Thread (5 messages) flat view 5 messages, 4 authors, 16d ago

Re: [PATCH v3] ata: ahci_da850: switch to DEFINE_SIMPLE_DEV_PM_OPS

From: sashiko-bot@kernel.org
Date: 2026-09-08 06:20:38

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

Pre-existing issues:
- [High] The ahci_da850 driver hooks into standard ahci_platform_resume, but fails to restore SoC-specific SATA PHY registers after suspend, causing SATA to be uninitialized on resume.
--

commit d5b66599e27a81a0e92aca89e46cdd92303b81e7
Author: Li Jun [off-list ref]

ata: ahci_da850: switch to DEFINE_SIMPLE_DEV_PM_OPS

This patch replaces the deprecated SIMPLE_DEV_PM_OPS macro with the
recommended DEFINE_SIMPLE_DEV_PM_OPS. It also uses pm_ptr() to complete
the API migration and allow the pm_ops structure to be dropped when
CONFIG_PM is disabled.
quoted hunk ↗ jump to hunk
diff --git a/drivers/ata/ahci_da850.c b/drivers/ata/ahci_da850.c
--- a/drivers/ata/ahci_da850.c
+++ b/drivers/ata/ahci_da850.c
[ ... ]
quoted hunk ↗ jump to hunk
@@ -227,8 +227,8 @@ static int ahci_da850_probe(struct platform_device *pdev)
 	return rc;
 }
 
-static SIMPLE_DEV_PM_OPS(ahci_da850_pm_ops, ahci_platform_suspend,
-			 ahci_platform_resume);
+static DEFINE_SIMPLE_DEV_PM_OPS(ahci_da850_pm_ops, ahci_platform_suspend,
+								ahci_platform_resume);
[Severity: High]
This is a pre-existing issue, but by examining the PM operations being wired
here, it appears the driver hooks into the generic ahci_platform_resume
without restoring SoC-specific SATA PHY registers.

During probe in drivers/ata/ahci_da850.c, ahci_da850_probe() initializes
the SATA PHY registers by calling da850_sata_init() with local parameters.
However, the resume callback is wired directly to the generic
ahci_platform_resume, which only restores standard AHCI state and bulk
clocks. It is completely unaware of the custom DA850 PHY initialization.

Furthermore, the necessary parameters for PHY initialization (pwrdn_reg and
mpy) are local to the ahci_da850_probe() function and not saved for later
use.

Could a system suspend followed by a resume cycle leave the SATA controller
uninitialized, causing connected drives to become inaccessible?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260908061025.3580261-1-lijun01@kylinos.cn?part=1
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help