[PATCH] Resolve Windows build issues

Andrew gnu.andrew at redhat.com
Mon Dec 2 07:08:09 PST 2013



----- Original Message -----
> Hi,
> 
> On 11/28/2013 04:07 AM, Andrew wrote:
> >
> >
> > ----- Original Message -----
> >> * Andrew <gnu.andrew at redhat.com> [2013-11-27 14:17]:
> >>> Webrev: http://cr.openjdk.java.net/~andrew/openjdk6/20131015/winmerge/
> >>>
> >>> This contains Alex's fix for the SA along with a fix for the merge issues
> >>> in compile.make.  The latter is different to both previous versions
> >>> published
> >>> by Alex and, I believe, the correct version, which I obtained by diffing
> >>> against
> >>> the version in 7 and checking all if blocks were matched.
> >>>
> >>> The new patch:
> >>>
> >>> 1.  Moves the /SAFESEH flag into the correct section (out of the
> >>> COMPILER_NAME block and
> >>> into the compiler settings for VS2008, like the other instances in that
> >>> file).
> >>> 2.  Removes the broken if block in the VS2010 settings.
> >>> 3.  Places the full if block with other linker settings further down.
> >>
> >> I am not familiar with windows build systems at all, but the approach
> >> certainly looks correct to me. If this makes more things match upstream
> >> 7u, then this is definitely a step in the right direction.
> >>
> >> Thanks,
> >> Omair
> >>
> >> --
> >> PGP Key: 66484681 (http://pgp.mit.edu/)
> >> Fingerprint = F072 555B 0A17 3957 4E95  0056 F286 F14F 6648 4681
> >>
> >
> > I'll wait for Alex to test before pushing.
> 
> amd64 windows build works fine. And I realised that my jdk6-windows-i586
> build stand is broken badly - some kind of virus that rewrites
> executables. So please disregard my notes about SAFESEH. I won't be able
> to test i586 build for some time (until I recreate the stand) so I think
> it is better to push compile.make changes and leave sa.make intact (with
> SAFESEH flag enabled). Sorry for the noise.
> 

Well, the flag is x86 only so you wouldn't see anything on x86_64 (it's effectively
always on, as far as I can tell).

I'll drop the SA changes and push, if that's ok with Omair.

> >
> > The only remaining differences with 7 are some checks from another patch
> > and the kernel section which is only present in 6.  The above fixes what
> > were clearly issues from a bad merge (lines added in the wrong places).
> >
> 
> 
> --
> Regards,
> Alex Kasko
> 
> 

-- 
Andrew :)

Free Java Software Engineer
Red Hat, Inc. (http://www.redhat.com)

PGP Key: 248BDC07 (https://keys.indymedia.org/)
Fingerprint = EC5A 1F5E C0AD 1D15 8F1F  8F91 3B96 A578 248B DC07



More information about the jdk6-dev mailing list