[PATCH] drm/i915/vrr: partial revert of dda7dcd9da73c5327aef42b89f0519bb51e84217

[Benjamin Tissoires]

Commit dda7dcd9da73 ("drm/i915/vrr: Use fixed timings for platforms that
support VRR") introduced a regression on my Ice Lake laptop with an Iris
Plus G7.

The screen can go blank (screen lock or just idle), but if I try to wake
it up, there is an immediate kernel lock. The netconsole was really
unhelpful here, and a bisect gave me the commit above to be the culprit.

For reference, the netconsole I could get was:
[  103.553110] netconsole: network logging started
[  125.642395] mce: [Hardware Error]: CPU 0: Machine Check Exception: 5 Bank 6: ba00000011000402
[  125.642400] mce: [Hardware Error]: RIP !INEXACT! 33:<00007fa4471665ec>
[  125.642403] mce: [Hardware Error]: TSC 38282e339d
[  125.642404] mce: [Hardware Error]: PROCESSOR 0:706e5 TIME 1771365482 SOCKET 0 APIC 0 microcode ca
[  125.642406] mce: [Hardware Error]: Run the above through 'mcelog --ascii'
[  125.642406] mce: [Hardware Error]: Machine check: Processor context corrupt
[  125.642407] Kernel panic - not syncing: Fatal machine check
[  125.642415] Kernel Offset: 0x1d400000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff)

and then the backtrace lead to the netconsole in itself.

If I restore the early return when vmin >= vmax (in my case vmin==vmax),
then I can use my machine again after it goes idle.

I have strictly no ideas if this is the correct solution, so please take
this with a grain of salt, but this is effectively preventing me to use
distro kernel since v6.16.

Signed-off-by: Benjamin Tissoires <bentiss@xxxxxxxxxx>
---
I've been experiencing this bug since kernel v6.16.

System Information
        Manufacturer: Dell Inc.
        Product Name: XPS 13 7390 2-in-1

cpu: Intel(R) Core(TM) i7-1065G7 CPU @ 1.30GHz

lspci:
00:00.0 Host bridge [0600]: Intel Corporation Ice Lake-LP Processor Host Bridge/DRAM Registers [8086:8a12] (rev 03)
	Subsystem: Dell Device [1028:08b0]
	Flags: bus master, fast devsel, latency 0, IOMMU group 2
	Capabilities: <access denied>
	Kernel driver in use: icl_uncore

00:02.0 VGA compatible controller [0300]: Intel Corporation Iris Plus Graphics G7 [8086:8a52] (rev 07) (prog-if 00 [VGA controller])
	DeviceName: To Be Filled by O.E.M.
	Subsystem: Dell Device [1028:08b0]
	Flags: bus master, fast devsel, latency 0, IRQ 188, IOMMU group 1
	Memory at 603d000000 (64-bit, non-prefetchable) [size=16M]
	Memory at 4000000000 (64-bit, prefetchable) [size=256M]
	I/O ports at 4000 [size=64]
	Expansion ROM at 000c0000 [virtual] [disabled] [size=128K]
	Capabilities: <access denied>
	Kernel driver in use: i915
	Kernel modules: i915

Again, I have really no idea if this is the correct fix, but it would be
nice if the issue could be solved in v7.0 and backported to older
kernels.

Please tell me if I need to provide more information.

Cheers,
Benjamin
---
 drivers/gpu/drm/i915/display/intel_vrr.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/gpu/drm/i915/display/intel_vrr.c b/drivers/gpu/drm/i915/display/intel_vrr.c
index b92c42fde937..3d16dd4611d9 100644
--- a/drivers/gpu/drm/i915/display/intel_vrr.c
+++ b/drivers/gpu/drm/i915/display/intel_vrr.c
@@ -377,6 +377,9 @@ intel_vrr_compute_config(struct intel_crtc_state *crtc_state,
 		vmax = vmin;
 	}
 
+	if (vmin >= vmax)
+		return;
+
 	if (crtc_state->uapi.vrr_enabled && vmin < vmax)
 		intel_vrr_compute_vrr_timings(crtc_state, vmin, vmax);
 	else if (is_cmrr_frac_required(crtc_state) && is_edp)

---
base-commit: 05f7e89ab9731565d8a62e3b5d1ec206485eeb0b
change-id: 20260218-wip-fix-iris-plus-28a916221bdc

Best regards,
-- 
Benjamin Tissoires <bentiss@xxxxxxxxxx>