Skip to contentSkip to conversation

ata: libahci_platform: Do not set mask_port_map when not needed

PATCH Superseded Damien Le Moal 3 participants 3 messages Feb 7, 2025

Subsystems (2)
libata libata SATA AHCI platform devices support

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore

Damien Le Moal

author maintainer
#0 lore raw

PATCH 1/1 ata: libahci_platform: Do not set mask_port_map when not needed

Commit 8c87215dd3a2 ("ata: libahci_platform: support non-consecutive port numbers") modified ahci_platform_get_resources() to allow identifying the ports of a controller that are defined as child nodes of the controller node in order to support non-consecutive port numbers (as defined by the platform device tree).

However, this commit also erroneously sets bits 0 of hpriv->mask_port_map when the platform devices tree does not define port child nodes, to match the fact that the temporary default number of ports used in that case is 1. Doing so causes ahci_platform_init_host() to initialize and probe only the first port, even if the controller has multiple ports (that are not defined through the platform DT).

Fix this by removing setting bit 0 of hpriv->mask_port_map when the platform devices tree does not define port child nodes.

Reported-by Klaus Kudielka <[email protected]>
Fixes 8c87215dd3a2 ("ata: libahci_platform: support non-consecutive port numbers")
Cc [email protected]
Tested-by Klaus Kudielka <[email protected]>
Signed-off-by Damien Le Moal <[email protected]>
---
 drivers/ata/libahci_platform.c | 2 --
 1 file changed, 2 deletions(-)
drivers/ata/libahci_platform.c +0 -2
index 53b2c7719dc5..91d44302eac9 100644
@@ -651,8 +651,6 @@ struct ahci_host_priv *ahci_platform_get_resources(struct platform_device *pdev,
 		 * If no sub-node was found, keep this for device tree
 		 * compatibility
 		 */
-		hpriv->mask_port_map |= BIT(0);
-
 		rc = ahci_platform_get_phy(hpriv, 0, dev, dev->of_node);
 		if (rc)
 			goto err_out;
Signature · 1 line
2.48.1
maintainer
#1 lore raw
Quoted 7 lines · Damien Le Moal

Commit 8c87215dd3a2 ("ata: libahci_platform: support non-consecutive port numbers") modified ahci_platform_get_resources() to allow identifying the ports of a controller that are defined as child nodes of the controller node in order to support non-consecutive port numbers (as defined by the platform device tree).

However, this commit also erroneously sets bits 0 of

s/bits/bit/

Quoted 8 lines

hpriv->mask_port_map when the platform devices tree does not define port child nodes, to match the fact that the temporary default number of ports used in that case is 1. Doing so causes ahci_platform_init_host() to initialize and probe only the first port, even if the controller has multiple ports (that are not defined through the platform DT).

Fix this by removing setting bit 0 of hpriv->mask_port_map when the platform devices tree does not define port child nodes.

s/devices/device/

Quoted 4 lines
Reported-by Klaus Kudielka <[email protected]>
Fixes 8c87215dd3a2 ("ata: libahci_platform: support non-consecutive port numbers")
Cc [email protected]

Cc: stable is used then the fix should be backported.
The commit referenced in Fixes: is only in v6.14-rc1, so no need to backport, and thus no need to Cc: stable.

Kind regards,
Niklas

#2 lore raw
In reply to Damien Le Moal

Hi Damien,

Am 07.02.25 um 08:48 schrieb Damien Le Moal:

Quoted 10 lines

Commit 8c87215dd3a2 ("ata: libahci_platform: support non-consecutive port numbers") modified ahci_platform_get_resources() to allow identifying the ports of a controller that are defined as child nodes of the controller node in order to support non-consecutive port numbers (as defined by the platform device tree).

However, this commit also erroneously sets bits 0 of hpriv->mask_port_map when the platform devices tree does not define port child nodes, to match the fact that the temporary default number of ports used in that case is 1.

And to match the fact that only index 0 of hpriv->phys[0] was initialized from calling into ahci_platform_get_phy.

Quoted 21 lines

Doing so causes ahci_platform_init_host()
to initialize and probe only the first port, even if the controller has multiple ports (that are not defined through the platform DT).

Fix this by removing setting bit 0 of hpriv->mask_port_map when the platform devices tree does not define port child nodes.

Reported-by Klaus Kudielka <[email protected]>
Fixes 8c87215dd3a2 ("ata: libahci_platform: support non-consecutive port numbers")
Cc [email protected]
Tested-by Klaus Kudielka <[email protected]>
Signed-off-by Damien Le Moal <[email protected]>
---
 drivers/ata/libahci_platform.c | 2 --
 1 file changed, 2 deletions(-)
drivers/ata/libahci_platform.c +0 -0
index 53b2c7719dc5..91d44302eac9 100644

drivers/ata/libahci_platform.c struct ahci_host_priv *ahci_platform_get_resources(struct platform_device *pdev, · line 654

Quoted 6 lines
drivers/ata/libahci_platform.c +0 -2
@@ -651,8 +651,6 @@ struct ahci_host_priv *ahci_platform_get_resources(struct platform_device *pdev,
 		 * If no sub-node was found, keep this for device tree
 		 * compatibility
 		 */
-		hpriv->mask_port_map |= BIT(0);
-
 		rc = ahci_platform_get_phy(hpriv, 0, dev, dev->of_node);

This function initializes hpriv->phys[0], and only at index 0.

I am not aware whether or not this causes any issue, therefore no strong objection to restoring original behaviour.

drivers/ata/libahci_platform.c struct ahci_host_priv *ahci_platform_get_resources(struct platform_device *pdev, · line 656

Quoted 2 lines
drivers/ata/libahci_platform.c +0 -0
 		if (rc)
 			goto err_out;

sincerely
Josua Mayer