Re: [PATCH v2 2/3] commands: implement devlookup to find device behind device file

[Ahmad Fatoum <a.fatoum@pengutronix.de>]

On 05.07.22 13:59, Sascha Hauer wrote:
> On Tue, Jul 05, 2022 at 01:55:37PM +0200, Ahmad Fatoum wrote:
>> For OF-enabled platforms with aliases, device file naming is pretty much
>> solved: If there is mmc2 = &something, then we have a mmc2 device and
>> a /dev/mmc2 device file. For other platforms like x86, EFI-provided
>> devices are harder to get ahold of. Add a command to make this
>> straight-forward to do in scripts. The main use of this is probably to
>> access parameters like nt_signature or guid:
>>
>>   devlookup /dev/disk0 guid
>>
>> This would print to console, but we have no output capture yet, so add
>> an optional -v VARIABLE parameter as well to allow easy use from
>> scripts.
>>
>> Signed-off-by: Ahmad Fatoum <a.fatoum@pengutronix.de>
>> ---
>> v1 -> v2:
>>   - fix typo in commit message (Michael)
>>   - accept cdev without /dev/ prefix (Sascha)
>>   - early exit on cdev without dev, e.g. /dev/zero
>> ---
>>  commands/Kconfig     | 13 ++++++++
>>  commands/Makefile    |  1 +
>>  commands/devlookup.c | 78 ++++++++++++++++++++++++++++++++++++++++++++
>>  3 files changed, 92 insertions(+)
>>  create mode 100644 commands/devlookup.c
>>
>> diff --git a/commands/Kconfig b/commands/Kconfig
>> index c5505321cfd9..037b41155695 100644
>> --- a/commands/Kconfig
>> +++ b/commands/Kconfig
>> @@ -75,6 +75,19 @@ config CMD_DEVINFO
>>  	  If called with a device path being the argument, devinfo shows more
>>  	  default information about this device and its parameters.
>>  
>> +config CMD_DEVLOOKUP
>> +	tristate
>> +	prompt "devlookup"
>> +	help
>> +	  Look up device behind device file and its parameters
>> +
>> +	  devlookup [-v VAR] /dev/DEVICE [parameter]
>> +
>> +          Detects the device behind a device file and outputs it,
>> +          unless a second argument is given. In that case the device
>> +          parameter with that name is looked up. Specifying -v VARIABLE
>> +          will write output to VARIABLE instead of printing it.
>> +
>>  config CMD_DEVUNBIND
>>  	tristate
>>  	prompt "devunbind"
>> diff --git a/commands/Makefile b/commands/Makefile
>> index b3b7bafe6b64..0aae8893d696 100644
>> --- a/commands/Makefile
>> +++ b/commands/Makefile
>> @@ -110,6 +110,7 @@ obj-$(CONFIG_CMD_DETECT)	+= detect.o
>>  obj-$(CONFIG_CMD_BOOT)		+= boot.o
>>  obj-$(CONFIG_CMD_DEVINFO)	+= devinfo.o
>>  obj-$(CONFIG_CMD_DEVUNBIND)	+= devunbind.o
>> +obj-$(CONFIG_CMD_DEVLOOKUP)	+= devlookup.o
>>  obj-$(CONFIG_CMD_DRVINFO)	+= drvinfo.o
>>  obj-$(CONFIG_CMD_READF)		+= readf.o
>>  obj-$(CONFIG_CMD_MENUTREE)	+= menutree.o
>> diff --git a/commands/devlookup.c b/commands/devlookup.c
>> new file mode 100644
>> index 000000000000..c361d5b58860
>> --- /dev/null
>> +++ b/commands/devlookup.c
>> @@ -0,0 +1,78 @@
>> +// SPDX-License-Identifier: GPL-2.0-only
>> +
>> +#include <common.h>
>> +#include <command.h>
>> +#include <fs.h>
>> +#include <getopt.h>
>> +#include <malloc.h>
>> +#include <linux/stat.h>
>> +#include <linux/ctype.h>
>> +#include <environment.h>
>> +
>> +static int report(const char *variable, const char *val)
>> +{
>> +	if (!val)
>> +		return -(errno ?: EINVAL);
>> +
>> +	if (variable)
>> +		return setenv(variable, val);
>> +
>> +	printf("%s\n", val);
>> +	return 0;
>> +}
>> +
>> +static int do_devlookup(int argc, char *argv[])
>> +{
>> +	const char *variable = NULL, *devicefile, *paramname;
>> +	struct cdev *cdev;
>> +	int opt;
>> +
>> +	while ((opt = getopt(argc, argv, "v:")) > 0) {
>> +		switch(opt) {
>> +		case 'v':
>> +			variable = optarg;
>> +			break;
>> +		}
>> +	}
>> +
>> +	if (argc - optind == 0 || argc - optind > 2)
>> +		return COMMAND_ERROR_USAGE;
>> +
>> +	devicefile = argv[optind];
>> +	paramname  = argv[optind+1];
>> +
>> +	if (strstarts(devicefile, "/dev/"))
>> +		devicefile += sizeof("/dev/") - 1;
> 
> Forgot to mention: We have devpath_to_name() which does exactly that.
> Also I merged your last series with that change, so unless you have
> made other changes we should be fine.

Yes, I added a check below for deviceless cdevs, which otherwise
would crash on devlookup /dev/full

Do you need me to create a v3 or will you fix this up?

Cheers,
Ahmad

> 
> Sascha
> 


-- 
Pengutronix e.K.                           |                             |
Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |