Control: fixed 828241 7.2.6-1
On 2016-11-10 05:03:17, Adrian Bunk wrote:
> On Mon, Aug 01, 2016 at 07:18:38PM -0400, Antoine Beaupré wrote:
>> On 2016-06-26 06:21:03, Kurt Roeckx wrote:
>> > OpenSSL 1.1.0 is about to released. During a rebuild of all packages using
>> > OpenSSL this package fail to build. A log of that build can be found at:
>> > https://breakpoint.cc/openssl-1.1-rebuild-2016-05-29/Attempted/atheme-services_7.0.7-2_amd64-20160529-1405
>> >
>> > On https://wiki.openssl.org/index.php/1.1_API_Changes you can see various
>> > of the
>> > reasons why it might fail. There are also updated man pages at
>> > https://www.openssl.org/docs/manmaster/ that should contain useful
>> > information.
>> >
>> > There is a libssl-dev package available in experimental that contains a
>> > recent
>> > snapshot, I suggest you try building against that to see if everything
>> > works.
>> >
>> > If you have problems making things work, feel free to contact us.
>>
>> Hi!
>>
>> I have just uploaded 7.2.6 to sid. Hopefully I didn't mess *that* up but
>> since it is a more recent, I wonder if this bug will just be fixed on
>> its own. Are rebuilds done regularly?
>>
>> I do not have time right now to investigate this further, but let me
>> know if I do need to run a rebuild on my own.
>
> 7.2.6-1 fails to build in unstable with OpenSSL 1.1.0:
>
> pbkdf2.c: In function 'PKCS5_PBKDF2_HMAC':
> pbkdf2.c:47:11: error: storage size of 'hctx' isn't known
> HMAC_CTX hctx;
> ^~~~
> pbkdf2.c:51:2: error: implicit declaration of function 'HMAC_CTX_init'
> [-Werror=implicit-function-declaration]
> HMAC_CTX_init(&hctx);
> ^~~~~~~~~~~~~
> pbkdf2.c:87:2: error: implicit declaration of function
> 'HMAC_CTX_cleanup' [-Werror=implicit-function-declaration]
> HMAC_CTX_cleanup(&hctx);
> ^~~~~~~~~~~~~~~~
For the record, upstream released two patches that fix the issues and
that I have uploaded to unstable as part of a new upstream release as
well.
I have tested that the package still compiles correctly under OpenSSL
1.0, so I believe this bug is *also* fixed in stretch, because:
1. if 7.2.7 doesn't migrate down to stretch because OpenSSL 1.1 doesn't
make it, the stretch version (7.2.6) still builds correctly with
OpenSSL 1.0, fixing the bug there
2. if 7.2.7 migrates down to stretch it will compile correctly against
1.0 *or* 1.1, fixing the bug as well
So I am marking this as fixed in 7.2.6-1 even though it is technically
not directly fixed there, to make sure that 7.2.7-1 migrates correctly.
Otherwise atheme would be removed from stretch, and I don't know how
else to keep that from happening.
Let me know if this is the right approach.
A.
--
Je viens d'un pays où engagé veut dire que tu t'es trouvé une job.
- Patrice Desbiens
