This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: Why was the reallocarray function not added to glibc?
- From: Dennis Wölfing <denniswoelfing at gmx dot de>
- To: Zack Weinberg <zackw at panix dot com>, GNU C Library <libc-alpha at sourceware dot org>
- Cc: Szabolcs Nagy <szabolcs dot nagy at arm dot com>
- Date: Fri, 7 Apr 2017 22:38:39 +0200
- Subject: Re: Why was the reallocarray function not added to glibc?
- Authentication-results: sourceware.org; auth=none
- References: <d3ddd2fb-7f6c-3a85-5c00-666200275997@gmx.de> <58E26EF7.5070008@arm.com> <7afa6179-0e3b-0efe-023b-9c3cdd6da442@gmx.de> <CAKCAbMi9CvNVM-7s4eA_eTFNMR0O=H5-aieP2z4ru+B2tG1raw@mail.gmail.com>
On 07.04.2017 22:09, Zack Weinberg wrote:
> On Fri, Apr 7, 2017 at 3:38 PM, Dennis Wölfing <denniswoelfing@gmx.de> wrote:
>>
>> I think I have fixed the problems with that patch.
>
> Thank you for doing this work.
>
> Do you have a copyright assignment on file with the FSF? We can't
> take patches from anyone without one.
I have sent the signed paper to assign@gnu.org today and am awaiting
their answer.
>> My changes include:
>> - Updating the version number in malloc/Versions
>> - Adding reallocarray to the abilist files
>> - Wrapping the reallocarray declaration in #ifdef __USE_GNU
>> - Changing the reallocarray strong_alias into a weak_alias
>> Now the conform tests (and also the other tests) pass.
>
> This definitely sounds like you're on the right track, at least.
>
>> However because most of the patch was not written by me I am not sure what
>> is the correct way to submit the updated patch.
>>
>> Of course the patch should include the original ChangeLog entry by Rüdiger
>> Sonderfeld. And perhaps I should add another ChangeLog for my changes to the
>> abilists. But do I also need to create additional ChangeLog entries for
>> files where I just changed a few lines of the original patch?
>
> ChangeLog entries need to mention every file that changed. Yes, it's
> tedious, but it's also enormously valuable when you are trying to
> figure out *why* something changed, many years later.
What I meant with this question was not whether every file should have a
ChangeLog but rather whether I should add a new ChangeLog entry when
there already is a ChangeLog entry from the original patch.
For example the ChangeLog from the original patch already says that a
new declaration was added to stdlib.h, do I also need to create a new
ChangeLog entry saying that I added #ifdef __USE_GNU?
>> Also the original patch added new testcase file. Should the patch list
>> 2014-2017 as the copyright date for that file because 2014 is year when it
>> was written or should it be changed to say 2017?
>
> I think we would do 2014-2017 for this.
>
> zw
>