AU: Make proxy resolution asynchronous.

AU: Make proxy resolution asynchronous.

This doesn't change proxy resolution overall (we still use settings
stored in the session manager), but it changes the implementation in
the updater to be asynchronous. The clients of the proxy resolver now
give a callback to be called when the proxies are known.

This is anticipation of a switch to using Chrome to resolve proxies,
which will need to be asynchronous.

BUG=chromium-os:12079
TEST=unittests; tested update on device w/ and w/o proxy settings

Committed: http://chrome-svn/viewvc/chromeos?view=rev&revision=f3ed8e7

[petkov, 2011-02-14 22:13:50]

mostly white-space nits, LGTM otherwise. honestly, this CL made my head spin :/

http://codereview.chromium.org/6516026/diff/1/http_fetcher.h
File http_fetcher.h (right):

http://codereview.chromium.org/6516026/diff/1/http_fetcher.h#newcode41
http_fetcher.h:41: no_resolver_idle_id_(0),
s/tab/spaces/

http://codereview.chromium.org/6516026/diff/1/http_fetcher.h#newcode43
http_fetcher.h:43: virtual ~HttpFetcher();
s/tab/spaces/

http://codereview.chromium.org/6516026/diff/1/http_fetcher.h#newcode121
http_fetcher.h:121: // The ID of the idle callback, used when we have no proxy
resolver.
indentation is off

http://codereview.chromium.org/6516026/diff/1/proxy_resolver.h
File proxy_resolver.h (right):

http://codereview.chromium.org/6516026/diff/1/proxy_resolver.h#newcode50
proxy_resolver.h:50: virtual ~DirectProxyResolver();
s/tab/spaces/

http://codereview.chromium.org/6516026/diff/1/proxy_resolver.h#newcode57
proxy_resolver.h:57: guint idle_callback_id_;
fix indentation

http://codereview.chromium.org/6516026/diff/1/utils.h
File utils.h (right):

http://codereview.chromium.org/6516026/diff/1/utils.h#newcode247
utils.h:247: gboolean GlibRunClosure(gpointer data);
Why don't you take a Closure* as input then?

[adlr, 2011-02-16 18:18:22]

Yeah, I know it's tricky. I basically changed a "leaf" function to be async, and
thus had to walk up the chain of callers and make them async as well.

FIxed the whitespace issues (sorry, been working w/ a lot of 3rd party code
lately) and will submit after another round of testing.

http://codereview.chromium.org/6516026/diff/1/http_fetcher.h
File http_fetcher.h (right):

http://codereview.chromium.org/6516026/diff/1/http_fetcher.h#newcode41
http_fetcher.h:41: no_resolver_idle_id_(0),
On 2011/02/14 22:13:50, petkov wrote:
> s/tab/spaces/

Sorry, been working w/ some 3rd party/non-google-style code lately. Done.

http://codereview.chromium.org/6516026/diff/1/http_fetcher.h#newcode43
http_fetcher.h:43: virtual ~HttpFetcher();
On 2011/02/14 22:13:50, petkov wrote:
> s/tab/spaces/

Done.

http://codereview.chromium.org/6516026/diff/1/http_fetcher.h#newcode121
http_fetcher.h:121: // The ID of the idle callback, used when we have no proxy
resolver.
On 2011/02/14 22:13:50, petkov wrote:
> indentation is off

Done.

http://codereview.chromium.org/6516026/diff/1/proxy_resolver.h
File proxy_resolver.h (right):

http://codereview.chromium.org/6516026/diff/1/proxy_resolver.h#newcode50
proxy_resolver.h:50: virtual ~DirectProxyResolver();
On 2011/02/14 22:13:50, petkov wrote:
> s/tab/spaces/

Done.

http://codereview.chromium.org/6516026/diff/1/proxy_resolver.h#newcode57
proxy_resolver.h:57: guint idle_callback_id_;
On 2011/02/14 22:13:50, petkov wrote:
> fix indentation

Done.

http://codereview.chromium.org/6516026/diff/1/utils.h
File utils.h (right):

http://codereview.chromium.org/6516026/diff/1/utils.h#newcode247
utils.h:247: gboolean GlibRunClosure(gpointer data);
On 2011/02/14 22:13:50, petkov wrote:
> Why don't you take a Closure* as input then?

This is the type that g_idle_add takes for a function pointer. If i use Closure
here, I would have to cast the pointer to this function to a different type of
function pointer (kinda gross)

[petkov, 2011-02-16 18:32:04]

Yeah, it's fine. It almost feels like if you didn't use protobuf's Closure and
other abstractions the code might have been simpler but I didn't look/follow
deeply enough to be sure.