Fixing loop to count up from the bottom instead of the top.

Fixing loop to count up from the bottom instead of the top.

The loop that adds prefixes from 1 billion to 3 billion has been broken
since it was modified in Feb 27, 2014, change 2b59b0a. It didn't survive
the change from signed to unsigned integers meaning that this test
function has effectively been disabled.

This bug was found by VC++'s /analyze which said:
src\chrome\browser\safe_browsing\prefix_set_unittest.cc(280) :
warning C6294: Ill-defined for-loop:
initial condition does not satisfy test.  Loop body not executed.

BUG=427616
Committed: https://crrev.com/0ebc0a391d9aca92e45bd3c674bd2c140e4fac0e
Cr-Commit-Position: refs/heads/master@{#307071}

[brucedawson, 2014-12-04 22:57:42]

brucedawson@chromium.org changed reviewers:
+ tsepez@chromium.org

[Scott Hess - ex-Googler, 2014-12-05 03:39:17]

shess@chromium.org changed reviewers:
+ shess@chromium.org

[Scott Hess - ex-Googler, 2014-12-05 03:39:18]

Man, that's sad, I thought I went over those cases like 30 times.

LGTM.

[brucedawson, 2014-12-05 18:35:54]

The CQ bit was checked by brucedawson@chromium.org

[brucedawson, 2014-12-05 18:36:07]

The CQ bit was unchecked by brucedawson@chromium.org

[brucedawson, 2014-12-05 18:37:46]

On 2014/12/05 03:39:18, Scott Hess - On Leave wrote:
> Man, that's sad, I thought I went over those cases like 30 times.
> 
> LGTM.

I thought you were on leave? That's why I just put you on the CC line.

It is depressing sometimes how easy it is to miss errors like this. All
syntactically correct errors that can be made will be made -- that must be
somebody's law.

[Scott Hess - ex-Googler, 2014-12-05 19:05:54]

On 2014/12/05 18:37:46, brucedawson wrote:
> On 2014/12/05 03:39:18, Scott Hess - On Leave wrote:
> > Man, that's sad, I thought I went over those cases like 30 times.
> > 
> > LGTM.
> 
> I thought you were on leave? That's why I just put you on the CC line.

Back in a week, so I'm re-engaging, cleaning up inbox, etc.  Or maybe it's rainy
outside and I'm bored.

> It is depressing sometimes how easy it is to miss errors like this. All
> syntactically correct errors that can be made will be made -- that must be
> somebody's law.

Yeah, but part of why I waited so long to flip this bit (ha!) is because I was
super paranoid about having unexpected breaks, so I _really_ tried to pay
attention :-).  Glad it was a test break rather than a functionality break!

[brucedawson, 2014-12-05 19:48:26]

The CQ bit was checked by brucedawson@chromium.org

[commit-bot: I haz the power, 2014-12-05 19:49:18]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/779163002/1

[commit-bot: I haz the power, 2014-12-05 20:28:48]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2014-12-05 20:29:37]

Patchset 1 (id:??) landed as
https://crrev.com/0ebc0a391d9aca92e45bd3c674bd2c140e4fac0e
Cr-Commit-Position: refs/heads/master@{#307071}