[gentoo-dev] -Werror unwanted?

[gentoo-dev] -Werror unwanted?

[hasufell]

https://bugs.gentoo.org/show_bug.cgi?id=260867

However, I don't see references to ebuild policy (in devmanual or
howtos) how to handle Werror.

Is there a common opinion on that. And shouldn't we add that to the
documentation then?

Re: [gentoo-dev] -Werror unwanted?

[Chí-Thanh Christopher Nguyễn]

hasufell schrieb:
> https://bugs.gentoo.org/show_bug.cgi?id=260867
>
> However, I don't see references to ebuild policy (in devmanual or
> howtos) how to handle Werror.
>
> Is there a common opinion on that. And shouldn't we add that to the
> documentation then?
>

-Werror is basically saying that it is not safe to ship code which
produces warnings.

I personally think that if an upstream says that no warnings must be
produced by the code, and a developer should look at them before
declaring any warnings safe, then that is best followed.

However this causes a heavy maintenance burden and will frequently break
compilation, so the majority opinion is to remove -Werror from compiler
flags.


Best regards,
Chí-Thanh Christopher Nguyễn

Re: [gentoo-dev] -Werror unwanted?

[Mike Frysinger]

[-- Attachment #1: Type: Text/Plain, Size: 419 bytes --]

On Monday 14 May 2012 11:44:17 hasufell wrote:
> https://bugs.gentoo.org/show_bug.cgi?id=260867
> 
> However, I don't see references to ebuild policy (in devmanual or
> howtos) how to handle Werror.
> 
> Is there a common opinion on that. And shouldn't we add that to the
> documentation then?

the common opinion is that no package in the tree should ever allow upstream 
to add -Werror to the build
-mike

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

Re: [gentoo-dev] -Werror unwanted?

[Alexandre Rostovtsev]

On Mon, 2012-05-14 at 17:44 +0200, hasufell wrote:
> https://bugs.gentoo.org/show_bug.cgi?id=260867
> 
> However, I don't see references to ebuild policy (in devmanual or
> howtos) how to handle Werror.
> 
> Is there a common opinion on that. And shouldn't we add that to the
> documentation then?

-Werror is unwanted in anything that links to glib, gtk+, or other gnome
libraries. This is because gnome upstream developers have been adding
compiler warnings for usage of deprecated API which, despite being
deprecated, will in all likelihood remain supported for years; -Werror
turns those warnings into fatal build errors, and tracking down all
instances of deprecated API use twice a year (after a new version of
gnome is released) increases maintenance burden for little benefit.

-Alexandre.

Re: [gentoo-dev] -Werror unwanted?

[hasufell]

On 05/14/2012 06:13 PM, Alexandre Rostovtsev wrote:
> On Mon, 2012-05-14 at 17:44 +0200, hasufell wrote:
>> https://bugs.gentoo.org/show_bug.cgi?id=260867
>>
>> However, I don't see references to ebuild policy (in devmanual or
>> howtos) how to handle Werror.
>>
>> Is there a common opinion on that. And shouldn't we add that to the
>> documentation then?
> 
> -Werror is unwanted in anything that links to glib, gtk+, or other gnome
> libraries. This is because gnome upstream developers have been adding
> compiler warnings for usage of deprecated API which, despite being
> deprecated, will in all likelihood remain supported for years; -Werror
> turns those warnings into fatal build errors, and tracking down all
> instances of deprecated API use twice a year (after a new version of
> gnome is released) increases maintenance burden for little benefit.
> 
> -Alexandre.
> 
> 

So, I will file a documentation bug unless someone can point me in the
right direction. I didn't find a reference to that issue.

Re: [gentoo-dev] -Werror unwanted?

[Jeroen Roovers]

On Mon, 14 May 2012 18:01:22 +0200
Chí-Thanh Christopher Nguyễn <chithanh@gentoo.org> wrote:

> -Werror is basically saying that it is not safe to ship code which
> produces warnings.

An upstream demanding -Werror should work means upstream would need to
test rather a lot more than their own favourite
distro/architecture/library versions/kernel/userland, which isn't
going to happen.

> I personally think that if an upstream says that no warnings must be
> produced by the code, and a developer should look at them before
> declaring any warnings safe, then that is best followed.

Upstream does not need to take into account warnings produced by
compilers for lesser known architectures, as explained above.

As an upstream development aid to check code that has just been added
or changed, -Werror is fine, but not in the wild jungle that is Gentoo.
You might as well just look at the warnings themselves instead of
breaking the build system by making them fatal. In other words, for
upstream development it's convenient, but never for our users out there.

Also, bug reports based on *FLAGS=-Werror will be closed as INVALID.
(Perhaps we should document that too.)


     jer

Re: [gentoo-dev] -Werror unwanted?

[Michał Górny]

[-- Attachment #1: Type: text/plain, Size: 624 bytes --]

On Mon, 14 May 2012 17:44:17 +0200
hasufell <hasufell@gentoo.org> wrote:

> https://bugs.gentoo.org/show_bug.cgi?id=260867
> 
> However, I don't see references to ebuild policy (in devmanual or
> howtos) how to handle Werror.
> 
> Is there a common opinion on that. And shouldn't we add that to the
> documentation then?

Upstream which enforces a particular warning flags on users is a dumb
upstream. Necessary warning flags should be set locally by devs /
distro maintainers rather than through autoconf. If they can't handle
that, someone should probably be replaced.

-- 
Best regards,
Michał Górny

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 316 bytes --]

Re: [gentoo-dev] -Werror unwanted?

[Pacho Ramos]

[-- Attachment #1: Type: text/plain, Size: 1490 bytes --]

El lun, 14-05-2012 a las 20:24 +0200, Jeroen Roovers escribió:
> On Mon, 14 May 2012 18:01:22 +0200
> Chí-Thanh Christopher Nguyễn <chithanh@gentoo.org> wrote:
> 
> > -Werror is basically saying that it is not safe to ship code which
> > produces warnings.
> 
> An upstream demanding -Werror should work means upstream would need to
> test rather a lot more than their own favourite
> distro/architecture/library versions/kernel/userland, which isn't
> going to happen.
> 
> > I personally think that if an upstream says that no warnings must be
> > produced by the code, and a developer should look at them before
> > declaring any warnings safe, then that is best followed.
> 
> Upstream does not need to take into account warnings produced by
> compilers for lesser known architectures, as explained above.
> 
> As an upstream development aid to check code that has just been added
> or changed, -Werror is fine, but not in the wild jungle that is Gentoo.
> You might as well just look at the warnings themselves instead of
> breaking the build system by making them fatal. In other words, for
> upstream development it's convenient, but never for our users out there.
> 
> Also, bug reports based on *FLAGS=-Werror will be closed as INVALID.
> (Perhaps we should document that too.)
> 
> 
>      jer
> 
> 

I fully agree with Jeroen on this, -Werror problems should be reported
directly to upstream if people want to help them on fixing warnings.

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-dev] -Werror unwanted?

[Markos Chandras]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

On 05/14/2012 06:03 PM, hasufell wrote:
> On 05/14/2012 06:13 PM, Alexandre Rostovtsev wrote:
>> On Mon, 2012-05-14 at 17:44 +0200, hasufell wrote:
>>> https://bugs.gentoo.org/show_bug.cgi?id=260867
>>> 
>>> However, I don't see references to ebuild policy (in devmanual
>>> or howtos) how to handle Werror.
>>> 
>>> Is there a common opinion on that. And shouldn't we add that to
>>> the documentation then?
>> 
>> -Werror is unwanted in anything that links to glib, gtk+, or
>> other gnome libraries. This is because gnome upstream developers
>> have been adding compiler warnings for usage of deprecated API
>> which, despite being deprecated, will in all likelihood remain
>> supported for years; -Werror turns those warnings into fatal
>> build errors, and tracking down all instances of deprecated API
>> use twice a year (after a new version of gnome is released)
>> increases maintenance burden for little benefit.
>> 
>> -Alexandre.
>> 
>> 
> 
> So, I will file a documentation bug unless someone can point me in
> the right direction. I didn't find a reference to that issue.
> 

Open a bug, write a devmanual patch and I will be happy to apply it

- -- 
Regards,
Markos Chandras / Gentoo Linux Developer / Key ID: B4AFF2C2
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.19 (GNU/Linux)
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=471E
-----END PGP SIGNATURE-----

Re: [gentoo-dev] -Werror unwanted?

[Chí-Thanh Christopher Nguyễn]

Jeroen Roovers schrieb:
>> -Werror is basically saying that it is not safe to ship code which
>> produces warnings.
> 
> An upstream demanding -Werror should work means upstream would need to
> test rather a lot more than their own favourite
> distro/architecture/library versions/kernel/userland, which isn't
> going to happen.

No. -Werror just means that if a warning is encountered, the user should
be prevented from installing the software. Then a developer looks at the
issue and determines whether it is safe to ignore or needs to be addressed.

>> I personally think that if an upstream says that no warnings must be
>> produced by the code, and a developer should look at them before
>> declaring any warnings safe, then that is best followed.
> 
> Upstream does not need to take into account warnings produced by
> compilers for lesser known architectures, as explained above.

These warnings could be harmless or introduce silent breakage. The user
often can't tell.

> As an upstream development aid to check code that has just been added
> or changed, -Werror is fine, but not in the wild jungle that is Gentoo.
> You might as well just look at the warnings themselves instead of
> breaking the build system by making them fatal. In other words, for
> upstream development it's convenient, but never for our users out there.

-Werror is not convenient for anybody. When the developer has looked at
the issue, then the particular warning could be made non-fatal. hasufell
mentioned in another post the GTK+ deprecated warnings.

Note that I don't propose the current policy to be changed. I can
totally live with filtering -Werror in order to reduce maintenance work,
at the small cost mentioned above.


Best regards,
Chí-Thanh Christopher Nguyễn

[gentoo-dev] Re: -Werror unwanted?

[Nikos Chantziaras]

On 14/05/12 23:42, Chí-Thanh Christopher Nguyễn wrote:
>>> I personally think that if an upstream says that no warnings must be
>>> produced by the code, and a developer should look at them before
>>> declaring any warnings safe, then that is best followed.
>>
>> Upstream does not need to take into account warnings produced by
>> compilers for lesser known architectures, as explained above.
>
> These warnings could be harmless or introduce silent breakage. The user
> often can't tell.

You can have breakage without any warnings being emitted, and you can 
have warnings that result in no breakage whatsoever.

Furthermore, -Werror on Gentoo makes zero sense; portage will already 
produce a QA notice with warnings that have the potential to result in 
breakage.  -Werror is not needed.

Re: [gentoo-dev] -Werror unwanted?

[Tony "Chainsaw" Vroon]

On 14/05/12 16:44, hasufell wrote:
> However, I don't see references to ebuild policy (in devmanual or
> howtos) how to handle Werror.

As can be judged by the title of my patches on the subject, I consider
-Werror to be short-sighted at best and idiotic at worst. The next GCC
version, which will add *loads* of warnings to anything that compiled
cleanly before, is going to kill you.
Remove it from the build system. It is one of those patches that will
probably live downstream until the end of time, but that is acceptable.

Regards,
Tony V.

Re: [gentoo-dev] -Werror unwanted?

[Kacper Kowalik]

[-- Attachment #1: Type: text/plain, Size: 782 bytes --]

On 15.05.2012 13:29, Tony "Chainsaw" Vroon wrote:
> On 14/05/12 16:44, hasufell wrote:
>> However, I don't see references to ebuild policy (in devmanual or
>> howtos) how to handle Werror.
> 
> As can be judged by the title of my patches on the subject, I consider
> -Werror to be short-sighted at best and idiotic at worst. The next GCC
> version, which will add *loads* of warnings to anything that compiled
> cleanly before, is going to kill you.
> Remove it from the build system. It is one of those patches that will
> probably live downstream until the end of time, but that is acceptable.

That's why IMHO the best way to fix those bugs is to make -Werror
optional. It the hardest path, but both upstream and downstream should
be satisfied.
Cheers,
Kacper


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 900 bytes --]

Re: [gentoo-dev] -Werror unwanted?

[Mike Frysinger]

[-- Attachment #1: Type: Text/Plain, Size: 890 bytes --]

On Tuesday 15 May 2012 07:29:36 Tony "Chainsaw" Vroon wrote:
> On 14/05/12 16:44, hasufell wrote:
> > However, I don't see references to ebuild policy (in devmanual or
> > howtos) how to handle Werror.
> 
> As can be judged by the title of my patches on the subject, I consider
> -Werror to be short-sighted at best and idiotic at worst. The next GCC
> version, which will add *loads* of warnings to anything that compiled
> cleanly before, is going to kill you.

to clarify, having -Werror in upstream packages and getting enabled by default 
when doing development is not short-sighted or idiotic at all, but in fact 
makes a lot of sense for a lot of setups.  shipping it enabled by default in a 
release could be considered those things though.

a good compromise is what toolchain (and a few other) packages do: provide a 
configure flag like --disable-werror.
-mike

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

[gentoo-dev] Re: -Werror unwanted?

[Ryan Hill]

[-- Attachment #1: Type: text/plain, Size: 544 bytes --]

On Mon, 14 May 2012 19:58:13 +0100
Markos Chandras <hwoarang@gentoo.org> wrote:

> On 05/14/2012 06:03 PM, hasufell wrote:
> > So, I will file a documentation bug unless someone can point me in
> > the right direction. I didn't find a reference to that issue.

> Open a bug, write a devmanual patch and I will be happy to apply it

FWIW I started a compiler flag policy / best-practice guide a while back.
One of these years I'll get around to actually writing it. :p


-- 
fonts, gcc-porting
toolchain, wxwidgets
@ gentoo.org

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

[gentoo-dev] Re: -Werror unwanted?

[Duncan]

Ryan Hill posted on Mon, 25 Jun 2012 22:57:30 -0600 as excerpted:

> On Mon, 14 May 2012 19:58:13 +0100 Markos Chandras <hwoarang@gentoo.org>
> wrote:
> 
>> On 05/14/2012 06:03 PM, hasufell wrote:
>> > So, I will file a documentation bug unless someone can point me in
>> > the right direction. I didn't find a reference to that issue.
> 
>> Open a bug, write a devmanual patch and I will be happy to apply it
> 
> FWIW I started a compiler flag policy / best-practice guide a while
> back.
> One of these years I'll get around to actually writing it. :p

FWIW, flameeyes' blog has a lot of this sort of content, and with his 
autotools mythbuster and similar content as well, I gather there's at 
least some upstream eyes following it.

So that's a bit of a start for anyone looking for such information right 
now, as well as reasonable research for anyone else considering writing 
such a guide who might want some other references or quote-worthy 
material.

I believe he has specifically covered -Werror too.  I know he mentioned 
it (with a link to /something/, of his or not I didn't check) in a recent 
entry.

-- 
Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman

Re: [gentoo-dev] Re: -Werror unwanted?

[hasufell]

On 06/26/2012 06:59 AM, Duncan wrote:
> Ryan Hill posted on Mon, 25 Jun 2012 22:57:30 -0600 as excerpted:
> 
>> On Mon, 14 May 2012 19:58:13 +0100 Markos Chandras <hwoarang@gentoo.org>
>> wrote:
>>
>>> On 05/14/2012 06:03 PM, hasufell wrote:
>>>> So, I will file a documentation bug unless someone can point me in
>>>> the right direction. I didn't find a reference to that issue.
>>
>>> Open a bug, write a devmanual patch and I will be happy to apply it
>>
>> FWIW I started a compiler flag policy / best-practice guide a while
>> back.
>> One of these years I'll get around to actually writing it. :p
> 
> FWIW, flameeyes' blog has a lot of this sort of content, and with his 
> autotools mythbuster and similar content as well, I gather there's at 
> least some upstream eyes following it.
> 
> So that's a bit of a start for anyone looking for such information right 
> now, as well as reasonable research for anyone else considering writing 
> such a guide who might want some other references or quote-worthy 
> material.
> 
> I believe he has specifically covered -Werror too.  I know he mentioned 
> it (with a link to /something/, of his or not I didn't check) in a recent 
> entry.
> 

Then he should write it. +1 for that

Re: [gentoo-dev] Re: -Werror unwanted?

[Michał Górny]

[-- Attachment #1: Type: text/plain, Size: 1049 bytes --]

On Tue, 26 Jun 2012 04:59:17 +0000 (UTC)
Duncan <1i5t5.duncan@cox.net> wrote:

> Ryan Hill posted on Mon, 25 Jun 2012 22:57:30 -0600 as excerpted:
> 
> > On Mon, 14 May 2012 19:58:13 +0100 Markos Chandras
> > <hwoarang@gentoo.org> wrote:
> > 
> >> On 05/14/2012 06:03 PM, hasufell wrote:
> >> > So, I will file a documentation bug unless someone can point me
> >> > in the right direction. I didn't find a reference to that issue.
> > 
> >> Open a bug, write a devmanual patch and I will be happy to apply it
> > 
> > FWIW I started a compiler flag policy / best-practice guide a while
> > back.
> > One of these years I'll get around to actually writing it. :p
> 
> FWIW, flameeyes' blog has a lot of this sort of content, and with his 
> autotools mythbuster and similar content as well, I gather there's at 
> least some upstream eyes following it.

Thinking about it, it probably would be great to have some organized
catalog of flameeyes' posts. Or maybe he should write a book :).

-- 
Best regards,
Michał Górny

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 316 bytes --]