HTTPTransportMac: CFStream Read() must always set at_eof

HTTPTransportMac: CFStream Read() must always set at_eof

CFStream’s CFReadStreamGetBuffer() calls the Read() callback without
initializing at_eof. The callback function is responsible for setting it
on any successful read operation. See 10.10.2 CF-1152.14/CFStream.c.

By chance, at_eof seems to always have an initial value of false on
x86_64, but true on 32-bit x86. Crashpad’s Read() callback assumed that
the initial value was always false. The discrepancy caused truncation
and possibly hangs when a 32-bit process attempted to upload a request
body larger than 32kB, the buffer size used by NSMutableURLRequest or
something between it and CFReadStream.

A new test with more than 32kB of data is added.

As discussed in:
https://groups.google.com/a/chromium.org/d/topic/crashpad-dev/Vz--qMZJRPU

TEST=crashpad_util_test HTTPTransport.Upload33k
BUG=
R=rsesek@chromium.org

Committed: https://chromium.googlesource.com/crashpad/crashpad/+/f496130fd594a0467d7dacdf923a9677cafe280b

[Mark Mentovai, 2015-08-18 19:19:48]

mark@chromium.org changed reviewers:
+ rsesek@chromium.org

[Robert Sesek, 2015-08-18 19:35:07]

LGTM

https://codereview.chromium.org/1304433004/diff/1/util/net/http_transport_tes...
File util/net/http_transport_test.cc (right):

https://codereview.chromium.org/1304433004/diff/1/util/net/http_transport_tes...
util/net/http_transport_test.cc:292: headers[kContentLength] =
base::StringPrintf("%zu", request_string.size());
PRIuS instead of %zu.

[Mark Mentovai, 2015-08-18 19:42:38]

Committed patchset #2 (id:20001) manually as
f496130fd594a0467d7dacdf923a9677cafe280b (presubmit successful).