Intel Alder Lake GPU performance much lower in 6.1/6.2 compared to 5.19

Bug #2011817 reported by Daniel van Vugt
14
This bug affects 3 people
Affects Status Importance Assigned to Milestone
Linux
Unknown
Unknown
linux (Ubuntu)
Fix Released
High
Unassigned

Bug Description

Intel Alder Lake GPU performance is much lower in 6.1 compared to 5.19.

For example, benchmarking gnome-shell main running on Xorg:

6.1.0-16-generic
30.02 FPS, average: 15.5ms, peak: 16.4ms
intel_gpu_top says 211/208 MHz despite gt_min_freq_mhz==300

5.19.0-21-generic
60.00 FPS, average: 8.3ms, peak: 8.7ms
intel_gpu_top says 542/560 MHz

The issue is so severe in 6.1.0-16-generic that even triple buffering can't fix it for Xorg sessions. It does for Wayland though.

I can work around it with `cp gt_max_freq_mhz gt_min_freq_mhz` but probably shouldn't need to.

ProblemType: Bug
DistroRelease: Ubuntu 23.04
Package: linux-image-6.1.0-16-generic 6.1.0-16.16
ProcVersionSignature: Ubuntu 6.1.0-16.16-generic 6.1.6
Uname: Linux 6.1.0-16-generic x86_64
ApportVersion: 2.26.0-0ubuntu2
Architecture: amd64
AudioDevicesInUse:
 USER PID ACCESS COMMAND
 /dev/snd/controlC0: dan 1730 F.... wireplumber
 /dev/snd/seq: dan 1727 F.... pipewire
CRDA: N/A
CasperMD5CheckResult: pass
Date: Thu Mar 16 19:31:50 2023
InstallationDate: Installed on 2022-11-28 (108 days ago)
InstallationMedia: Ubuntu 23.04 "Lunar Lobster" - Alpha amd64 (20221126)
MachineType: Intel(R) Client Systems NUC12DCMi7
ProcFB: 0 i915drmfb
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-6.1.0-16-generic root=UUID=3a6db1e7-0341-452a-80e1-2895838feeac ro quiet splash vt.handoff=7
PulseList: Error: command ['pacmd', 'list'] failed with exit code 1: No PulseAudio daemon running, or not running as session daemon.
RelatedPackageVersions:
 linux-restricted-modules-6.1.0-16-generic N/A
 linux-backports-modules-6.1.0-16-generic N/A
 linux-firmware 20230302.git60971a64-0ubuntu1
SourcePackage: linux
UpgradeStatus: No upgrade log present (probably fresh install)
dmi.bios.date: 12/20/2021
dmi.bios.release: 5.24
dmi.bios.vendor: Intel Corp.
dmi.bios.version: EDADL579.0046.2021.1220.2351
dmi.board.name: NUC12EDBi7
dmi.board.vendor: Intel Corporation
dmi.board.version: M27908-302
dmi.chassis.type: 35
dmi.chassis.vendor: Intel Corporation
dmi.chassis.version: 2.0
dmi.ec.firmware.release: 3.7
dmi.modalias: dmi:bvnIntelCorp.:bvrEDADL579.0046.2021.1220.2351:bd12/20/2021:br5.24:efr3.7:svnIntel(R)ClientSystems:pnNUC12DCMi7:pvrM30143-302:rvnIntelCorporation:rnNUC12EDBi7:rvrM27908-302:cvnIntelCorporation:ct35:cvr2.0:skuRNUC12DCMi70000:
dmi.product.family: DC
dmi.product.name: NUC12DCMi7
dmi.product.sku: RNUC12DCMi70000
dmi.product.version: M30143-302
dmi.sys.vendor: Intel(R) Client Systems

Revision history for this message
Daniel van Vugt (vanvugt) wrote :
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

I know this is probably power savings, but if triple buffering can't kick the GPU up to a sufficient clock speed then it feels like something is wrong.

Revision history for this message
Ubuntu Kernel Bot (ubuntu-kernel-bot) wrote : Status changed to Confirmed

This change was made by a bot.

Changed in linux (Ubuntu):
status: New → Confirmed
Revision history for this message
Daniel van Vugt (vanvugt) wrote : Re: Intel GPU performance much lower in 6.1/6.2 compared to 5.19

Still slow in 6.2.0-18-generic

summary: - Intel GPU performance much lower in 6.1 compared to 5.19
+ Intel GPU performance much lower in 6.1/6.2 compared to 5.19
Changed in linux (Ubuntu):
importance: Undecided → High
Revision history for this message
Daniel van Vugt (vanvugt) wrote :
description: updated
summary: - Intel GPU performance much lower in 6.1/6.2 compared to 5.19
+ Intel Alder Lake GPU performance much lower in 6.1/6.2 compared to 5.19
description: updated
tags: added: rls-ll-incoming
Changed in linux (Ubuntu):
milestone: none → ubuntu-23.04
Changed in linux (Ubuntu):
status: Confirmed → In Progress
assignee: nobody → Daniel van Vugt (vanvugt)
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Yes the kernel issue with gt_min_freq_mhz not being honoured is fixed in both:

  6.3.0-060300rc7drmtip20230419-generic
  6.2.0-20-generic

My new issue is one or maybe two mutter 44.0 regressions.

Changed in linux (Ubuntu):
assignee: Daniel van Vugt (vanvugt) → nobody
status: In Progress → Fix Released
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

The first of the new issues is bug 2017137

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.