Ilya Chernykh wrote:

On Thursday 23 December 2010 16:12:14 Ludwig Nussel wrote:

...

...

I cannot disable an rpmlint check. It says "no-return-in-nonvoid-function". I added an rpmlint file with content [...] Is it a bug in OBS?

No, it's a bug in your package. You are supposed to fix the cause of the problem by patching the sources of the package. Ie add a proper return statement at the place the check tells you.

So this message is impossible to disable?

Nothing is impossible :-) It really doesn't make sense in this case though. The error isn't from rpmlint anyways so trying to disable it by means of rpmlint can't work of course. cu Ludwig -- (o_ Ludwig Nussel //\ V_/_ http://www.suse.de/ SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nuernberg) -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org

On Thu, Dec 23, 2010 at 12:47 PM, Ilya Chernykh wrote:

On Thursday 23 December 2010 16:28:18 Ludwig Nussel wrote:

...

...

So this message is impossible to disable?

Nothing is impossible :-) It really doesn't make sense in this case though. The error isn't from rpmlint anyways so trying to disable it by means of rpmlint can't work of course.

The error is displayed during rpmlint check. How can it be not from rpmlint?

It seems to me you need to do some homework and understand how things work here ;) This error comes from post-build-checks and (luckily) cannot be disabled. Do not attempt to bypass this, if you do, your submissions are likely gonna be rejected. -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org

Am 23.12.2010 15:56, schrieb Cristian Rodríguez:

...

So this message is impossible to disable?

Why would you want the function to return random crap instead of a defined value ?

There are reasons for doing that ;-) I've seen the following examples in the past: - the undefined "return" was never reached. Sometimes the compiler cannot recognize that it's not possible to get there apparently at build time. - people expect to get a random pointer from that function Some developers have interesting ideas but it's not really forbidden so what? But yes, it makes sense to check these errors/warnings/strange things. Wolfgang -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org

On Thu, Dec 23, 2010 at 10:45 AM, Ilya Chernykh wrote:

On Thursday 23 December 2010 17:56:42 Cristian Rodríguez wrote:

...

...

So this message is impossible to disable?

Why would you want the function to return random crap instead of a defined value ?

Simply because returning zero makes not much more sense. I am not the author of that software and I see no reason why returning 0 is better than some other random value. If it is not used, why to patch?

Patching requires efforts and maintenance so that the patch to remain compatible with future versions. Also the program can have no-return problem in multiple places. Just mechanically adding "return 0" or "return 10" does not make the program's behavior more predictable.

Just guessing, but it sounds like the right fix is to change the function return type to "void". Greg -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org

2010/12/26 Juergen Weigert :

On Dec 23, 10 16:15:10 +0300, Ilya Chernykh wrote:

...

On Thursday 23 December 2010 16:12:14 Ludwig Nussel wrote:

...

...

I cannot disable an rpmlint check. It says "no-return-in-nonvoid-function". I added an rpmlint file with content [...] Is it a bug in OBS?

No, it's a bug in your package. You are supposed to fix the cause of the problem by patching the sources of the package. Ie add a proper return statement at the place the check tells you.

So this message is impossible to disable?

Impossible, I am afraid. In addition to rpmlint, the SUSE BuildService also runs several post-build-checks, which also can trigger warnings or abort the build with errors. Often it is not obvious, if an error message came from rpmlint or from post-build-checks. Anything you do in your .rpmlintrc will only affect rpmlint.

If someone finds a way how to disarm fatal errors from the post-build-checks, please add this info to http://en.opensuse.org/openSUSE:Packaging_checks#Disarming_Fatal_Errors Maybe it is *intentionally* impossible, ... should the wiki say that more clearly?

IIRC you could do something like BuildRequires: -post-build-checks ...and there was an alternative syntax more compatible? Perhaps #!BuildRequires: post-build-checks ?

From memory, reading this ML. Never saw it documented and should be tested before adding it to the wiki. The first time I found this brp vs post-build-checks vs rpmlint thing someone said me the idea was to port all the checks to rpmlint...

Anyway I agree, this case should be patched. And obviously submitted upstream to avoid the "Patching requires efforts and maintenance so that the patch to remain compatible with future versions" part. Could be an stupid patch, but nobody ever rejected me a patch to remove a gcc warning (the thing the post-build-check is looking at). -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org

On Sun, Dec 26, 2010 at 10:57:42PM +0100, Cristian Morales Vega wrote:

2010/12/26 Cristian Morales Vega :

...

2010/12/26 Juergen Weigert :

...

On Dec 23, 10 16:15:10 +0300, Ilya Chernykh wrote:

...

On Thursday 23 December 2010 16:12:14 Ludwig Nussel wrote:

...

...

I cannot disable an rpmlint check. It says "no-return-in-nonvoid-function". I added an rpmlint file with content [...] Is it a bug in OBS?

No, it's a bug in your package. You are supposed to fix the cause of the problem by patching the sources of the package. Ie add a proper return statement at the place the check tells you.

So this message is impossible to disable?

Impossible, I am afraid. In addition to rpmlint, the SUSE BuildService also runs several post-build-checks, which also can trigger warnings or abort the build with errors. Often it is not obvious, if an error message came from rpmlint or from post-build-checks. Anything you do in your .rpmlintrc will only affect rpmlint.

If someone finds a way how to disarm fatal errors from the post-build-checks, please add this info to http://en.opensuse.org/openSUSE:Packaging_checks#Disarming_Fatal_Errors Maybe it is *intentionally* impossible, ... should the wiki say that more clearly?

IIRC you could do something like

BuildRequires: -post-build-checks

...and there was an alternative syntax more compatible? Perhaps

#!BuildRequires: post-build-checks

?

Similar... #!BuildIgnore: post-build-checks

http://lists.opensuse.org/opensuse-buildservice/2008-12/msg00022.html

Please just fix the bug, If you need help, please ask. Ciao, Marcus -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org

On 2010-12-23 16:16:42 +0300, Ilya Chernykh wrote:

On Thursday 23 December 2010 16:12:14 Ludwig Nussel wrote:

...

...

I cannot disable an rpmlint check. It says "no-return-in-nonvoid-function". I added an rpmlint file with content [...] Is it a bug in OBS?

No, it's a bug in your package. You are supposed to fix the cause of the problem by patching the sources of the package. Ie add a proper return statement at the place the check tells you.

By the way, it behaves differently for 32-bit and 64-bit cases.For 64-bit case there is no error reported. So I think there is anyway some bug there in rpmlint.

or just arch specific code paths. double check the code and fix it. darix -- openSUSE - SUSE Linux is my linux openSUSE is good for you www.opensuse.org -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org

On Friday 31 December 2010 17:42:38 Dirk Müller wrote:

...

I cannot disable an rpmlint check. It says "no-return-in-nonvoid-function". I added an rpmlint file with content

The reason for that is simple: it is not a rpmlint check. it is from post- build-checks which can not be overwritten.

Thank you. I know it can be disabled, but I rather fixed the source. I just wondered whether can I disable it in theory or not. -- To unsubscribe, e-mail: opensuse-packaging+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-packaging+help@opensuse.org