Revert "Enable out-of-line constant pool for Arm."

Revert "Enable out-of-line constant pool for Arm."

This reverts commit r23323 for letting arm timeout.

Bisect build without the CL:
http://build.chromium.org/p/client.v8/builders/V8%20Arm/builds/94

Build with the CL:
http://build.chromium.org/p/client.v8/builders/V8%20Arm/builds/95

TBR=rmcilroy@chromium.org, ulan@chromium.org, yangguo@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=23325

[Michael Achenbach, 2014-08-23 10:27:34]

FYI

[Michael Achenbach, 2014-08-23 18:38:48]

Committed patchset #1 manually as 23325 (tree was closed).

[rmcilroy, 2014-08-26 10:49:38]

On 2014/08/23 18:38:48, Michael Achenbach wrote:
> Committed patchset #1 manually as 23325 (tree was closed).

Thanks for reverting (apologies for not spotting this, the bot took a while to
go red and it was after I'd gone home).

I've just tried running the command on the bot which timedout and it works fine
on my local machine.  Is there anything special which the Bot is doing which I'm
missing?

[Michael Achenbach, 2014-08-27 17:54:51]

On 2014/08/26 10:49:38, rmcilroy wrote:
> On 2014/08/23 18:38:48, Michael Achenbach wrote:
> > Committed patchset #1 manually as 23325 (tree was closed).
> 
> Thanks for reverting (apologies for not spotting this, the bot took a while to
> go red and it was after I'd gone home).
> 
> I've just tried running the command on the bot which timedout and it works
fine
> on my local machine.  Is there anything special which the Bot is doing which
I'm
> missing?

You need arm hardware. The bot is a lucas board with an arm cpu. You can also
download the executables the bot used from GS to your device and try to repro
with that first.

[ulan, 2014-08-28 08:30:11]

On 2014/08/27 17:54:51, Michael Achenbach wrote:
> On 2014/08/26 10:49:38, rmcilroy wrote:
> > On 2014/08/23 18:38:48, Michael Achenbach wrote:
> > > Committed patchset #1 manually as 23325 (tree was closed).
> > 
> > Thanks for reverting (apologies for not spotting this, the bot took a while
to
> > go red and it was after I'd gone home).
> > 
> > I've just tried running the command on the bot which timedout and it works
> fine
> > on my local machine.  Is there anything special which the Bot is doing which
> I'm
> > missing?
> 
> You need arm hardware. The bot is a lucas board with an arm cpu. You can also
> download the executables the bot used from GS to your device and try to repro
> with that first.

If you don't find such device, I can take a look at it in the second half of the
next week (working on other urgent stuff until then).

[rmcilroy, 2014-08-28 09:35:21]

On 2014/08/28 08:30:11, ulan wrote:
> On 2014/08/27 17:54:51, Michael Achenbach wrote:
> > On 2014/08/26 10:49:38, rmcilroy wrote:
> > > On 2014/08/23 18:38:48, Michael Achenbach wrote:
> > > > Committed patchset #1 manually as 23325 (tree was closed).
> > > 
> > > Thanks for reverting (apologies for not spotting this, the bot took a
while
> to
> > > go red and it was after I'd gone home).
> > > 
> > > I've just tried running the command on the bot which timedout and it works
> > fine
> > > on my local machine.  Is there anything special which the Bot is doing
which
> > I'm
> > > missing?
> > 
> > You need arm hardware. The bot is a lucas board with an arm cpu. You can
also
> > download the executables the bot used from GS to your device and try to
repro
> > with that first.
> 
> If you don't find such device, I can take a look at it in the second half of
the
> next week (working on other urgent stuff until then).

Thanks for the info, I didn't realise the bot used real Arm hardware.  I don't
think I have access to that board and I'm going on vacation from tomorrow so I
probably won't get a chance to get back to this before I get back.  Ulan, if you
manage to have a look before I get back that would be awesome, but if not no
worries, I'll try and sort it out once I get back.

[ulan, 2014-09-04 13:52:58]

On 2014/08/28 09:35:21, rmcilroy_OOO_until_Sept_17 wrote:
> On 2014/08/28 08:30:11, ulan wrote:
> > On 2014/08/27 17:54:51, Michael Achenbach wrote:
> > > On 2014/08/26 10:49:38, rmcilroy wrote:
> > > > On 2014/08/23 18:38:48, Michael Achenbach wrote:
> > > > > Committed patchset #1 manually as 23325 (tree was closed).
> > > > 
> > > > Thanks for reverting (apologies for not spotting this, the bot took a
> while
> > to
> > > > go red and it was after I'd gone home).
> > > > 
> > > > I've just tried running the command on the bot which timedout and it
works
> > > fine
> > > > on my local machine.  Is there anything special which the Bot is doing
> which
> > > I'm
> > > > missing?
> > > 
> > > You need arm hardware. The bot is a lucas board with an arm cpu. You can
> also
> > > download the executables the bot used from GS to your device and try to
> repro
> > > with that first.
> > 
> > If you don't find such device, I can take a look at it in the second half of
> the
> > next week (working on other urgent stuff until then).
> 
> Thanks for the info, I didn't realise the bot used real Arm hardware.  I don't
> think I have access to that board and I'm going on vacation from tomorrow so I
> probably won't get a chance to get back to this before I get back.  Ulan, if
you
> manage to have a look before I get back that would be awesome, but if not no
> worries, I'll try and sort it out once I get back.

I took a look and uploaded a fix for test failures seen in the logs for r23323:
https://codereview.chromium.org/541733002/

I hope that it fill also fix bot timeout issue. If not I'll look at it again.