Generate debug information by linker with win_z7=1 even if fastbuild=1

Generate debug information by linker with win_z7=1 even if fastbuild=1

without this CL
 fastbuild=1 => no debug information
 win_z7=1 => debug information available (with /Z7)
 fastbuild=1 win_z7 => *.obj has debug info, but linker doesn't generate debug info in *.exe. it would be useless, and just makes build slower

This CL makes win_z7 wins against fastbuild.

R=scottmg,arthurhsu
BUG=none

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=152953

[scottmg, 2012-08-22 05:17:58]

https://chromiumcodereview.appspot.com/10866016/diff/1/build/common.gypi
File build/common.gypi (right):

https://chromiumcodereview.appspot.com/10866016/diff/1/build/common.gypi#newc...
build/common.gypi:1558: 'GenerateDebugInformation': 'true',
Does that merge correctly with line 1452? I'm not sure what happens when there's
conflicting settings, it might depend on file or hashing order.

If it works as desired, lgtm.

[ukai, 2012-08-22 06:10:13]

https://chromiumcodereview.appspot.com/10866016/diff/1/build/common.gypi
File build/common.gypi (right):

https://chromiumcodereview.appspot.com/10866016/diff/1/build/common.gypi#newc...
build/common.gypi:1558: 'GenerateDebugInformation': 'true',
On 2012/08/22 05:17:58, scottmg wrote:
> Does that merge correctly with line 1452? I'm not sure what happens when
there's
> conflicting settings, it might depend on file or hashing order.

yes.
without this CL
 fastbuild=1 => no debug information
 win_z7=1 => debug information available (with /Z7)
 fastbuild=1 win_z7 => *.obj has debug info, but linker doesn't generate debug
info. it would be useless, and just make build slower

this CL makes win_z7 wins against fastbuild.

> 
> If it works as desired, lgtm.

[scottmg, 2012-08-22 15:35:19]

https://chromiumcodereview.appspot.com/10866016/diff/1/build/common.gypi
File build/common.gypi (right):

https://chromiumcodereview.appspot.com/10866016/diff/1/build/common.gypi#newc...
build/common.gypi:1558: 'GenerateDebugInformation': 'true',
On 2012/08/22 06:10:13, ukai wrote:
> On 2012/08/22 05:17:58, scottmg wrote:
> > Does that merge correctly with line 1452? I'm not sure what happens when
> there's
> > conflicting settings, it might depend on file or hashing order.
> 
> yes.
> without this CL
>  fastbuild=1 => no debug information
>  win_z7=1 => debug information available (with /Z7)
>  fastbuild=1 win_z7 => *.obj has debug info, but linker doesn't generate debug
> info. it would be useless, and just make build slower
> 
> this CL makes win_z7 wins against fastbuild.

I understand the goal, but line 1452 sets GenerateDebugInfo: false, and line
1558 sets GenerateDebugInfo: true when (fastbuild!=0 and win_z7!=0 and
OS=="win").

I'm not sure whether gyp has defined behaviour for repeated settings. If the
setting you're adding here always "wins" because it's lower in the file, then
please add a comment to that effect.

(Personally I think it'd be better to just error out if both fastbuild=1 and
win_z7=1, and get the user to be explicit and choose one, but I don't think gyp
can do that easily)


> 
> > 
> > If it works as desired, lgtm.
>

[ukai, 2012-08-23 02:19:19]

https://chromiumcodereview.appspot.com/10866016/diff/1/build/common.gypi
File build/common.gypi (right):

https://chromiumcodereview.appspot.com/10866016/diff/1/build/common.gypi#newc...
build/common.gypi:1558: 'GenerateDebugInformation': 'true',
On 2012/08/22 15:35:19, scottmg wrote:
> On 2012/08/22 06:10:13, ukai wrote:
> > On 2012/08/22 05:17:58, scottmg wrote:
> > > Does that merge correctly with line 1452? I'm not sure what happens when
> > there's
> > > conflicting settings, it might depend on file or hashing order.
> > 
> > yes.
> > without this CL
> >  fastbuild=1 => no debug information
> >  win_z7=1 => debug information available (with /Z7)
> >  fastbuild=1 win_z7 => *.obj has debug info, but linker doesn't generate
debug
> > info. it would be useless, and just make build slower
> > 
> > this CL makes win_z7 wins against fastbuild.
> 
> I understand the goal, but line 1452 sets GenerateDebugInfo: false, and line
> 1558 sets GenerateDebugInfo: true when (fastbuild!=0 and win_z7!=0 and
> OS=="win").
> 
> I'm not sure whether gyp has defined behaviour for repeated settings. If the
> setting you're adding here always "wins" because it's lower in the file, then
> please add a comment to that effect.

I believe it is merged
http://code.google.com/p/gyp/wiki/InputFormatReference#Merging
> 
> (Personally I think it'd be better to just error out if both fastbuild=1 and
> win_z7=1, and get the user to be explicit and choose one, but I don't think
gyp
> can do that easily)
> 
> 
> > 
> > > 
> > > If it works as desired, lgtm.
> > 
>

[commit-bot: I haz the power, 2012-08-23 02:21:09]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/ukai@chromium.org/10866016/6001

[commit-bot: I haz the power, 2012-08-23 05:49:10]

Change committed as 152953