crazy. I used a Nvidia GTX 1060 before and at least ATM the whole system uses 50 Watt more then before even when in idle mode (measured with a power energy usage meter). I don't know if "sensors" command provides numbers that of any use (at least some of the are obviously wrong ;-) ) but it prints out the following values:

amdgpu-pci-0c00
Adapter: PCI adapter
vddgfx:       +0.72 V  
fan1:         128 RPM  (min =    0 RPM, max = 3200 RPM)
edge:         +58.0°C  (crit = +118.0°C, hyst =  +0.0°C)
                       (emerg = +80000.0°C)
junction:     +59.0°C  (crit = +80000.0°C, hyst =  +0.0°C)
                       (emerg = +80000.0°C)
mem:          +64.0°C  (crit = +80000.0°C, hyst =  +0.0°C)
                       (emerg = +80000.0°C)
power1:       33.00 W  (cap = 195.00 W)

asus-isa-0000
Adapter: ISA adapter
cpu_fan:        0 RPM

33 Watt in idle mode is pretty high - if the number is correct. I guess I need to look around if I can adjust a few settings to get this down. I would expect more like 8-10 Watt here. But even then power consumption in idle would be much higher just because I replaced the GTX 1060 with a RX 5700 XT. X570 chipset + Ryzen 9 3900X + RX 5700 XT seems to be pretty power hungry in general - at least ATM. I already expected more power usage mainly because of the X570 chipset (which was reported that it uses about 10-15 Watt alone) but even then current power usage is way too high.

The Sapphire Pulse has a second (silent) bios I'll try out after I tested kernel 5.3rc5. Also WattmanGTK might be something to try out.

ineedmana · 2019-08-21 00:38:38

I've just checked. Without amdgpu-pro etc I get lots of glitches in Vulkan (DXVK).

ledbettj · 2019-08-21 15:19:30

I installed some additional packages from mesa-git repository (libdrm-git, lib32-libdrm-git xf86-video-amdgpu-git), switched back to the latest 8/19 firmware, and upgraded to kernel 5.3-rc5; I also switched from Xorg to Wayland based Gnome.

No clue which of these solved the problem, but everything seems solid now! Appreciate everyone weighing in here with their experience and suggestions.

sirn · 2019-08-21 18:39:20

archixxx wrote:

33 Watt in idle mode is pretty high - if the number is correct. I guess I need to look around if I can adjust a few settings to get this down.

I believe this is due to a workaround in DCN20 that force UCLK to run at maximum clock all the time. This has been fixed in drm-next but I don't think it will be in 5.3. You could try drm-next-5.4 which contains the fix (but also means we're stuck with 30W idle in mainline until Nov...)

Edit: drm-next-5.4-2019-08-09 on Radeon RX 5700 XT and 2019-08-19 firmware is able to get 8W idle:

Adapter: PCI adapter
vddgfx:       +0.72 V  
fan1:        1124 RPM  (min =    0 RPM, max = 4950 RPM)
edge:         +52.0°C  (crit = +118.0°C, hyst =  +0.0°C)
                       (emerg = +80000.0°C)
junction:     +52.0°C  (crit = +80000.0°C, hyst =  +0.0°C)
                       (emerg = +80000.0°C)
mem:          +54.0°C  (crit = +80000.0°C, hyst =  +0.0°C)
                       (emerg = +80000.0°C)
power1:        8.00 W  (cap = 180.00 W)

Last edited by sirn (2019-08-22 09:49:59)

archixxx · 2019-08-23 23:12:06

@sirn Thanks for the hint! I installed "linux-amd-staging-drm-next-git 5.3.841285.f1f7ad1b3b98-1" + latest firmware from 2019-08-21. If I don't launch KDE and just stay in the console after booting I get quite good values (the fan1 value is still wrong. The fan doesn't spin anymore in this state):

amdgpu-pci-0c00
Adapter: PCI adapter
vddgfx:       +0.72 V  
fan1:          82 RPM  (min =    0 RPM, max = 3200 RPM)
edge:         +46.0°C  (crit = +118.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
junction:     +46.0°C  (crit = +99.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
mem:          +50.0°C  (crit = +99.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
power1:        8.00 W  (cap = 180.00 W)

asus-isa-0000
Adapter: ISA adapter
cpu_fan:        0 RPM

But as soon as I launch KDE via SDDM everything is as before. At first I thought it was my fault as I still used Legacy BIOS instead of UEFI because I migrated from an older Intel system but continued to use the same old Arch installation. So I started from scratch by installing a fresh Archlinux and used UEFI. Power consumption goes up by 32 W just by launching KDE. And even if I stop SDDM/KDE and go back to the console power consumption stays the same. With a Nvidia GTX 1060 my system uses 90 W while KDE is running (without monitor). With the RX 5700 XT it uses 95 W while in the console and goes up to 135 W if using KDE. So I guess regarding power consumption the driver is still in a very bad shape IMHO. After all that years I hoped that AMD finally comes close to Intel and Nvidia but it seems this doesn't hold true :-(

Does someone know where such an issue can be posted? Is this something for the linux mailing list? But since the patches are not in the official tree I guess it doesn't really fit there.

loqs · 2019-08-23 23:20:05

@archixxx https://bugs.freedesktop.org/buglist.cg … esolution=---

sirn · 2019-08-24 06:42:40

@archixxx that's strange, I'm able to get 8W even within KDE Plasma 5 session (with OpenGL 3.1, though I'm using a reference RX 5700 XT). I'm running from drm-next-5.4-2019-08-23 which should be mostly identical to amd-staging-drm-next branch. Can you try 2019-08-19 firmware? I've uploaded it at https://files.grid.in.th/pub/firmwares/ … 019-08-19/

archixxx · 2019-08-24 07:56:34

@loqs Thanks! I'll playing around a little bit more and if nothing else helps I guess I'll give it a try.

@sirn Also thanks! :-) How do I get "drm-next-5.4-2019-08-23"? If I search AUR regarding something like "5.4" I don't find anything. I tried your firmware files but sadly no luck. As soon as I start KDE Plasma 5 power consumption raises about the already mentioned 30-35 W. I also figured out while in the console "sensors" provides some useful values. But if I run "sensors" while KDE Plasma is running trouble begins. The command prints some values very slowly and after that KDE Plasma basically becomes unusable:

amdgpu-pci-0c00
Adapter: PCI adapter
vddgfx:       +0.72 V  
fan1:             N/A  (min =    0 RPM, max = 3200 RPM)
edge:             N/A  (crit = +118.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
junction:         N/A  (crit = +99.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
mem:              N/A  (crit = +99.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
power1:           N/A  (cap = 180.00 W)

asus-isa-0000
Adapter: ISA adapter
cpu_fan:        0 RPM

And if I've a look at "dmesg" output a lot of errors appear:

[   64.489797] amdgpu: [powerplay] failed send message: TransferTableSmu2Dram (18)      param: 0x00000006 response 0xffffffc2
[   64.489799] amdgpu: [powerplay] Failed to export SMU metrics table!
[   66.698826] amdgpu: [powerplay] failed send message: SetDriverDramAddrHigh (14)      param: 0x00000080 response 0xffffffc2
[   68.370796] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xffffffc2
[   68.894830] amdgpu: [powerplay] failed send message: SetDriverDramAddrHigh (14)      param: 0x00000080 response 0xffffffc2
[   68.894831] amdgpu: [powerplay] Failed to export SMU metrics table!
[   70.566288] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xffffffc2
[   71.090721] amdgpu: [powerplay] failed send message: SetDriverDramAddrHigh (14)      param: 0x00000080 response 0xffffffc2
[   72.762220] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xffffffc2
[   73.286666] amdgpu: [powerplay] failed send message: SetDriverDramAddrHigh (14)      param: 0x00000080 response 0xffffffc2
[   73.286667] amdgpu: [powerplay] Failed to export SMU metrics table!
[   74.957571] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xffffffc2
[   75.482237] amdgpu: [powerplay] failed send message: SetDriverDramAddrHigh (14)      param: 0x00000080 response 0xffffffc2
[   77.154662] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xffffffc2
[   77.679254] amdgpu: [powerplay] failed send message: SetDriverDramAddrHigh (14)      param: 0x00000080 response 0xffffffc2
[   77.679255] amdgpu: [powerplay] Failed to export SMU metrics table!
[   79.349495] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xffffffc2
[   79.874565] amdgpu: [powerplay] failed send message: SetDriverDramAddrHigh (14)      param: 0x00000080 response 0xffffffc2
[   81.544721] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xffffffc2
[   82.069700] amdgpu: [powerplay] failed send message: SetDriverDramAddrHigh (14)      param: 0x00000080 response 0xffffffc2
[   82.069702] amdgpu: [powerplay] Failed to export SMU metrics table!
[   83.751439] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xffffffc2
[   86.510260] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xffffffc2
[   88.724407] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xffffffc2
[   90.939521] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xffffffc2
[   93.154104] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xffffffc2
[   95.369700] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xffffffc2
[   97.577417] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xffffffc2
[   99.786910] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xffffffc2
[  101.991865] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xffffffc2
[  104.202538] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xffffffc2
[  106.411645] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xffffffc2
[  108.623987] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xffffffc2
[  110.833675] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xffffffc2
[  113.050165] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xffffffc2
[  113.050171] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xfffffffb
[  113.050176] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  113.050181] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  113.050205] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xfffffffb
[  113.050214] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xfffffffb
[  113.050220] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  113.050226] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  113.050245] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xfffffffb
[  113.050253] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xfffffffb
[  113.050257] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  113.050263] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  113.050341] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xfffffffb
[  113.050347] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xfffffffb
[  113.050349] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  113.050354] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  126.175275] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xfffffffb
[  126.175282] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xfffffffb
[  126.175284] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  126.175289] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  126.175403] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xfffffffb
[  126.175408] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00000000 response 0xfffffffb
[  126.175411] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  126.175416] amdgpu: [powerplay] failed send message: GetMaxDpmFreq (31)      param: 0x00020000 response 0xfffffffb
[  126.275941] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xfffffffb
[  126.275948] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xfffffffb
[  126.277751] amdgpu 0000:0c:00.0: [mmhub] page fault (src_id:0 ring:169 vmid:0 pasid:0, for process  pid 0 thread  pid 0)
[  126.277755] amdgpu 0000:0c:00.0:   in page starting at address 0x00000834001c6000 from client 18
[  126.277757] amdgpu 0000:0c:00.0: GCVM_L2_PROTECTION_FAULT_STATUS:0x00041152
[  126.277758] amdgpu 0000:0c:00.0:      MORE_FAULTS: 0x0
[  126.277759] amdgpu 0000:0c:00.0:      WALKER_ERROR: 0x1
[  126.277760] amdgpu 0000:0c:00.0:      PERMISSION_FAULTS: 0x5
[  126.277762] amdgpu 0000:0c:00.0:      MAPPING_ERROR: 0x1
[  126.277763] amdgpu 0000:0c:00.0:      RW: 0x1
[  132.922762] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  135.142849] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  137.356260] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  144.498635] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  146.712357] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  148.954945] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  151.168615] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  156.466479] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  158.679673] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  160.907929] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  163.120301] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  165.346024] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  167.558322] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  169.791383] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2
[  172.003757] amdgpu: [powerplay] failed send message: NumOfDisplays (64)      param: 0x00000001 response 0xffffffc2

I assume that power regulation in general works as I also see

[    1.987180] amdgpu: [powerplay] SMU is initialized successfully!

in "dmesg" output after a reboot. Maybe I should try to remove "powerdevil". I guess it won't help that much but maybe I'll give it a try later. I also tried the latest firmware again but there is no difference in using your files and the latest firmware.

BTW: Do you have any "amdgpu" related configuration files in "/etc/X11/"? I didn't created any because besides the power consumption issue everything just works out of the box.

sirn · 2019-08-24 08:21:24

@archixxx basically, get the linux-amd-staging-drm-next-git snapshot from AUR and modify the following lines in PKGBUILD:

_branch=amd-staging-drm-next

Replace it with:

_tag=drm-next-5.4-2019-08-23

And this line:

source=("${pkgbase}::git://people.freedesktop.org/~agd5f/linux#branch=${_branch}"

Replace it with:

source=("${pkgbase}::git://people.freedesktop.org/~agd5f/linux#tag=${_tag}"

Then run makepkg as usual. It will build using the given tag instead of tip of the branch (which, arguably, is more stable, as staging-drm-next may contains few changes not yet ready for merging into mainline). drm-next-VER-DATE tags are source code that are intended to be merged to corresponding kernel VER; drm-fixes-VER-DATE are a small bugfixes to drm-next-VER-DATE that already has been merged into mainline.

In this case, drm-next-5.4-DATE means they're changes pending to be merged into mainline 5.4.

I don't have any special configurations in /etc/X11.

archixxx · 2019-08-24 10:20:02

@sirn Many thanks for the detailed description! But sadly also no luck with that version. So I guess the Sapphire Pulse and the AMD reference card behave differently in this regard. So I'll open a ticket at https://bugs.freedesktop.org/buglist.cg … esolution= and hope for the best ;-)

archixxx · 2019-08-25 08:48:24

FYI: I created a bug ticket at freedesktop's Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111482 . Maybe it's of any interest for other people.

ledbettj · 2019-08-25 16:08:48

@sirn Many thanks for the detailed description! But sadly also no luck with that version. So I guess the Sapphire Pulse and the AMD reference card behave differently in this regard. So I'll open a ticket at https://bugs.freedesktop.org/buglist.cg … esolution= and hope for the best ;-)

I have the same card as you -- I just built the tag referenced above:

john@argus:~$ uname -r
5.3.0-rc3-amd-staging-drm-next-git-b8cd95e15410

And while I don't use KDE, I am running Gnome 3 under Wayland and I see 8W at idle now (whereas on 5.3-rc5 mainline I was seeing 32W)

amdgpu-pci-2b00
Adapter: PCI adapter
vddgfx:       +0.72 V  
fan1:          95 RPM  (min =    0 RPM, max = 3200 RPM)
edge:         +48.0°C  (crit = +118.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
junction:     +48.0°C  (crit = +99.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
mem:          +48.0°C  (crit = +99.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
power1:        8.00 W  (cap = 195.00 W)

k10temp-pci-00c3
Adapter: PCI adapter
Tdie:         +37.8°C  (high = +70.0°C)
Tctl:         +37.8°C

Edit: I do see some minor/occasional tearing in fullscreen games with this kernel that I wasn't seeing with 5.3-rc5, though.

Last edited by ledbettj (2019-08-25 16:32:03)

archixxx · 2019-08-25 18:32:41

Just verified that I've the same kernel:

[root@sulu]# uname -r
5.3.0-rc3-amd-staging-drm-next-git-b8cd95e15410

[root@sulu]# sensors
amdgpu-pci-0c00
Adapter: PCI adapter
vddgfx:       +0.72 V  
fan1:         532 RPM  (min =    0 RPM, max = 3200 RPM)
edge:         +52.0°C  (crit = +118.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
junction:     +53.0°C  (crit = +99.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
mem:          +64.0°C  (crit = +99.0°C, hyst = -273.1°C)
                       (emerg = +99.0°C)
power1:       32.00 W  (cap = 180.00 W)

asus-isa-0000
Adapter: ISA adapter
cpu_fan:        0 RPM

Before I had a Nvidia GTX 1060 from Asus (which was additionally overclocked by default) in exactly the same host and with the Nvidia binary driver the card used around 8-10W in console and also while running KDE. So at least with that configuration I got the expected results. I don't know if this proves anything

Can you tell me which versions of llvm-minimal-git/llvm-git, mesa-git, firmware and amd-ucode you're using currently? Maybe there is some difference. Thanks!

ledbettj · 2019-08-25 20:59:44

archixxx wrote:

Just verified that I've the same kernel:

Before I had a Nvidia GTX 1060 from Asus (which was additionally overclocked by default) in exactly the same host and with the Nvidia binary driver the card used around 8-10W in console and also while running KDE. So at least with that configuration I got the expected results. I don't know if this proves anything
Can you tell me which versions of llvm-minimal-git/llvm-git, mesa-git, firmware and amd-ucode you're using currently? Maybe there is some difference. Thanks!

Sure!

mesa-git 1:19.2.0_devel.114550.6a09405368a-1
llvm-git 10.0.0_r324692.12cbbab9d93-1
amd-ucode 20190815.07b925b-1
libdrm-git 2.4.99.r16.g14922551-1

There are newer versions of mesa/vulkan available in the mesa-git repository, but I haven't upgraded. I'm using the navi10 firmware from August 19; it looks like there is also a newer version available of that as well.

archixxx · 2019-08-27 07:36:18

@ledbettj Out of curiosity: May I ask you what chipset/board you're using? I've a Asus ROG STRIX X570-E GAMING board so it has a X570 chipset. In the freedesktop bug I opened a user is reporting that he has the same problem with a MSI Evoke 5700 XT (https://bugs.freedesktop.org/show_bug.cgi?id=111482#c2).

Meanwhile I updated to Mesa 19.2-rc1. I suspect if it is really Mesa related and the bug is still there then it will be even in the final release of Mesa 19.2. I also switched from "llvm-minimal-git" to "llvm-git 10.0.0_r325224.1c9aa70b042-1" and compiled Mesa again afterwards but the problem still persists.

ledbettj · 2019-08-27 17:00:35

archixxx wrote:

@ledbettj Out of curiosity: May I ask you what chipset/board you're using? I've a Asus ROG STRIX X570-E GAMING board so it has a X570 chipset. In the freedesktop bug I opened a user is reporting that he has the same problem with a MSI Evoke 5700 XT (https://bugs.freedesktop.org/show_bug.cgi?id=111482#c2).
Meanwhile I updated to Mesa 19.2-rc1. I suspect if it is really Mesa related and the bug is still there then it will be even in the final release of Mesa 19.2. I also switched from "llvm-minimal-git" to "llvm-git 10.0.0_r325224.1c9aa70b042-1" and compiled Mesa again afterwards but the problem still persists.

I have an "MSI B350 Gaming Plus" with the latest "beta" bios.

gabmus · 2019-08-28 06:37:29

Wanted to add a couple of things to the conversation.

This is my configuration:

linux-git from AUR

These packages from mesa-git

amdvlk
clang-git
libclc-git
libdrm-git
llvm-git
llvm-libs-git
mesa-git
opencl-mesa-git
vulkan-mesa-layer-git
vulkan-radeon-git
xf86-video-amdgpu-git

And the firmware from https://people.freedesktop.org/~agd5f/r … de/navi10/

First, as for the firmware, I made an AUR package out of it to so that it's easier to install and track. Here you go, hope it's useful: https://aur.archlinux.org/packages/linu … on-navi10/

Second, I am having a pretty good experience so far with performance and functionality, the only big problem I'm having right now is X11 DEs are kinda choppy. Games work fine, but the DE itself feels slow. This happens with both GNOME on Xorg and XFCE. Weirdly enough, the performance on Wayland is perfect, and I don't experience the same choppiness. Any idea what's going on?

Last edited by gabmus (2019-08-28 06:38:09)

giuge · 2019-08-31 08:21:02

gabmus wrote:

I made an AUR package out of it to so that it's easier to install and track. Here you go, hope it's useful

Thank you for this, way cleaner than copy pasting them in!

I got the card to work using:

# From AUR
linux-mainline
linux-firmware-agd5f-radeon-navi10

# From MESA-GIT
amdvlk
clang-git
lib32-llvm-libs-git
lib32-mesa-git
lib32-vulkan-radeon-git
libclc-git
llvm-git
llvm-libs-git
mesa-git
opencl-mesa-git
vulkan-radeon-git
xf86-video-amdgpu-git

The only issue I have is the fan wasn't spinning at all. I have an XFX 5700 XT THICC II, wich has a "ZERO DB AUTO LOAD SENSING FANS", so maybe the driver is not up to date enough to support this kind of feature.

For now I'm using

amdgpu-fan

with the default fan curve and it works flowlessly. No choppy wm (I use dwm tho) and the few games I tested worked perfectly.

Wondering what will be the procedure to replace the firmware once we have full support... I guess everything will be ready once 5.3 ships?
Do you know what we'll do with the firmware then?

Last edited by giuge (2019-08-31 08:28:57)

Malstrond · 2019-08-31 21:52:23

I'm also using a 5700 XT with linux-mainline, mesa-git, xf86-video-amdgpu-git and linux-firmware-agd5f-radeon-navi10, but apparently the driver attempts to divide by zero. Happened while watching a .webm video in Firefox.
Has anyone seen this one before?
Log:

Aug 31 19:40:33 korhal kernel: divide error: 0000 [#1] PREEMPT SMP NOPTI
Aug 31 19:40:33 korhal kernel: CPU: 7 PID: 1789 Comm: Xorg Not tainted 5.3.0-rc6-mainline #1
Aug 31 19:40:33 korhal kernel: Hardware name: Gigabyte Technology Co., Ltd. X570 AORUS ELITE/X570 AORUS ELITE, BIOS F4j 08/02/2019
Aug 31 19:40:33 korhal kernel: RIP: 0010:CalculateVMAndRowBytes.constprop.0+0x458/0xa90 [amdgpu]
Aug 31 19:40:33 korhal kernel: Code: 83 e0 ef 83 e8 05 83 f8 01 0f 87 30 04 00 00 46 8d 04 b5 00 00 00 00 41 bb 00 10 00 00 44 89 d8 31 d2 48 8b bc 24 d0 00 00 00 <f7> b4 24 80 00 00 00 31 d2 41 f7 f0 89 07 45 84 e4 0f 84 87 fd ff
Aug 31 19:40:33 korhal kernel: RSP: 0018:ffffa334409a7790 EFLAGS: 00010246
Aug 31 19:40:33 korhal kernel: RAX: 0000000000010000 RBX: 0000000000000008 RCX: 0000000000000008
Aug 31 19:40:33 korhal kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff999c286f77a8
Aug 31 19:40:33 korhal kernel: RBP: 000000000000001f R08: 0000000000000080 R09: 000000000000001f
Aug 31 19:40:33 korhal kernel: R10: 0000000000000000 R11: 0000000000010000 R12: 0000000000000000
Aug 31 19:40:33 korhal kernel: R13: 0000000000000000 R14: 0000000000000008 R15: ffff999c286f1ed8
Aug 31 19:40:33 korhal kernel: FS:  00007f3578820dc0(0000) GS:ffff999cae7c0000(0000) knlGS:0000000000000000
Aug 31 19:40:33 korhal kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Aug 31 19:40:33 korhal kernel: CR2: 00007f0f32413000 CR3: 00000007e81a4000 CR4: 0000000000340ee0
Aug 31 19:40:33 korhal kernel: Call Trace:
Aug 31 19:40:33 korhal kernel:  dml20_ModeSupportAndSystemConfigurationFull+0x3be2/0x5f70 [amdgpu]
Aug 31 19:40:33 korhal kernel:  dml_get_voltage_level+0x12f/0x1c0 [amdgpu]
Aug 31 19:40:33 korhal kernel:  dcn20_validate_bandwidth+0x2de/0x1670 [amdgpu]
Aug 31 19:40:33 korhal kernel:  dc_validate_global_state+0x28a/0x310 [amdgpu]
Aug 31 19:40:33 korhal kernel:  amdgpu_dm_atomic_check+0x577/0x6e0 [amdgpu]
Aug 31 19:40:33 korhal kernel:  drm_atomic_check_only+0x571/0x7a0 [drm]
Aug 31 19:40:33 korhal kernel:  ? drm_mode_object_put.part.0+0x1f/0x40 [drm]
Aug 31 19:40:33 korhal kernel:  ? drm_atomic_set_property+0x8b/0xa00 [drm]
Aug 31 19:40:33 korhal kernel:  drm_atomic_commit+0x13/0x50 [drm]
Aug 31 19:40:33 korhal kernel:  drm_mode_obj_set_property_ioctl+0x126/0x2c0 [drm]
Aug 31 19:40:33 korhal kernel:  ? drm_mode_obj_find_prop_id+0x40/0x40 [drm]
Aug 31 19:40:33 korhal kernel:  drm_ioctl_kernel+0xb8/0x100 [drm]
Aug 31 19:40:33 korhal kernel:  drm_ioctl+0x23d/0x3d0 [drm]
Aug 31 19:40:33 korhal kernel:  ? drm_mode_obj_find_prop_id+0x40/0x40 [drm]
Aug 31 19:40:33 korhal kernel:  amdgpu_drm_ioctl+0x49/0x80 [amdgpu]
Aug 31 19:40:33 korhal kernel:  do_vfs_ioctl+0x43d/0x6c0
Aug 31 19:40:33 korhal kernel:  ? __sys_recvmsg+0xaf/0xd0
Aug 31 19:40:33 korhal kernel:  ksys_ioctl+0x5e/0x90
Aug 31 19:40:33 korhal kernel:  __x64_sys_ioctl+0x16/0x20
Aug 31 19:40:33 korhal kernel:  do_syscall_64+0x5f/0x1c0
Aug 31 19:40:33 korhal kernel:  entry_SYSCALL_64_after_hwframe+0x44/0xa9
Aug 31 19:40:33 korhal kernel: RIP: 0033:0x7f3579c5221b
Aug 31 19:40:33 korhal kernel: Code: 0f 1e fa 48 8b 05 75 8c 0c 00 64 c7 00 26 00 00 00 48 c7 c0 ff ff ff ff c3 66 0f 1f 44 00 00 f3 0f 1e fa b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 45 8c 0c 00 f7 d8 64 89 01 48
Aug 31 19:40:33 korhal kernel: RSP: 002b:00007ffc0cb6f2a8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
Aug 31 19:40:33 korhal kernel: RAX: ffffffffffffffda RBX: 00007ffc0cb6f2e0 RCX: 00007f3579c5221b
Aug 31 19:40:33 korhal kernel: RDX: 00007ffc0cb6f2e0 RSI: 00000000c01864ba RDI: 000000000000000a
Aug 31 19:40:33 korhal kernel: RBP: 00000000c01864ba R08: 000000000000006c R09: 00000000cccccccc
Aug 31 19:40:33 korhal kernel: R10: 000055c61fdc5094 R11: 0000000000000246 R12: 000055c61e4f72c0
Aug 31 19:40:33 korhal kernel: R13: 000000000000000a R14: 0000000000000fff R15: 0000000000000003
Aug 31 19:40:33 korhal kernel: Modules linked in: uinput udf crc_itu_t loop rfcomm msr cpufreq_ondemand cfg80211 8021q cmac garp mrp stp algif_hash llc bnep fuse btusb btrtl btbcm btintel bluetooth blowfish_generic ecdh_generic blowfish_x86_64 rfkill blowfish_common usblp mousedev input_leds ecc des3_ede_x86_64 des_generic cast5_avx_x86_64 cast5_generic cast_common nls_iso8859_1 nls_cp437 vfat fat lrw camellia_generic amdgpu camellia_aesni_avx2 camellia_aesni_avx_x86_64 camellia_x86_64 twofish_generic twofish_avx_x86_64 twofish_x86_64_3way twofish_x86_64 amd_iommu_v2 twofish_common snd_hda_codec_generic gpu_sched ledtrig_audio ttm snd_hda_codec_hdmi drm_kms_helper snd_hda_intel serpent_avx2 serpent_avx_x86_64 edac_mce_amd drm snd_hda_codec kvm_amd serpent_sse2_x86_64 snd_hda_core serpent_generic snd_hwdep algif_skcipher af_alg wmi_bmof snd_pcm kvm agpgart syscopyarea snd_timer sysfillrect snd sysimgblt sp5100_tco fb_sys_fops irqbypass soundcore pcspkr i2c_piix4 igb i2c_algo_bit dca wmi pinctrl_amd evdev
Aug 31 19:40:33 korhal kernel:  mac_hid acpi_cpufreq sg ip_tables x_tables ext4 crc32c_generic crc16 mbcache jbd2 dm_crypt dm_mod sr_mod cdrom sd_mod hid_generic usbhid hid crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel aesni_intel ahci libahci aes_x86_64 crypto_simd ccp libata cryptd xhci_pci glue_helper rng_core xhci_hcd scsi_mod
Aug 31 19:40:33 korhal kernel: ---[ end trace 17c66dd07c513819 ]---
Aug 31 19:40:33 korhal kernel: RIP: 0010:CalculateVMAndRowBytes.constprop.0+0x458/0xa90 [amdgpu]
Aug 31 19:40:33 korhal kernel: Code: 83 e0 ef 83 e8 05 83 f8 01 0f 87 30 04 00 00 46 8d 04 b5 00 00 00 00 41 bb 00 10 00 00 44 89 d8 31 d2 48 8b bc 24 d0 00 00 00 <f7> b4 24 80 00 00 00 31 d2 41 f7 f0 89 07 45 84 e4 0f 84 87 fd ff
Aug 31 19:40:33 korhal kernel: RSP: 0018:ffffa334409a7790 EFLAGS: 00010246
Aug 31 19:40:33 korhal kernel: RAX: 0000000000010000 RBX: 0000000000000008 RCX: 0000000000000008
Aug 31 19:40:33 korhal kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff999c286f77a8
Aug 31 19:40:33 korhal kernel: RBP: 000000000000001f R08: 0000000000000080 R09: 000000000000001f
Aug 31 19:40:33 korhal kernel: R10: 0000000000000000 R11: 0000000000010000 R12: 0000000000000000
Aug 31 19:40:33 korhal kernel: R13: 0000000000000000 R14: 0000000000000008 R15: ffff999c286f1ed8
Aug 31 19:40:33 korhal kernel: FS:  00007f3578820dc0(0000) GS:ffff999cae7c0000(0000) knlGS:0000000000000000
Aug 31 19:40:33 korhal kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Aug 31 19:40:33 korhal kernel: CR2: 00007f0f32413000 CR3: 00000007e81a4000 CR4: 0000000000340ee0

gabmus · 2019-09-01 07:04:44

I made a post in my blog, detailing (again) the installation and some problems I found in some games.

https://gabmus.gitlab.io/2019/08/31/Ear … rch-Linux/

Hopefully you have an idea on how to fix these problems.

Last edited by gabmus (2019-09-01 07:04:58)

MadByteDE · 2019-09-01 15:57:27

Second, I am having a pretty good experience so far with performance and functionality, the only big problem I'm having right now is X11 DEs are kinda choppy. Games work fine, but the DE itself feels slow. This happens with both GNOME on Xorg and XFCE. Weirdly enough, the performance on Wayland is perfect, and I don't experience the same choppiness. Any idea what's going on?

The latest linux-amd-staging-drm-next-git fixed the power consumption and the choppy DE on my system (tested gnome and XFCE). Everything seems to work fine with the Powercolor 5700XT Red Devil.
The only bit left is AMDVLK because it causes weird screen artifacts in DXVK games, which don't occur using vulkan-radeon-git. I read something about a bug with vertex shaders.. I'd love to try the latest git version of amdvlk but the aur package is outdated

Last edited by MadByteDE (2019-09-01 16:08:20)

Arch Linux

#26 2019-08-17 08:17:54

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#27 2019-08-20 14:05:15

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#28 2019-08-20 19:33:26

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#29 2019-08-20 22:08:44

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#30 2019-08-20 22:24:04

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#31 2019-08-21 00:38:38

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#32 2019-08-21 15:19:30

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#33 2019-08-21 18:39:20

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#34 2019-08-23 23:12:06

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#35 2019-08-23 23:20:05

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#36 2019-08-24 06:42:40

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#37 2019-08-24 07:56:34

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#38 2019-08-24 08:21:24

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#39 2019-08-24 10:20:02

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#40 2019-08-25 08:48:24

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#41 2019-08-25 16:08:48

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#42 2019-08-25 18:32:41

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#43 2019-08-25 20:59:44

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#44 2019-08-27 07:36:18

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#45 2019-08-27 17:00:35

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#46 2019-08-28 06:37:29

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#47 2019-08-31 08:21:02

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#48 2019-08-31 21:52:23

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#49 2019-09-01 07:04:44

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

#50 2019-09-01 15:57:27

Re: [RESOLVED] Anyone got AMD RX 5700 (XT) to work?

Board footer