Am 2021-01-20 15:23, schrieb Rob Herring:
On Wed, Jan 20, 2021 at 4:53 AM Michael Walle <michael@xxxxxxxx> wrote:
fw_devlink will defer the probe until all suppliers are ready. We
can't
use builtin_platform_driver_probe() because it doesn't retry after
probe
deferral. Convert it to builtin_platform_driver().
If builtin_platform_driver_probe() doesn't work with fw_devlink, then
shouldn't it be fixed or removed? Then we're not fixing drivers later
when folks start caring about deferred probe and devlink.
I'd really prefer if we convert this to a module instead. (And all the
other PCI host drivers.)
I briefly looked into adding a remove() op. But I don't know what will
have to be undo of the ls_pcie_host_init(). That would be up to the
maintainers of this driver.
_But_ I'd really like to see PCI working on my board again ;) At
least until the end of this development cycle.
-michael
Fixes: e590474768f1 ("driver core: Set fw_devlink=on by default")
This happened!?
Signed-off-by: Michael Walle <michael@xxxxxxxx>
---
drivers/pci/controller/dwc/pci-layerscape.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/pci/controller/dwc/pci-layerscape.c
b/drivers/pci/controller/dwc/pci-layerscape.c
index 44ad34cdc3bc..5b9c625df7b8 100644
--- a/drivers/pci/controller/dwc/pci-layerscape.c
+++ b/drivers/pci/controller/dwc/pci-layerscape.c
@@ -232,7 +232,7 @@ static const struct of_device_id
ls_pcie_of_match[] = {
{ },
};
-static int __init ls_pcie_probe(struct platform_device *pdev)
+static int ls_pcie_probe(struct platform_device *pdev)
{
struct device *dev = &pdev->dev;
struct dw_pcie *pci;
@@ -271,10 +271,11 @@ static int __init ls_pcie_probe(struct
platform_device *pdev)
}
static struct platform_driver ls_pcie_driver = {
+ .probe = ls_pcie_probe,
.driver = {
.name = "layerscape-pcie",
.of_match_table = ls_pcie_of_match,
.suppress_bind_attrs = true,
},
};
-builtin_platform_driver_probe(ls_pcie_driver, ls_pcie_probe);
+builtin_platform_driver(ls_pcie_driver);
--
2.20.1
