GCC 4.7 OpenJDK Build Failure (original) (raw)

Coleen Phillimore coleen.phillimore at oracle.com
Thu Nov 15 07:23:33 PST 2012


On 11/15/2012 2:13 AM, Mikael Vidstedt wrote:

Agreed, and it would obviously be interesting to know out why the original fix was lost.

Jon's freelist change was a big change, so it's not surprising this was lost. I hope my new ubuntu 12.04 system can get a later gcc compiler on it so we can find these earlier.

this-> seemed like the simplest fix so that's why I checked it in. I didn't know what to use with "using" at 11pm my time.

Coleen

/Mikael On 2012-11-14 20:56, David Holmes wrote: IMHO Mikaels fix should be restored.

David On 15/11/2012 1:40 PM, Coleen Phillimore wrote: On 11/14/2012 8:17 PM, Andrew Hughes wrote: I hear this has been reported already, but I couldn't see anything obvious in the archives for the various HotSpot lists. The build of jdk8 (currently b64) fails on gcc 4.7. I'm not sure when it exactly regressed, but a number of people have mentioned over the past week or so. The following webrev fixes the build (following GCC's advice):

http://cr.openjdk.java.net/~andrew/hotspot/webrev.05/ If you've already committed something and it's in the pipes, feel free to ignore this. Otherwise, we could really do with this going in so people can build OpenJDK8 again. Thanks, I thought someone was already going to check this in. It's reported as CR 8003259. <https://jbs.oracle.com/bugs/browse/JDK-8003259> It looks like it was also contributed by peter.levart at gmail.com. Jon's on vacation. We have reviewed this. I'll check it in now. Thanks, Coleen



More information about the hotspot-dev mailing list