Problem: testing for serious compiler warnings

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Problem: testing for serious compiler warnings

Christian
hi,

when building I get:
[ 1703s] ... running 01-check-debuginfo
[ 1703s] ... testing for empty debuginfo packages
[ 1703s] ... running 02-check-gcc-output
[ 1703s] ... testing for serious compiler warnings
[ 1703s]     (using /usr/lib/build/checks-data/check_gcc_output)
[ 1703s]     (using /srv/oscbuild/openSUSE_42.3-x86_64/.build.log)
[ 1703s]
[ 1703s] I: Expression compares a char* pointer with a string literal.
[ 1703s]    Usually a strcmp() was intended by the programmer
[ 1703s] E: palemoon stringcompare
/home/abuild/rpmbuild/BUILD/Pale-Moon-27.7.1_Release/toolkit/xre/nsAppRunner.cpp:3597

I did report it upstream and got this:
https://github.com/MoonchildProductions/Pale-Moon/issues/1581

AFAIK I am not a developer.
So who is right ?
wolfbeast or /usr/lib/build/checks-data/check_gcc_output

Thank you

--

Christian
------------------------------------------------------------
   https://join.worldcommunitygrid.org?recruiterId=177038
------------------------------------------------------------
           http://www.sc24.de - Sportbekleidung
------------------------------------------------------------


signature.asc (484 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Problem: testing for serious compiler warnings

Brüns, Stefan
On Freitag, 19. Januar 2018 20:08:10 CET Christian wrote:

> hi,
>
> when building I get:
> [ 1703s] ... running 01-check-debuginfo
> [ 1703s] ... testing for empty debuginfo packages
> [ 1703s] ... running 02-check-gcc-output
> [ 1703s] ... testing for serious compiler warnings
> [ 1703s]     (using /usr/lib/build/checks-data/check_gcc_output)
> [ 1703s]     (using /srv/oscbuild/openSUSE_42.3-x86_64/.build.log)
> [ 1703s]
> [ 1703s] I: Expression compares a char* pointer with a string literal.
> [ 1703s]    Usually a strcmp() was intended by the programmer
> [ 1703s] E: palemoon stringcompare
> /home/abuild/rpmbuild/BUILD/Pale-Moon-27.7.1_Release/toolkit/xre/nsAppRunner
> .cpp:3597
>
> I did report it upstream and got this:
> https://github.com/MoonchildProductions/Pale-Moon/issues/1581
>
> AFAIK I am not a developer.
> So who is right ?
> wolfbeast or /usr/lib/build/checks-data/check_gcc_output

The author apparently lacks any C/C++ knowledge ...

http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2017/n4659.pdf

5.13.5 String literals, paragraph 16
 Evaluating a string-literal results in a string literal object with static
storage duration, initialized from the given characters as specified above.
Whether all string literals are distinct (that is, are stored in
nonoverlapping objects) and whether successive evaluations of a string-literal
yield the same or a different object is unspecified. [ Note: The effect of
attempting to modify a string literal is undefined. — end note ]

Note the "unspecified" in the paragraph above.

Kind regards,

Stefan
N�����r��y隊Z)z{.���Wlz��qﮞ˛���m�)z{.��+�:�{Zr�az�'z��j)h���Ǜ�)]���Ǿ� ޮ�^�ˬz��
Reply | Threaded
Open this post in threaded view
|

Re: Problem: testing for serious compiler warnings

Christian
Am 19.01.2018 um 22:56 schrieb Brüns, Stefan:
> The author apparently lacks any C/C++ knowledge ...
Looks like wolfbeast is not right.
But he is behaving as if he 'wrote' C/C++ himself ...

>
> Note the "unspecified" in the paragraph above.
Thanks for the doc snippet, but as a non-developer I don't understand it.

Thank you

--

Christian
------------------------------------------------------------
   https://join.worldcommunitygrid.org?recruiterId=177038
------------------------------------------------------------
           http://www.sc24.de - Sportbekleidung
------------------------------------------------------------


signature.asc (484 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Problem: testing for serious compiler warnings

Neal Gompa
On Sat, Jan 20, 2018 at 12:21 PM, Christian <[hidden email]> wrote:
> Am 19.01.2018 um 22:56 schrieb Brüns, Stefan:
>> The author apparently lacks any C/C++ knowledge ...
> Looks like wolfbeast is not right.
> But he is behaving as if he 'wrote' C/C++ himself ...
>
>>
>> Note the "unspecified" in the paragraph above.
> Thanks for the doc snippet, but as a non-developer I don't understand it.
>

"unspecified" and "undefined" basically mean that the behavior is not
set on how to handle such a case. That means that it's more or less up
to the compiler. And while the compiler will attempt to resolve the
issue, the way it does is not guaranteed to be consistent,
reproducible, or even sane.

Markus Straver (the author of Pale Moon) develops primarily on Windows
with the lazy MSVC compiler (MSVC 2012[1]), so he won't notice much
from stricter compilers. In fact, last I checked, the project actively
advises against building Pale Moon with anything newer than GCC 4.9
anyway[2]. They don't have CI or bother to try to make modern
compilers like GCC or Clang part of their testing, and do not actively
try to resolve issues detected by the compiler.

I'm honestly surprised Pale Moon is available somewhere for openSUSE,
given how much they go out of their way to make things better on
Linux.

[1]: https://developer.palemoon.org/Developer_Guide:Build_Instructions/Pale_Moon/Windows
[2]: https://developer.palemoon.org/Developer_Guide:Build_Instructions/Pale_Moon/Linux


--
真実はいつも一つ!/ Always, there's only one truth!
--
To unsubscribe, e-mail: [hidden email]
To contact the owner, e-mail: [hidden email]