Skip to contentSkip to conversation

cxl/pci: Honor -EPROBE_DEFER from component register setup

PATCH Applied Dave Jiang 3 participants 4 messages Jul 6 – Jul 14 CXL

lore
Share & tools
lore

Dave Jiang

author maintainer
#0 lore raw

PATCH v2 1/1 cxl/pci: Honor -EPROBE_DEFER from component register setup

cxl_pci_setup_regs() for CXL_REGLOC_RBI_COMPONENT can return -EPROBE_DEFER on a Restricted CXL Host (RCD) when the upstream port has not yet been enumerated and the Component Registers must be extracted from the RCRB. cxl_pci_probe() treats every non-zero return from that call as the benign "component registers not found" case, logs a warning, and continues. The rc is then immediately overwritten by the subsequent cxl_pci_type3_init_mailbox() call, so the deferral is silently swallowed.

Return -EPROBE_DEFER instead of continuing so the probe is retried once the upstream port is available.

Fixes 733b57f262b0 ("cxl/pci: Early setup RCH dport component registers from RCRB")
Reported-by [email protected]
Assisted-by Claude:claude-opus-4-8
---
 drivers/cxl/pci.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
drivers/cxl/pci.c +5 -2
index 267c679b0b3c..6a132a6c4c5b 100644
@@ -829,10 +829,13 @@ static int cxl_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	 */
 	rc = cxl_pci_setup_regs(pdev, CXL_REGLOC_RBI_COMPONENT,
 				&cxlds->reg_map);
-	if (rc)
+	if (rc) {
+		if (rc == -EPROBE_DEFER)
+			return rc;
 		dev_warn(&pdev->dev, "No component registers (%d)\n", rc);
-	else if (!cxlds->reg_map.component_map.ras.valid)
+	} else if (!cxlds->reg_map.component_map.ras.valid) {
 		dev_dbg(&pdev->dev, "RAS registers not found\n");
+	}
 
 	rc = cxl_pci_type3_init_mailbox(cxlds);
 	if (rc)

base-commit: 8cdeaa50eae8dad34885515f62559ee83e7e8dda

Signature · 1 line
2.54.0
#1 lore raw
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] cxl_pci_type3_init_mailbox() incorrectly evaluates component_map.ras.valid instead of device_map.mbox.valid to check mailbox presence, leading to kernel panics or functional regressions. --

--- Patch [1]: [PATCH v2] cxl/pci: Honor -EPROBE_DEFER from component register setup --- commit 04637d8b82b95fbc354ce623ffbeb09e3d57ae3e
Author: Dave Jiang <[email protected]>

cxl/pci: Honor -EPROBE_DEFER from component register setup

cxl_pci_setup_regs() can return -EPROBE_DEFER on a Restricted CXL Host (RCD) when the upstream port has not yet been enumerated. This patch ensures that cxl_pci_probe() returns -EPROBE_DEFER instead of continuing so the probe is retried once the upstream port is available.

Quoted 10 lines
drivers/cxl/pci.c +1 -1
index 267c679b0b3c2..6a132a6c4c5b3 100644
@@ -829,10 +829,13 @@ static int cxl_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	 */
 	rc = cxl_pci_setup_regs(pdev, CXL_REGLOC_RBI_COMPONENT,
 				&cxlds->reg_map);
-	if (rc)
+	if (rc) {

drivers/cxl/pci.c static int cxl_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) · line 841

Quoted 10 lines
drivers/cxl/pci.c +4 -1
+		if (rc == -EPROBE_DEFER)
+			return rc;
 		dev_warn(&pdev->dev, "No component registers (%d)\n", rc);
-	else if (!cxlds->reg_map.component_map.ras.valid)
+	} else if (!cxlds->reg_map.component_map.ras.valid) {
 		dev_dbg(&pdev->dev, "RAS registers not found\n");
+	}
 
 	rc = cxl_pci_type3_init_mailbox(cxlds);
 	if (rc)

[Severity: High]
This isn't a bug introduced by this patch, but looking at the mailbox initialization, cxl_pci_type3_init_mailbox() checks the mbox valid flag:

drivers/cxl/pci.c:cxl_pci_type3_init_mailbox() {
    if (!cxlds->reg_map.device_map.mbox.valid)
        return -ENODEV;
    ...
}

Since cxl_pci_setup_regs() was just called above with CXL_REGLOC_RBI_COMPONENT and populated cxlds->reg_map, does this mailbox check inadvertently evaluate component_map.ras.valid due to the union in struct cxl_register_map?

If a device lacks a mailbox but has RAS registers, could the presence check incorrectly pass?

drivers/cxl/pci.c:cxl_pci_setup_mailbox() {
    ...
    const int cap = readl(cxlds->regs.mbox + CXLDEV_MBOX_CAPS_OFFSET);
    ...
}

Would this lead to a null pointer dereference in cxl_pci_setup_mailbox() because cxlds->regs.mbox is null?

Conversely, if a device has a valid mailbox but no RAS registers, would cxl_pci_type3_init_mailbox() incorrectly return -ENODEV and fail the probe of a functional device?

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
maintainer Reviewed-by on v2 1/1
#2 lore raw
In reply to Dave Jiang
Quoted 11 lines · Dave Jiang

cxl_pci_setup_regs() for CXL_REGLOC_RBI_COMPONENT can return -EPROBE_DEFER on a Restricted CXL Host (RCD) when the upstream port has not yet been enumerated and the Component Registers must be extracted from the RCRB. cxl_pci_probe() treats every non-zero return from that call as the benign "component registers not found" case, logs a warning, and continues. The rc is then immediately overwritten by the subsequent cxl_pci_type3_init_mailbox() call, so the deferral is silently swallowed.

Return -EPROBE_DEFER instead of continuing so the probe is retried once the upstream port is available.

Let's get this one merged, and go after the Sashiko pre-existing issue in new patch.

Quoted 36 lines
Fixes 733b57f262b0 ("cxl/pci: Early setup RCH dport component registers from RCRB")
Reported-by [email protected]
Assisted-by Claude:claude-opus-4-8
---
 drivers/cxl/pci.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
drivers/cxl/pci.c +5 -2
index 267c679b0b3c..6a132a6c4c5b 100644
@@ -829,10 +829,13 @@ static int cxl_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	 */
 	rc = cxl_pci_setup_regs(pdev, CXL_REGLOC_RBI_COMPONENT,
 				&cxlds->reg_map);
-	if (rc)
+	if (rc) {
+		if (rc == -EPROBE_DEFER)
+			return rc;
 		dev_warn(&pdev->dev, "No component registers (%d)\n", rc);
-	else if (!cxlds->reg_map.component_map.ras.valid)
+	} else if (!cxlds->reg_map.component_map.ras.valid) {
 		dev_dbg(&pdev->dev, "RAS registers not found\n");
+	}
 
 	rc = cxl_pci_type3_init_mailbox(cxlds);
 	if (rc)

base-commit: 8cdeaa50eae8dad34885515f62559ee83e7e8dda

Signature · 1 line
2.54.0
author maintainer Applied
#3 lore raw
In reply to Dave Jiang
Quoted 9 lines · Dave Jiang

cxl_pci_setup_regs() for CXL_REGLOC_RBI_COMPONENT can return -EPROBE_DEFER on a Restricted CXL Host (RCD) when the upstream port has not yet been enumerated and the Component Registers must be extracted from the RCRB. cxl_pci_probe() treats every non-zero return from that call as the benign "component registers not found" case, logs a warning, and continues. The rc is then immediately overwritten by the subsequent cxl_pci_type3_init_mailbox() call, so the deferral is silently swallowed.

Quoted 8 lines

Return -EPROBE_DEFER instead of continuing so the probe is retried once the upstream port is available.

Fixes 733b57f262b0 ("cxl/pci: Early setup RCH dport component registers from RCRB")
Reported-by [email protected]
Assisted-by Claude:claude-opus-4-8

Applied to cxl/next
430c502c80e5

Quoted 26 lines
---
 drivers/cxl/pci.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
drivers/cxl/pci.c +5 -2
index 267c679b0b3c..6a132a6c4c5b 100644
@@ -829,10 +829,13 @@ static int cxl_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	 */
 	rc = cxl_pci_setup_regs(pdev, CXL_REGLOC_RBI_COMPONENT,
 				&cxlds->reg_map);
-	if (rc)
+	if (rc) {
+		if (rc == -EPROBE_DEFER)
+			return rc;
 		dev_warn(&pdev->dev, "No component registers (%d)\n", rc);
-	else if (!cxlds->reg_map.component_map.ras.valid)
+	} else if (!cxlds->reg_map.component_map.ras.valid) {
 		dev_dbg(&pdev->dev, "RAS registers not found\n");
+	}
 
 	rc = cxl_pci_type3_init_mailbox(cxlds);
 	if (rc)

base-commit: 8cdeaa50eae8dad34885515f62559ee83e7e8dda