linux-stable.git/drivers/gpu/drm/amd, branch linux-6.14.y

Revert "drm/amd/display: more liberal vmin/vmax update for freesync"

2025-06-10T11:15:58+00:00

commit 1b824eef269db44d068bbc0de74c94a8e8f9ce02 upstream.

This reverts commit cfb2d41831ee5647a4ae0ea7c24971a92d5dfa0d since it
causes regressions on certain configs. Revert until the issue can be
isolated and debugged.

Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4238
Signed-off-by: Aurabindo Pillai 
Acked-by: Alex Deucher 
Signed-off-by: Alex Deucher 
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman

drm/amd/display: check stream id dml21 wrapper to get plane_id

2025-06-04T12:45:10+00:00

[ Upstream commit 2ddac70fed50485aa4ae49cdb7478ce41d8d4715 ]

[Why & How]
Fix a false positive warning which occurs due to lack of correct checks
when querying plane_id in DML21. This fixes the warning when performing a
mode1 reset (cat /sys/kernel/debug/dri/1/amdgpu_gpu_recover):

[   35.751250] WARNING: CPU: 11 PID: 326 at /tmp/amd.PHpyAl7v/amd/amdgpu/../display/dc/dml2/dml2_dc_resource_mgmt.c:91 dml2_map_dc_pipes+0x243d/0x3f40 [amdgpu]
[   35.751434] Modules linked in: amdgpu(OE) amddrm_ttm_helper(OE) amdttm(OE) amddrm_buddy(OE) amdxcp(OE) amddrm_exec(OE) amd_sched(OE) amdkcl(OE) drm_suballoc_helper drm_ttm_helper ttm drm_display_helper cec rc_core i2c_algo_bit rfcomm qrtr cmac algif_hash algif_skcipher af_alg bnep amd_atl intel_rapl_msr intel_rapl_common snd_hda_codec_hdmi snd_hda_intel edac_mce_amd snd_intel_dspcfg snd_intel_sdw_acpi snd_hda_codec kvm_amd snd_hda_core snd_hwdep snd_pcm kvm snd_seq_midi snd_seq_midi_event snd_rawmidi crct10dif_pclmul polyval_clmulni polyval_generic btusb ghash_clmulni_intel sha256_ssse3 btrtl sha1_ssse3 snd_seq btintel aesni_intel btbcm btmtk snd_seq_device crypto_simd sunrpc cryptd bluetooth snd_timer ccp binfmt_misc rapl snd i2c_piix4 wmi_bmof gigabyte_wmi k10temp i2c_smbus soundcore gpio_amdpt mac_hid sch_fq_codel msr parport_pc ppdev lp parport efi_pstore nfnetlink dmi_sysfs ip_tables x_tables autofs4 hid_generic usbhid hid crc32_pclmul igc ahci xhci_pci libahci xhci_pci_renesas video wmi
[   35.751501] CPU: 11 UID: 0 PID: 326 Comm: kworker/u64:9 Tainted: G           OE      6.11.0-21-generic #21~24.04.1-Ubuntu
[   35.751504] Tainted: [O]=OOT_MODULE, [E]=UNSIGNED_MODULE
[   35.751505] Hardware name: Gigabyte Technology Co., Ltd. X670E AORUS PRO X/X670E AORUS PRO X, BIOS F30 05/22/2024
[   35.751506] Workqueue: amdgpu-reset-dev amdgpu_debugfs_reset_work [amdgpu]
[   35.751638] RIP: 0010:dml2_map_dc_pipes+0x243d/0x3f40 [amdgpu]
[   35.751794] Code: 6d 0c 00 00 8b 84 24 88 00 00 00 41 3b 44 9c 20 0f 84 fc 07 00 00 48 83 c3 01 48 83 fb 06 75 b3 4c 8b 64 24 68 4c 8b 6c 24 40 <0f> 0b b8 06 00 00 00 49 8b 94 24 a0 49 00 00 89 c3 83 f8 07 0f 87
[   35.751796] RSP: 0018:ffffbfa3805d7680 EFLAGS: 00010246
[   35.751798] RAX: 0000000000010000 RBX: 0000000000000006 RCX: 0000000000000000
[   35.751799] RDX: 0000000000000000 RSI: 0000000000000005 RDI: 0000000000000000
[   35.751800] RBP: ffffbfa3805d78f0 R08: 0000000000000000 R09: 0000000000000000
[   35.751801] R10: 0000000000000000 R11: 0000000000000000 R12: ffffbfa383249000
[   35.751802] R13: ffffa0e68f280000 R14: ffffbfa383249658 R15: 0000000000000000
[   35.751803] FS:  0000000000000000(0000) GS:ffffa0edbe580000(0000) knlGS:0000000000000000
[   35.751804] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   35.751805] CR2: 00005d847ef96c58 CR3: 000000041de3e000 CR4: 0000000000f50ef0
[   35.751806] PKRU: 55555554
[   35.751807] Call Trace:
[   35.751810]  
[   35.751816]  ? show_regs+0x6c/0x80
[   35.751820]  ? __warn+0x88/0x140
[   35.751822]  ? dml2_map_dc_pipes+0x243d/0x3f40 [amdgpu]
[   35.751964]  ? report_bug+0x182/0x1b0
[   35.751969]  ? handle_bug+0x6e/0xb0
[   35.751972]  ? exc_invalid_op+0x18/0x80
[   35.751974]  ? asm_exc_invalid_op+0x1b/0x20
[   35.751978]  ? dml2_map_dc_pipes+0x243d/0x3f40 [amdgpu]
[   35.752117]  ? math_pow+0x48/0xa0 [amdgpu]
[   35.752256]  ? srso_alias_return_thunk+0x5/0xfbef5
[   35.752260]  ? math_pow+0x48/0xa0 [amdgpu]
[   35.752400]  ? srso_alias_return_thunk+0x5/0xfbef5
[   35.752403]  ? math_pow+0x11/0xa0 [amdgpu]
[   35.752524]  ? srso_alias_return_thunk+0x5/0xfbef5
[   35.752526]  ? core_dcn4_mode_programming+0xe4d/0x20d0 [amdgpu]
[   35.752663]  ? srso_alias_return_thunk+0x5/0xfbef5
[   35.752669]  dml21_validate+0x3d4/0x980 [amdgpu]

Reviewed-by: Austin Zheng 
Signed-off-by: Aurabindo Pillai 
Signed-off-by: Ray Wu 
Tested-by: Daniel Wheeler 
Signed-off-by: Alex Deucher 
(cherry picked from commit f8ad62c0a93e5dd94243e10f1b742232e4d6411e)
Signed-off-by: Sasha Levin

drm/amd/display: fix link_set_dpms_off multi-display MST corner case

2025-06-04T12:45:10+00:00

[ Upstream commit 3c1a467372e0c356b1d3c59f6d199ed5a6612dd1 ]

[Why & How]
When MST config is unplugged/replugged too quickly, it can potentially
result in a scenario where previous DC state has not been reset before
the HPD link detection sequence begins. In this case, driver will
disable the streams/link prior to re-enabling the link for link
training.

There is a bug in the current logic that does not account for the fact
that current_state can be released and cleared prior to swapping to a
new state (resulting in the pipe_ctx stream pointers to be cleared) in
between disabling streams.

To resolve this, cache the original streams prior to committing any
stream updates.

Reviewed-by: Wenjing Liu 
Signed-off-by: George Shen 
Signed-off-by: Ray Wu 
Tested-by: Daniel Wheeler 
Signed-off-by: Alex Deucher 
(cherry picked from commit 1561782686ccc36af844d55d31b44c938dd412dc)
Signed-off-by: Sasha Levin

drm/amdgpu: read back register after written for VCN v4.0.5

2025-05-29T09:14:08+00:00

commit ee7360fc27d6045510f8fe459b5649b2af27811a upstream.

On VCN v4.0.5 there is a race condition where the WPTR is not
updated after starting from idle when doorbell is used. Adding
register read-back after written at function end is to ensure
all register writes are done before they can be used.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/12528
Signed-off-by: David (Ming Qiang) Wu 
Reviewed-by: Mario Limonciello 
Tested-by: Mario Limonciello 
Reviewed-by: Alex Deucher 
Reviewed-by: Ruijing Dong 
Signed-off-by: Alex Deucher 
(cherry picked from commit 07c9db090b86e5211188e1b351303fbc673378cf)
Cc: stable@vger.kernel.org
Tested-by: Eric Naim 
Signed-off-by: Eric Naim 
Signed-off-by: Greg Kroah-Hartman

drm/amdgpu/vcn4.0.5: split code along instances

2025-05-29T09:14:08+00:00

commit ecc9ab4e924b7eb9e2c4a668162aaa1d9d60d08c upstream.

Split the code on a per instance basis.  This will allow
us to use the per instance functions in the future to
handle more things per instance.

v2: squash in fix for stop() from Boyuan

Reviewed-by: Boyuan Zhang 
Signed-off-by: Alex Deucher 
Tested-by: Eric Naim 
Signed-off-by: Eric Naim 
Signed-off-by: Greg Kroah-Hartman

Revert "drm/amd: Keep display off while going into S4"

2025-05-29T09:14:04+00:00

commit 7e7cb7a13c81073d38a10fa7b450d23712281ec4 upstream.

commit 68bfdc8dc0a1a ("drm/amd: Keep display off while going into S4")
attempted to keep displays off during the S4 sequence by not resuming
display IP.  This however leads to hangs because DRM clients such as the
console can try to access registers and cause a hang.

Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4155
Fixes: 68bfdc8dc0a1a ("drm/amd: Keep display off while going into S4")
Reviewed-by: Alex Deucher 
Link: https://lore.kernel.org/r/20250522141328.115095-1-mario.limonciello@amd.com
Signed-off-by: Mario Limonciello 
Signed-off-by: Alex Deucher 
(cherry picked from commit e485502c37b097b0bd773baa7e2741bf7bd2909a)
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman

drm/amd/display: Call FP Protect Before Mode Programming/Mode Support

2025-05-29T09:13:54+00:00

[ Upstream commit eba692ca3abca258b3214a6e4126afefad1822f0 ]

[Why]
Memory allocation occurs within dml21_validate() for adding phantom planes.
May cause kernel to be tainted due to usage of FP Start.

[How]
Move FP start from dml21_validate to before mode programming/mode support.
Calculations requiring floating point are all done within mode programming
or mode support.

Reviewed-by: Alvin Lee 
Signed-off-by: Austin Zheng 
Signed-off-by: Ray Wu 
Tested-by: Daniel Wheeler 
Signed-off-by: Alex Deucher 
(cherry picked from commit fe3250f10819b411808ab9ae1d824c5fc9b59170)
Signed-off-by: Sasha Levin

drm/amdkfd: Fix pasid value leak

2025-05-29T09:13:49+00:00

[ Upstream commit 10e08943caedfb4b0b95933d248503a6f6b9fef6 ]

Curret kfd does not allocate pasid values, instead uses pasid value for each
vm from graphic driver. So should not prevent graphic driver from releasing
pasid values since the values are allocated by graphic driver, not kfd driver
anymore. This patch does not stop graphic driver release pasid values.

Fixes: 8544374c0f82 ("drm/amdkfd: Have kfd driver use same PASID values from graphic driver")
Signed-off-by: Xiaogang Chen 
Reviewed-by: Felix Kuehling 
Signed-off-by: Alex Deucher 
Signed-off-by: Sasha Levin

drm/amdkfd: Fix error handling for missing PASID in 'kfd_process_device_init_vm'

2025-05-29T09:13:49+00:00

[ Upstream commit 2b04d04de956b44cc140d45cf8ebccfb378ce3bf ]

In the kfd_process_device_init_vm function, a valid error code is now
returned when the associated Process Address Space ID (PASID) is not
present.

If the address space virtual memory (avm) does not have an associated
PASID, the function sets the ret variable to -EINVAL before proceeding
to the error handling section. This ensures that the calling function,
such as kfd_ioctl_acquire_vm, can appropriately handle the error,
thereby preventing any issues during virtual memory initialization.

Fixes the below:
drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_process.c:1694 kfd_process_device_init_vm()
warn: missing error code 'ret'

drivers/gpu/drm/amd/amdgpu/../amdkfd/kfd_process.c
    1647 int kfd_process_device_init_vm(struct kfd_process_device *pdd,
    1648                                struct file *drm_file)
    1649 {
    ...
    1690
    1691         if (unlikely(!avm->pasid)) {
    1692                 dev_warn(pdd->dev->adev->dev, "WARN: vm %p has no pasid associated",
    1693                                  avm);
--> 1694                 goto err_get_pasid;

ret = -EINVAL?

    1695         }

Fixes: 8544374c0f82 ("drm/amdkfd: Have kfd driver use same PASID values from graphic driver")
Reported by: Dan Carpenter 
Cc: Xiaogang Chen 
Cc: Felix Kuehling 
Cc: Christian König 
Cc: Alex Deucher 
Signed-off-by: Srinivasan Shanmugam 
Reviewed-by: Felix Kuehling 
Signed-off-by: Alex Deucher 
Signed-off-by: Sasha Levin

drm/amd/display: Exit idle optimizations before accessing PHY

2025-05-29T09:13:49+00:00

[ Upstream commit c488967488d7eff7b9c527d5469c424c15377502 ]

[why & how]
By default, DCN HW is in idle optimized state which does not allow access
to PHY registers. If BIOS powers up the DCN, it is fine because they will
power up everything. Only exit idle optimized state when not taking control
from VBIOS.

Fixes: be704e5ef4bd ("Revert "drm/amd/display: Exit idle optimizations before attempt to access PHY"")
Reviewed-by: Charlene Liu 
Signed-off-by: Ovidiu Bunea 
Signed-off-by: Roman Li 
Tested-by: Daniel Wheeler 
Signed-off-by: Alex Deucher 
Signed-off-by: Sasha Levin