Re: [PATCH 4/8] platform/x86/intel/pmc: Allow reenabling LTRs

[Ilpo Järvinen]

On Wed, 27 Dec 2023, Ilpo Järvinen wrote:

> On Fri, 22 Dec 2023, David E. Box wrote:
> 
> > Commit 804951203aa5 ("platform/x86:intel/pmc: Combine core_init() and
> > core_configure()") caused a network performance regression due to the GBE
> > LTR ignore that it added during probe. The fix will move the ignore to
> > occur at suspend-time (so as to not affect suspend power). This will
> > require the ability to enable the LTR again on resume. Modify
> > pmc_core_send_ltr_ignore() to allow enabling an LTR.
> > 
> > Fixes: 804951203aa5 ("platform/x86:intel/pmc: Combine core_init() and core_configure()")
> > Signed-off-by: David E. Box <david.e.box@xxxxxxxxxxxxxxx>
> > ---
> >  drivers/platform/x86/intel/pmc/adl.c  | 2 +-
> >  drivers/platform/x86/intel/pmc/cnp.c  | 2 +-
> >  drivers/platform/x86/intel/pmc/core.c | 9 ++++++---
> >  drivers/platform/x86/intel/pmc/core.h | 3 ++-
> >  drivers/platform/x86/intel/pmc/mtl.c  | 2 +-
> >  drivers/platform/x86/intel/pmc/tgl.c  | 2 +-
> >  6 files changed, 12 insertions(+), 8 deletions(-)
> > 
> > diff --git a/drivers/platform/x86/intel/pmc/adl.c b/drivers/platform/x86/intel/pmc/adl.c
> > index 882f2d5d8937..fbe0678f766c 100644
> > --- a/drivers/platform/x86/intel/pmc/adl.c
> > +++ b/drivers/platform/x86/intel/pmc/adl.c
> > @@ -327,7 +327,7 @@ int adl_core_init(struct pmc_dev *pmcdev)
> >  	 * when a cable is attached. Tell the PMC to ignore it.
> >  	 */
> >  	dev_dbg(&pmcdev->pdev->dev, "ignoring GBE LTR\n");
> > -	pmc_core_send_ltr_ignore(pmcdev, 3);
> > +	pmc_core_send_ltr_ignore(pmcdev, 3, 1);
> >  
> >  	return 0;
> >  }
> > diff --git a/drivers/platform/x86/intel/pmc/cnp.c b/drivers/platform/x86/intel/pmc/cnp.c
> > index 59298f184d0e..80f73242f9dd 100644
> > --- a/drivers/platform/x86/intel/pmc/cnp.c
> > +++ b/drivers/platform/x86/intel/pmc/cnp.c
> > @@ -220,7 +220,7 @@ int cnp_core_init(struct pmc_dev *pmcdev)
> >  	 * when a cable is attached. Tell the PMC to ignore it.
> >  	 */
> >  	dev_dbg(&pmcdev->pdev->dev, "ignoring GBE LTR\n");
> > -	pmc_core_send_ltr_ignore(pmcdev, 3);
> > +	pmc_core_send_ltr_ignore(pmcdev, 3, 1);
> >  
> >  	return 0;
> >  }
> > diff --git a/drivers/platform/x86/intel/pmc/core.c b/drivers/platform/x86/intel/pmc/core.c
> > index 31905003bb05..aa44c6612af9 100644
> > --- a/drivers/platform/x86/intel/pmc/core.c
> > +++ b/drivers/platform/x86/intel/pmc/core.c
> > @@ -476,7 +476,7 @@ static int pmc_core_pll_show(struct seq_file *s, void *unused)
> >  }
> >  DEFINE_SHOW_ATTRIBUTE(pmc_core_pll);
> >  
> > -int pmc_core_send_ltr_ignore(struct pmc_dev *pmcdev, u32 value)
> > +int pmc_core_send_ltr_ignore(struct pmc_dev *pmcdev, u32 value, int ignore)
> 
> Nothing seems to need int for anything and it's used like a bool. So 

Note that I've merged this change without changing these into bool. Please 
consider making a follow up change to the type after next -rc1 when 
things have settled down.

> > +++ b/drivers/platform/x86/intel/pmc/core.h
> > @@ -567,7 +567,8 @@ extern const struct pmc_reg_map arl_pchs_reg_map;
> >  
> >  extern void pmc_core_get_tgl_lpm_reqs(struct platform_device *pdev);
> >  extern int pmc_core_ssram_get_lpm_reqs(struct pmc_dev *pmcdev);
> > -extern int pmc_core_send_ltr_ignore(struct pmc_dev *pmcdev, u32 value);
> > +extern int
> > +pmc_core_send_ltr_ignore(struct pmc_dev *pmcdev, u32 value, int ignore);
> 
> extern is unnecessary.

While merging, I dropped this one as I had to resolve a conflict here 
anyway.

Patches 3-5 are now in review-ilpo towards fixes branch. Please do double 
check they're correct because the conflicts were quite large.


-- 
 i.