Hi,
On Wed, 2015-11-04 at 12:12 +0100, Volker Simonis wrote:
> Hi Erik,
>
> thanks a lot for your understanding :)
>
> Your suggestion is perfectly fine for me and we can live with both
> versions of the fix.
I went for the change in the globalDefinitions* file.
New webrevs:
At Google, we've also noticed crashes in Bits.c with recent gcc.
Adding nio-dev, since this is not really a build problem.
It appears that Bits.c is resorting to undefined behavior, even though
there is a long tradition of unaligned access on x86 being permitted by the
hardware.
My colleague
Please review at your convenience.
Issue: https://bugs.openjdk.java.net/browse/JDK-8140630
Patch: Included below
The change is to compile src/java.base/share/native/libjava/Bits.c only into
libjava.so with normal optimization level (-O2) instead of high optimization
level (-O3). The other
I would suggest you try doing exactly what the error message suggests as
that will most likely work fine.
/Erik
On 2015-11-04 21:03, Naveen Kumar wrote:
Hi,
I am getting the following error (below) while running configure
on jdk9 (openjdk) on cygwin on windows7.
Should I download
Hi,
can somebody please review and sponsor the following tiny build fix:
http://cr.openjdk.java.net/~simonis/webrevs/2015/8141416/
https://bugs.openjdk.java.net/browse/JDK-8141416
Building hotspot on certain systems results in a series of:
expr: syntax error
expr: syntax error
expr: syntax
Hi,
I am getting the following error (below) while running configure
on jdk9 (openjdk) on cygwin on windows7.
Should I download freetype & build it with VS(2013)?
If yes, then what are special build instructions.
If no then how to fix this.
On Nov 4, 2015, at 2:12 PM, Martin Buchholz wrote:
> At Google, we've also noticed crashes in Bits.c with recent gcc.
> Adding nio-dev, since this is not really a build problem.
Agreed. The posted patch is not intended to be a final solution.
> It appears that Bits.c is
Alright, it's a much better situation if we're not giving up on fixing the
underlying problem.
I don't like temporary fixes. They tend to become permanent.
On Wed, Nov 4, 2015 at 8:13 PM, Brian Burkhalter <
brian.burkhal...@oracle.com> wrote:
> The problem is due to a non-alignment issue as you
The problem is due to a non-alignment issue as you previously suggested.
Suppressing it via the build configuration change is a procedural matter: work
to find the root problem and a real solution continues nonetheless.
On Nov 4, 2015, at 7:46 PM, Martin Buchholz wrote:
>
Hi Volker,
looks fine and builds fine on Ubuntu 14.4.
...Thomas
On Wed, Nov 4, 2015 at 7:21 PM, Volker Simonis
wrote:
> Hi,
>
> can somebody please review and sponsor the following tiny build fix:
>
> http://cr.openjdk.java.net/~simonis/webrevs/2015/8141416/
>
The JDK launchers have been built by a macro which is using positional
arguments instead of named argument. This needs to be fixed to be able
to properly track what is happening.
Some additional TLC for launchers is also needed.
To verify the fix, I've used the new COMPARE_BUILD feature.
This is just so nice! Only one small comment. In LauncherCommon.gmk
there are two ifeq for OS macosx. Perhaps those should be combined?
/Erik
On 2015-11-05 08:27, Magnus Ihse Bursie wrote:
The JDK launchers have been built by a macro which is using positional
arguments instead of named
Looks good to me.
/Erik
On 2015-11-04 14:49, Magnus Ihse Bursie wrote:
It turned out that the fix in JDK-8136813 was only tested with -2dirs
on compare.sh. When using -o, the compare script now fails to
correctly determine it's own location. This patch remedies that.
Bug:
Hi Thomas,
the change looks good. Thomas St. is just checking the VS2010 build.
We'll let you new the results once it's finished.
Thank you and best regards,
Volker
On Wed, Nov 4, 2015 at 4:04 PM, Erik Helin wrote:
> On 2015-11-04, Thomas Schatzl wrote:
>> Hi,
>>
>> On
In the initial build-infra setup, it was possible to selectively replace
individual repos (e.g. corba). This has never been widely used, and has
not been tested or supported for a long time. It recently turned out to
be broken on hotspot.
We should deprecate this functionality.
Bug:
On 2015-11-04 11:46, Erik Helin wrote:
Maybe we can come up with something that removes these #pragmas from the
GC code while still allowing you to compile with VS2010. What do you
think about the following suggestions:
- We change the Makefiles to add the flag /Wd4355 if we use VS2010 or
Hi Eric,
On Wed, Nov 4, 2015 at 11:46 AM, Erik Helin wrote:
> (adding build-dev as well)
>
> On 2015-11-03, Volker Simonis wrote:
> > Hi Erik,
> >
> > while I agree that this is a nice clean up I want to second Thomas in
> > his attempt to save the VS 2010 build. I
On 2015-11-04 12:04, Volker Simonis wrote:
Hi Magnus,
thanks for checking, but here comes the new (and improved :) version:
http://cr.openjdk.java.net/~simonis/webrevs/2015/8141290_toplevel/
http://cr.openjdk.java.net/~simonis/webrevs/2015/8141290_jdk/
Looks good.
In the toplevel change I've
Volker,
Thanks for the great analysis! Your trivial fix seems reasonable. I
opened https://bugs.openjdk.java.net/browse/JDK-8141416 to track this.
It's up for grabs to add that line. :) (Otherwise I'll fix it when I
have the time.)
/Magnus
On 2015-11-04 08:49, Volker Simonis wrote:
Hi
Thanks for the opening the bug report.
I've triaged it but as I can not push it anyway, I leave the fix for
the interested reader :)
On Wed, Nov 4, 2015 at 12:58 PM, Magnus Ihse Bursie
wrote:
> Volker,
>
> Thanks for the great analysis! Your trivial fix seems
It turned out that the fix in JDK-8136813 was only tested with -2dirs on
compare.sh. When using -o, the compare script now fails to correctly
determine it's own location. This patch remedies that.
Bug: https://bugs.openjdk.java.net/browse/JDK-8141439
WebRev:
Hello Magnus:
It turned out that the fix in JDK-8136813 was only tested with -2dirs
on compare.sh. When using -o, the compare script now fails to
correctly determine it's own location. This patch remedies that.
Bug: https://bugs.openjdk.java.net/browse/JDK-8141439
WebRev:
Hi Eric,
your change builds fine on VS 2010.
Thank you!
Thomas
On Wed, Nov 4, 2015 at 4:04 PM, Erik Helin wrote:
> On 2015-11-04, Thomas Schatzl wrote:
> > Hi,
> >
> > On Wed, 2015-11-04 at 12:12 +0100, Volker Simonis wrote:
> > > Hi Erik,
> > >
> > > thanks a lot for
On 2015-11-03 18:20, Magnus Ihse Bursie wrote:
On 2015-11-03 17:20, Volker Simonis wrote:
Hi,
unfortunately, change '8140661: Rename LDFLAGS_SUFFIX to LIBS' broke
the AIX build.
I'm sorry for that. :-(
Can somebody please review this small change which fixes the build?
Hi Magnus,
On 2015-11-03 18:23, Magnus Ihse Bursie wrote:
On 2015-11-02 15:41, Mikael Gerdin wrote:
Hi build-devs,
I was trying to do a fancy setup with multiple different hotspot
directories in a single source tree to create some builds for
performance measurements.
I used the configure
Hi Johnathan,
this is a known problem - at least to me :)
It is caused by a peculiarity of the gcc version on Ubuntu where "gcc
-dumpversion" doesn't print a micro-version:
Ubuntu:
$ gcc -dumpversion
4.6
Any other Linux:
$ gcc -dumpversion
4.8.3
This "feature" is tracked under
Hi,
Could I +1 a request for this to be fixed. I appreciate it is harmless, but it
is an annoyance and it is just going to be a source of bug reports forever:-)
If you wish I can author a changeset for this.
Regards,
Ed.
On Wed, 2015-11-04 at 08:49 +0100, Volker Simonis wrote:
> Hi
On Tue, Nov 3, 2015 at 6:20 PM, Magnus Ihse Bursie
wrote:
> On 2015-11-03 17:20, Volker Simonis wrote:
>>
>> Hi,
>>
>> unfortunately, change '8140661: Rename LDFLAGS_SUFFIX to LIBS' broke
>> the AIX build.
>
>
> I'm sorry for that. :-(
>>
>>
>> Can somebody please
(adding build-dev as well)
On 2015-11-03, Volker Simonis wrote:
> Hi Erik,
>
> while I agree that this is a nice clean up I want to second Thomas in
> his attempt to save the VS 2010 build. I understand that the official
> Oracle build tool chain for jdk9 on Windows is VS 2013, but this may
> be
On Wed, Nov 4, 2015 at 11:46 AM, Erik Helin wrote:
> (adding build-dev as well)
>
> On 2015-11-03, Volker Simonis wrote:
>> Hi Erik,
>>
>> while I agree that this is a nice clean up I want to second Thomas in
>> his attempt to save the VS 2010 build. I understand that the
Hi Magnus,
thanks for checking, but here comes the new (and improved :) version:
http://cr.openjdk.java.net/~simonis/webrevs/2015/8141290_toplevel/
http://cr.openjdk.java.net/~simonis/webrevs/2015/8141290_jdk/
In the toplevel change I've simply removed '-liconv' from JDKLIB_LIBS on AIX.
In the
31 matches
Mail list logo