Re: [LTP] OpenPOSIX: Add common entry point function for all tests

Re: [LTP] OpenPOSIX: Add common entry point function for all tests

[linuxtestproject.agent]

Hi Martin,

On Thu, 16 Apr 2026, Martin Doucha wrote:
> OpenPOSIX: Add common entry point function for all tests
>
> Add a new shared source file implementing the main() function and rename
> the existing main() functions in test case sources to test_main(). This
> allows adding custom debug code to all test cases at once.

[...]

> diff --git a/testcases/open_posix_testsuite/conformance/interfaces/pthread_create/assertions.xml b/...
> -  NOTE: that the thread in which main() was originally invoked is different
> +  NOTE: that the thread in which test_main() was originally invoked is different

This changes verbatim POSIX specification text. The assertions.xml files
quote the POSIX standard, which defines behaviour relative to main() — the C
entry point, not an LTP-internal function name. Revert these hunks in both
pthread_create/assertions.xml and pthread_exit/assertions.xml.

> diff --git a/.../aio_write/9-2.c b/.../aio_write/9-2.c
> +int test_main(int argc, char** argv)

s/char** argv/char **argv/ to match the style used everywhere else.

---
Note:

Our agent completed the review of the patch. The full review can be
found at: https://patchwork.ozlabs.org/project/ltp/list/?series=500173

The agent can sometimes produce false positives although often its
findings are genuine. If you find issues with the review, please
comment this email or ignore the suggestions.

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH] OpenPOSIX: Add common entry point function for all tests

[Petr Vorel]

Hi all,

> Add a new shared source file implementing the main() function and rename
> the existing main() functions in test case sources to test_main(). This
> allows adding custom debug code to all test cases at once.

Reviewed-by: Petr Vorel <pvorel@suse.cz>
Martin, thanks!

Li, Jan, hopefully you don't mind. We'd prefer not having to keep rebasing.

> Signed-off-by: Martin Doucha <mdoucha@suse.cz>
> ---
>  testcases/open_posix_testsuite/Makefile       | 13 +++---
>  .../conformance/behavior/WIFEXITED/1-1.c      |  2 +-
>  .../conformance/interfaces/nanosleep/3-2.c    |  2 +-
...
>  testcases/open_posix_testsuite/lib/Makefile   | 18 ++++++++
>  testcases/open_posix_testsuite/lib/common.c   | 13 ++++++
>  .../scripts/generate-makefiles.sh             | 10 ++++-
...
>  1695 files changed, 1974 insertions(+), 1920 deletions(-)
>  create mode 100644 testcases/open_posix_testsuite/lib/Makefile
>  create mode 100644 testcases/open_posix_testsuite/lib/common.c

...
> diff --git a/testcases/open_posix_testsuite/conformance/behavior/WIFEXITED/1-1.c b/testcases/open_posix_testsuite/conformance/behavior/WIFEXITED/1-1.c
> index ae021917c..6f7044b4a 100644
> --- a/testcases/open_posix_testsuite/conformance/behavior/WIFEXITED/1-1.c
> +++ b/testcases/open_posix_testsuite/conformance/behavior/WIFEXITED/1-1.c
> @@ -11,7 +11,7 @@
>  #include <unistd.h>
>  #include "posixtest.h"

> -int main()
> +int test_main(int argc PTS_ATTRIBUTE_UNUSED, char **argv PTS_ATTRIBUTE_UNUSED)
>  {
>  	int s;
...
> new file mode 100644
> index 000000000..ad6a9d7e6
> --- /dev/null
> +++ b/testcases/open_posix_testsuite/lib/Makefile
> @@ -0,0 +1,18 @@
> +top_srcdir?=		..
> +subdir=			lib
> +
> +AR?=			ar
> +RANLIB?=		ranlib
> +CFLAGS+=		-I$(top_srcdir)/include
> +
> +
> +vpath %.c $(top_srcdir)/$(subdir)
> +
> +all: libcommon.a
> +
> +clean:
> +	rm -f libcommon.a *.o
> +
> +libcommon.a: common.o
> +	$(AR) -rc "$@" $^
> +	$(RANLIB) "$@"
> diff --git a/testcases/open_posix_testsuite/lib/common.c b/testcases/open_posix_testsuite/lib/common.c
> new file mode 100644
> index 000000000..71e9afb2b
> --- /dev/null
> +++ b/testcases/open_posix_testsuite/lib/common.c
> @@ -0,0 +1,13 @@
> +// SPDX-License-Identifier: GPL-2.0-or-later
> +/*
> + * Copyright (C) 2026 SUSE LLC <mdoucha@suse.cz>
> + *
> + * OpenPOSIX test bootstrap
> + */
> +
> +int test_main(int argc, char **argv);
> +
> +int main(int argc, char **argv)
> +{
> +	return test_main(argc, argv);
> +}

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] [PATCH] OpenPOSIX: Add common entry point function for all tests

[Li Wang via ltp]

Hi Petr, Martin,

Petr Vorel <pvorel@suse.cz> wrote:

> > Add a new shared source file implementing the main() function and rename
> > the existing main() functions in test case sources to test_main(). This
> > allows adding custom debug code to all test cases at once.
>
> Reviewed-by: Petr Vorel <pvorel@suse.cz>
> Martin, thanks!
>
> Li, Jan, hopefully you don't mind. We'd prefer not having to keep rebasing.

Never mind :). Feel free to add:

Reviewed-by: Li Wang <wangli.ahau@gmail.com>


--
Regards,
Li Wang


-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] OpenPOSIX: Add common entry point function for all tests

[Petr Vorel]

Hi Martin,

> Hi Martin,

> On Thu, 16 Apr 2026, Martin Doucha wrote:
> > OpenPOSIX: Add common entry point function for all tests

> > Add a new shared source file implementing the main() function and rename
> > the existing main() functions in test case sources to test_main(). This
> > allows adding custom debug code to all test cases at once.

> [...]

> > diff --git a/testcases/open_posix_testsuite/conformance/interfaces/pthread_create/assertions.xml b/...
> > -  NOTE: that the thread in which main() was originally invoked is different
> > +  NOTE: that the thread in which test_main() was originally invoked is different

> This changes verbatim POSIX specification text. The assertions.xml files
> quote the POSIX standard, which defines behaviour relative to main() — the C
> entry point, not an LTP-internal function name. Revert these hunks in both
> pthread_create/assertions.xml and pthread_exit/assertions.xml.

+1

> > diff --git a/.../aio_write/9-2.c b/.../aio_write/9-2.c
> > +int test_main(int argc, char** argv)

> s/char** argv/char **argv/ to match the style used everywhere else.

+1

@Martin, will you please send v2?

Kind regards,
Petr

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] OpenPOSIX: Add common entry point function for all tests

[Martin Doucha]

Hi!

On 4/17/26 15:09, Petr Vorel wrote:
> Hi Martin,
> 
>> Hi Martin,
> 
>> On Thu, 16 Apr 2026, Martin Doucha wrote:
>>> OpenPOSIX: Add common entry point function for all tests
> 
>>> Add a new shared source file implementing the main() function and rename
>>> the existing main() functions in test case sources to test_main(). This
>>> allows adding custom debug code to all test cases at once.
> 
>> [...]
> 
>>> diff --git a/testcases/open_posix_testsuite/conformance/interfaces/pthread_create/assertions.xml b/...
>>> -  NOTE: that the thread in which main() was originally invoked is different
>>> +  NOTE: that the thread in which test_main() was originally invoked is different
> 
>> This changes verbatim POSIX specification text. The assertions.xml files
>> quote the POSIX standard, which defines behaviour relative to main() — the C
>> entry point, not an LTP-internal function name. Revert these hunks in both
>> pthread_create/assertions.xml and pthread_exit/assertions.xml.
> 
> +1

This "issue" is nonsense. I've checked the XML files and the assertions 
are still correct as is regardless of renaming a few functions. The 
pthread_create() and pthread_exit() specification applies special 
treatment to the main thread and adding one more function call between 
the main() function and the test code does not change that.

>>> diff --git a/.../aio_write/9-2.c b/.../aio_write/9-2.c
>>> +int test_main(int argc, char** argv)
> 
>> s/char** argv/char **argv/ to match the style used everywhere else.

This typo is actually in conformance/interfaces/strcpy/1-1.c (line 23793 
of the patch). Please change the test_main() signature to this during merge:
int test_main(int argc PTS_ATTRIBUTE_UNUSED, char **argv 
PTS_ATTRIBUTE_UNUSED)

I don't think it makes sense to send another 1MB patch to fix one line.

-- 
Martin Doucha   mdoucha@suse.cz
SW Quality Engineer
SUSE LINUX, s.r.o.
CORSO IIa
Krizikova 148/34
186 00 Prague 8
Czech Republic

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] OpenPOSIX: Add common entry point function for all tests

[Petr Vorel]

Hi Martin,

> Hi Martin,

> > Hi Martin,

> > On Thu, 16 Apr 2026, Martin Doucha wrote:
> > > OpenPOSIX: Add common entry point function for all tests

> > > Add a new shared source file implementing the main() function and rename
> > > the existing main() functions in test case sources to test_main(). This
> > > allows adding custom debug code to all test cases at once.

> > [...]

> > > diff --git a/testcases/open_posix_testsuite/conformance/interfaces/pthread_create/assertions.xml b/...
> > > -  NOTE: that the thread in which main() was originally invoked is different
> > > +  NOTE: that the thread in which test_main() was originally invoked is different

> > This changes verbatim POSIX specification text. The assertions.xml files
> > quote the POSIX standard, which defines behaviour relative to main() — the C
> > entry point, not an LTP-internal function name. Revert these hunks in both
> > pthread_create/assertions.xml and pthread_exit/assertions.xml.

> +1

> > > diff --git a/.../aio_write/9-2.c b/.../aio_write/9-2.c
> > > +int test_main(int argc, char** argv)

> > s/char** argv/char **argv/ to match the style used everywhere else.

> +1

> @Martin, will you please send v2?


@Martin, or I can apply the following diff if you agree.

Kind regards,
Petr

> Kind regards,
> Petr

diff --git testcases/open_posix_testsuite/conformance/interfaces/pthread_create/assertions.xml testcases/open_posix_testsuite/conformance/interfaces/pthread_create/assertions.xml
index 9d3cda4d87..7a5e60557d 100644
--- testcases/open_posix_testsuite/conformance/interfaces/pthread_create/assertions.xml
+++ testcases/open_posix_testsuite/conformance/interfaces/pthread_create/assertions.xml
@@ -33,12 +33,12 @@
   If the 'start_routine' returns, the effect shall be as if there was an
   implicit call to pthread_exit() using the return value of 'start_routine'
   as the exit status.
-  NOTE: that the thread in which test_main() was originally invoked is different
+  NOTE: that the thread in which main() was originally invoked is different
   from this
   </assertion>
   <assertion id="7" tag="ref:XSH6:32848:32849">
-  When it returns from test_main(), the effect shall be as if there was an implicit
-  call to exit() using the return value of test_main() as the exit status.
+  When it returns from main(), the effect shall be as if there was an implicit
+  call to exit() using the return value of main() as the exit status.
   </assertion>
   <assertion id="8" tag="ref:XSH6:32850:32852">
   The signal state of the new thread will be initialized as so:
diff --git testcases/open_posix_testsuite/conformance/interfaces/pthread_exit/assertions.xml testcases/open_posix_testsuite/conformance/interfaces/pthread_exit/assertions.xml
index 15d2a09a11..f34f6a197e 100644
--- testcases/open_posix_testsuite/conformance/interfaces/pthread_exit/assertions.xml
+++ testcases/open_posix_testsuite/conformance/interfaces/pthread_exit/assertions.xml
@@ -24,7 +24,7 @@
   </assertion>
   <assertion id="5" tag="ref:XSH6:33031:33033">
   An implicit call to pthread_exit() is made when a thread other than the
-  thread in which test_main() was first invoked returns from the start routine that
+  thread in which main() was first invoked returns from the start routine that
   was used to create it.  The function's return value shall serve as the
   thread's exit status.
   </assertion>
diff --git testcases/open_posix_testsuite/conformance/interfaces/strcpy/1-1.c testcases/open_posix_testsuite/conformance/interfaces/strcpy/1-1.c
index 441654ed6c..e783fc92ea 100644
--- testcases/open_posix_testsuite/conformance/interfaces/strcpy/1-1.c
+++ testcases/open_posix_testsuite/conformance/interfaces/strcpy/1-1.c
@@ -44,7 +44,7 @@ static char *random_string(int len)
     return output_string;
 }
 
-int test_main(int argc, char** argv)
+int test_main(int argc PTS_ATTRIBUTE_UNUSED, char **argv PTS_ATTRIBUTE_UNUSED)
 {
     char *ret_str;
     int i;

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] OpenPOSIX: Add common entry point function for all tests

[Petr Vorel]

Hi Martin,

> Hi!
I overlooked your reply (please my following one with diff).

> On 4/17/26 15:09, Petr Vorel wrote:
> > Hi Martin,

> > > Hi Martin,

> > > On Thu, 16 Apr 2026, Martin Doucha wrote:
> > > > OpenPOSIX: Add common entry point function for all tests

> > > > Add a new shared source file implementing the main() function and rename
> > > > the existing main() functions in test case sources to test_main(). This
> > > > allows adding custom debug code to all test cases at once.

> > > [...]

> > > > diff --git a/testcases/open_posix_testsuite/conformance/interfaces/pthread_create/assertions.xml b/...
> > > > -  NOTE: that the thread in which main() was originally invoked is different
> > > > +  NOTE: that the thread in which test_main() was originally invoked is different

> > > This changes verbatim POSIX specification text. The assertions.xml files
> > > quote the POSIX standard, which defines behaviour relative to main() — the C
> > > entry point, not an LTP-internal function name. Revert these hunks in both
> > > pthread_create/assertions.xml and pthread_exit/assertions.xml.

> > +1

> This "issue" is nonsense. I've checked the XML files and the assertions are
> still correct as is regardless of renaming a few functions. The
> pthread_create() and pthread_exit() specification applies special treatment
> to the main thread and adding one more function call between the main()
> function and the test code does not change that.

I understood that spec talks about main() as a specific C startup function.
But ok, I was probably wrong.

> > > > diff --git a/.../aio_write/9-2.c b/.../aio_write/9-2.c
> > > > +int test_main(int argc, char** argv)

> > > s/char** argv/char **argv/ to match the style used everywhere else.

> This typo is actually in conformance/interfaces/strcpy/1-1.c (line 23793 of
> the patch). Please change the test_main() signature to this during merge:
> int test_main(int argc PTS_ATTRIBUTE_UNUSED, char **argv
> PTS_ATTRIBUTE_UNUSED)

OK, fixed and merged. Thanks!

> I don't think it makes sense to send another 1MB patch to fix one line.

That's why I sent another reply with diff.

Kind regards,
Petr

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] OpenPOSIX: Add common entry point function for all tests

[Martin Doucha]

On 4/17/26 16:44, Petr Vorel wrote:
>>>> This changes verbatim POSIX specification text. The assertions.xml files
>>>> quote the POSIX standard, which defines behaviour relative to main() — the C
>>>> entry point, not an LTP-internal function name. Revert these hunks in both
>>>> pthread_create/assertions.xml and pthread_exit/assertions.xml.
> 
>>> +1
> 
>> This "issue" is nonsense. I've checked the XML files and the assertions are
>> still correct as is regardless of renaming a few functions. The
>> pthread_create() and pthread_exit() specification applies special treatment
>> to the main thread and adding one more function call between the main()
>> function and the test code does not change that.
> 
> I understood that spec talks about main() as a specific C startup function.
> But ok, I was probably wrong.

The spec DOES talk about main() as the C process startup function. But 
the AI for some reason believes that my patch can somehow rewrite the 
POSIX standard itself.

-- 
Martin Doucha   mdoucha@suse.cz
SW Quality Engineer
SUSE LINUX, s.r.o.
CORSO IIa
Krizikova 148/34
186 00 Prague 8
Czech Republic

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp

Re: [LTP] OpenPOSIX: Add common entry point function for all tests

[Petr Vorel]

> On 4/17/26 16:44, Petr Vorel wrote:
> > > > > This changes verbatim POSIX specification text. The assertions.xml files
> > > > > quote the POSIX standard, which defines behaviour relative to main() — the C
> > > > > entry point, not an LTP-internal function name. Revert these hunks in both
> > > > > pthread_create/assertions.xml and pthread_exit/assertions.xml.

> > > > +1

> > > This "issue" is nonsense. I've checked the XML files and the assertions are
> > > still correct as is regardless of renaming a few functions. The
> > > pthread_create() and pthread_exit() specification applies special treatment
> > > to the main thread and adding one more function call between the main()
> > > function and the test code does not change that.

> > I understood that spec talks about main() as a specific C startup function.
> > But ok, I was probably wrong.

> The spec DOES talk about main() as the C process startup function. But the
> AI for some reason believes that my patch can somehow rewrite the POSIX
> standard itself.

I understand it differently: keep main() because that's the C entry function
and any reader of assertions.xml does not really care about our implementation
that we wrap main() with other function for our internal purposes of other test
scenario.

Kind regards,
Petr

-- 
Mailing list info: https://lists.linux.it/listinfo/ltp