When marking dependent code for deoptimization print the group that is being deoptimized.

When marking dependent code for deoptimization print the group that is being deoptimized.

Otherwise it is impossible to figure out from the --trace-deoptimization output what is going on.

R=yangguo@chromium.org

Committed: https://code.google.com/p/v8/source/detail?r=23191

[Vyacheslav Egorov (Google), 2014-08-13 12:14:06]

Please take a look. 

This is a usability improvement for --trace-deopt.

[Yang, 2014-08-14 07:05:08]

On 2014/08/13 12:14:06, Vyacheslav Egorov (Google wrote:
> Please take a look. 
> 
> This is a usability improvement for --trace-deopt.

lgtm.

[Vyacheslav Egorov (Google), 2014-08-14 11:20:40]

Thanks! 

I will sit on this until Chrome is branched, as the tree is throttled right now.

[Vyacheslav Egorov (Google), 2014-08-19 10:56:58]

Committed patchset #2 manually as 23191 (presubmit successful).

[Toon Verwaest, 2014-08-19 11:08:05]

https://codereview.chromium.org/467183002/diff/20001/src/objects.cc
File src/objects.cc (right):

https://codereview.chromium.org/467183002/diff/20001/src/objects.cc#newcode11878
src/objects.cc:11878: default:
Any reason why you added a default case? That kinda ensures we'll definitely
forget to update this list when we add a new case.

[Vyacheslav Egorov (Google), 2014-08-19 11:14:41]

https://codereview.chromium.org/467183002/diff/20001/src/objects.cc
File src/objects.cc (right):

https://codereview.chromium.org/467183002/diff/20001/src/objects.cc#newcode11878
src/objects.cc:11878: default:
On 2014/08/19 11:08:04, Toon Verwaest wrote:
> Any reason why you added a default case? That kinda ensures we'll definitely
> forget to update this list when we add a new case.

Well, main reason is that enumeration has parasitic kGroupCount right now.

Right way would be to move kGroupCount out and then I could kill default: case
here, because I do agree - it's nice to not have it, and get a warning for
uncovered cases.

If you want I can submit a follow up CL for this.

[Yang, 2014-08-19 11:19:06]

On 2014/08/19 11:14:41, Vyacheslav Egorov (Google wrote:
> https://codereview.chromium.org/467183002/diff/20001/src/objects.cc
> File src/objects.cc (right):
> 
>
https://codereview.chromium.org/467183002/diff/20001/src/objects.cc#newcode11878
> src/objects.cc:11878: default:
> On 2014/08/19 11:08:04, Toon Verwaest wrote:
> > Any reason why you added a default case? That kinda ensures we'll definitely
> > forget to update this list when we add a new case.
> 
> Well, main reason is that enumeration has parasitic kGroupCount right now.
> 
> Right way would be to move kGroupCount out and then I could kill default: case
> here, because I do agree - it's nice to not have it, and get a warning for
> uncovered cases.
> 
> If you want I can submit a follow up CL for this.

or we could turn that default case into an unreachable case for kGroupCount.

[Sven Panne, 2014-08-25 11:49:00]

[ Just wading through my email backlog after vacation... ]

On 2014/08/19 11:19:06, Yang wrote:
> or we could turn that default case into an unreachable case for kGroupCount.

This would have been the right way IMHO. "default" is a maintenance nightmare
most of the time.

Regarding the count within an enum: I consider this an anti-pattern and we
should remove those enum constants, there has been a discussion about this on
chromium-dev some time ago, BTW. Iterating over enum values or even doing
arithmetic with them is often a bit obscure.

[Yang, 2014-08-25 11:57:13]

On 2014/08/25 11:49:00, Sven Panne wrote:
> [ Just wading through my email backlog after vacation... ]
> 
> On 2014/08/19 11:19:06, Yang wrote:
> > or we could turn that default case into an unreachable case for kGroupCount.
> 
> This would have been the right way IMHO. "default" is a maintenance nightmare
> most of the time.
> 
> Regarding the count within an enum: I consider this an anti-pattern and we
> should remove those enum constants, there has been a discussion about this on
> chromium-dev some time ago, BTW. Iterating over enum values or even doing
> arithmetic with them is often a bit obscure.

https://codereview.chromium.org/502933003/

[Vyacheslav Egorov (Google), 2014-08-25 12:11:31]

How about this one instead 

https://codereview.chromium.org/488993004

?