Mailing List Archive

PRIV_TASK and PRIV_TOP changes in 6.1
Hi all,

I assume https://github.com/varnishcache/varnish-cache/commit/f03948f48445be4dbea5f389ac4168b4a6befc6c is related with the PRIV_TASK and PRIV_TOP change mentioned in section 'Changes for developers' in https://github.com/varnishcache/varnish-cache/blob/master/doc/sphinx/whats-new/upgrading-6.1.rst.

Problem is I'm not sure I'm understanding what are the implications of that change for VMOD developers. For example, I checked implementation of xyzzy_test_priv_task() in lib/libvmod_debug/vmod_debug.c, both in 'master' and 'varnish-6.0.1', and I don't see any relevant changes: no assertions on 'priv', 'priv->priv' initialized as expected, etc. It seems I'm missing something here but I don't know what. Any clues? :)

Thanks,

--
Carlos Abalde

_______________________________________________
varnish-dev mailing list
varnish-dev@varnish-cache.org
https://www.varnish-cache.org/lists/mailman/listinfo/varnish-dev
Re: PRIV_TASK and PRIV_TOP changes in 6.1 [ In reply to ]
On 09/10/2018 12:53 PM, Carlos Abalde wrote:
>
> I assume
> https://github.com/varnishcache/varnish-cache/commit/f03948f48445be4dbea5f389ac4168b4a6befc6c
> is related with the PRIV_TASK and PRIV_TOP change mentioned in section
> 'Changes for developers' in
> https://github.com/varnishcache/varnish-cache/blob/master/doc/sphinx/whats-new/upgrading-6.1.rst.

The draft in 'Upgrading for 6.1' is currently very new (since
yesterday), has not been reviewed and could very well be in error.

We're about to have bugwash (in just a few minutes) and will probably
talk about reviewing and polishing the release notes, among other
things. We'll have quality-assured documentation by the time the release
goes public a week from today.


Best,
Geoff
--
** * * UPLEX - Nils Goroll Systemoptimierung

Scheffelstraße 32
22301 Hamburg

Tel +49 40 2880 5731
Mob +49 176 636 90917
Fax +49 40 42949753

http://uplex.de
Re: PRIV_TASK and PRIV_TOP changes in 6.1 [ In reply to ]
On 10/09/18 12:53, Carlos Abalde wrote:
> Problem is I'm not sure I'm understanding what are the implications of that change for VMOD developers.

now documented here:

https://github.com/varnishcache/varnish-cache/commit/2423a578478c69cc78e52c8900a5c68797e9f998

in short: there is no change from a vmod developer perspective.
_______________________________________________
varnish-dev mailing list
varnish-dev@varnish-cache.org
https://www.varnish-cache.org/lists/mailman/listinfo/varnish-dev
Re: PRIV_TASK and PRIV_TOP changes in 6.1 [ In reply to ]
On 09/11/2018 09:36 AM, Nils Goroll wrote:
> On 10/09/18 12:53, Carlos Abalde wrote:
>> Problem is I'm not sure I'm understanding what are the
>> implications of that change for VMOD developers.
>
> now documented here:
>
> https://github.com/varnishcache/varnish-cache/commit/2423a578478c69cc78e52c8900a5c68797e9f998
>
> in short: there is no change from a vmod developer perspective.

I think it's wise for a VMOD implementation to check if PRIV_*
parameters are NULL, in case of the out-of-workspace condition.

--
** * * UPLEX - Nils Goroll Systemoptimierung

Scheffelstraße 32
22301 Hamburg

Tel +49 40 2880 5731
Mob +49 176 636 90917
Fax +49 40 42949753

http://uplex.de
Re: PRIV_TASK and PRIV_TOP changes in 6.1 [ In reply to ]
On 11/09/18 09:40, Geoff Simmons wrote:
> I think it's wise for a VMOD implementation to check if PRIV_*
> parameters are NULL, in case of the out-of-workspace condition.

We had discussed that option and decided against it. VCC guarantees that the
PRIV_* parameters are never NULL - a VCL failure happens before calling the vmod
function if a PRIV Parameter would be NULL.

Nils
_______________________________________________
varnish-dev mailing list
varnish-dev@varnish-cache.org
https://www.varnish-cache.org/lists/mailman/listinfo/varnish-dev