[PATCH 3/3] Do not reset bereq on retry. Add rollback_bereq for this, purpose.

[Federico Schwindt]

Attached.

OK?




On Wed, Jul 30, 2014 at 9:01 AM, Poul-Henning Kamp <phk at phk.freebsd.dk>
wrote:

> --------
> In message <CAJV_h0bR-h=i0e-GKwZrd5fE9q-1AsF=_sJ=gFJQ=
> 6CbiF0kow at mail.gmail.com>
> , Federico Schwindt writes:
>
> Two things, on trivial-ish and the other serious-ish.
>
> The trivial thing is that I think we should put the actual
> rollback in VRT_Rollback() (give it an VCL_HTTP argument) so that
> std.rollback just wraps that.
>
> The other thing is that I'm not even close to being convinced that
> resetting the workspace is safe anymore, and I have no idea how I
> can convince myself that it is, in a world full of VMODs.
>
> We may have to give up on that, and only rollback the struct http.
>
> --
> Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
> phk at FreeBSD.ORG         | TCP/IP since RFC 956
> FreeBSD committer       | BSD since 4.3-tahoe
> Never attribute to malice what can adequately be explained by incompetence.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.varnish-cache.org/lists/pipermail/varnish-dev/attachments/20140730/a0beb461/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: r1512v2.patch
Type: text/x-diff
Size: 7716 bytes
Desc: not available
URL: <https://www.varnish-cache.org/lists/pipermail/varnish-dev/attachments/20140730/a0beb461/attachment-0001.patch>