Hello,

I've compiled the current gridengine source rpm with the Red Hat
buildsystem koji and put the rpm packages here:

    http://jur-linux.org/rpms/el-updates/5/
    http://jur-linux.org/rpms/el-updates/6/

For RHEL6 the following cludgy changes did make it compile:
--- gridengine.spec
+++ gridengine.spec
@@ -152,7 +152,7 @@

 # -O2/-O3 gives warnings about type puns.  It's not clear whether
 # they're serious, but -fno-strict-aliasing just in case.
-export SGE_INPUT_CFLAGS="$RPM_OPT_FLAGS -fno-strict-aliasing"
+export SGE_INPUT_CFLAGS="$RPM_OPT_FLAGS -fno-strict-aliasing -I /usr/include/freetype2"
 [ -n "$RPM_BUILD_NCPUS" ] && parallel_flags="-parallel $RPM_BUILD_NCPUS"
 %if %{without java}
 JAVA_BUILD_OPTIONS="-no-java -no-jni"
@@ -182,8 +182,8 @@
   rm man/man8/SGE_Helper_Service.exe.8
   rm -r util/gui-installer util/sgeSMF
   rm start_gui_installer
-  for l in lib/*/libdrmaa.so.1; do
-    ( cd $(dirname $l); ln -sf libdrmaa.so.1 libdrmaa.so; )
+  for l in lib/*/libdrmaa.so.1.0; do
+    ( cd $(dirname $l); ln -sf libdrmaa.so.1.0 libdrmaa.so; )
   done
   gzip man/man*/*
 )


Please let me know if you run these rpms and they work out ok for you.

best regards,

Florian La Roche

Hi!

I am experiencing a problem on 8.0.0c and later code. Where

the qmaster will suddenly increase the amount of ram it using and then be killed by the oom killer. On 8.0.0c this removes the parrallel environments causing all processing with PE's to stop, even when we detect the failure and recover.

Working backwards to determine when this starts happening.

The attached patch fixes build on Fedora aarch64