-
Bug
-
Resolution: Fixed
-
P4
-
9, 10, 11
-
b11
-
generic
-
linux
| Issue | Fix Version | Assignee | Priority | Status | Resolution | Resolved In Build |
|---|---|---|---|---|---|---|
| JDK-8202964 | 10u-cpu | Severin Gehwolf | P4 | Resolved | Fixed | master |
| JDK-8202925 | 10.0.2 | Severin Gehwolf | P4 | Resolved | Fixed | b07 |
| JDK-8202779 | 10u-open | Severin Gehwolf | P4 | Resolved | Fixed | master |
We've discovered that bootcycle builds may run out of resources due to bad jobs config.
In make/Main.gmk around line 324 where the bootcycle-images target is defined we see:
+$(MAKE) $(MAKE_ARGS) -f $(TOPDIR)/make/Init.gmk PARALLEL_TARGETS=$(BOOTCYCLE_TARGET) \
JOBS= SPEC=$(dir $(SPEC))bootcycle-spec.gmk main
So it's using 'JOBS=' which corresponds to 'make -j' invokation without an argument unconditionally. That is, make is free to use any number of parallel jobs for the bootcycle-images target even though configure might have had option "--with-num-cores=<NUM>".
For background info see this mail thread:
http://mail.openjdk.java.net/pipermail/build-dev/2017-April/018929.html
In make/Main.gmk around line 324 where the bootcycle-images target is defined we see:
+$(MAKE) $(MAKE_ARGS) -f $(TOPDIR)/make/Init.gmk PARALLEL_TARGETS=$(BOOTCYCLE_TARGET) \
JOBS= SPEC=$(dir $(SPEC))bootcycle-spec.gmk main
So it's using 'JOBS=' which corresponds to 'make -j' invokation without an argument unconditionally. That is, make is free to use any number of parallel jobs for the bootcycle-images target even though configure might have had option "--with-num-cores=<NUM>".
For background info see this mail thread:
http://mail.openjdk.java.net/pipermail/build-dev/2017-April/018929.html
- backported by
-
JDK-8202779 Number of make jobs wrong for bootcycle-images target
-
- Resolved
-
-
-
- Resolved
-
-
-
- Resolved
-