On 09/14/2010 06:34 PM, David Lehman wrote:
On Tue, 2010-09-14 at 18:03 +0200, Radek Vykydal wrote:
On 09/14/2010 05:32 PM, Jesse Keating wrote:
Didn't we start using urlgrabber.reset_curl_obj() specifically because
isys.resetResolver wasn't doing enough, and the curl object wasn't
getting the new settings and thus failing to connect to things? What
changed?
Yes, that's what I am thinking now too. I thought that
c-ares is doing complete resolver reset for curl when
doing reset_curl_obj(), but it doesn't seem so. Removing
of isys.resetResolver call is due to my changes in networking
code some months ago, and now I wonder if it ever worked
since, network enablement in stage 2 is not very frequent
use/test case.
Should we combine the two? We could add a call to
urlgrabber.grabber.reset__curl_obj inside isys.resetResolv, then go back
and replace all the explicit calls to urlgrabber.grabber.reset_curl_obj
with calls to isys.resetResolv. Will that help clean things up a bit?
+1, something for master, I'll insert it into my todo list.
Radek
_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
