[PATCH] sandbox: solve undefined reference to pthread_kill symbol

[Patrice CHOTARD]

On 6/4/25 15:56, Tom Rini wrote:
> On Wed, Jun 04, 2025 at 03:33:31PM +0530, Anshul Dalal wrote:
>> Hi Raymond,
>>
>> On Wed Jun 4, 2025 at 2:30 AM IST, Raymond Mao wrote:
>>> Hi Tom,
>>>
>>> On Tue, 3 Jun 2025 at 15:23, Tom Rini <trini at konsulko.com> wrote:
>>>>
>>>> On Tue, Jun 03, 2025 at 02:00:23PM -0400, Raymond Mao wrote:
>>>>> Hi Tom,
>>>>>
>>>>> On Tue, 3 Jun 2025 at 11:47, Tom Rini <trini at konsulko.com> wrote:
>>>>>>
>>>>>> On Tue, Jun 03, 2025 at 08:13:58AM -0700, Raymond Mao wrote:
>>>>>>
>>>>>>> This patch is to solve the sandbox building error:
>>>>>>> $ make O=build-sandbox -s sandbox_defconfig
>>>>>>> $ make O=build-sandbox -s -j2
>>>>>>> /usr/bin/ld: /tmp/u-boot.27rzOu.ltrans58.ltrans.o: undefined reference to symbol 'pthread_kill@@GLIBC_2.2.5'
>>>>>>> /usr/bin/ld: /lib/x86_64-linux-gnu/libpthread.so.0: error adding symbols: DSO missing from command line
>>>>>>> collect2: error: ld returned 1 exit status
>>>>>>> [...]
>>>>>>>
>>>>>>> Signed-off-by: Raymond Mao <raymond.mao at linaro.org>
>>>>>>> ---
>>>>>>>  arch/sandbox/config.mk | 2 +-
>>>>>>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>>>>>
>>>>>> OK, but how do you get this failure to start with? What host?
>>>>> It failed when I started any sandbox unit test like:
>>>>> ./test/py/test.py --bd sandbox --build -k <any_test>
>>>>> My local host: Ubuntu v20.04, x86_64
>>>>
>>>> We don't see that normally, is there something odd about your compilers?
>>>> This is the kind of thing which should be failing everywhere, but isn't.
>>>> Thanks.
>>>>
>>> This failure occured in my host since v2025.07-rc1, but with the same
>>> host, same setup and same instructions it worked fine in v2025.04-rc5.
>>> Seems that some impacts were introduced during v2025.04-rc5 ~
>>> v2025.07-rc1, which might be worth investigating.
>>>
>>
>> I am unable to reproduce the issue on gcc 15.1.1 and glibc 2.41. I think
>> the issue is likely on your host end.  What compiler and libc are you
>> using for the build?
> 
> This feels like if it's generic, CI should be blowing up too, even if
> it's Ubuntu 22.04 not 20.04 these days. Please bisect down to the
> failure, since you have known good/bad commits and 'git bisect run' can
> use the test case you have.
> 
Hi 

I got exactly the same issue on my Ubuntu 20.04 when trying to compile sandbox.
b989f9ed9fe1 ("test: lib: add initjmp() test") is the the commit which is causing this issue.

Until now, i am still using the default gcc (Ubuntu 10.5.0-1ubuntu1~20.04) 10.5.0.

Thanks
Patrice