cl : Command line warning D9025: overriding '/Oy' with '/Oy-'

[Gabriel Charette]

I'm seeing

cl : Command line warning D9025: overriding '/Oy' with '/Oy-'

all over the build off an LKGR sync this morning with VS2010.

According to msdn this happens because we specify /Oy and then /Oy- which overwrites the former.

Seems to be a link warning for every project now, anyone else seeing this?

Gab

[Thiago Farina]

On May 24, 12:02 pm, Gabriel Charette <g...@chromium.org> wrote:
> I'm seeing
> cl : Command line warning D9025: overriding '/Oy' with '/Oy-'
> all over the build off an LKGR sync this morning with VS2010.
>

> According to msdn<http://msdn.microsoft.com/en-us/library/8k3f51f1(v=vs.80).aspx>

> this
> happens because we specify /Oy and then /Oy- which overwrites the former.
>
> Seems to be a link warning for every project now, anyone else seeing this?
>

Yes, I saw this problem too in my tries.

http://build.chromium.org/p/tryserver.chromium/builders/win_rel/builds/32203/steps/compile/logs/stdio

> Gab

[Alex Pakhunov]

Well, this piece from src/build/common.gpyi seems to be related:

              ['win_debug_OmitFramePointers==0', {
                'OmitFramePointers': 'false',
                # The above is not sufficient (http://crbug.com/106711): it
                # simply eliminates an explicit "/Oy", but both /O2 and /Ox
                # perform FPO regardless, so we must explicitly disable.
                # We still want the false setting above to avoid having
                # "/Oy /Oy-" and warnings about overriding.
                'AdditionalOptions': ['/Oy-'],
              }],

Alex.

> Gab

--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
   http://groups.google.com/a/chromium.org/group/chromium-dev

--

Alex.

[Scott Graham]

I can't reproduce this locally. Gab, did you see it locally, or on a try bot?

The gyp settings and gyp don't seem to have changed recently.

On Fri, May 25, 2012 at 10:02 AM, Thiago Farina <tfa...@chromium.org> wrote:
>
>

[Scott Graham]

I take that back, I can reproduce it locally.

I think it started here: https://chromiumcodereview.appspot.com/10423010

Looking at common.gypi, Release seems to default FPO to on via
win_release_OmitFramePointers=1, but then internal/release_impl.gypi
wants to turn it off again.

Do we actually want FPO on or off in Release? It looks like most
recently we felt off was correct, so win_release_OmitFramePointers
should default to 0.

[Scott Graham]

On Fri, May 25, 2012 at 11:28 AM, Scott Graham <sco...@chromium.org> wrote:
> I take that back, I can reproduce it locally.
>
> I think it started here: https://chromiumcodereview.appspot.com/10423010
>
> Looking at common.gypi, Release seems to default FPO to on via
> win_release_OmitFramePointers=1, but then internal/release_impl.gypi
> wants to turn it off again.
>
> Do we actually want FPO on or off in Release? It looks like most
> recently we felt off was correct, so win_release_OmitFramePointers
> should default to 0.

https://chromiumcodereview.appspot.com/10453040

[Gabriel Charette]

sgtm, but I have no knowledge of whether this is the correct thing to do (i.e. whether we want to OmitFramePointers or not).

[Sigurður Ásgeirsson]

I believe we had decided FPO should be off, as otherwise crash reporting gets pretty messed up.

Anthony, do you remember the details?

[Gabriel Charette]

Ping?

[Anthony LaForge]

Siggi sized it up correctly,  ultimately we disabled it to improve our ability to analyze crashes.  There were some dumps that FPO mangled too badly to diagnose (I don't recall if the problems were walking the dumps via breakpad, or if too much info was removed).

Kind Regards,

Anthony Laforge
Technical Program Manager
Mountain View, CA