Looks good to me.
/Erik
On 2015-05-12 03:44, Magnus Ihse Bursie wrote:
JDK-8079891 caused this behavior:
/scratch/dh198349/jdk9-dev > mkdir mybuild
/scratch/dh198349/jdk9-dev > cd mybuild
/scratch/dh198349/jdk9-dev/mybuild > bash ../configure
Running custom generated-configure.sh
configure: Co
On 12/05/2015 11:44 AM, Magnus Ihse Bursie wrote:
JDK-8079891 caused this behavior:
/scratch/dh198349/jdk9-dev > mkdir mybuild
/scratch/dh198349/jdk9-dev > cd mybuild
/scratch/dh198349/jdk9-dev/mybuild > bash ../configure
Running custom generated-configure.sh
configure: Configuration created at
JDK-8079891 caused this behavior:
/scratch/dh198349/jdk9-dev > mkdir mybuild
/scratch/dh198349/jdk9-dev > cd mybuild
/scratch/dh198349/jdk9-dev/mybuild > bash ../configure
Running custom generated-configure.sh
configure: Configuration created at Mon May 11 19:09:12 EDT 2015.
configure: configure
Hi Erik,
On 11/05/2015 10:31 PM, Erik Joelsson wrote:
Hello,
Please review this small fix to CFLAGS when compiling with GCC on a
32-bit system. Hotspot currently sets -march=i586 and the jdk libraries
should do the same for consistency.
Bug: https://bugs.openjdk.java.net/browse/JDK-8039426
Web
On 2015-05-11 14:31, Erik Joelsson wrote:
Hello,
Please review this small fix to CFLAGS when compiling with GCC on a
32-bit system. Hotspot currently sets -march=i586 and the jdk
libraries should do the same for consistency.
Bug: https://bugs.openjdk.java.net/browse/JDK-8039426
Webrev: http:
Hello,
Please review this small fix to CFLAGS when compiling with GCC on a
32-bit system. Hotspot currently sets -march=i586 and the jdk libraries
should do the same for consistency.
Bug: https://bugs.openjdk.java.net/browse/JDK-8039426
Webrev: http://cr.openjdk.java.net/~erikj/8039426/webrev
On 2015-05-08 18:49, Alexander Smundak wrote:
And I guess the successor of the 8u60 is going to be the last JDK8 release?
As far as I know, 8u60 is the last JDK8 feature release being done by
Oracle.
/Magnus
On Fri, May 8, 2015 at 2:00 AM, Magnus Ihse Bursie
wrote:
On 2015-05-08 04:46, Ale
Looks good to me.
/Erik
On 2015-05-11 14:05, Magnus Ihse Bursie wrote:
JDK-8074859, which turns on warnings-as-errors, unfortunately was not
properly tested with open-only, just the Oracle closed build. An extra
warning present in freetype2 code causes the build to fail.
Bug: https://bugs.op
JDK-8074859, which turns on warnings-as-errors, unfortunately was not
properly tested with open-only, just the Oracle closed build. An extra
warning present in freetype2 code causes the build to fail.
Bug: https://bugs.openjdk.java.net/browse/JDK-8080013
WebRev inline:
diff --git a/make/lib/Aw
On 2015-05-11 13:02, Roger Riggs wrote:
Hi,
Great to see the log output to a file.
I don't find the blast of trivial output from configure to be useful.
It would be a separate change, but can the output be limited to just
error and warnings or only the notable intermediate results.
Unfortunat
Note that SDT and dtrace4linux are different things. As far as I know, no one
has tried dtrace4linux with Hotspot so you are in uncharted territories. SDT
should work (if installed). Oracle Linux also has an implementation of Dtrace
that I think is different from dtrace4linux. This is also untes
On 11/05/2015 8:51 PM, Mahesh Pujari wrote:
Thanks for replying back Erik.
I ran as you stated i.e. using "make DTRACE_ENABLED=true" but then I realized
that the configure that I was running would do the same i.e. by having
CFLAGS=-DDTRACE_ENABLED=1.
No it isn't the same. The linux/makefiles/
Hi,
Great to see the log output to a file.
I don't find the blast of trivial output from configure to be useful.
It would be a separate change, but can the output be limited to just
error and warnings or only the notable intermediate results.
$.02, Roger
On 5/11/2015 5:05 AM, Magnus Ihse Burs
Looks good to me.
/Erik
On 2015-05-11 12:55, Marcus Lagergren wrote:
+1
On 11 May 2015, at 12:47, A. Sundararajan
wrote:
Please review fix for http://cr.openjdk.java.net/~sundar/8079900/webrev.00/
for https://bugs.openjdk.java.net/browse/JDK-8079900#comment-13640520
Note that jdk.nashor
+1
> On 11 May 2015, at 12:47, A. Sundararajan
> wrote:
>
> Please review fix for http://cr.openjdk.java.net/~sundar/8079900/webrev.00/
> for https://bugs.openjdk.java.net/browse/JDK-8079900#comment-13640520
>
> Note that jdk.nashorn.api.tree is part of nashorn API as specified by Nashorn
>
Thanks for replying back Erik.
I ran as you stated i.e. using "make DTRACE_ENABLED=true" but then I realized
that the configure that I was running would do the same i.e. by having
CFLAGS=-DDTRACE_ENABLED=1.
When I make (using your approach and passing enable in configure), I end up
with lots and
Please review fix for
http://cr.openjdk.java.net/~sundar/8079900/webrev.00/ for
https://bugs.openjdk.java.net/browse/JDK-8079900#comment-13640520
Note that jdk.nashorn.api.tree is part of nashorn API as specified by
Nashorn Parser API JEP http://openjdk.java.net/jeps/236
Thanks,
-Sundar
Hello,
I don't have any experience building with dtrace support, but the
Hotspot makefiles seem to imply that it can be achieved by running:
make DTRACE_ENABLED=true
That variable will trigger the appropriate preprocessor flags to be
added at least. See hotspot/make/linux/makefiles/dtrace.ma
Hi all,
I am trying make OpenJdk 9 with dtrace enabled, on my Ubuntu machine (installed
with https://github.com/dtrace4linux/linux), with configuration as below (and I
am noob).
./configure --enable-option-checking=fatal --with-debug-level=slowdebug
--with-target-bits=64 --disable-zip-debug-info
Looks good to me.
/Erik
On 2015-05-11 11:05, Magnus Ihse Bursie wrote:
It is trivial to save the log file of configure in the build output
directory. We should do that.
I also noted that the repeat of warnings at the end of the configure
run had stopped working since we moved config.log to
It is trivial to save the log file of configure in the build output
directory. We should do that.
I also noted that the repeat of warnings at the end of the configure run
had stopped working since we moved config.log to configure-support.
(Also, it was incorrectly indented.) I fixed this as we
On 2015-05-10 04:27, Phil Race wrote:
I guess this was my merge .. I know it was 2 months ago but
I do not remember being asked to merge this manually so it
looks like an odd auto-merge anomaly as it seemed to get
all the other 6 or so cases right.
Merges of code you haven't written yourself ca
22 matches
Mail list logo
