Add d8 to the chromium_builder_asan target.

Add d8 to the chromium_builder_asan target.

BUG=
R=inferno@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=249758

[Martin Barbella, 2014-02-07 01:43:19]

Could you take a look at this when you get a chance?

The previous version was causing errors when trying to build chrome for Chrome
OS from chrome-sdk (on gclient runhooks). Both building and gclient runhooks
from chrome-sdk seem to work properly with these changes.

[inferno, 2014-02-07 02:16:52]

lgtm

[Martin Barbella, 2014-02-07 16:11:51]

The CQ bit was checked by mbarbella@chromium.org

[commit-bot: I haz the power, 2014-02-07 16:12:08]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/mbarbella@chromium.org/157143003/1

[commit-bot: I haz the power, 2014-02-07 20:32:03]

Change committed as 249758

[maheshkk, 2014-02-10 17:59:56]

On 2014/02/07 01:43:19, mbarbella wrote:
> Could you take a look at this when you get a chance?
> 
> The previous version was causing errors when trying to build chrome for Chrome
> OS from chrome-sdk (on gclient runhooks). Both building and gclient runhooks
> from chrome-sdk seem to work properly with these changes.

This change adds d8.gyp to build config but however when I "gclient sync -f" on
my env (ubuntu chrome browser desktop), the command executes file but however
d8.gyp isn't present at src/v8/src/ do you think you need more changes for this
fix?

[Martin Barbella, 2014-02-10 18:09:42]

On 2014/02/10 17:59:56, maheshkk wrote:
> On 2014/02/07 01:43:19, mbarbella wrote:
> > Could you take a look at this when you get a chance?
> > 
> > The previous version was causing errors when trying to build chrome for
Chrome
> > OS from chrome-sdk (on gclient runhooks). Both building and gclient runhooks
> > from chrome-sdk seem to work properly with these changes.
> 
> This change adds d8.gyp to build config but however when I "gclient sync -f"
on
> my env (ubuntu chrome browser desktop), the command executes file but however
> d8.gyp isn't present at src/v8/src/ do you think you need more changes for
this
> fix?

I'm taking a look now to see if I can find a way to reproduce this error, but I
was just able to do a gclient sync -f from a similar setup. Is there anything
unusual about your configuration that might be relevant for reproducing this?

Abhishek, do you think we should revert this again to be safe and set an
additional define similar to how the IPC fuzzer target is set? I'm not sure what
might be causing this in this case.

[aarya, 2014-02-10 18:13:04]

On 2014/02/10 18:09:42, mbarbella wrote:
> On 2014/02/10 17:59:56, maheshkk wrote:
> > On 2014/02/07 01:43:19, mbarbella wrote:
> > > Could you take a look at this when you get a chance?
> > > 
> > > The previous version was causing errors when trying to build chrome for
> Chrome
> > > OS from chrome-sdk (on gclient runhooks). Both building and gclient
runhooks
> > > from chrome-sdk seem to work properly with these changes.
> > 
> > This change adds d8.gyp to build config but however when I "gclient sync -f"
> on
> > my env (ubuntu chrome browser desktop), the command executes file but
however
> > d8.gyp isn't present at src/v8/src/ do you think you need more changes for
> this
> > fix?
> 
> I'm taking a look now to see if I can find a way to reproduce this error, but
I
> was just able to do a gclient sync -f from a similar setup. Is there anything
> unusual about your configuration that might be relevant for reproducing this?
> 
> Abhishek, do you think we should revert this again to be safe and set an
> additional define similar to how the IPC fuzzer target is set? I'm not sure
what
> might be causing this in this case.

I don't think we should revert for these one-off cases. We should find what
config they are using and fix it quick.

[maheshkk, 2014-02-10 19:26:19]

On 2014/02/10 18:09:42, mbarbella wrote:
> On 2014/02/10 17:59:56, maheshkk wrote:
> > On 2014/02/07 01:43:19, mbarbella wrote:
> > > Could you take a look at this when you get a chance?
> > > 
> > > The previous version was causing errors when trying to build chrome for
> Chrome
> > > OS from chrome-sdk (on gclient runhooks). Both building and gclient
runhooks
> > > from chrome-sdk seem to work properly with these changes.
> > 
> > This change adds d8.gyp to build config but however when I "gclient sync -f"
> on
> > my env (ubuntu chrome browser desktop), the command executes file but
however
> > d8.gyp isn't present at src/v8/src/ do you think you need more changes for
> this
> > fix?
> 
> I'm taking a look now to see if I can find a way to reproduce this error, but
I
> was just able to do a gclient sync -f from a similar setup. Is there anything
> unusual about your configuration that might be relevant for reproducing this?
> 
> Abhishek, do you think we should revert this again to be safe and set an
> additional define similar to how the IPC fuzzer target is set? I'm not sure
what
> might be causing this in this case.

Just pulled latest blink codebase. My .gclient_entries says v8/src is at
bb8234d89692f5088ce3fe3ff5a8e8da2f038cfe.

when I do build/gyp_chromium, I get 

gyp: Dependency '/home/blahblah/work/trunk/src/v8/src/d8.gyp:d8#target' not
found while trying to load target
/home/blahblah/work/trunk/src/build/all.gyp:chromium_builder_asan#target


src/v8/src/d8.gyp doesn't exist atleast in the gclient entry I have. If v8/src
is updated to add d8.gyp, how do I update my .gclient_entries to point to it?

[Martin Barbella, 2014-02-10 20:45:05]

Any idea what might be going on here, dpranke? Abhishek mentioned that you might
have some ideas.

[Dirk Pranke, 2014-02-10 20:56:08]

On 2014/02/10 20:45:05, mbarbella wrote:
> Any idea what might be going on here, dpranke? Abhishek mentioned that you
might
> have some ideas.

It seems like there's something wrong w/ maheshkk's checkout. d8.gyp should be
found in that version of v8:

https://chromium.googlesource.com/external/v8.git/+/bb8234d89692f5088ce3fe3ff...

(and I see it locally).

maheshkk: if you do a 'git status' in src/v8/src, does anything look off?

[maheshkk, 2014-02-10 22:29:30]

On 2014/02/10 20:56:08, Dirk Pranke wrote:
> On 2014/02/10 20:45:05, mbarbella wrote:
> > Any idea what might be going on here, dpranke? Abhishek mentioned that you
> might
> > have some ideas.
> 
> It seems like there's something wrong w/ maheshkk's checkout. d8.gyp should be
> found in that version of v8:
> 
>
https://chromium.googlesource.com/external/v8.git/+/bb8234d89692f5088ce3fe3ff...
> 
> (and I see it locally).
> 
> maheshkk: if you do a 'git status' in src/v8/src, does anything look off?

Dirk, I have no idea what was wrong with my env. I pulled the latest again and
there I could find d8.gyp file all of sudden. I did git pull many times this
morning but did not see that file. Anyway, false alarm mbarbella. Please ignore
this error

[Martin Barbella, 2014-02-10 22:59:56]

On 2014/02/10 22:29:30, maheshkk wrote:
> On 2014/02/10 20:56:08, Dirk Pranke wrote:
> > On 2014/02/10 20:45:05, mbarbella wrote:
> > > Any idea what might be going on here, dpranke? Abhishek mentioned that you
> > might
> > > have some ideas.
> > 
> > It seems like there's something wrong w/ maheshkk's checkout. d8.gyp should
be
> > found in that version of v8:
> > 
> >
>
https://chromium.googlesource.com/external/v8.git/+/bb8234d89692f5088ce3fe3ff...
> > 
> > (and I see it locally).
> > 
> > maheshkk: if you do a 'git status' in src/v8/src, does anything look off?
> 
> Dirk, I have no idea what was wrong with my env. I pulled the latest again and
> there I could find d8.gyp file all of sudden. I did git pull many times this
> morning but did not see that file. Anyway, false alarm mbarbella. Please
ignore
> this error

No problem, I'm glad it worked out. Thanks for taking a look, Dirk.

[Raphael Kubo da Costa (rakuco), 2014-02-11 14:43:46]

I'm getting the same issue here.

Looking at the dependency_nodes when the error is shown, I do have d8.gyp and a
d8 target, but it's d8.gyp:d8#host, not d8.gyp:d8#target. This is a regular
Linux build, with no cross-compilation involved. Should I file a bug report?

[petarj, 2014-02-11 15:37:02]

On 2014/02/11 14:43:46, Raphael Kubo da Costa (rakuco) wrote:
> I'm getting the same issue here.
> 
> Looking at the dependency_nodes when the error is shown, I do have d8.gyp and
a
> d8 target, but it's d8.gyp:d8#host, not d8.gyp:d8#target. This is a regular
> Linux build, with no cross-compilation involved. Should I file a bug report?

This change broke the build for MIPS.

Check:
http://www.rt-rk.com/mips-buildbot/builders/build_and_test_chromium_trunk_for...

Should this maybe be changed to:
-                '../v8/src/d8.gyp:d8',
+                '../v8/src/d8.gyp:d8#host',
?

[Martin Barbella, 2014-02-11 16:26:17]

On 2014/02/11 15:37:02, petarj wrote:
> On 2014/02/11 14:43:46, Raphael Kubo da Costa (rakuco) wrote:
> > I'm getting the same issue here.
> > 
> > Looking at the dependency_nodes when the error is shown, I do have d8.gyp
and
> a
> > d8 target, but it's d8.gyp:d8#host, not d8.gyp:d8#target. This is a regular
> > Linux build, with no cross-compilation involved. Should I file a bug report?
> 
> This change broke the build for MIPS.
> 
> Check:
>
http://www.rt-rk.com/mips-buildbot/builders/build_and_test_chromium_trunk_for...
> 
> Should this maybe be changed to:
> -                '../v8/src/d8.gyp:d8',
> +                '../v8/src/d8.gyp:d8#host',
> ?

Yes, good catch. It does seem like this should be ../v8/src/d8.gyp:d8#host. I'm
putting a CL together to fix this now.

[Martin Barbella, 2014-02-11 16:41:58]

On 2014/02/11 16:26:17, mbarbella wrote:
> On 2014/02/11 15:37:02, petarj wrote:
> > On 2014/02/11 14:43:46, Raphael Kubo da Costa (rakuco) wrote:
> > > I'm getting the same issue here.
> > > 
> > > Looking at the dependency_nodes when the error is shown, I do have d8.gyp
> and
> > a
> > > d8 target, but it's d8.gyp:d8#host, not d8.gyp:d8#target. This is a
regular
> > > Linux build, with no cross-compilation involved. Should I file a bug
report?
> > 
> > This change broke the build for MIPS.
> > 
> > Check:
> >
>
http://www.rt-rk.com/mips-buildbot/builders/build_and_test_chromium_trunk_for...
> > 
> > Should this maybe be changed to:
> > -                '../v8/src/d8.gyp:d8',
> > +                '../v8/src/d8.gyp:d8#host',
> > ?
> 
> Yes, good catch. It does seem like this should be ../v8/src/d8.gyp:d8#host.
I'm
> putting a CL together to fix this now.

Working on a fix at https://codereview.chromium.org/156283004/