On 07/06/16 15:33, Paul Gortmaker wrote:
> [[PATCH] serial: mps2-uart: make driver explicitly non-modular] On 07/06/2016 (Tue 14:47) Vladimir Murzin wrote:
>
>> mps2-uart is currently selected by the arch code, which implies is not
>> being built as a module by anyone.
>
> Maybe "depends on" is more accurate, since we have:
>
>
> config SERIAL_MPS2_UART
> bool "MPS2 UART port"
> depends on ARM || COMPILE_TEST
>
> ...so even if people are not using ARM, they can still build it if they
> have selected COMPILE_TEST.
>
>>
>> Follow 89ebc27 "drivers/tty: make serial/mvebu-uart.c explicitly
>> non-modular" as an example of moving modular code to non-modular.
>
> It is good to reference other commits, but the commit message should
> give enough detail to have the full meaning stand-alone even if someone
> doesn't have reference to cited ones.
>
Ok. Just wanted to avoid copy-and-paste, since there is nothing new
except s/module_init/arch_initcall :)
> Also it is good to give at least 12/40 chars of the existing commit ID
> since the kernel has so many commits now that any less than 12 has a
> risk of being ambiguous. There is ./scripts/checkpatch.pl that some
> people find useful for detecting common oversights.
Hmm... it is what I got from checkpatch
total: 0 errors, 0 warnings, 61 lines checked
Anyway, thanks for feedback - I'll update and send v2 shortly.
Cheers
Vladimir
>
>>
>> Reported-by: Paul Gortmaker <paul.gortmaker@xxxxxxxxxxxxx>
>> Signed-off-by: Vladimir Murzin <vladimir.murzin@xxxxxxx>
>> ---
>> drivers/tty/serial/mps2-uart.c | 29 ++++-------------------------
>> 1 file changed, 4 insertions(+), 25 deletions(-)
>>
>> diff --git a/drivers/tty/serial/mps2-uart.c b/drivers/tty/serial/mps2-uart.c
>> index da9e27d..492ec4b 100644
>> --- a/drivers/tty/serial/mps2-uart.c
>> +++ b/drivers/tty/serial/mps2-uart.c
>> @@ -1,4 +1,6 @@
>> /*
>> + * MPS2 UART driver
>> + *
>> * Copyright (C) 2015 ARM Limited
>> *
>> * Author: Vladimir Murzin <vladimir.murzin@xxxxxxx>
>> @@ -17,7 +19,6 @@
>> #include <linux/console.h>
>> #include <linux/io.h>
>> #include <linux/kernel.h>
>> -#include <linux/module.h>
>
> You should replace this with init.h since your driver uses "__init" and
> the file does not have that include already. We shouldn't rely on any
> implicit include paths, even if they just happen to work.
>
>> #include <linux/of_device.h>
>> #include <linux/of.h>
>> #include <linux/platform_device.h>
>> @@ -569,30 +570,20 @@ static int mps2_serial_probe(struct platform_device *pdev)
>> return 0;
>> }
>>
>> -static int mps2_serial_remove(struct platform_device *pdev)
>> -{
>> - struct mps2_uart_port *mps_port = platform_get_drvdata(pdev);
>> -
>> - uart_remove_one_port(&mps2_uart_driver, &mps_port->port);
>> -
>> - return 0;
>> -}
>> -
>> #ifdef CONFIG_OF
>> static const struct of_device_id mps2_match[] = {
>> { .compatible = "arm,mps2-uart", },
>> {},
>> };
>> -MODULE_DEVICE_TABLE(of, mps2_match);
>> #endif
>>
>> static struct platform_driver mps2_serial_driver = {
>> .probe = mps2_serial_probe,
>> - .remove = mps2_serial_remove,
>>
>> .driver = {
>> .name = DRIVER_NAME,
>> .of_match_table = of_match_ptr(mps2_match),
>> + .suppress_bind_attrs = true,
>> },
>> };
>>
>> @@ -610,16 +601,4 @@ static int __init mps2_uart_init(void)
>>
>> return ret;
>> }
>> -module_init(mps2_uart_init);
>> -
>> -static void __exit mps2_uart_exit(void)
>> -{
>> - platform_driver_unregister(&mps2_serial_driver);
>> - uart_unregister_driver(&mps2_uart_driver);
>> -}
>> -module_exit(mps2_uart_exit);
>> -
>> -MODULE_AUTHOR("Vladimir Murzin <vladimir.murzin@xxxxxxx>");
>> -MODULE_DESCRIPTION("MPS2 UART driver");
>> -MODULE_LICENSE("GPL v2");
>> -MODULE_ALIAS("platform:" DRIVER_NAME);
>> +arch_initcall(mps2_uart_init);
>
> The "module_init" maps onto "device_initcall" and since this is a device
> driver, that seems like the appropriate mapping. By moving it to
> arch_initcall, you've changed it to be earlier and that may not be good.
>
> The arch_initcall is more meant for things like poking registers to
> enable some arch/platform specific MMIO bus mapping of devices or
> similar. If you look at include/linux/init.h and the ordering of the
> calls there, it will probably make more sense to you then to leave it as
> device_initcall.
>
>
> #define pure_initcall(fn) __define_initcall(fn, 0)
> #define core_initcall(fn) __define_initcall(fn, 1)
> #define postcore_initcall(fn) __define_initcall(fn, 2)
> #define arch_initcall(fn) __define_initcall(fn, 3) <---
> #define subsys_initcall(fn) __define_initcall(fn, 4)
> #define fs_initcall(fn) __define_initcall(fn, 5)
> #define device_initcall(fn) __define_initcall(fn, 6) <---
> #define late_initcall(fn) __define_initcall(fn, 7)
>
>
> Thanks.
> Paul.
> --
>
>> --
>> 1.7.9.5
>>
>
--
To unsubscribe from this list: send the line "unsubscribe linux-serial" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
