From: Alex Williamson <alex.williamson@redhat.com>
To: linux-pci@vger.kernel.org
Cc: bhelgaas@google.com, iommu@lists.linux-foundation.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH 3/5] PCI: Extract link retraining from pcie_aspm_configure_common_clock()
Date: Wed, 26 Oct 2016 14:42:46 -0600 [thread overview]
Message-ID: <20161026144246.5bfdf25c@t450s.home> (raw)
In-Reply-To: <20161026180128.23495.19403.stgit@gimli.home>
On Wed, 26 Oct 2016 12:01:28 -0600
Alex Williamson <alex.williamson@redhat.com> wrote:
> Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
> ---
> drivers/pci/pci.c | 29 +++++++++++++++++++++++++++++
> drivers/pci/pcie/aspm.c | 17 +----------------
> include/linux/pci.h | 1 +
> 3 files changed, 31 insertions(+), 16 deletions(-)
>
> diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
> index 6d6cf89..4d327d4 100644
> --- a/drivers/pci/pci.c
> +++ b/drivers/pci/pci.c
> @@ -4729,6 +4729,35 @@ int pcie_set_mps(struct pci_dev *dev, int mps)
> }
> EXPORT_SYMBOL(pcie_set_mps);
>
> +int pcie_retrain_link(struct pci_dev *dev)
> +{
> + int ret;
> + u16 lnksta;
> + unsigned long timeout;
> +
> + /* Can only retrain from downstream ports */
> + if (!pci_is_pcie(dev) ||
> + (pci_pcie_type(dev) != PCI_EXP_TYPE_DOWNSTREAM &&
> + pci_pcie_type(dev) != PCI_EXP_TYPE_ROOT_PORT))
> + return -EINVAL;
> +
> + ret = pcie_capability_set_word(dev, PCI_EXP_LNKCTL, PCI_EXP_LNKCTL_RL);
> + if (ret)
> + return ret;
> +
> + timeout = jiffies + HZ; /* Retraining timeout */
> + for (;;) {
> + pcie_capability_read_word(dev, PCI_EXP_LNKSTA, &lnksta);
> + if (!(lnksta & PCI_EXP_LNKSTA_LT) ||
> + time_after(jiffies, timeout))
> + break;
> + msleep(1);
> + }
> +
> + return (lnksta & PCI_EXP_LNKSTA_LT) ? -EBUSY : 0;
> +}
> +EXPORT_SYMBOL(pcie_retrain_link);
> +
> int pcie_get_link(struct pci_dev *dev, enum pci_bus_speed *speed,
> enum pcie_link_width *width)
> {
> diff --git a/drivers/pci/pcie/aspm.c b/drivers/pci/pcie/aspm.c
> index 0ec649d..482d60c 100644
> --- a/drivers/pci/pcie/aspm.c
> +++ b/drivers/pci/pcie/aspm.c
> @@ -91,8 +91,6 @@ struct pcie_link_state {
> [POLICY_POWERSAVE] = "powersave"
> };
>
> -#define LINK_RETRAIN_TIMEOUT HZ
> -
> static int policy_to_aspm_state(struct pcie_link_state *link)
> {
> switch (aspm_policy) {
> @@ -222,20 +220,7 @@ static void pcie_aspm_configure_common_clock(struct pcie_link_state *link)
> pcie_capability_write_word(parent, PCI_EXP_LNKCTL, reg16);
>
> /* Retrain link */
> - reg16 |= PCI_EXP_LNKCTL_RL;
> - pcie_capability_write_word(parent, PCI_EXP_LNKCTL, reg16);
> -
> - /* Wait for link training end. Break out after waiting for timeout */
> - start_jiffies = jiffies;
Zero day builders remind me that I forgot to remove the declaration of
start_jiffies from this function and that it's now unused. I'll refrain
from immediately spamming the list with a v2 for that minor change, but
note that this patch will need a respin or fixup on commit, whichever
is preferred. Thanks,
Alex
> - for (;;) {
> - pcie_capability_read_word(parent, PCI_EXP_LNKSTA, ®16);
> - if (!(reg16 & PCI_EXP_LNKSTA_LT))
> - break;
> - if (time_after(jiffies, start_jiffies + LINK_RETRAIN_TIMEOUT))
> - break;
> - msleep(1);
> - }
> - if (!(reg16 & PCI_EXP_LNKSTA_LT))
> + if (!pcie_retrain_link(parent))
> return;
>
> /* Training failed. Restore common clock configurations */
> diff --git a/include/linux/pci.h b/include/linux/pci.h
> index fbfbb40..eb0e9be 100644
> --- a/include/linux/pci.h
> +++ b/include/linux/pci.h
> @@ -1024,6 +1024,7 @@ static inline int pci_is_managed(struct pci_dev *pdev)
> int pcie_set_readrq(struct pci_dev *dev, int rq);
> int pcie_get_mps(struct pci_dev *dev);
> int pcie_set_mps(struct pci_dev *dev, int mps);
> +int pcie_retrain_link(struct pci_dev *dev);
> int pcie_get_minimum_link(struct pci_dev *dev, enum pci_bus_speed *speed,
> enum pcie_link_width *width);
> int pcie_get_link(struct pci_dev *dev, enum pci_bus_speed *speed,
>
next prev parent reply other threads:[~2016-10-26 20:42 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-10-26 18:01 [PATCH 0/5] PCI: ACS enable quirk for link balancing switches Alex Williamson
2016-10-26 18:01 ` [PATCH 1/5] PCI: Make pci_std_enable_acs() non-static Alex Williamson
2016-11-14 20:59 ` Bjorn Helgaas
2016-10-26 18:01 ` [PATCH 2/5] PCI: Extract link speed & width retrieval from pcie_get_minimum_link() Alex Williamson
2016-11-14 21:02 ` Bjorn Helgaas
2016-10-26 18:01 ` [PATCH 3/5] PCI: Extract link retraining from pcie_aspm_configure_common_clock() Alex Williamson
2016-10-26 20:42 ` Alex Williamson [this message]
2016-10-26 18:01 ` [PATCH 4/5] iommu: Move REQ_ACS_FLAGS out to header and rename Alex Williamson
2016-11-10 12:27 ` Joerg Roedel
2016-11-11 22:57 ` Bjorn Helgaas
2016-11-14 12:48 ` Joerg Roedel
2016-10-26 18:01 ` [PATCH 5/5] PCI: Balance ports to avoid ACS errata on Pericom switches Alex Williamson
2016-11-14 21:03 ` Bjorn Helgaas
2016-11-14 21:21 ` Alex Williamson
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=20161026144246.5bfdf25c@t450s.home \
--to=alex.williamson@redhat.com \
--cc=bhelgaas@google.com \
--cc=iommu@lists.linux-foundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pci@vger.kernel.org \
/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;
as well as URLs for NNTP newsgroup(s).