This reverts commit 31d9930a.

Nat's suggested improvement to LLThreadSafeQueue that takes account of queue being empty as well as the status flag condition

Get rid of that ugly '#pragma clang diagnostic ignored "-Wdelete-incomplete"' by making sure the delete always happens inside llview.cpp, where the type of LLView is known.

Updated nghttp2 to codeticket version 545354
Updated boost to codeticket version 545361
Updated colladadom to codeticket version 545362
Updated googlemock to codeticket version 545363
Updated curl to codeticket version 545369

LLExtStat had been a S32, this wasn't right, as some of the constants lead to integer overflow: const LLExtStat LL_EXSTAT_RES_RESULT = 2L<<30; const LLExtStat LL_EXSTAT_VFS_RESULT = 3L<<30; This shifts into the sign bit and clang gets (rightfully) upset about this.

LLExtStatus needs to be at least of type U32 to remedy this problem, but
while at it it makes sense to turn it into what it is: An enum. Turning
it into a class enum has the added benefit we get type safety for mostly
free.
Which incidentally turned up a problem right away:
A call to removeAndCallbackPendingDownloads had status and extstatus
reversed and thus was wrong.

There was an idiom to test with LL_DARWIN if the compiler used is clang. This is not correct, as clang is available on all 3 platforms. The correct way to check for this is via __clang__ (https://clang.llvm.org/docs/LanguageExtensions.html#builtin-macros)

# Conflicts:
#	autobuild.xml
#	indra/newview/llimprocessing.cpp

following promotion of DRTVWR-476

Ever since February 2010, the body of the login coroutine function has been
enclosed in try/catch (...), with an llerrs message to try to crash more
informatively than the runtime's unhandled-exception termination. Over the
years this evolved to LL_ERRS and then to CRASH_ON_UNHANDLED_EXCEPTION.

This persisted despite the August 2016 addition of generic catch clauses in
the LLCoros::toplevel() function to serve the same purpose, and despite the
subsequent introduction of the LLCoros::Stop family of exceptions to
deliberately throw into waiting coroutines on viewer shutdown.

That's exactly what was happening. When the user closed the viewer while
waiting for the response from login.cgi, the waiting operation threw
LLCoros::Stopping, which was caught by that CRASH_ON_UNHANDLED_EXCEPTION,
which crashed the viewer with LL_ERRS rather than propagating up to the
toplevel() and cleanly terminating the coroutine.

Change CRASH_ON_UNHANDLED_EXCEPTION() to LOG_UNHANDLED_EXCEPTION() and
re-throw so toplevel() can handle.

SL-13540: Do not fail if binary bucket is too large, attempt to extract the asset type from the old style bucket. Notification still not shown.

SL-13540: Do not fail if binary bucket is too large, attempt to extract the asset type from the old style bucket. Notification still not shown.

The llappviewerwin32.cpp create_console() function called by
LLAppViewerWin32::initConsole() used to assign *stderr = *(new FILE* value),
and so forth for stdout and stdin. That dubious tactic no longer works with
the new Windows CRT introduced with VS 2015. freopen_s() works much better.

(cherry picked from commit 0b61150e)

There are separate stanzas in llappviewerwin32.cpp's create_console() function
for each of STD_INPUT_HANDLE, STD_OUTPUT_HANDLE and STD_ERROR_HANDLE. SL-13361
wants to add more code to each. Factor out new local set_stream() function and
make create_console() call it three times.

(cherry picked from commit 13b78a0c)