[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20251021153940.GA1191128@bhelgaas>
Date: Tue, 21 Oct 2025 10:39:40 -0500
From: Bjorn Helgaas <helgaas@...nel.org>
To: Manivannan Sadhasivam <mani@...nel.org>
Cc: Manivannan Sadhasivam <manivannan.sadhasivam@....qualcomm.com>,
Christian Zigotzky <chzigotzky@...osoft.de>,
FUKAUMI Naoki <naoki@...xa.com>, linux-rockchip@...r.kernel.org,
linux-pci@...r.kernel.org, linux-kernel@...r.kernel.org,
Bjorn Helgaas <bhelgaas@...gle.com>
Subject: Re: [PATCH] PCI/ASPM: Enable only L0s and L1 for devicetree platforms
On Tue, Oct 21, 2025 at 06:05:07PM +0530, Manivannan Sadhasivam wrote:
> On Tue, Oct 21, 2025 at 09:31:32AM +0530, Manivannan Sadhasivam wrote:
> > On Mon, Oct 20, 2025 at 05:12:07PM -0500, Bjorn Helgaas wrote:
> > > + if (link->aspm_support & PCIE_LINK_STATE_L0S)
> > > + link->aspm_default |= PCIE_LINK_STATE_L0S;
> > > + if (link->aspm_support & PCIE_LINK_STATE_L1)
> > > + link->aspm_default |= PCIE_LINK_STATE_L1;
> >
> > Not sure if it is worth setting these states conditionally. Link
> > state enablement code should make use of the cached ASPM cap in
> > 'link->aspm_capable'.
>
> I see the point now. Without the check, we falsely claim that the
> ASPM states are getting enabled. But we cannot be sure until the
> register write to LNKCTL happens, which will depend on many factors
> (own device capability, upstream/downstream port capability,...)
>
> To avoid ambiguity, can we reword the log to something like,
>
> "ASPM: Overridding default states %s%s\n"
I think aspm_support already includes the capabilities from both ends
of the link:
if (parent_lnkcap & child_lnkcap & PCI_EXP_LNKCAP_ASPM_L0S)
link->aspm_support |= PCIE_LINK_STATE_L0S;
if (parent_lnkcap & child_lnkcap & PCI_EXP_LNKCAP_ASPM_L1)
link->aspm_support |= PCIE_LINK_STATE_L1;
Powered by blists - more mailing lists