AMD General > -----Original Message----- > From: Alex Deucher <[email protected]> > Sent: Wednesday, July 29, 2026 21:25 > To: Teodor Bostan <[email protected]>; Wang, Yang(Kevin) > <[email protected]> > Cc: [email protected]; Feng, Kenneth <[email protected]>; > Deucher, Alexander <[email protected]>; Koenig, Christian > <[email protected]>; Limonciello, Mario <[email protected]>; > [email protected] > Subject: Re: [PATCH] drm/amd/pm: Add PPTable power limit quirk for ASUS G14 > GA402R > > + Kevin > > On Wed, Jul 29, 2026 at 9:23 AM Teodor Bostan <[email protected]> wrote: > > > > ASUS ROG Zephyrus G14 (2022) GA402RJ and GA402RK laptops use custom > AC > > power limits of 100W (RX 6700S) and 105W (RX 6800S). On resume from > > suspend, the SMU firmware resets to the reference vBIOS default of > > 80W, hard-capping GPU performance. > > > > Override SocketPowerLimitAc in sienna_cichlid_patch_pptable_quirk for > > these subsystem IDs so the driver restores the platform's intended AC > > power limit when reloading PPTable during boot and resume. > > > > Signed-off-by: Teodor Bostan <[email protected]> > > --- > > .../gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c | 13 > > +++++++++++++ > > 1 file changed, 13 insertions(+) > > > > diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c > > b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c > > index cef628752..10d330edb 100644 > > --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c > > +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c > > @@ -479,6 +479,7 @@ static int sienna_cichlid_patch_pptable_quirk(struct > smu_context *smu) > > struct amdgpu_device *adev = smu->adev; > > uint32_t *board_reserved; > > uint16_t *freq_table_gfx; > > + uint16_t *power_limit; > > uint32_t i; > > > > /* Fix some OEM SKU specific stability issues */ @@ -501,6 > > +502,18 @@ static int sienna_cichlid_patch_pptable_quirk(struct smu_context > *smu) > > } > > } > > > > + /* Fix some OEM SKU specific power limit issues */ > > + if ((adev->pdev->device == 0x73EF) && > > + (adev->pdev->subsystem_vendor == 0x1043)) { > > + if (adev->pdev->subsystem_device == 0x1DEC) { > > + GET_PPTABLE_MEMBER(SocketPowerLimitAc, > > &power_limit); > > + power_limit[PPT_THROTTLER_PPT0] = 100; > > + } else if (adev->pdev->subsystem_device == 0x1DCC) { > > + GET_PPTABLE_MEMBER(SocketPowerLimitAc, > > &power_limit); > > + power_limit[PPT_THROTTLER_PPT0] = 105; > > + } > > + }
Aha, I believe this is not root cause and the amdgpu driver should avoid adding these hardcode value for specific device. and the amdgpu kernel driver restores power limit settings after system resume in smu_late_init() stage. I will check and review this sequence. Fyi. smu_restore_dpm_user_profile(). Best regards, Kevin > > + > > return 0; > > } > > > > -- > > 2.55.0 > >
