Radeon RX7900M support?

Message

KShots · Post by **KShots** » Sat Dec 02, 2023 2:23 am

So, I just purchased an Alienware m18 r1 with the AMD branch of hardware (Ryzen 9 7945HX and the subject GPU)... and found some quirks. First off, as of kernel 6.1.57, I get a non-fatal kernel oops on boot:

Code: Select all

[    5.393054] WARNING: CPU: 4 PID: 563 at drivers/gpu/drm/amd/amdgpu/../display
/dc/dcn20/dcn20_dsc.c:273 dsc2_disable+0x147/0x160 [amdgpu]
[    5.393177] Modules linked in: amdgpu(+) drm_ttm_helper ttm agpgart iommu_v2 
gpu_sched i2c_algo_bit drm_buddy drm_display_helper crc32_pclmul drm_kms_helper 
crc32c_intel polyval_clmulni syscopyarea polyval_generic sysfillrect ghash_clmul
ni_intel rtsx_pci_sdmmc sysimgblt sha512_ssse3 fb_sys_fops ccp mmc_core aesni_in
tel crypto_simd xhci_pci cryptd drm sha1_generic nvme rtsx_pci xhci_hcd rng_core
 cec mfd_core ucsi_acpi nvme_core typec_ucsi roles typec i2c_hid_acpi i2c_hid dm
_mod efivarfs
[    5.393190] CPU: 4 PID: 563 Comm: (udev-worker) Not tainted 6.1.57-gentoo #10
[    5.393191] Hardware name: Alienware Alienware m18 R1 AMD/049YP9, BIOS 1.7.0 
09/21/2023
[    5.393192] RIP: 0010:dsc2_disable+0x147/0x160 [amdgpu]
[    5.393273] Code: 00 00 00 75 2c 48 83 c4 18 5b e9 a4 59 e7 e5 8b 4b 10 44 8b
 44 24 0c be 02 00 00 00 31 ff 48 c7 c2 30 f1 e1 c0 e8 09 8f 94 ff <0f> 0b e9 6e
 ff ff ff e8 ed 4b ac e5 66 66 2e 0f 1f 84 00 00 00 00
[    5.393274] RSP: 0018:ffffacc180a3f290 EFLAGS: 00010246
[    5.393275] RAX: 0000000000000000 RBX: ffff90c5ccf77600 RCX: 0000000000000000
[    5.393275] RDX: ffffffffc0e1f130 RSI: 0000000000000002 RDI: 0000000000000000
[    5.393276] RBP: 0000000000000001 R08: 0000000000000000 R09: 0000000000000004
[    5.393276] R10: 0000000000000001 R11: 000000005b29c82f R12: ffff90c5f3bc01e8
[    5.393276] R13: ffff90c5c2c71800 R14: ffff90c5dda4e800 R15: ffff90c5f3bc01e8
[    5.393277] FS:  00007f88d46e1340(0000) GS:ffff90d4fd700000(0000) knlGS:0000000000000000
[    5.393277] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    5.393278] CR2: 00007f88d53464c8 CR3: 0000000104640000 CR4: 0000000000750ee0
[    5.393278] PKRU: 55555554
[    5.393279] Call Trace:
[    5.393280]  <TASK>
[    5.393281]  ? dsc2_disable+0x147/0x160 [amdgpu]
[    5.393351]  ? __warn+0x74/0xc0
[    5.393354]  ? dsc2_disable+0x147/0x160 [amdgpu]
[    5.393417]  ? report_bug+0xe6/0x170
[    5.393420]  ? handle_bug+0x3a/0x70
[    5.393422]  ? exc_invalid_op+0x13/0x60
[    5.393423]  ? asm_exc_invalid_op+0x16/0x20
[    5.393424]  ? dsc2_disable+0x147/0x160 [amdgpu]
[    5.393486]  dp_set_dsc_on_stream+0x317/0x3e0 [amdgpu]
[    5.393563]  ? dm_helpers_dp_write_dsc_enable+0x25b/0x5b0 [amdgpu]
[    5.393647]  dp_set_dsc_enable+0x6c/0x80 [amdgpu]
[    5.393718]  core_link_disable_stream+0x155/0x550 [amdgpu]
[    5.393786]  ? dc_dmub_srv_is_restore_required+0x33/0x80 [amdgpu]
[    5.393876] usb 5-1.3: New USB device found, idVendor=0d62, idProduct=bab0, bcdDevice=33.09
[    5.393877] usb 5-1.3: New USB device strings: Mfr=1, Product=2, SerialNumber=0
[    5.393878] usb 5-1.3: Product: Keyboard
[    5.393878] usb 5-1.3: Manufacturer: DELL Technologies
[    5.393853]  ? dm_read_reg_func+0x32/0xa0 [amdgpu]
[    5.393923]  dc_commit_state_no_check+0xc8e/0xd10 [amdgpu]
[    5.393992]  dc_commit_state+0x102/0x120 [amdgpu]
[    5.394055]  amdgpu_dm_atomic_commit_tail+0x619/0x31d0 [amdgpu]
[    5.394120]  ? dsc2_validate_stream+0x32/0x50 [amdgpu]
[    5.394191]  ? dcn20_validate_dsc+0x18d/0x1b0 [amdgpu]
[    5.394264]  ? dcn30_internal_validate_bw+0x911/0x9f0 [amdgpu]
[    5.394336]  ? kernel_fpu_end+0x1d/0x40
[    5.394338]  ? dcn31_validate_bandwidth+0x126/0x2b0 [amdgpu]
[    5.394402]  ? dc_validate_global_state+0x283/0x340 [amdgpu]
[    5.394472]  ? dma_resv_iter_first_unlocked+0x5d/0x70
[    5.394474]  ? dma_resv_get_fences+0x58/0x220
[    5.394475]  ? dm_plane_helper_prepare_fb+0x17c/0x290 [amdgpu]
[    5.394545]  ? dm_plane_helper_prepare_fb+0x1bf/0x290 [amdgpu]
[    5.394606]  commit_tail+0x8f/0x120 [drm_kms_helper]
[    5.394612]  drm_atomic_helper_commit+0x111/0x140 [drm_kms_helper]
[    5.394617]  drm_atomic_commit+0x91/0xc0 [drm]
[    5.394628]  ? drm_plane_get_damage_clips.cold+0x1c/0x1c [drm]
[    5.394637]  drm_client_modeset_commit_atomic+0x1f5/0x230 [drm]
[    5.394646]  drm_client_modeset_commit_locked+0x51/0x150 [drm]
[    5.394655]  drm_client_modeset_commit+0x1c/0x40 [drm]
[    5.394662]  drm_fb_helper_set_par+0x99/0xd0 [drm_kms_helper]
[    5.394667]  fbcon_init+0x273/0x630
[    5.394670]  visual_init+0xc7/0x120
[    5.394672]  do_bind_con_driver.isra.0+0x1b5/0x2c0
[    5.394673]  do_take_over_console+0x111/0x180
[    5.394674]  do_fbcon_takeover+0x55/0xc0
[    5.394675]  fbcon_fb_registered+0x12e/0x140
[    5.394676]  register_framebuffer+0x1cf/0x310
[    5.394677]  __drm_fb_helper_initial_config_and_unlock+0x383/0x4e0 [drm_kms_helper]
[    5.394682]  drm_fbdev_client_hotplug+0x181/0x210 [drm_kms_helper]
[    5.394686]  drm_client_register+0x58/0x90 [drm]
[    5.394695]  amdgpu_pci_probe+0x35e/0x370 [amdgpu]
[    5.394764]  local_pci_probe+0x3c/0x80
[    5.394766]  pci_device_probe+0xa7/0x1b0
[    5.394768]  really_probe+0xd4/0x280
[    5.394770]  ? pm_runtime_barrier+0x4b/0x80
[    5.394772]  __driver_probe_device+0x6e/0xe0
[    5.394773]  driver_probe_device+0x1a/0xe0
[    5.394774]  __driver_attach+0x7f/0x180
[    5.394775]  ? __device_attach_driver+0x110/0x110
[    5.394776]  bus_for_each_dev+0x82/0xc0
[    5.394777]  bus_add_driver+0x181/0x1d0
[    5.394778]  driver_register+0x84/0xe0
[    5.394779]  ? 0xffffffffc1916000
[    5.394780]  do_one_initcall+0x54/0x200
[    5.394782]  do_init_module+0x45/0x1c0
[    5.394784]  __do_sys_finit_module+0x9e/0xf0
[    5.394785]  do_syscall_64+0x3b/0x90
[    5.394787]  entry_SYSCALL_64_after_hwframe+0x64/0xce
[    5.394788] RIP: 0033:0x7f88d52716a9
[    5.394789] Code: 08 89 e8 5b 5d c3 66 2e 0f 1f 84 00 00 00 00 00 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 3f c7 0c 00 f7 d8 64 89 01 48
[    5.394789] RSP: 002b:00007ffe59a2c478 EFLAGS: 00000246 ORIG_RAX: 0000000000000139
[    5.394790] RAX: ffffffffffffffda RBX: 000055a67d63c500 RCX: 00007f88d52716a9
[    5.394790] RDX: 0000000000000000 RSI: 00007f88d535bb95 RDI: 000000000000001b
[    5.394791] RBP: 00007f88d535bb95 R08: 0000000000000000 R09: 000055a67d611560
[    5.394791] R10: 000000000000001b R11: 0000000000000246 R12: 0000000000020000
[    5.394791] R13: 0000000000000000 R14: 000055a67d64fd50 R15: 0000000000000000
[    5.394792]  </TASK>
[    5.394792] ---[ end trace 0000000000000000 ]---

This appears to occur shortly after initializing the EFI framebuffer. It successfully gets into gnome (not sure yet if it's xorg or wayland), but it doesn't see my displays (other than the built-in screen). My external displays are undetected.

I've also attempted to update my linux-firmware to 20231111-r1 (latest as of today), but that still has the same oops and symptoms. Any hope of getting the display output working?

EDIT: I also tried updating to the latest kernel in the tree as of today (6.6.3) and not only got the same problem, but a whole slew of other problems... so, apparently the newest kernel in the tree wouldn't fix this...

New at Gentoo · Post by **pietinger** » Sat Dec 02, 2023 7:58 am

To see more we would need:
- your kernel .config
- a complete dmesg after boot
- output of "lspci -nnk"
- output of "emerge --info"

logrusx · Post by **logrusx** » Sat Dec 02, 2023 8:41 am

KShots wrote:I also tried updating to the latest kernel in the tree as of today (6.6.3) and not only got the same problem, but a whole slew of other problems... so, apparently the newest kernel in the tree wouldn't fix this...

I suggest you try 6.5. There are a bunch of gentoo-kernel-bin's in portage tree from that line, I suggest trying it as it would be easier and faster and also with the most generic configuration, supporting the largest amount of hardware. The newest kernels can have problems and 6.6.3 is pretty new. I haven't even tried it yet. Maybe I'll wait for it to become stable on portage tree. I have mixed experience with trying the mainline kernels early on, including problems with drivers.

Also we're expecting what pietinger asked for.

Best Regards,
Georgi

KShots · Post by **KShots** » Sat Dec 02, 2023 6:53 pm

pietinger wrote:To see more we would need:
- your kernel .config
- a complete dmesg after boot
- output of "lspci -nnk"
- output of "emerge --info"

Ok, easy enough.

New at Gentoo · Post by **pietinger** » Sat Dec 02, 2023 7:56 pm

First of all I will do the simple part: Touchpad

Please enable both as module:

Code: Select all

# CONFIG_I2C_PIIX4 is not set
# CONFIG_HID_MULTITOUCH is not set

If this is not sufficient also:

Code: Select all

# CONFIG_I2C_DESIGNWARE_SLAVE is not set
# CONFIG_I2C_DESIGNWARE_PLATFORM is not set

Now to your Oops ... I would like to say in advance that I am not a dell expert.

I have seen in your dmesg:

Code: Select all

[    0.000000] DMI: Alienware Alienware m18 R1 AMD/049YP9, BIOS 1.7.0 09/21/2023

[    0.003176] Secure boot enabled

[    0.222762] smpboot: CPU0: AMD Ryzen 9 7945HX with Radeon Graphics (family: 0x19, model: 0x61, stepping: 0x2)

[    3.306809] amdgpu: ATOM BIOS: 113-BRT121633-001
[    3.306910] Loading firmware: amdgpu/gc_11_0_0_mes_2.bin
[    3.306933] Loading firmware: amdgpu/gc_11_0_0_mes1.bin
[    3.306947] amdgpu 0000:03:00.0: amdgpu: Trusted Memory Zone (TMZ) feature not supported
[    3.306980] [drm] vm size is 262144 GB, 4 levels, block size is 9-bit, fragment size is 9-bit
[    3.306983] amdgpu 0000:03:00.0: amdgpu: VRAM: 16368M 0x0000008000000000 - 0x00000083FEFFFFFF (16368M used)
[    3.306984] amdgpu 0000:03:00.0: amdgpu: GART: 512M 0x0000000000000000 - 0x000000001FFFFFFF
[    3.306989] [drm] Detected VRAM RAM=16368M, BAR=16384M
[    3.306989] [drm] RAM width 256bits GDDR6
[    3.307012] [drm] amdgpu: 16368M of VRAM memory ready
[    3.307012] [drm] amdgpu: 31747M of GTT memory ready.
[    3.307123] Loading firmware: amdgpu/psp_13_0_0_sos.bin
[    3.307142] Loading firmware: amdgpu/psp_13_0_0_ta.bin
[    3.307284] Loading firmware: amdgpu/smu_13_0_0.bin
[    3.307341] Loading firmware: amdgpu/dcn_3_2_0_dmcub.bin
[    3.307365] Loading firmware: amdgpu/gc_11_0_0_imu.bin
[    3.307378] Loading firmware: amdgpu/gc_11_0_0_pfp.bin
[    3.307392] Loading firmware: amdgpu/gc_11_0_0_me.bin
[    3.307409] Loading firmware: amdgpu/gc_11_0_0_rlc.bin
[    3.307425] Loading firmware: amdgpu/gc_11_0_0_mec.bin
[    3.307574] Loading firmware: amdgpu/sdma_6_0_0.bin
[    3.307596] Loading firmware: amdgpu/vcn_4_0_0.bin

[    5.034652] [drm] Detected VRAM RAM=512M, BAR=512M
[    5.034652] [drm] RAM width 128bits DDR5
[    5.034665] [drm] amdgpu: 512M of VRAM memory ready
[    5.034665] [drm] amdgpu: 31747M of GTT memory ready.
[    5.034671] [drm] GART: num cpu pages 262144, num gpu pages 262144
[    5.034758] [drm] PCIE GART of 1024M enabled (table at 0x000000F41FC00000).
[    5.034819] Loading firmware: amdgpu/psp_13_0_5_toc.bin
[    5.034832] Loading firmware: amdgpu/psp_13_0_5_ta.bin
[    5.034927] Loading firmware: amdgpu/dcn_3_1_5_dmcub.bin
[    5.034953] Loading firmware: amdgpu/gc_10_3_6_pfp.bin
[    5.034986] Loading firmware: amdgpu/gc_10_3_6_me.bin
[    5.035011] Loading firmware: amdgpu/gc_10_3_6_ce.bin
[    5.035035] Loading firmware: amdgpu/gc_10_3_6_rlc.bin
[    5.035055] Loading firmware: amdgpu/gc_10_3_6_mec.bin
[    5.035080] Loading firmware: amdgpu/gc_10_3_6_mec2.bin
[    5.035466] Loading firmware: amdgpu/sdma_5_2_6.bin

My interpretation is: You have a CPU with integrated GPU + an additional graphics card ... both AMD ... kernel tries to load for BOTH necessary drivers ... and after this you have a problem with /gpu/drm/amd/amdgpu/ ... so, maybe this is the problem. Have you tried to disable the integrated GPU of your CPU in your BIOS ? Are there any options in your BIOS ?

Last but not least: I have seen this in your kernel .config:

Code: Select all

a)
CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE=y
b)
# CONFIG_X86_AMD_PSTATE is not set
c)
# CONFIG_USB_STORAGE is not set
d)
CONFIG_AMD_IOMMU_V2=m
# CONFIG_IRQ_REMAP is not set

a) Please read: https://wiki.gentoo.org/wiki/User:Pieti ... Must_Haves
b) Enable it static <*>
c) If you want use an USB stick or USB harddisk
d) Enable both static <*>

KShots · Post by **KShots** » Sun Dec 03, 2023 12:42 am

pietinger wrote:First of all I will do the simple part: Touchpad

Please enable both as module:
Code: Select all
# CONFIG_I2C_PIIX4 is not set
# CONFIG_HID_MULTITOUCH is not set
If this is not sufficient also:
Code: Select all
# CONFIG_I2C_DESIGNWARE_SLAVE is not set
# CONFIG_I2C_DESIGNWARE_PLATFORM is not set

Got it... Got the following after modprobing the PIIX module

Code: Select all

[82443.761970] piix4_smbus 0000:00:14.0: SMBus Host Controller at 0xb00, revision 0
[82443.761973] piix4_smbus 0000:00:14.0: Using register 0x02 for SMBus port selection
[82443.798511] piix4_smbus 0000:00:14.0: Auxiliary SMBus Host Controller at 0xb20

... but some of those require a reboot, and without rebooting it still isn't working (even after restarting gdm). I'll test again after rebooting (currently transferring a large amount of data from the old machine to this machine and don't want to interrupt it).

pietinger wrote:Now to your Oops ... I would like to say in advance that I am not a dell expert.

I have seen in your dmesg:

Code: Select all

[    0.000000] DMI: Alienware Alienware m18 R1 AMD/049YP9, BIOS 1.7.0 09/21/2023

[    0.003176] Secure boot enabled

[    0.222762] smpboot: CPU0: AMD Ryzen 9 7945HX with Radeon Graphics (family: 0x19, model: 0x61, stepping: 0x2)

[    3.306809] amdgpu: ATOM BIOS: 113-BRT121633-001
[    3.306910] Loading firmware: amdgpu/gc_11_0_0_mes_2.bin
[    3.306933] Loading firmware: amdgpu/gc_11_0_0_mes1.bin
[    3.306947] amdgpu 0000:03:00.0: amdgpu: Trusted Memory Zone (TMZ) feature not supported
[    3.306980] [drm] vm size is 262144 GB, 4 levels, block size is 9-bit, fragment size is 9-bit
[    3.306983] amdgpu 0000:03:00.0: amdgpu: VRAM: 16368M 0x0000008000000000 - 0x00000083FEFFFFFF (16368M used)
[    3.306984] amdgpu 0000:03:00.0: amdgpu: GART: 512M 0x0000000000000000 - 0x000000001FFFFFFF
[    3.306989] [drm] Detected VRAM RAM=16368M, BAR=16384M
[    3.306989] [drm] RAM width 256bits GDDR6
[    3.307012] [drm] amdgpu: 16368M of VRAM memory ready
[    3.307012] [drm] amdgpu: 31747M of GTT memory ready.
[    3.307123] Loading firmware: amdgpu/psp_13_0_0_sos.bin
[    3.307142] Loading firmware: amdgpu/psp_13_0_0_ta.bin
[    3.307284] Loading firmware: amdgpu/smu_13_0_0.bin
[    3.307341] Loading firmware: amdgpu/dcn_3_2_0_dmcub.bin
[    3.307365] Loading firmware: amdgpu/gc_11_0_0_imu.bin
[    3.307378] Loading firmware: amdgpu/gc_11_0_0_pfp.bin
[    3.307392] Loading firmware: amdgpu/gc_11_0_0_me.bin
[    3.307409] Loading firmware: amdgpu/gc_11_0_0_rlc.bin
[    3.307425] Loading firmware: amdgpu/gc_11_0_0_mec.bin
[    3.307574] Loading firmware: amdgpu/sdma_6_0_0.bin
[    3.307596] Loading firmware: amdgpu/vcn_4_0_0.bin

[    5.034652] [drm] Detected VRAM RAM=512M, BAR=512M
[    5.034652] [drm] RAM width 128bits DDR5
[    5.034665] [drm] amdgpu: 512M of VRAM memory ready
[    5.034665] [drm] amdgpu: 31747M of GTT memory ready.
[    5.034671] [drm] GART: num cpu pages 262144, num gpu pages 262144
[    5.034758] [drm] PCIE GART of 1024M enabled (table at 0x000000F41FC00000).
[    5.034819] Loading firmware: amdgpu/psp_13_0_5_toc.bin
[    5.034832] Loading firmware: amdgpu/psp_13_0_5_ta.bin
[    5.034927] Loading firmware: amdgpu/dcn_3_1_5_dmcub.bin
[    5.034953] Loading firmware: amdgpu/gc_10_3_6_pfp.bin
[    5.034986] Loading firmware: amdgpu/gc_10_3_6_me.bin
[    5.035011] Loading firmware: amdgpu/gc_10_3_6_ce.bin
[    5.035035] Loading firmware: amdgpu/gc_10_3_6_rlc.bin
[    5.035055] Loading firmware: amdgpu/gc_10_3_6_mec.bin
[    5.035080] Loading firmware: amdgpu/gc_10_3_6_mec2.bin
[    5.035466] Loading firmware: amdgpu/sdma_5_2_6.bin

My interpretation is: You have a CPU with integrated GPU + an additional graphics card ... both AMD ... kernel tries to load for BOTH necessary drivers ... and after this you have a problem with /gpu/drm/amd/amdgpu/ ... so, maybe this is the problem. Have you tried to disable the integrated GPU of your CPU in your BIOS ? Are there any options in your BIOS ?

Unfortunately, this is perhaps the simplest BIOS I have yet seen.. very few toggles available, and none of them include integrated GPU options

I'll check again after the prior-mentioned reboot, but pretty sure no such option existed.

pietinger wrote:Last but not least: I have seen this in your kernel .config:
Code: Select all
a)
CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE=y
b)
# CONFIG_X86_AMD_PSTATE is not set
c)
# CONFIG_USB_STORAGE is not set
d)
CONFIG_AMD_IOMMU_V2=m
# CONFIG_IRQ_REMAP is not set
a) Please read: https://wiki.gentoo.org/wiki/User:Pieti ... Must_Haves
b) Enable it static <*>
c) If you want use an USB stick or USB harddisk
d) Enable both static <*>

Thanks, good catches.

logrusx · Post by **logrusx** » Sun Dec 03, 2023 6:24 am

From what I read on the topic, your setup should work out of the box and you should be able to offload to the dGPU with either switcheroo-control or through DRI_PRIME. Could you please provide dmesg with the changes suggested by pietinger?

Also it would be interesting to see what switcherooctl outputs.

I also compiled and booted 6.6.3 and it looks like working stable so far, but sometimes there's a little artifact left where the cursor was when deleting text, so I don't trust it to be really stable. I'd wait a few more iterations before I switch to 6.6.

Best Regards,
Georgi

KShots · Post by **KShots** » Sun Dec 03, 2023 6:40 am

The problems I had with 6.6.3 was that the oops was still there and that my ZFS out-of-kernel build didn't like it much - it kept halting on disk IO. I had to upgrade ZFS way past stable too to get it to even try compiling... given that the oops was still there, I didn't think it worth the effort to continue down that road just yet.

I did get the machine rebooted after the data transfer. The touchpad fired up right away, and during the reboot I actually found a BIOS option to disable the integrated GPU (disable hybrid GPU)... but that said, I still get an oops (pastebin).

Code: Select all

[    4.680849] ------------[ cut here ]------------
[    4.680851] WARNING: CPU: 6 PID: 572 at drivers/gpu/drm/amd/amdgpu/../display/dc/dcn20/dcn20_dsc.c:273 dsc2_disable+0x147/0x160 [amdgpu]
[    4.680993] Modules linked in: amdgpu(+) drm_ttm_helper ttm agpgart gpu_sched i2c_algo_bit drm_buddy drm_display_helper drm_kms_helper syscopyarea sysfillrect crc32_pclmul rtsx_pci_sdmmc sysimgblt fb_sys_fops crc32c_intel ccp polyval_clmulni mmc_core polyval_generic drm ghash_clmulni_intel sha1_generic xhci_pci sha512_ssse3 aesni_intel crypto_simd cryptd rtsx_pci rng_core xhci_hcd cec mfd_core nvme ucsi_acpi nvme_core typec_ucsi roles typec i2c_hid_acpi i2c_hid dm_mod
[    4.681011] CPU: 6 PID: 572 Comm: (udev-worker) Tainted: G                T  6.1.57-gentoo #18
[    4.681013] Hardware name: Alienware Alienware m18 R1 AMD/049YP9, BIOS 1.7.0 09/21/2023
[    4.681014] RIP: 0010:dsc2_disable+0x147/0x160 [amdgpu]
[    4.681093] Code: 00 00 00 75 2c 48 83 c4 18 5b e9 d4 a3 3b c6 8b 4b 10 44 8b 44 24 0c be 02 00 00 00 31 ff 48 c7 c2 a0 a1 ed c0 e8 39 19 b5 ff <0f> 0b e9 6e ff ff ff e8 1d e0 02 c6 66 66 2e 0f 1f 84 00 00 00 00
[    4.681094] RSP: 0018:ffffb6ea40bd3148 EFLAGS: 00010246
[    4.681096] RAX: 0000000000000000 RBX: ffff9a554d3e7000 RCX: 0000000000000000
[    4.681097] RDX: ffffffffc0eda1a0 RSI: 0000000000000002 RDI: 0000000000000000
[    4.681097] RBP: 0000000000000001 R08: 0000000000000000 R09: 0000000000000004
[    4.681098] R10: 0000000000000001 R11: 0000000000000163 R12: ffff9a5570b001e8
[    4.681099] R13: ffff9a555fc69000 R14: ffff9a5544514c00 R15: ffff9a5570b001e8
[    4.681099] FS:  00007f3546215340(0000) GS:ffff9a643d780000(0000) knlGS:0000000000000000
[    4.681100] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[    4.681100] CR2: 00007f354614f000 CR3: 000000010ce52000 CR4: 0000000000750ee0
[    4.681101] PKRU: 55555554
[    4.681101] Call Trace:
[    4.681104]  <TASK>
[    4.681105]  ? dsc2_disable+0x147/0x160 [amdgpu]
[    4.681167]  ? __warn+0x74/0xc0
[    4.681172]  ? dsc2_disable+0x147/0x160 [amdgpu]
[    4.681228]  ? report_bug+0xe6/0x170
[    4.681232]  ? handle_bug+0x3a/0x70
[    4.681236]  ? exc_invalid_op+0x13/0x60
[    4.681237]  ? asm_exc_invalid_op+0x16/0x20
[    4.681240]  ? dsc2_disable+0x147/0x160 [amdgpu]
[    4.681295]  ? dsc2_disable+0x147/0x160 [amdgpu]
[    4.681349]  dp_set_dsc_on_stream+0x30d/0x3e0 [amdgpu]
[    4.681438]  ? dm_helpers_dp_write_dsc_enable+0x25b/0x5b0 [amdgpu]
[    4.681524]  dp_set_dsc_enable+0x6c/0x80 [amdgpu]
[    4.681598]  core_link_disable_stream+0x155/0x550 [amdgpu]
[    4.681664]  ? dc_validate_with_context+0x473/0x4f0 [amdgpu]
[    4.681721]  ? dm_read_reg_func+0x33/0xa0 [amdgpu]
[    4.681786]  dc_commit_state_no_check+0xc8d/0xd00 [amdgpu]
[    4.681847]  dc_commit_streams+0x2aa/0x410 [amdgpu]
[    4.681904]  dc_commit_state+0xe6/0x120 [amdgpu]
[    4.681959]  amdgpu_dm_atomic_commit_tail+0x619/0x31d0 [amdgpu]
[    4.682017]  ? vscnprintf+0xe/0x30
[    4.682019]  ? printk_sprint+0x105/0x120
[    4.682023]  ? update_load_avg+0x5f/0x6e0
[    4.682026]  ? newidle_balance.isra.0+0x1fe/0x3d0
[    4.682027]  ? __cgroup_account_cputime+0x47/0x70
[    4.682030]  ? _raw_spin_unlock+0xd/0x30
[    4.682031]  ? finish_task_switch.isra.0+0x93/0x2c0
[    4.682034]  ? __schedule+0x2bf/0x10b0
[    4.682036]  ? dma_resv_iter_next_unlocked+0x7c/0x90
[    4.682040]  ? dma_resv_get_fences+0xa1/0x220
[    4.682041]  ? schedule+0x62/0xb0
[    4.682042]  ? schedule_timeout+0x113/0x150
[    4.682043]  ? preempt_count_add+0x65/0xa0
[    4.682043]  ? _raw_spin_lock_irqsave+0x12/0x40
[    4.682044]  ? _raw_spin_unlock_irqrestore+0x16/0x30
[    4.682045]  ? dma_fence_default_wait+0x169/0x210
[    4.682046]  ? __bpf_trace_dma_fence+0x10/0x10
[    4.682048]  commit_tail+0x8f/0x120 [drm_kms_helper]
[    4.682058]  drm_atomic_helper_commit+0x111/0x140 [drm_kms_helper]
[    4.682063]  drm_atomic_commit+0x91/0xc0 [drm]
[    4.682081]  ? drm_plane_get_damage_clips.cold+0x1c/0x1c [drm]
[    4.682091]  drm_client_modeset_commit_atomic+0x1f5/0x230 [drm]
[    4.682100]  drm_client_modeset_commit_locked+0x51/0x160 [drm]
[    4.682109]  drm_client_modeset_commit+0x1c/0x40 [drm]
[    4.682117]  drm_fb_helper_set_par+0x99/0xd0 [drm_kms_helper]
[    4.682122]  fbcon_init+0x273/0x630
[    4.682126]  visual_init+0xc7/0x120
[    4.682130]  do_bind_con_driver.isra.0+0x1b5/0x2c0
[    4.682131]  do_take_over_console+0x111/0x180
[    4.682133]  do_fbcon_takeover+0x55/0xc0
[    4.682134]  fbcon_fb_registered+0x12e/0x140
[    4.682135]  register_framebuffer+0x1cf/0x310
[    4.682139]  __drm_fb_helper_initial_config_and_unlock+0x383/0x4e0 [drm_kms_helper]
[    4.682144]  drm_fbdev_client_hotplug+0x181/0x210 [drm_kms_helper]
[    4.682148]  drm_client_register+0x58/0x90 [drm]
[    4.682157]  amdgpu_pci_probe+0x35e/0x370 [amdgpu]
[    4.682220]  local_pci_probe+0x3c/0x80
[    4.682225]  pci_device_probe+0xa7/0x1b0
[    4.682226]  really_probe+0xd4/0x280
[    4.682229]  ? pm_runtime_barrier+0x4b/0x80
[    4.682232]  __driver_probe_device+0x6e/0xe0
[    4.682233]  driver_probe_device+0x1a/0xe0
[    4.682234]  __driver_attach+0x7f/0x180
[    4.682235]  ? __device_attach_driver+0x110/0x110
[    4.682235]  bus_for_each_dev+0x82/0xc0
[    4.682238]  bus_add_driver+0x181/0x1d0
[    4.682239]  driver_register+0x84/0xe0
[    4.682240]  ? 0xffffffffc1155000
[    4.682240]  do_one_initcall+0x54/0x200
[    4.682244]  do_init_module+0x42/0x1d0
[    4.682247]  __do_sys_finit_module+0x9e/0xf0
[    4.682249]  do_syscall_64+0x3b/0x90
[    4.682252]  entry_SYSCALL_64_after_hwframe+0x64/0xce
[    4.682253] RIP: 0033:0x7f3546da56a9
[    4.682255] Code: 08 89 e8 5b 5d c3 66 2e 0f 1f 84 00 00 00 00 00 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 3f c7 0c 00 f7 d8 64 89 01 48
[    4.682255] RSP: 002b:00007ffdb8d204b8 EFLAGS: 00000246 ORIG_RAX: 0000000000000139
[    4.682256] RAX: ffffffffffffffda RBX: 000055be8b63ed30 RCX: 00007f3546da56a9
[    4.682256] RDX: 0000000000000000 RSI: 00007f3546e8fb95 RDI: 000000000000001a
[    4.682257] RBP: 00007f3546e8fb95 R08: 0000000000000000 R09: 000055be8b613560
[    4.682257] R10: 000000000000001a R11: 0000000000000246 R12: 0000000000020000
[    4.682257] R13: 0000000000000000 R14: 000055be8b64b790 R15: 0000000000000000
[    4.682258]  </TASK>
[    4.682258] ---[ end trace 0000000000000000 ]---

KShots · Post by **KShots** » Sun Dec 03, 2023 6:44 am

logrusx wrote:Also it would be interesting to see what switcherooctl outputs.

With the BIOS option disabled, I get the following:

Code: Select all

succubus ~ # switcherooctl
Device: 0
  Name:        Advanced Micro Devices, Inc. [AMD®/ATI] Navi 31 [Radeon RX 7900 XT/7900 XTX]
  Default:     yes
  Environment: DRI_PRIME=pci-0000_03_00_0

logrusx · Post by **logrusx** » Sun Dec 03, 2023 7:35 am

KShots wrote:
logrusx wrote:Also it would be interesting to see what switcherooctl outputs.
With the BIOS option disabled, I get the following:
Code: Select all
succubus ~ # switcherooctl
Device: 0
  Name:        Advanced Micro Devices, Inc. [AMD®/ATI] Navi 31 [Radeon RX 7900 XT/7900 XTX]
  Default:     yes
  Environment: DRI_PRIME=pci-0000_03_00_0

It would be more interesting to see it with both adapters enabled.

Best Regards,
Georgi

KShots · Post by **KShots** » Sun Dec 03, 2023 2:54 pm

logrusx wrote:
KShots wrote:
logrusx wrote:Also it would be interesting to see what switcherooctl outputs.
With the BIOS option disabled, I get the following:
Code: Select all
succubus ~ # switcherooctl
Device: 0
  Name:        Advanced Micro Devices, Inc. [AMD®/ATI] Navi 31 [Radeon RX 7900 XT/7900 XTX]
  Default:     yes
  Environment: DRI_PRIME=pci-0000_03_00_0
It would be more interesting to see it with both adapters enabled.

Best Regards,
Georgi

Yeah, given that disabling the integrated GPU didn't fix anything, easy enough:

Code: Select all

succubus ~ # switcherooctl 
Device: 0
  Name:        Advanced Micro Devices, Inc. [AMD®/ATI] Raphael
  Default:     yes
  Environment: DRI_PRIME=pci-0000_0c_00_0

Device: 1
  Name:        Advanced Micro Devices, Inc. [AMD®/ATI] Navi 31 [Radeon RX 7900 XT/7900 XTX]
  Default:     no
  Environment: DRI_PRIME=pci-0000_03_00_0

logrusx · Post by **logrusx** » Sun Dec 03, 2023 4:30 pm

Does it still not see your external displays? It should now. You should also be able to run anything you want on the dGPU with switcherooctl.

Best Regards,
Georgi

KShots · Post by **KShots** » Sun Dec 03, 2023 5:35 pm

logrusx wrote:Does it still not see your external displays? It should now. You should also be able to run anything you want on the dGPU with switcherooctl.

Best Regards,
Georgi

Afraid not - the external displays are still not detected under gnome. I went ahead and verified they at least worked under Windows (because... new laptop, coulda been defective... but no, it works there).

logrusx · Post by **logrusx** » Sun Dec 03, 2023 6:46 pm

KShots wrote:
logrusx wrote:Does it still not see your external displays? It should now. You should also be able to run anything you want on the dGPU with switcherooctl.

Best Regards,
Georgi
Afraid not - the external displays are still not detected under gnome. I went ahead and verified they at least worked under Windows (because... new laptop, coulda been defective... but no, it works there).

Try Wayland too, also when I connect my TV to the laptop through HDMI, I need to first open Gnome Settings and navigate to the Displays section. Sometimes it sees them only after I disconnect and reconnect the cable. That's for Wayland, I don't know how it's with X, maybe it has the same bug.

Can anyone of the more knowledgeable fellows here help with commands that'll tell if at all the displays are connected? Now I remember something related to EDID data and tools that could read it, in another thread about AMD graphics, I think it was 6700. I'll try to find it.

EDIT: found it: viewtopic-t-1164907.html
The tool is x11-misc/read-edid. See the thread for additional i2c kernel settings as well.

Best Regards,
Georgi

KShots · Post by **KShots** » Sun Dec 03, 2023 11:54 pm

logrusx wrote:
KShots wrote:
logrusx wrote:Does it still not see your external displays? It should now. You should also be able to run anything you want on the dGPU with switcherooctl.

Best Regards,
Georgi
Afraid not - the external displays are still not detected under gnome. I went ahead and verified they at least worked under Windows (because... new laptop, coulda been defective... but no, it works there).
Try Wayland too, also when I connect my TV to the laptop through HDMI, I need to first open Gnome Settings and navigate to the Displays section. Sometimes it sees them only after I disconnect and reconnect the cable. That's for Wayland, I don't know how it's with X, maybe it has the same bug.

According to XDG_SESSION_TYPE, I'm running wayland

logrusx wrote:Can anyone of the more knowledgeable fellows here help with commands that'll tell if at all the displays are connected? Now I remember something related to EDID data and tools that could read it, in another thread about AMD graphics, I think it was 6700. I'll try to find it.

EDIT: found it: viewtopic-t-1164907.html
The tool is x11-misc/read-edid. See the thread for additional i2c kernel settings as well.

Best Regards,
Georgi

Got it... not good, though... with or without (meaning it doesn't change) a screen plugged into the mini-DP, I get the following from get-edid:

Code: Select all

succubus ~ # get-edid
This is read-edid version 3.0.2. Prepare for some fun.
Attempting to use i2c interface
No EDID on bus 0
No EDID on bus 1
No EDID on bus 2
No EDID on bus 3
No EDID on bus 4
No EDID on bus 5
No EDID on bus 6
No EDID on bus 7
No EDID on bus 8
No EDID on bus 9
No EDID on bus 10
No EDID on bus 11
No EDID on bus 12
No EDID on bus 13
No EDID on bus 14
No EDID on bus 15
No EDID on bus 17
No EDID on bus 18
No EDID on bus 19
No EDID on bus 21
No EDID on bus 23
No EDID on bus 24
3 potential busses found: 16 20 22
Will scan through until the first EDID is found.
Pass a bus number as an option to this program to go only for that one.
256-byte EDID successfully retrieved from i2c bus 16
 �'x��RC�&PT�t�Vp�(@0 �����DN9CJ�180JME
A�A
Looks like i2c was successful. Have a good day.
  �p y"(�7�U/�'	�7U/��	%	�7�7x���r�x�SQSQ��

Just for the hell of it, I also ran xrandr:

Code: Select all

Screen 0: minimum 16 x 16, current 1920 x 1200, maximum 32767 x 32767
eDP-2 connected primary 1920x1200+0+0 (normal left inverted right x axis y axis) 390mm x 240mm
   1920x1200    479.85*+
   1600x1200    479.85  
   1440x1080    479.76  
   1400x1050    479.81  
   1280x1024    479.70  
   1280x960     479.79  
   1152x864     479.97  
   1024x768     479.75  
   800x600      479.37  
   640x480      479.04  
   320x240      477.30  
   1680x1050    479.85  
   1440x900     479.89  
   1280x800     479.53  
   1152x720     479.53  
   960x600      479.55  
   928x580      479.46  
   800x500      479.39  
   768x480      479.20  
   720x480      479.42  
   640x400      479.12  
   320x200      478.06  
   1920x1080    479.79  
   1600x900     479.93  
   1368x768     479.70  
   1280x720     479.67  
   1024x576     479.80  
   864x486      479.51  
   720x400      479.29  
   640x350      479.56

It only sees the integrated display

EDIT: Also of note, I went ahead and added (as modules) every single I2C hardware device, as well as all LCD panels found uder device drivers->graphics. I note that until I added the rest of the I2C modules, the above only displayed that it couldn't find any buses.

KShots · Post by **KShots** » Mon Dec 04, 2023 1:07 am

Ok, some more progress. The Gentoo live DVD actually supports the external DP output. Here's the corresponding dmesg output from that session.

logrusx · Post by **logrusx** » Mon Dec 04, 2023 6:08 am

KShots wrote:Ok, some more progress. The Gentoo live DVD actually supports the external DP output. Here's the corresponding dmesg output from that session.

Lsmod would give you better idea which drivers you need. Well not sure about better, but it's certainly easier. Also, you still haven't tried the bin kernel, have you?

Best Regards,
Georgi

KShots · Post by **KShots** » Tue Dec 05, 2023 9:44 pm

Odd... I took the kernel .config from the live USB and used it to generate a kernel on the install... no external displays when I do that. What else could be causing that? Also, I never got around to testing the HDMI output previously (I was just using the mini-DP)... it appears the HDMI doesn't work either (either under live (even when mini-DP does work) or install). *sigh*

New at Gentoo · Post by **pietinger** » Tue Dec 05, 2023 10:00 pm

KShots wrote:Also, I never got around to testing the HDMI output previously (I was just using the mini-DP)... it appears the HDMI doesn't work either (either under live (even when mini-DP does work) or install). *sigh*

If your kernel is able to send signals to your DisplayPort then the kernel is automatically able to use also the HDMI port (the kernel does not differ in its configuration) ... if HDMI is not disabled in the BIOS. This means: Maybe ... only maybe your HDMI cable has a problem (or your monitor has a faulty HDMI port).

KShots · Post by **KShots** » Sat Dec 09, 2023 5:40 pm

So, I updated to kernel 6.1.66... and the displayport seems to be working intermittently. I managed to get it working this morning and started using the machine... and it fell flat on its face. Upon playing an h265 video, the desktop corrupts with graphical artifacts and I get the below (as a sample, full linked):

Code: Select all

[  595.375512] ACPI BIOS Error (bug): Failure creating named object [\_SB.ALIB.****], AE_ALREADY_EXISTS (20220331/dswload2-326)
[  595.375518] ACPI Error: AE_ALREADY_EXISTS, During name lookup/catalog (20220331/psobject-220)
[  595.375520] ACPI Error: Aborting method \_SB.ALIB due to previous error (AE_ALREADY_EXISTS) (20220331/psparse-529)
[  595.375522] ACPI Error: Aborting method \_SB.PCI0.GPP2.M629 due to previous error (AE_ALREADY_EXISTS) (20220331/psparse-529)
[  595.375524] ACPI Error: Aborting method \_SB.PCI0.GPP2.PPS3 due to previous error (AE_ALREADY_EXISTS) (20220331/psparse-529)
[  595.375526] ACPI Error: Aborting method \_SB.PCI0.GPP2._PS3 due to previous error (AE_ALREADY_EXISTS) (20220331/psparse-529)
[  665.454315] ACPI BIOS Error (bug): The DSDT has been corrupted or replaced - old, new headers below (20220331/tbutils-87)
[  665.454321] ACPI: DSDT 0x0000000000000000 00D416 (v02 ALWARE WN09     00000002 ASL  00040000)
[  665.454323] ACPI: ???? 0x0000000000000000 80808080 (v128 ?????? ???????? 80808080 ???? 80808080)
[  665.454325] ACPI Error: Please send DMI info to linux-acpi@vger.kernel.org
               If system does not work as expected, please boot with acpi=copy_dsdt (20220331/tbutils-94)
[  665.454331] ACPI Error: Null stack entry at 00000000923dad19 (20220331/exresop-139)
[  665.454333] ACPI Error: AE_AML_INTERNAL, While resolving operands for [OpcodeName unavailable] (20220331/dswexec-431)
[  665.454335] ACPI Error: Aborting method \_SB.PCI0.GP19.XHC2._S0W due to previous error (AE_AML_INTERNAL) (20220331/psparse-529)
[  665.454340] ACPI Error: Null stack entry at 00000000923dad19 (20220331/exresop-139)
[  665.454341] ACPI Error: AE_AML_INTERNAL, While resolving operands for [OpcodeName unavailable] (20220331/dswexec-431)
[  665.454342] ACPI Error: Aborting method \_SB.PCI0.GP19.XHC2._S0W due to previous error (AE_AML_INTERNAL) (20220331/psparse-529)
[  665.466358] ACPI Error: Null stack entry at 00000000923dad19 (20220331/exresop-139)
[  665.466360] ACPI Error: AE_AML_INTERNAL, While resolving operands for [OpcodeName unavailable] (20220331/dswexec-431)

... and under no circumstances does the HDMI output ever work (still).

KShots · Post by **KShots** » Sat Dec 09, 2023 5:44 pm

pietinger wrote:
KShots wrote:Also, I never got around to testing the HDMI output previously (I was just using the mini-DP)... it appears the HDMI doesn't work either (either under live (even when mini-DP does work) or install). *sigh*
If your kernel is able to send signals to your DisplayPort then the kernel is automatically able to use also the HDMI port (the kernel does not differ in its configuration) ... if HDMI is not disabled in the BIOS. This means: Maybe ... only maybe your HDMI cable has a problem (or your monitor has a faulty HDMI port).

... you might be right - windows isn't showing it either. That piece of it sounds like a message to dell support... HDMI should work under windows... and a call later, traced it down to an intermittent HDMI cable (wow, amazon display cables are getting rather shoddy)

EDIT: After more research, I ordered the wrong cable - that was a DisplayPort to HDMI cable, not an HDMI to DisplayPort cable (Apparently these are directional). Looking at the horror stories of going from HDMI to DP, I've decided to order a USB-C to DisplayPort cable instead (dell support claims the USB-C ports have DP Alt Mode enabled, even though they're not Thunderbolt). God this thing's turning into a nightmare

EDIT2: Got the Type-C to DP cable - it works... but it flickers on & off, sometime as often as once/5 seconds, sometimes as far as once/several minutes (not much consistency between boots). Not sure there's a reasonable means of connecting two high-res displays to this laptop :/

KShots · Post by **KShots** » Fri Dec 15, 2023 3:03 am

An update - found a means of stabilizing the machine somewhat rather than waiting for an update. If I disable the Ryzen GPU in the BIOS, the machine stops freezing and having graphical artifacts at this point. I still get the early kernel oops, and thus far I get rather miserable (<5 FPS) in pretty much any steam/lutris game... but at least it's not freezing anymore. Any insight on the miserable potato performance?

For now, my test game is Dyson Sphere Program under Steam... if someone has another preference (maybe some form of Vulkan test app) I can certainly use that for tracking this...

KShots · Post by **KShots** » Wed Dec 20, 2023 3:59 pm

Update for kernel 6.1.67 and linux-firmware-20231211 - still has the oops

KShots · Post by **KShots** » Fri Dec 29, 2023 10:53 pm

Update for media-libs/mesa-23.3.2 - I seem to get reasonable framerates now - went from <5 FPS in dyson sphere program to >140 FPS. Only issue left is that I still get that OOPS during boot.

nvaert1986 · Post by **nvaert1986** » Wed Jan 10, 2024 11:06 am

For the AMD Radeon RX7000 series I'd suggest using kernel 6.2 or newer as many bugs have been fixed in more recent kernels. My PC build had a black screen on cold boot using the 6.1.x series kernel. Upgrading it to 6.6.9 seems to have resolved that issue (so far). Also be sure to configure the kernel correctly acccording to: https://wiki.gentoo.org/wiki/AMDGPU as there's a few gotchas like HSA support and Unaddressable memory / Device Memory support.

What were the errors you received on 6.6.3?

Radeon RX7900M support?

Radeon RX7900M support?

Re: Radeon RX7900M support?