| Message ID | 20240217131943.2472659-1-jonas@kwiboo.se |
|---|---|
| State | Accepted |
| Commit | d07d868be38f2316e4e8818d4d8c30442b8a4fb0 |
| Delegated to: | Tom Rini |
| Headers | show |
| Series | dm: core: Fix address translation in devfdt_get_addr_index() | expand |
On Sat, Feb 17, 2024 at 01:19:41PM +0000, Jonas Karlman wrote: > During address translation #address/size-cells props are expected to > apply to child nodes. However, devfdt_get_addr_index() incorrectly use > the parent offset of the udevice parent instead of getting the offset of > the node parent. This will work in most cases when there is only one > udevice tied to a node. > > On e.g. an Orange Pi R1 Plus TLS board the dwc3-generic (parent) and > dwc3-generic-host (child) udevice is tied to the same node. > > In that case both the offset and parent values end up being the same. > As a result, the #address/size-cells props intended for child nodes > incorrectly gets applied to the node itself resulting in wrong addr > being returned, 0x0 instead of 0xff600000. > > The following can be seen on console: > > dwc3-generic-host usb@ff600000: this is not a DesignWare USB3 DRD Core > dwc3-generic-host usb@ff600000: failed to initialize core > > Fix this by using the offset of the parent node and not the offset to > the node the parent udevice is tied to. > > Signed-off-by: Jonas Karlman <jonas@kwiboo.se> Applied to u-boot/next, thanks!
diff --git a/drivers/core/fdtaddr.c b/drivers/core/fdtaddr.c index 8e774d49ce6e..5f27d2511485 100644 --- a/drivers/core/fdtaddr.c +++ b/drivers/core/fdtaddr.c @@ -23,7 +23,7 @@ fdt_addr_t devfdt_get_addr_index(const struct udevice *dev, int index) { #if CONFIG_IS_ENABLED(OF_REAL) int offset = dev_of_offset(dev); - int parent = dev_of_offset(dev->parent); + int parent = fdt_parent_offset(gd->fdt_blob, offset); fdt_addr_t addr; if (CONFIG_IS_ENABLED(OF_TRANSLATE)) {
During address translation #address/size-cells props are expected to apply to child nodes. However, devfdt_get_addr_index() incorrectly use the parent offset of the udevice parent instead of getting the offset of the node parent. This will work in most cases when there is only one udevice tied to a node. On e.g. an Orange Pi R1 Plus TLS board the dwc3-generic (parent) and dwc3-generic-host (child) udevice is tied to the same node. In that case both the offset and parent values end up being the same. As a result, the #address/size-cells props intended for child nodes incorrectly gets applied to the node itself resulting in wrong addr being returned, 0x0 instead of 0xff600000. The following can be seen on console: dwc3-generic-host usb@ff600000: this is not a DesignWare USB3 DRD Core dwc3-generic-host usb@ff600000: failed to initialize core Fix this by using the offset of the parent node and not the offset to the node the parent udevice is tied to. Signed-off-by: Jonas Karlman <jonas@kwiboo.se> --- I think the following unit test dts and code should be able to detect this issue. Not sure on how to add a test with a different dts, most existing ones seemed to use #address/size-cells = <1>. Please advide on how following could be turned into a proper unit test. /dts-v1/; / { #address-cells = <2>; #size-cells = <2>; some-bus { compatible = "denx,u-boot-test-bus"; reg = <0x0 0xffff0000 0x0 0x1000>; #address-cells = <1>; #size-cells = <0>; device@0 { compatible = "denx,u-boot-fdt-test"; reg = <0>; }; }; }; struct udevice *bus, *dev; ut_assertok(uclass_get_device(UCLASS_TEST_BUS, 0, &bus)); ut_asserteq(0xffff0000, devfdt_get_addr(bus)); ut_assertok(device_bind(bus, DM_DRIVER_GET(denx_u_boot_fdt_test), "test", NULL, dev_ofnode(bus), &dev)); ut_asserteq(0xffff0000, devfdt_get_addr(dev)); --- drivers/core/fdtaddr.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)