handler/mac: Respect the uploads-enabled user preference and rate-limit upload attempts

handler/mac: Respect the uploads-enabled user preference and rate-limit
upload attempts to no more than 1 per hour.

The rate limiting is simplistic but duplicates the existing Breakpad
client’s behavior, and is suitable for the time being.

TEST=run_with_crashpad --handler crashpad_handler \
         -a --database=/tmp/crashpad_db \
         -a --url=https://clients2.google.com/cr/staging_report \
         -a --annotation=prod=crashpad \
         -a --annotation=ver=0.7.0 \
         crashy_program

R=rsesek@chromium.org

Committed: https://chromium.googlesource.com/crashpad/crashpad/+/42db629ee334fa8cd63f085e4adaab5f4fe938eb

[Mark Mentovai, 2015-03-10 19:00:18]

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

[Mark Mentovai, 2015-03-10 19:00:18]

The pieces keep getting smaller and smaller…

[Robert Sesek, 2015-03-10 19:20:54]

https://codereview.chromium.org/992303002/diff/1/handler/mac/crash_report_upl...
File handler/mac/crash_report_upload_thread.cc (right):

https://codereview.chromium.org/992303002/diff/1/handler/mac/crash_report_upl...
handler/mac/crash_report_upload_thread.cc:245: // it “happened” at least one day
in the future, assume that the last
This makes me think that saturating the last report time is maybe not the best
option and that reseting to 0 is better. Would you still handle this case that
way if so? (Maybe yes because of bad clocks).

[Mark Mentovai, 2015-03-10 19:28:24]

https://codereview.chromium.org/992303002/diff/1/handler/mac/crash_report_upl...
File handler/mac/crash_report_upload_thread.cc (right):

https://codereview.chromium.org/992303002/diff/1/handler/mac/crash_report_upl...
handler/mac/crash_report_upload_thread.cc:245: // it “happened” at least one day
in the future, assume that the last
Robert Sesek wrote:
> This makes me think that saturating the last report time is maybe not the best
> option and that reseting to 0 is better. Would you still handle this case that
> way if so? (Maybe yes because of bad clocks).

Yes, exactly, this is primarily for bad clocks. Saturation’s just a very tiny
part of it by comparison.

[Robert Sesek, 2015-03-10 19:30:21]

LGTM

[Mark Mentovai, 2015-03-10 19:31:30]

Committed patchset #1 (id:1) manually as
42db629ee334fa8cd63f085e4adaab5f4fe938eb (presubmit successful).