commit 557c93c3070e306ea9b685632dddc7c8c8fd8d4b
author Findit <findit-for-me@appspot.gserviceaccount.com> Sun Oct 21 05:02:08 2018
committer Findit <findit-for-me@appspot.gserviceaccount.com> Sun Oct 21 05:02:08 2018
tree 29ecd440b203fccd429234aa7c3a6b96426d26a1
parent 3d1759c957a0b1420e4638fe907b82611b880b9c

Revert "Migrate chrome/test/chromedriver/net/net_util.cc using SimpleURLLoader"

This reverts commit c7e36b09013c046e265d70dfed0ca40b73bc23da.

Reason for revert:

Findit (https://goo.gl/kROfz5) identified CL at revision 601422 as the
culprit for failures in the build cycles as shown on:
https://findit-for-me.appspot.com/waterfall/culprit?key=ag9zfmZpbmRpdC1mb3ItbWVyRAsSDVdmU3VzcGVjdGVkQ0wiMWNocm9taXVtL2M3ZTM2YjA5MDEzYzA0NmUyNjVkNzBkZmVkMGNhNDBiNzNiYzIzZGEM

Sample Failed Build: https://ci.chromium.org/buildbot/chromium/mac-dbg/1239

Sample Failed Step: compile

Original change's description:
> Migrate chrome/test/chromedriver/net/net_util.cc using SimpleURLLoader
> 
> URLFetcher will stop working with advent of Network Service, and
> SimpleURLLoader is the replacement API for most clients.
> This CL migrates chromedriver and its respective unittests away from
> URLFetcher.
> 
> Note that because of the multithreaded nature of the code, the natural
> approach was to simply call SharedURLLoaderFactory::Clone and pass
> the "info" mojo handle accross threads, until the operational thread.
> However, the structure holding this mojo handle is called/passed from
> 
>   (i) an indirect chain of BindRepeating calls,
>   (ii) various threads.
> 
> Given that network::mojom::URLLoaderFactoryPtrInfo and
> network::SharedURLLoaderFactoryInfo are both move-only object
> (which does not comply with BingRepeating) and that raw
> a network::SharedURLoaderLoader instance must operate on the same
> thread it was created on, this CL comes up with a local thread-agnostic
> "wrapper" to mojom::URLLoaderFactory, named WrapperURLLoaderFactory
> (see chrome/test/chromedriver/server/http_handler.cc).
> Its instance replaces URLRequestContextGetter instead across threads
> and BindRepeating chains just fine.
> 
> This CL is based on the original work of Sergio Villar (svillar@igalia.com)
> on [1].
> 
> TBR=johnchen@chromium.org (John +1'ed the original incarnation of this CL at [1]).
> 
> [1] https://crrev.com/c/1221256
> 
> Bug: 872887
> 
> Change-Id: I21386edb04b25438ba3aa40dd5ecc1461342aecf
> Reviewed-on: https://chromium-review.googlesource.com/c/1289849
> Commit-Queue: Antonio Gomes <tonikitoo@igalia.com>
> Reviewed-by: Matt Menke <mmenke@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#601422}

Change-Id: I0bb5f07fde222dacb7e34892d907458c8b8535a5
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 872887
Reviewed-on: https://chromium-review.googlesource.com/c/1293041
Cr-Commit-Position: refs/heads/master@{#601430}