Hello,
I plan to do the following to the definition of i2c_device_id:
diff --git a/include/linux/mod_devicetable.h b/include/linux/mod_devicetable.h
index 23ff24080dfd..aebd3a5e90af 100644
--- a/include/linux/mod_devicetable.h
+++ b/include/linux/mod_devicetable.h
@@ -477,7 +477,11 @@ struct rpmsg_device_id {
struct i2c_device_id {
char name[I2C_NAME_SIZE];
- kernel_ulong_t driver_data; /* Data private to the driver */
+ union {
+ /* Data private to the driver */
+ kernel_ulong_t driver_data;
+ const void *driver_data_ptr;
+ };
};
/* pci_epf */
. This requires that .driver_data is assigned via a named initializer
for static data. This requirement isn't a bad one because named
initializers are also much better readable than list initializers.
The union added to struct i2c_device_id enables further cleanups like:
diff --git a/drivers/regulator/ad5398.c b/drivers/regulator/ad5398.c
index 0123ca8157a8..dfb0b07500a7 100644
--- a/drivers/regulator/ad5398.c
+++ b/drivers/regulator/ad5398.c
@@ -207,8 +207,8 @@ struct ad5398_current_data_format {
static const struct ad5398_current_data_format df_10_4_120 = {10, 4, 0, 120000};
static const struct i2c_device_id ad5398_id[] = {
- { .name = "ad5398", .driver_data = (kernel_ulong_t)&df_10_4_120 },
- { .name = "ad5821", .driver_data = (kernel_ulong_t)&df_10_4_120 },
+ { .name = "ad5398", .driver_data_ptr = &df_10_4_120 },
+ { .name = "ad5821", .driver_data_ptr = &df_10_4_120 },
{ }
};
MODULE_DEVICE_TABLE(i2c, ad5398_id);
@@ -219,8 +219,7 @@ static int ad5398_probe(struct i2c_client *client)
struct regulator_init_data *init_data = dev_get_platdata(&client->dev);
struct regulator_config config = { };
struct ad5398_chip_info *chip;
- const struct ad5398_current_data_format *df =
- (struct ad5398_current_data_format *)id->driver_data;
+ const struct ad5398_current_data_format *df = id->driver_data;
chip = devm_kzalloc(&client->dev, sizeof(*chip), GFP_KERNEL);
if (!chip)
that are an improvement for readability (again!) and it keeps some
properties of the pointers (here: being const) without having to pay
attention for that. (I didn't find a extcon driver that benefits, so
this is "only" a regulator driver example.)
My additional motivation for this effort is CHERI[1]. This is a hardware
extension that uses 128 bit pointers but unsigned long is still 64 bit.
So with CHERI you cannot store pointers in unsigned long variables.
The first patch simplifies one extcon driver to not use .driver_data,
the second converts all extcon drivers to initialize i2c_device_ids
using named initializers.
Best regards
Uwe
[1] https://cheri-alliance.org/discover-cheri/
https://lwn.net/Articles/1037974/
Uwe Kleine-König (The Capable Hub) (2):
extcon: rt8973: Drop unused driver data
extcon: Use named initializers for arrays of i2c_device_data
drivers/extcon/extcon-fsa9480.c | 4 ++--
drivers/extcon/extcon-lc824206xa.c | 2 +-
drivers/extcon/extcon-max14526.c | 2 +-
drivers/extcon/extcon-ptn5150.c | 2 +-
drivers/extcon/extcon-rt8973a.c | 2 +-
drivers/extcon/extcon-rt8973a.h | 4 ----
drivers/extcon/extcon-sm5502.c | 6 +++---
7 files changed, 9 insertions(+), 13 deletions(-)
base-commit: 254f49634ee16a731174d2ae34bc50bd5f45e731
--
2.47.3