[<prev] [next>] [day] [month] [year] [list]
Message-Id: <20201030141806.1272650-1-paul.kocialkowski@bootlin.com>
Date: Fri, 30 Oct 2020 15:18:06 +0100
From: Paul Kocialkowski <paul.kocialkowski@...tlin.com>
To: linux-kernel@...r.kernel.org
Cc: Lee Jones <lee.jones@...aro.org>,
Thomas Petazzoni <thomas.petazzoni@...tlin.com>,
Paul Kocialkowski <paul.kocialkowski@...tlin.com>
Subject: [PATCH] mfd: core: Restore error-free return for dt-disabled devices
When adding devices via mfd (mfd_add_device), a check on the device-tree
status of the device was performed to gracefully handle disabled devices
without errors.
This was done with a check using of_device_is_available, that was recently
moved to the mfd_match_of_node_to_dev function (by returning -ENODEV).
However, no specific case was added to handle -ENODEV, so it was handled
like any error.
At the end of the day, this causes the mfd_add_devices process to fail as
soon as one device is marked disabled. This is quite problematic and
definitely not what we want.
Introduce a special case to handle -ENODEV gracefully and restore original
behavior.
Fixes: 466a62d7642f ("mfd: core: Make a best effort attempt to match
devices with the correct of_nodes")
Signed-off-by: Paul Kocialkowski <paul.kocialkowski@...tlin.com>
---
drivers/mfd/mfd-core.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/drivers/mfd/mfd-core.c b/drivers/mfd/mfd-core.c
index c3651f06684f..be9900866f2d 100644
--- a/drivers/mfd/mfd-core.c
+++ b/drivers/mfd/mfd-core.c
@@ -215,6 +215,13 @@ static int mfd_add_device(struct device *parent, int id,
ret = mfd_match_of_node_to_dev(pdev, np, cell);
if (ret == -EAGAIN)
continue;
+
+ /* Ignore 'disabled' devices error free */
+ if (ret == -ENODEV) {
+ ret = 0;
+ goto fail_alias;
+ }
+
if (ret)
goto fail_alias;
--
2.28.0
Powered by blists - more mailing lists