From: "Ilpo Järvinen" <ilpo.jarvinen@linux.intel.com>
To: "Lukas Wunner" <lukas@wunner.de>,
"Bjorn Helgaas" <bhelgaas@google.com>,
"Ilpo Järvinen" <ilpo.jarvinen@linux.intel.com>,
linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: [PATCH 1/1] PCI: Update Link Speed after retraining
Date: Mon, 12 May 2025 16:01:24 +0300 [thread overview]
Message-ID: <20250512130125.9062-1-ilpo.jarvinen@linux.intel.com> (raw)
PCIe Link Retraining can alter Link Speed. While bwctrl listens for
Link Bandwidth Management Status (LBMS) to pick up changes in Link
Speed, there is a race between pcie_reset_lbms() clearing LBMS after
the Link Training and pcie_bwnotif_irq() reading the Link Status
register. If LBMS is already cleared when the irq handler reads the
register, the interrupt handler will return early with IRQ_NONE and
won't update the Link Speed.
When Link Speed update originates from bwctrl,
pcie_bwctrl_change_speed() ensures Link Speed is update after the
retraining. ASPM driver, however, calls pcie_retrain_link() but does
not update the Link Speed after retraining which can result in stale
Link Speed.
To ensure Link Speed is not left state after Link Training, move the
call to pcie_update_link_speed() from pcie_bwctrl_change_speed() into
pcie_retrain_link().
Suggested-by: Lukas Wunner <lukas@wunner.de>
Link: https://lore.kernel.org/linux-pci/aBCjpfyYmlkJ12AZ@wunner.de/
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
---
Based on top of pci/bwctrl.
drivers/pci/pci.c | 17 +++++++++++++++++
drivers/pci/pcie/bwctrl.c | 13 +------------
2 files changed, 18 insertions(+), 12 deletions(-)
diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index 3d94cf33c1b6..eb0c55078d5e 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -4718,6 +4718,11 @@ static int pcie_wait_for_link_status(struct pci_dev *pdev,
* @pdev: Device whose link to retrain.
* @use_lt: Use the LT bit if TRUE, or the DLLLA bit if FALSE, for status.
*
+ * Trigger retraining of the PCIe Link and wait for the completion of the
+ * retraining. As link retraining is known to asserts LBMS and may change
+ * the Link Speed, LBMS is cleared after the retraining and the Link Speed
+ * of the subordinate bus is updated.
+ *
* Retrain completion status is retrieved from the Link Status Register
* according to @use_lt. It is not verified whether the use of the DLLLA
* bit is valid.
@@ -4758,6 +4763,18 @@ int pcie_retrain_link(struct pci_dev *pdev, bool use_lt)
* in attempt to correct unreliable link operation.
*/
pcie_reset_lbms(pdev);
+
+ /*
+ * Ensure the Link Speed updates after retraining in case the Link
+ * Speed was changed because of the retraining. While the bwctrl's
+ * IRQ handler normally picks up the new Link Speed, clearing LBMS
+ * races with the IRQ handler reading the Link Status register and
+ * can result in the handler returning early without updating the
+ * Link Speed.
+ */
+ if (pdev->subordinate)
+ pcie_update_link_speed(pdev->subordinate);
+
return rc;
}
diff --git a/drivers/pci/pcie/bwctrl.c b/drivers/pci/pcie/bwctrl.c
index fdafa20e4587..790a935b34fd 100644
--- a/drivers/pci/pcie/bwctrl.c
+++ b/drivers/pci/pcie/bwctrl.c
@@ -125,18 +125,7 @@ static int pcie_bwctrl_change_speed(struct pci_dev *port, u16 target_speed, bool
if (ret != PCIBIOS_SUCCESSFUL)
return pcibios_err_to_errno(ret);
- ret = pcie_retrain_link(port, use_lt);
- if (ret < 0)
- return ret;
-
- /*
- * Ensure link speed updates also with platforms that have problems
- * with notifications.
- */
- if (port->subordinate)
- pcie_update_link_speed(port->subordinate);
-
- return 0;
+ return pcie_retrain_link(port, use_lt);
}
/**
base-commit: 0238f352a63a075ac1f35ea565b5bec3057ec8bd
--
2.39.5
next reply other threads:[~2025-05-12 13:02 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-05-12 13:01 Ilpo Järvinen [this message]
2025-05-12 13:43 ` [PATCH 1/1] PCI: Update Link Speed after retraining Lukas Wunner
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=20250512130125.9062-1-ilpo.jarvinen@linux.intel.com \
--to=ilpo.jarvinen@linux.intel.com \
--cc=bhelgaas@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pci@vger.kernel.org \
--cc=lukas@wunner.de \
/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