- Save the original conn->data before it's changed to the specified
  data transfer for the connection check and then restore it afterwards.

This is a follow-up to 38d8e1bd 2019-02-11.

History:

It was discovered a month ago that before checking whether to extract a
dead connection that that connection should be associated with a "live"
transfer for the check (ie original conn->data ignored and set to the
passed in data). A fix was landed in 54b201b4 which did that and also
cleared conn->data after the check. The original conn->data was not
restored, so presumably it was thought that a valid conn->data was no
longer needed.

Several days later it was discovered that a valid conn->data was needed
after the check and follow-up fix was landed in bbae24c3 which partially
reverted the original fix and attempted to limit the scope of when
conn->data was changed to only when pruning dead connections. In that
case conn->data was not cleared and the original conn->data not
restored.

A month later it was discovered that the original fix was somewhat
correct; a "live" transfer is needed for the check in all cases
because original conn->data could be null which could cause a bad deref
at arbitrary points in the check. A fix was landed in 38d8e1bd

 which
expanded the scope to all cases. conn->data was not cleared and the
original conn->data not restored.

A day later it was discovered that not restoring the original conn->data
may lead to busy loops in applications that use the event interface, and
given this observation it's a pretty safe assumption that there is some
code path that still needs the original conn->data. This commit is the
follow-up fix for that, it restores the original conn->data after the
connection check.

Assisted-by:  <tholin@users.noreply.github.com>
Reported-by:  <tholin@users.noreply.github.com>

Fixes https://github.com/curl/curl/issues/3542
Closes #3559

Used by debug builds with NSS.

Reverted from 05b100ae

On non-ascii platforms, the chunked hex header was measured for char code
conversion length, even for chunked trailers that do not have an hex header.
In addition, the efective length is already known: use it.
Since the hex length can be zero, only convert if needed.

Reported by valgrind.

Closes #2367

Detected by scan-build (false positive).

Closes #3569

Closes #2905

Closes #2719

Heimdal includes on FreeBSD spewed out lots of them. Less so now.

Closes #3566

Closes #3523

Closes #3109

Rephrase to satisfy scan-build.

This test doesn't link without FTP support after
fc7ab483, which made Curl_fnmatch
unavailable without FTP support.

Closes https://github.com/curl/curl/pull/3565

Nobody works on this now.

Closes #3378

Mostly a false positive, but this makes the code easier to read anyway.

Detected by scan-build.

Closes #3563

Detected by scan-build.

Follow-up to 84c10dc1 from earlier today which wrapped a lot of the noisy
debug output in DEBUGF but omitted a few lines.

Ref: https://github.com/curl/curl/commit/84c10dc#r32292900

- Pass an empty string to CURLOPT_ACCEPT_ENCODING to use the default
  supported encodings.

Prior to this change the specific encodings of gzip and deflate were set
but there's no guarantee they'd be supported by the user's libcurl.

... instead of allocating it separately and point to it. It is
fixed-size and always used for each part.

Closes #3561

Convert numerous infof() calls into debug-build only messages since they
are annoyingly verbose for regular applications. Removed a few.

Bug: https://curl.haxx.se/mail/lib-2019-02/0027.html
Reported-by: Volker Schmid
Closes #3552

Closes #3562

Closes #3557

Follow-up to cac0e4a6

Detected by scan-build
Closes #3560

... since that data won't be used in the request anyway.

Fixes #3548
Reported-by: Renaud Allard
Close #3549

Reported-by: James Brown
Bug: https://curl.haxx.se/mail/archive-2019-02/0013.html
Closes #3558

Reported-by: MAntoniak on github
Fixes #3553
Closes #3556

Fix two occurrences of s/convers/converts/ spotted while reading code.

Closes #3551

The code is more or less copied from the stdout comparison code, maybe
some better reuse is possible.

test 1457 is adjusted to make the output actually match (by using --silent)
test 506 used <stderr> without actually needing it, so that <stderr> block is removed

Closes #3536

Option -F generates an intermediate representation of the mime structure
that is used later to create the libcurl mime structure and generate
the --libcurl statements.

Reported-by: Daniel Stenberg
Fixes #3532
Closes #3546

and bump the version in progress to 7.64.1. If we merge any "change"
before the cut-off date, we update again.

Commit 3f16990e followed-up a bug in b49652ac

 but was
inadvertently introducing a new bug in the ternary expression.

Close #3555
Reviewed-by: Daniel Stenberg <daniel@haxx.se>

There is no benefit to holding the data sharelock when freeing the
addrinfo in case it fails, so ensure releaseing it as soon as we can
rather than holding on to it. This also aligns the code with other
consumers of sharelocks.

Closes #3516
Reviewed-by: Daniel Stenberg <daniel@haxx.se>

On FreeBSD, return non-zero on error otherwise zero.

Reported-by: Marcel Raad

... and added braces to two function calls which fixes warnings if they
are replace by empty macros at build-time.

Closes #3550

The http2 code for connection checking needs a transfer to use. Make
sure a working one is set before handler->connection_check() is called.

Reported-by: jnbr on github
Fixes #3541
Closes #3547

Closes #3544