Test AMD RX 7600 GPU

[geerlingguy]

Following on my testing with the AMD Radeon Pro W7700, I would like to see if an ASRock Challenger AMD Radeon RX 7600 8GB runs on the Pi 5.

See the W7700/7700 XT issue for setup instructions.

[geerlingguy]

Got an Oops! @Coreforge hunts these down like it's deer season.

[    4.688899] amdgpu 0000:03:00.0: amdgpu: RAS: optional ras ta ucode is not available
[    4.705344] amdgpu 0000:03:00.0: amdgpu: RAP: optional rap ta ucode is not available
[    4.705357] amdgpu 0000:03:00.0: amdgpu: SECUREDISPLAY: securedisplay ta ucode is not available
[    4.705405] amdgpu 0000:03:00.0: amdgpu: smu driver if version = 0x00000035, smu fw if version = 0x0000003c, smu fw program = 0, smu fw version = 0x00524500 (82.69.0)
[    4.705408] amdgpu 0000:03:00.0: amdgpu: SMU driver if version not matched
[    4.745485] amdgpu 0000:03:00.0: amdgpu: SMU is initialized successfully!
[    4.745963] [drm] Display Core v3.2.247 initialized on DCN 3.2.1
[    4.745968] [drm] DP-HDMI FRL PCON supported
[    4.748579] [drm] DMUB hardware initialized: version=0x07001900
[    5.009654] [drm] kiq ring mec 3 pipe 1 q 0
[    5.014614] [drm] VCN decode and encode initialized successfully(under DPG Mode).
[    5.015078] amdgpu 0000:03:00.0: [drm:jpeg_v4_0_hw_init [amdgpu]] JPEG decode initialized successfully.
[    5.016362] amdgpu 0000:03:00.0: amdgpu: SE 2, SH per SE 2, CU per SH 8, active_cu_number 32
[    5.016803] amdgpu 0000:03:00.0: amdgpu: ring gfx_0.0.0 uses VM inv eng 0 on hub 0
[    5.016806] amdgpu 0000:03:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 1 on hub 0
[    5.016808] amdgpu 0000:03:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 4 on hub 0
[    5.016809] amdgpu 0000:03:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 6 on hub 0
[    5.016811] amdgpu 0000:03:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 7 on hub 0
[    5.016812] amdgpu 0000:03:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 8 on hub 0
[    5.016814] amdgpu 0000:03:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 9 on hub 0
[    5.016816] amdgpu 0000:03:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 10 on hub 0
[    5.016817] amdgpu 0000:03:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 11 on hub 0
[    5.016819] amdgpu 0000:03:00.0: amdgpu: ring sdma0 uses VM inv eng 12 on hub 0
[    5.016820] amdgpu 0000:03:00.0: amdgpu: ring sdma1 uses VM inv eng 13 on hub 0
[    5.016822] amdgpu 0000:03:00.0: amdgpu: ring vcn_unified_0 uses VM inv eng 0 on hub 8
[    5.016823] amdgpu 0000:03:00.0: amdgpu: ring jpeg_dec uses VM inv eng 1 on hub 8
[    5.016825] amdgpu 0000:03:00.0: amdgpu: ring mes_kiq_3.1.0 uses VM inv eng 14 on hub 0
[    5.017964] Unable to handle kernel paging request at virtual address ffffffc08208d000
[    5.026457] Mem abort info:
[    5.029409]   ESR = 0x0000000096000061
[    5.029412]   EC = 0x25: DABT (current EL), IL = 32 bits
[    5.029414]   SET = 0, FnV = 0
[    5.029415]   EA = 0, S1PTW = 0
[    5.029416]   FSC = 0x21: alignment fault
[    5.029417] Data abort info:
[    5.029417]   ISV = 0, ISS = 0x00000061, ISS2 = 0x00000000
[    5.029418]   CM = 0, WnR = 1, TnD = 0, TagAccess = 0
[    5.029420]   GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0
[    5.029421] swapper pgtable: 4k pages, 39-bit VAs, pgdp=000000000137f000
[    5.029423] [ffffffc08208d000] pgd=1000000100165003, p4d=1000000100165003, pud=1000000100165003, pmd=100000010f037003, pte=00680001029a0f0f
[    5.029428] Internal error: Oops: 0000000096000061 [#1] PREEMPT SMP
[    5.029431] Modules linked in: bnep binfmt_misc spidev brcmfmac_wcc amdgpu(+) hci_uart btbcm aes_ce_blk aes_ce_cipher ghash_ce gf128mul bluetooth sha2_ce sha256_arm64 sha1_ce ecdh_generic vc4 ecc raspberrypi_hwmon libaes rpivid_hevc(C) pisp_be v4l2_mem2mem brcmfmac snd_soc_hdmi_codec videobuf2_dma_contig snd_usb_audio brcmutil cec i2c_brcmstb drm_dma_helper spi_bcm2835 cfg80211 snd_soc_core amdxcp snd_hwdep snd_usbmidi_lib drm_exec snd_compress videobuf2_memops i2c_algo_bit snd_rawmidi drm_buddy videobuf2_v4l2 snd_pcm_dmaengine drm_suballoc_helper snd_seq_device v3d drm_display_helper gpu_sched gpio_keys videodev snd_pcm rfkill drm_shmem_helper drm_ttm_helper snd_timer ttm videobuf2_common pwm_fan drm_kms_helper hid_apple mc snd raspberrypi_gpiomem rp1_adc nvmem_rmem joydev hid_multitouch sg uio_pdrv_genirq uio uinput cuse i2c_dev ledtrig_pattern drm fuse dm_mod drm_panel_orientation_quirks backlight ip_tables x_tables ipv6
[    5.029491] CPU: 1 PID: 336 Comm: (udev-worker) Tainted: G         C         6.6.60-v8-AMDGPU+ #1
[    5.029494] Hardware name: Raspberry Pi 5 Model B Rev 1.0 (DT)
[    5.029496] pstate: 40400009 (nZcv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[    5.029498] pc : __memset+0x16c/0x188
[    5.029505] lr : amdgpu_mes_ctx_alloc_meta_data+0x68/0xc0 [amdgpu]
[    5.029865] sp : ffffffc08093b520
[    5.029866] x29: ffffffc08093b520 x28: ffffff8102c904c8 x27: 0000000000000000
[    5.029869] x26: ffffff8102cb67b8 x25: ffffff8102c904c8 x24: ffffff8102cc0000
[    5.029871] x23: 0000000000000001 x22: 0000000000008000 x21: ffffff8102c80000
[    5.029874] x20: 0000000000000000 x19: ffffffc08093b628 x18: ffffffffffffffff
[    5.029876] x17: ffffff81fef97250 x16: ffffffd084ced580 x15: 0000000000000000
[    5.029878] x14: 0000000000000000 x13: 0000000000000000 x12: 0000000000000000
[    5.029880] x11: 0000000000000000 x10: ffffff81fef855c0 x9 : 0000000000000000
[    5.029883] x8 : ffffffc08208d000 x7 : 0000000000000000 x6 : 000000000000003f
[    5.029885] x5 : 0000000000000040 x4 : 0000000000000000 x3 : 0000000000000004
[    5.029887] x2 : 0000000000033fc0 x1 : 0000000000000000 x0 : ffffffc08208d000
[    5.029890] Call trace:
[    5.029891]  __memset+0x16c/0x188
[    5.029894]  amdgpu_mes_self_test+0xc0/0x450 [amdgpu]
[    5.030135]  mes_v11_0_late_init+0x58/0x78 [amdgpu]
[    5.030373]  amdgpu_device_ip_late_init+0x68/0x308 [amdgpu]
[    5.030608]  amdgpu_device_init+0x2038/0x2208 [amdgpu]
[    5.309586]  amdgpu_driver_load_kms+0x20/0x1a8 [amdgpu]
[    5.309875]  amdgpu_pci_probe+0x154/0x420 [amdgpu]
[    5.310099]  local_pci_probe+0x48/0xb8
[    5.310102]  pci_device_probe+0xac/0x1c8
[    5.310104]  really_probe+0x150/0x2c0
[    5.310107]  __driver_probe_device+0x80/0x140
[    5.310109]  driver_probe_device+0x44/0x170
[    5.310110]  __driver_attach+0x9c/0x1b0
[    5.310112]  bus_for_each_dev+0x80/0xe8
[    5.310116]  driver_attach+0x2c/0x40
[    5.310117]  bus_add_driver+0xec/0x218
[    5.310119]  driver_register+0x68/0x138
[    5.310121]  __pci_register_driver+0x54/0x68
[    5.310123]  amdgpu_init+0x6c/0xff8 [amdgpu]
[    5.310346]  do_one_initcall+0x60/0x2c0
[    5.310349]  do_init_module+0x60/0x218
[    5.310353]  load_module+0x1dd0/0x2080
[    5.310354]  __do_sys_init_module+0x19c/0x1e0
[    5.310356]  __arm64_sys_init_module+0x24/0x38
[    5.310357]  invoke_syscall+0x50/0x128
[    5.310360]  el0_svc_common.constprop.0+0xc8/0xf0
[    5.310363]  do_el0_svc+0x24/0x38
[    5.310365]  el0_svc+0x40/0xe8
[    5.310368]  el0t_64_sync_handler+0x100/0x130
[    5.310369]  el0t_64_sync+0x190/0x198
[    5.310372] Code: 91010108 54ffff4a 8b040108 cb050042 (d50b7428) 
[    5.310374] ---[ end trace 0000000000000000 ]---

[geerlingguy]

@DanaGoyette - For decoding the stack traces, I'm trying to use decode_stacktrace.sh, but it seems to be failing a little:

pi@pi5-pcie:~/linux/scripts $ ./decode_stacktrace.sh /boot/firmware/kernel_2712.img < log.txt
[    5.726629] __memset+0x16c/0x188 
readelf: Error: Not an ELF file - it has the wrong magic bytes at the start
WARNING! Modules path isn't set, but is needed to parse this symbol
[    5.726632] amdgpu_mes_self_test+0xc0/0x450 amdgpu
...

I assumed that's my vmlinuz as I compile and copy it with:

make -j6 Image.gz modules dtbs
sudo cp arch/arm64/boot/Image.gz /boot/firmware/$KERNEL.img

Or am I missing something obvious?

Ah... same thing with the right vmlinux:

pi@pi5-pcie:~/linux $ ./scripts/decode_stacktrace.sh /usr/lib/modules/6.6.60-v8-AMDGPU+/build/vmlinux /home/pi/linux < scripts/log.txt
[    5.726629] __memset (??:?) 
readelf: Error: Not an ELF file - it has the wrong magic bytes at the start
WARNING! Modules path isn't set, but is needed to parse this symbol

(that dir is just a symlink back to build -> /home/pi/linux)

[geerlingguy]

Ah... looking at https://github.com/torvalds/linux/blob/master/scripts/decode_stacktrace.sh, I noticed it's not unzipping (well, un-xzing) the module files, so it's plain readelf calls are erroring out.

# Doesn't work
$ readelf -WS /lib/modules/6.6.60-v8-AMDGPU+/kernel/drivers/gpu/drm/amd/amdgpu/amdgpu.ko.xz 
readelf: Error: Not an ELF file - it has the wrong magic bytes at the start

# Works
$ sudo unxz /lib/modules/6.6.60-v8-AMDGPU+/kernel/drivers/gpu/drm/amd/amdgpu/amdgpu.ko.xz
$ readelf -WS /lib/modules/6.6.60-v8-AMDGPU+/kernel/drivers/gpu/drm/amd/amdgpu/amdgpu.ko
There are 63 section headers, starting at offset 0xae1e50:
...

Thanks to this LKML message hinting at that.

Module compression is a compile-time option, and my Kernel was configured with MODULE_COMPRESS_XZ. In menuconfig I changed it to Enable loadable module support > Module compression mode > None and will recompile.

[geerlingguy]

Hmm... closer this time, but still no dice:

pi@pi5-pcie:~/linux $ ./scripts/decode_stacktrace.sh /home/pi/linux/vmlinux /lib/modules/6.6.60-v8-AMDGPU+ < trace.txt
[    6.527910] __memset (??:?) 
WARNING! Modules path isn't set, but is needed to parse this symbol
[    6.531234] amdgpu_mes_self_test+0xc0/0x450 amdgpu
WARNING! Modules path isn't set, but is needed to parse this symbol
[    6.536646] mes_v11_0_late_init+0x58/0x78 amdgpu

Heh, found scripts/decode_stacktrace.sh: remove find_module recursion and improve error reporting

Changed Kernel hacking > Compile-time checks and compiler options > Debug information to Rely on the toolchain's implicit default DWARF version (was on Disable debug information) and recompiling again...

[Coreforge]

I just pushed all of the 7000 series stuff, hopefully this oops should be fixed too.

[geerlingguy]

@Coreforge - Thanks! I assume it's all in your https://github.com/Coreforge/linux/tree/rpi-6.6.y-gpu branch? (I see latest commit 45 min ago). I will still do some testing for this issue before pulling that code because I'd like to get the 'paste the stack trace, get the line of code' debugging working :D

[geerlingguy]

That worked!

pi@pi5-pcie:~/linux $ ./scripts/decode_stacktrace.sh /home/pi/linux/vmlinux /lib/modules/6.6.60-v8-AMDGPU+ < trace.txt
[    5.542863] Call trace:
[    5.545310] __memset (/home/pi/linux/arch/arm64/lib/memset.S:197) 
[    5.548634] amdgpu_mes_self_test (/home/pi/linux/drivers/gpu/drm/amd/amdgpu/amdgpu_mes.c:1367) amdgpu
[    5.553950] mes_v11_0_late_init (/home/pi/linux/drivers/gpu/drm/amd/amdgpu/mes_v11_0.c:1320) amdgpu
[    5.559071] amdgpu_device_ip_late_init (/home/pi/linux/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:2650) amdgpu
[    5.564886] amdgpu_device_init (/home/pi/linux/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c:3858) amdgpu
[    5.570262] amdgpu_driver_load_kms (/home/pi/linux/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c:147) amdgpu
[    5.575724] amdgpu_pci_probe (/home/pi/linux/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c:2145) amdgpu
[    5.580749] local_pci_probe (/home/pi/linux/drivers/pci/pci-driver.c:325) 
[    5.584508] pci_device_probe (/home/pi/linux/drivers/pci/pci-driver.c:392 /home/pi/linux/drivers/pci/pci-driver.c:417 /home/pi/linux/drivers/pci/pci-driver.c:460) 
[    5.588439] really_probe (/home/pi/linux/drivers/base/dd.c:579 /home/pi/linux/drivers/base/dd.c:658) 
[    5.592110] __driver_probe_device (/home/pi/linux/drivers/base/dd.c:800) 
[    5.596478] driver_probe_device (/home/pi/linux/drivers/base/dd.c:830) 
[    5.600671] __driver_attach (/home/pi/linux/drivers/base/dd.c:1217 /home/pi/linux/drivers/base/dd.c:1156) 
[    5.604515] bus_for_each_dev (/home/pi/linux/drivers/base/bus.c:370) 
[    5.608360] driver_attach (/home/pi/linux/drivers/base/dd.c:1234) 
[    5.611942] bus_add_driver (/home/pi/linux/drivers/base/bus.c:675) 
[    5.615699] driver_register (/home/pi/linux/drivers/base/driver.c:246) 
[    5.619544] __pci_register_driver (/home/pi/linux/drivers/pci/pci-driver.c:1463) 
[    5.623824] amdgpu_init (/home/pi/linux/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c:2315) amdgpu
[    5.628327] do_one_initcall (/home/pi/linux/init/main.c:1238) 
[    5.632172] do_init_module (/home/pi/linux/kernel/module/main.c:2535) 
[    5.635931] load_module (/home/pi/linux/kernel/module/main.c:2986) 
[    5.639687] init_module_from_file (/home/pi/linux/kernel/module/main.c:3176) 
[    5.643967] __arm64_sys_finit_module (/home/pi/linux/kernel/module/main.c:3187 /home/pi/linux/kernel/module/main.c:3214 /home/pi/linux/kernel/module/main.c:3197 /home/pi/linux/kernel/module/main.c:3197) 
[    5.648684] invoke_syscall (/home/pi/linux/./arch/arm64/include/asm/current.h:19 /home/pi/linux/arch/arm64/kernel/syscall.c:56) 
[    5.652442] el0_svc_common.constprop.0 (/home/pi/linux/arch/arm64/kernel/syscall.c:141) 
[    5.657160] do_el0_svc (/home/pi/linux/arch/arm64/kernel/syscall.c:154) 
[    5.660481] el0_svc (/home/pi/linux/./arch/arm64/include/asm/daifflags.h:28 /home/pi/linux/arch/arm64/kernel/entry-common.c:133 /home/pi/linux/arch/arm64/kernel/entry-common.c:144 /home/pi/linux/arch/arm64/kernel/entry-common.c:679) 
[    5.663540] el0t_64_sync_handler (/home/pi/linux/arch/arm64/kernel/entry-common.c:697) 
[    5.667908] el0t_64_sync (/home/pi/linux/arch/arm64/kernel/entry.S:599) 
[ 5.671578] Code: 91010108 54ffff4a 8b040108 cb050042 (d50b7428)
All code
========
   0:	91010108 	add	x8, x8, #0x40
   4:	54ffff4a 	b.ge	0xffffffffffffffec  // b.tcont
   8:	8b040108 	add	x8, x8, x4
   c:	cb050042 	sub	x2, x2, x5
  10:*	d50b7428 	dc	zva, x8		<-- trapping instruction

Code starting with the faulting instruction
===========================================
   0:	d50b7428 	dc	zva, x8
[    5.677692] ---[ end trace 0000000000000000 ]---

I'm guessing it's this one: raspberrypi/linux@fad1d03#diff-6aa215669ed6d8f14a1d9949230b71ee7033128d587fd937a94e582dcdc15b7bL1131

Re-applying your latest branch against rpi-6.6.y and recompiling now!

[geerlingguy]

As an aside, when the driver gets locked up in whatever state it's in, the system won't shut down (for at least 5-10 minutes), and a sudo modprobe amdgpu hangs... all the while, the card is sucking down 160W of power, with the fans running full blast :)

[geerlingguy]

New error:

[    8.084752] Hardware name: Raspberry Pi 5 Model B Rev 1.0 (DT)
[    8.090603] pstate: 80400009 (Nzcv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[    8.097590] pc : sdma_v6_0_mqd_init+0xe4/0x148 [amdgpu]
[    8.103206] lr : sdma_v6_0_mqd_init+0xcc/0x148 [amdgpu]
[    8.108674] sp : ffffffc080983200
[    8.111994] x29: ffffffc080983200 x28: ffffff8110fc0da0 x27: ffffff81010cba00
[    8.119156] x26: ffffff81010cbb30 x25: ffffff81010cbb38 x24: ffffff81010cbb28
[    8.126318] x23: ffffff8110fc0c00 x22: ffffff81010cbb00 x21: ffffff8110f80000
[    8.133480] x20: ffffffc080983250 x19: ffffffc08106d000 x18: ffffffffffffffff
[    8.140642] x17: 0000000000001fd0 x16: ffffffd084ced580 x15: 0000000000000100
[    8.147804] x14: ffffff81a85f832f x13: 0000000000000000 x12: 7720322e322e3836
[    8.154966] x11: 0000000000000040 x10: ffffff8110fc0de8 x9 : ffffffd064283acc
[    8.162128] x8 : ffffffc080983298 x7 : 0000000000000000 x6 : 0000000000000228
[    8.169290] x5 : ffffff81077f16a8 x4 : 0000000010000000 x3 : 000000000000000f
[    8.176452] x2 : 0000000000004000 x1 : 0000000000000a00 x0 : 0000000000000000
[    8.183614] Call trace:
[    8.186062]  sdma_v6_0_mqd_init+0xe4/0x148 [amdgpu]
[    8.191181]  amdgpu_mes_queue_init_mqd+0xb4/0x190 [amdgpu]
[    8.196908]  amdgpu_mes_add_hw_queue+0x208/0x4e0 [amdgpu]
[    8.202548]  amdgpu_mes_add_ring+0x1ec/0x3a8 [amdgpu]
[    8.207837]  amdgpu_mes_self_test+0x1f8/0x450 [amdgpu]
[    8.213213]  mes_v11_0_late_init+0x58/0x78 [amdgpu]
[    8.218328]  amdgpu_device_ip_late_init+0x68/0x308 [amdgpu]
[    8.224139]  amdgpu_device_init+0x2038/0x2208 [amdgpu]
[    8.229514]  amdgpu_driver_load_kms+0x20/0x1a8 [amdgpu]
[    8.234975]  amdgpu_pci_probe+0x154/0x420 [amdgpu]
[    8.239999]  local_pci_probe+0x48/0xb8
[    8.243758]  pci_device_probe+0xac/0x1c8
[    8.247690]  really_probe+0x150/0x2c0
[    8.251360]  __driver_probe_device+0x80/0x140
[    8.255728]  driver_probe_device+0x44/0x170
[    8.259921]  __driver_attach+0x9c/0x1b0
[    8.263765]  bus_for_each_dev+0x80/0xe8
[    8.267611]  driver_attach+0x2c/0x40
[    8.271194]  bus_add_driver+0xec/0x218
[    8.274950]  driver_register+0x68/0x138
[    8.278795]  __pci_register_driver+0x54/0x68
[    8.283075]  amdgpu_init+0x6c/0xff8 [amdgpu]
[    8.287577]  do_one_initcall+0x60/0x2c0
[    8.291423]  do_init_module+0x60/0x218
[    8.295182]  load_module+0x1dd0/0x2080
[    8.298938]  init_module_from_file+0x8c/0xd8
[    8.303218]  __arm64_sys_finit_module+0x150/0x330
[    8.307935]  invoke_syscall+0x50/0x128
[    8.311693]  el0_svc_common.constprop.0+0xc8/0xf0
[    8.316412]  do_el0_svc+0x24/0x38
[    8.319733]  el0_svc+0x40/0xe8
[    8.322792]  el0t_64_sync_handler+0x100/0x130
[    8.327159]  el0t_64_sync+0x190/0x198
[    8.330830] Code: 52880002 528001e3 b9402a81 52800000 (f803c27f) 
[    8.336945] ---[ end trace 0000000000000000 ]---

This one doesn't look like a standard memset, maybe there's a pointer that's null or something here?

[Coreforge]

That's accesses to a struct (most of the time, including this one, it's mqd stuff). To my understanding, the issue is caused by compiler optimization, and so far, changing the type to volatile has worked fine (that won't work if the struct members aren't aligned or if the struct is at an odd address, but I haven't run into that so far).

I'm surprised this hasn't happened earlier though, as this isn't specific to the 7000 series (well, sdma_v6 _0 is).
I added a fix for it.

[geerlingguy]

Interesting. I recompiled with your latest commit, and I'm getting display output.

But glmark2-es2 won't run, it starts to open a window then it stops (no dmesg errors or anything):

pi@pi5-pcie:~ $ DISPLAY=:0 glmark2-es2
=======================================================
    glmark2 2023.01
=======================================================
    OpenGL Information
    GL_VENDOR:      AMD
    GL_RENDERER:    AMD Radeon RX 7600 (gfx1102, LLVM 15.0.6, DRM 3.54, 6.6.60-v8-AMDGPU+)
    GL_VERSION:     OpenGL ES 3.2 Mesa 23.2.1-1~bpo12+rpt3
    Surface Config: buf=32 r=8 g=8 b=8 a=8 depth=24 stencil=0 samples=0
    Surface Size:   800x600 windowed
=======================================================
[build] use-vbo=false: FPS: 37 FrameTime: 27.721 ms
=======================================================
                                  glmark2 Score: 36 
=======================================================

GravityMark works, and at 1440p windowed, I got a score of 20,809. (Result here, compared to W7700 result of 31,982)

(Edit: glmark2 issue maybe related to using labwc now? glmark2/glmark2#207)

labwc -d output while attempting glmark2-es2:

The XKEYBOARD keymap compiler (xkbcomp) reports:
> Warning:          Unsupported maximum keycode 708, clipping.
>                   X11 cannot support keycodes above 255.
Errors from xkbcomp are not fatal to the X server
00:00:13.689 [xwayland/xwm.c:1648] unhandled X11 event: 34
00:00:13.689 [xwayland/xwm.c:1648] unhandled X11 event: 34
00:00:13.800 [xwayland/xwm.c:882] unhandled X11 property 277 (WM_STATE) for window 4194306
00:00:13.800 [xwayland/xwm.c:1163] Failed to get window property
00:00:13.800 [types/wlr_compositor.c:692] New wlr_surface 0x5568ed1880 (res 0x5568ed1c30)
00:00:13.801 [xwayland/xwm.c:1738] New xwayland surface: 0x5568ed1880
00:00:13.803 [xwayland/xwm.c:1627] xcb error: op 18:0, code 3, sequence 122, value 4194306
00:00:13.804 [xwayland/xwm.c:1627] xcb error: op 19:0, code 3, sequence 123, value 4194306
xsettingsd: Loaded 19 settings from /home/pi/.config/xsettingsd/xsettingsd.conf
00:00:14.539 [xwayland/xwm.c:882] unhandled X11 property 322 (WM_LOCALE_NAME) for window 4194305
xsettingsd: Created window 0x400001 on screen 0 with timestamp 776651
xsettingsd: Selection _XSETTINGS_S0 is owned by 0x0
xsettingsd: Took ownership of selection _XSETTINGS_S0
00:00:14.540 xsettingsd: Ignoring event of type 33
[xwayland/xwm.c:882] unhandled X11 property 323 (_XSETTINGS_SETTINGS) for window 4194305
00:00:14.981 [render/swapchain.c:106] Allocating new swapchain buffer
00:00:14.981 [render/allocator/gbm.c:145] Allocated 19x28 GBM buffer with format AR24 (0x34325241), modifier GFX11,GFX9>
00:00:14.981 [render/gles2/renderer.c:159] Created GL FBO for buffer 19x28
00:00:14.982 [render/swapchain.c:106] Allocating new swapchain buffer
00:00:14.982 [render/allocator/gbm.c:145] Allocated 19x28 GBM buffer with format AR24 (0x34325241), modifier GFX11,GFX9>
00:00:14.982 [render/gles2/renderer.c:159] Created GL FBO for buffer 19x28