Linux PCI subsystem development
 help / color / mirror / Atom feed
From: Bjorn Helgaas <helgaas@kernel.org>
To: Lukas Wunner <lukas@wunner.de>
Cc: linux-pci@vger.kernel.org, Adria Vilanova Martinez <me@avm99963.com>
Subject: Re: [PATCH] PCI/ASPM: Fix pci_clear_and_set_config_dword() usage
Date: Tue, 17 Feb 2026 10:11:19 -0600	[thread overview]
Message-ID: <20260217161119.GA3378634@bhelgaas> (raw)
In-Reply-To: <5c1752d7512eed0f4ea57b84b12d7ee08ca61fc5.1771226659.git.lukas@wunner.de>

On Mon, Feb 16, 2026 at 08:46:13AM +0100, Lukas Wunner wrote:
> When aspm_calc_l12_info() programs the L1 PM Substates Control 1 register
> fields Common_Mode_Restore_Time, LTR_L1.2_THRESHOLD_Value and _Scale, it
> invokes pci_clear_and_set_config_dword() in an incorrect way:
> 
> For the bits to clear it selects those corresponding to the field.  So far
> so good.  But for the bits to set it passes a full register value.
> pci_clear_and_set_config_dword() performs a boolean OR operation which
> sets all bits of that value, not just the ones that were just cleared.
> 
> Thus, when setting the LTR_L1.2_THRESHOLD_Value and _Scale on the child of
> an ASPM link, aspm_calc_l12_info() also sets the Common_Mode_Restore_Time.
> That's a spec violation:  PCIe r7.0 sec 7.8.3.3 says this field is RsvdP
> for Upstream Ports.  On Adrià's Pixelbook Eve, Common_Mode_Restore_Time
> of the Intel 7265 "Stone Peak" wifi card is zero, yet aspm_calc_l12_info()
> does not preserve the zero bits but instead programs the value calculated
> for the Root Port into the wifi card.
> 
> Likewise, when setting the Common_Mode_Restore_Time on the Root Port,
> aspm_calc_l12_info() also changes the LTR_L1.2_THRESHOLD_Value and _Scale
> from the initial 163840 nsec to 237568 nsec (due to ORing those fields),
> only to reduce it afterwards to 106496 nsec.
> 
> Amend all invocations of pci_clear_and_set_config_dword() to only set bits
> which are cleared.
> 
> Finally, when setting the T_POWER_ON_Value and _Scale on the Root Port and
> the wifi card, aspm_calc_l12_info() fails to preserve bits declared RsvdP
> and instead overwrites them with zeroes.  Replace pci_write_config_dword()
> with pci_clear_and_set_config_dword() to avoid this.
> 
> Fixes: aeda9adebab8 ("PCI/ASPM: Configure L1 substate settings")
> Link: https://bugzilla.kernel.org/show_bug.cgi?id=220705#c22
> Tested-by: Adrià Vilanova Martínez <me@avm99963.com>
> Signed-off-by: Lukas Wunner <lukas@wunner.de>
> Cc: stable@vger.kernel.org # v4.11+

Applied to pci/aspm for v7.1, thanks!  Will be rebased after v7.0-rc1.

> ---
> Only a "Link" tag, not a "Closes" tag because this patch is just a
> byproduct that was created while working on the above-linked bugzilla.
> The actual root cause of the bugzilla seems to be an ASPM erratum
> of the Sunrise Point PCH which needs to be addressed in coreboot.
> 
>  drivers/pci/pcie/aspm.c | 17 ++++++++++++-----
>  1 file changed, 12 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/pci/pcie/aspm.c b/drivers/pci/pcie/aspm.c
> index cedea47..a1f2752 100644
> --- a/drivers/pci/pcie/aspm.c
> +++ b/drivers/pci/pcie/aspm.c
> @@ -706,22 +706,29 @@ static void aspm_calc_l12_info(struct pcie_link_state *link,
>  	}
>  
>  	/* Program T_POWER_ON times in both ports */
> -	pci_write_config_dword(parent, parent->l1ss + PCI_L1SS_CTL2, ctl2);
> -	pci_write_config_dword(child, child->l1ss + PCI_L1SS_CTL2, ctl2);
> +	pci_clear_and_set_config_dword(parent, parent->l1ss + PCI_L1SS_CTL2,
> +				       PCI_L1SS_CTL2_T_PWR_ON_VALUE |
> +				       PCI_L1SS_CTL2_T_PWR_ON_SCALE, ctl2);
> +	pci_clear_and_set_config_dword(child, child->l1ss + PCI_L1SS_CTL2,
> +				       PCI_L1SS_CTL2_T_PWR_ON_VALUE |
> +				       PCI_L1SS_CTL2_T_PWR_ON_SCALE, ctl2);
>  
>  	/* Program Common_Mode_Restore_Time in upstream device */
>  	pci_clear_and_set_config_dword(parent, parent->l1ss + PCI_L1SS_CTL1,
> -				       PCI_L1SS_CTL1_CM_RESTORE_TIME, ctl1);
> +				       PCI_L1SS_CTL1_CM_RESTORE_TIME,
> +				       ctl1 & PCI_L1SS_CTL1_CM_RESTORE_TIME);
>  
>  	/* Program LTR_L1.2_THRESHOLD time in both ports */
>  	pci_clear_and_set_config_dword(parent, parent->l1ss + PCI_L1SS_CTL1,
>  				       PCI_L1SS_CTL1_LTR_L12_TH_VALUE |
>  				       PCI_L1SS_CTL1_LTR_L12_TH_SCALE,
> -				       ctl1);
> +				       ctl1 & (PCI_L1SS_CTL1_LTR_L12_TH_VALUE |
> +					       PCI_L1SS_CTL1_LTR_L12_TH_SCALE));
>  	pci_clear_and_set_config_dword(child, child->l1ss + PCI_L1SS_CTL1,
>  				       PCI_L1SS_CTL1_LTR_L12_TH_VALUE |
>  				       PCI_L1SS_CTL1_LTR_L12_TH_SCALE,
> -				       ctl1);
> +				       ctl1 & (PCI_L1SS_CTL1_LTR_L12_TH_VALUE |
> +					       PCI_L1SS_CTL1_LTR_L12_TH_SCALE));
>  
>  	if (pl1_2_enables || cl1_2_enables) {
>  		pci_clear_and_set_config_dword(parent,
> -- 
> 2.51.0
> 

      reply	other threads:[~2026-02-17 16:11 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-02-16  7:46 [PATCH] PCI/ASPM: Fix pci_clear_and_set_config_dword() usage Lukas Wunner
2026-02-17 16:11 ` Bjorn Helgaas [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260217161119.GA3378634@bhelgaas \
    --to=helgaas@kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=lukas@wunner.de \
    --cc=me@avm99963.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox