Postgres Pro
Downloads
Home   >   mailing lists

Thread: [HACKERS] OpenSSL support in our back branches

[HACKERS] OpenSSL support in our back branches

From
Tom Lane
Date:
I see that buildfarm member anchovy has been failing in pre-9.5
branches for the last several weeks, with symptoms suggesting that
it's been updated to openssl 1.1.0 or later.  The failures are
unsurprising given that commit 593d4e47d wasn't back-patched further
than 9.5, but I wonder whether it isn't time to do that now.

Alternatively, we could ask anchovy's owner to remove --with-openssl
from its configuration for pre-9.5 branches, but there are going to
be more and more buildfarm critters with this issue.
        regards, tom lane

Re: [HACKERS] OpenSSL support in our back branches

From
Andres Freund
Date:
On 2017-04-15 18:04:00 -0400, Tom Lane wrote:
> I see that buildfarm member anchovy has been failing in pre-9.5
> branches for the last several weeks, with symptoms suggesting that
> it's been updated to openssl 1.1.0 or later.  The failures are
> unsurprising given that commit 593d4e47d wasn't back-patched further
> than 9.5, but I wonder whether it isn't time to do that now.

+1.

I think Andreas already did some of the required work in
http://archives.postgresql.org/message-id/0c817abb-3f7d-20fb-583a-58f7593a0bea%40proxel.se


> Alternatively, we could ask anchovy's owner to remove --with-openssl
> from its configuration for pre-9.5 branches, but there are going to
> be more and more buildfarm critters with this issue.

Not just critters, there'll likely also be distributions running into
the issues, too :(

- Andres

Re: [HACKERS] OpenSSL support in our back branches

From
Tom Lane
Date:
Andres Freund <andres@anarazel.de> writes:
> On 2017-04-15 18:04:00 -0400, Tom Lane wrote:
>> I see that buildfarm member anchovy has been failing in pre-9.5
>> branches for the last several weeks, with symptoms suggesting that
>> it's been updated to openssl 1.1.0 or later.  The failures are
>> unsurprising given that commit 593d4e47d wasn't back-patched further
>> than 9.5, but I wonder whether it isn't time to do that now.

> +1.

> I think Andreas already did some of the required work in
> http://archives.postgresql.org/message-id/0c817abb-3f7d-20fb-583a-58f7593a0bea%40proxel.se

Oh, right.  I'll see what I can do with that.
        regards, tom lane