diff mbox series

disk: part: change from hextoul to dectoul

Message ID 20220617100159.47538-1-jh80.chung@samsung.com
State Changes Requested
Delegated to: Tom Rini
Headers show
Series disk: part: change from hextoul to dectoul | expand

Commit Message

Jaehoon Chung June 17, 2022, 10:01 a.m. UTC
When 'ls' command is running with partition number, it's passed by a hex
value. For example, if want to check a 15th partition, it has to input
as 0xf.
Before applied
- ls mmc 0:f
After applied
- ls mmc 0:15

The using decimal number is more readable than passed by a hex value.

Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
---
 disk/part.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Tom Rini June 17, 2022, 12:27 p.m. UTC | #1
On Fri, Jun 17, 2022 at 07:01:59PM +0900, Jaehoon Chung wrote:

> When 'ls' command is running with partition number, it's passed by a hex
> value. For example, if want to check a 15th partition, it has to input
> as 0xf.
> Before applied
> - ls mmc 0:f
> After applied
> - ls mmc 0:15
> 
> The using decimal number is more readable than passed by a hex value.
> 
> Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>

NAK.  I agree it's a more obvious way of interacting.  But the CLI is
our API with our users and there's lots of stuff out there today that
knows (and then deals with) needing to pass hex not decimal for
partitions.  And given SoCs that have large numbers of partitions, you
can't assume that "12" isn't in intentional and referring to partition
18 in decimal.

We could maybe do this with a flag or environment variable to allow
people to opt-in, since that would at least allow for multi-platform
scripts to know what to expect and not have to guess / hard-code per
platform.
Jaehoon Chung June 20, 2022, 12:46 a.m. UTC | #2
Hi Tom,

On 6/17/22 21:27, Tom Rini wrote:
> On Fri, Jun 17, 2022 at 07:01:59PM +0900, Jaehoon Chung wrote:
> 
>> When 'ls' command is running with partition number, it's passed by a hex
>> value. For example, if want to check a 15th partition, it has to input
>> as 0xf.
>> Before applied
>> - ls mmc 0:f
>> After applied
>> - ls mmc 0:15
>>
>> The using decimal number is more readable than passed by a hex value.
>>
>> Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
> 
> NAK.  I agree it's a more obvious way of interacting.  But the CLI is
> our API with our users and there's lots of stuff out there today that
> knows (and then deals with) needing to pass hex not decimal for
> partitions.  And given SoCs that have large numbers of partitions, you
> can't assume that "12" isn't in intentional and referring to partition
> 18 in decimal.
> 
> We could maybe do this with a flag or environment variable to allow
> people to opt-in, since that would at least allow for multi-platform
> scripts to know what to expect and not have to guess / hard-code per
> platform.

Okay. Frankly, I were not sure that it needs to change or not.
In our boot script, I'm using the hex value and decimal value as other environment variable.
I agreed yours. 

Best Regards,
Jaehoon Chung


>
diff mbox series

Patch

diff --git a/disk/part.c b/disk/part.c
index 79955c7fb000..31a42d8a2312 100644
--- a/disk/part.c
+++ b/disk/part.c
@@ -540,7 +540,7 @@  int blk_get_device_part_str(const char *ifname, const char *dev_part_str,
 		part = PART_AUTO;
 	} else {
 		/* Something specified -> use exactly that */
-		part = (int)hextoul(part_str, &ep);
+		part = (int)dectoul(part_str, &ep);
 		/*
 		 * Less than whole string converted,
 		 * or request for whole device, but caller requires partition.