[PATCH v2 1/1] cmd/setexpr: support concatenation of direct strings

[Heinrich Schuchardt]

On 2/2/25 16:00, Tom Rini wrote:
> On Sun, Feb 02, 2025 at 01:01:00PM +0100, Heinrich Schuchardt wrote:
>> On 1/20/25 21:52, Tom Rini wrote:
>>> On Sun, Jan 05, 2025 at 11:46:14AM +0100, Heinrich Schuchardt wrote:
>>>
>>>> The setexpr.s command allows to concatenate two strings.
>>>>
>>>> According to the description in doc/usage/cmd/setexpr.rst the parameters
>>>> value1 and value2 can be either direct values or pointers to a
>>>> memory location holding the values.
>>>>
>>>> Unfortunately `setexpr.s <value1> + <value2>` fails if any of the values
>>>> is a direct value. $? is set to false.
>>>>
>>>> * Add support for direct values in setexpr.s.
>>>> * Correct the unit test for "setexpr.s fred 0".
>>>> * Add a new unit test for "setexpr.s fred '1' + '3'" giving '13'.
>>>>
>>>> Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt at canonical.com>
>>>
>>> This causes ut_setexpr_setexpr_test_str to fail.
>>>
>>
>> Hello Tom,
>>
>> I am not able to reproduce this problem in Gitlab.
>> https://source.denx.de/u-boot/custodians/u-boot-efi/-/pipelines/24424
>>
>> Please, provide a link to the failure.
> 
> The three failures in:
> https://source.denx.de/u-boot/u-boot/-/jobs/999497
> were from that test. Do we have some problem lurking perhaps similar to
> what Marek fixed in commit 082b41df8abd ("test/cmd/wget.c: Fix loadaddr
> rewrite") ?
> 

Hello Tom,

Thanks for the link.

With my patch

    setexpr.s fred 0

is expected to succeed and allocate a string "0" that is assigned to 
environment variable fred.

A few lines below Simon acknowledges that the assumption that replacing 
an environment string by a string of same size does not change the value 
of ut_check_free() does not hold true:

         /*
          * This fails in CI at present.
          *
          * ut_assertok(ut_check_delta(start_mem));
          */

I will update the patch accordingly.

@Simon:
In env_do_env_set() free() occurs after malloc(). Depending on the 
current memory layout the free() may or may not result in merging of 
chunks. So it is not expected that the available memory size stays the 
same when replacing the value of an environment variable by a string of 
the same size.

Best regards

Heinrich