> can you please provide some basic information that helps us find the reason: > > 1) What was the original compiler error that you got, before adding > "extern".
With the i5/OS C compiler, all vars defined in a .h are included in the objects (.o) using this include. Adding the extern solve the problem. > 2) What is the httpd and jk configuration that was used when the core > happened. Apache 2.0.5x (I didn't know which one) > 3) What did you have to do to get the core, e.g. which URL did you call, or > does it happen immediately during startup etc. During startup > 4) Can you provide the contents of the httpd error log and a JK log file > with log level trace containing startup and the core situation. I see : [Thu Mar 19 12:29:50.079 2009] [7590:3] [debug] jk_util.c (458): Pre-processed log time stamp format is '[%a %b %d %H:%M:%S.000 %Y] ' [Thu Mar 19 12:29:50.109 2009] [7590:3] [debug] jk_uri_worker_map.c (770): rule map size is 3 [Thu Mar 19 12:29:50.109 2009] [7590:3] [debug] jk_uri_worker_map.c (720): wildchar rule '/xylos-soa/*=xylos-soa-850' source 'JkMount' was added [Thu Mar 19 12:29:50.109 2009] [7590:3] [debug] jk_uri_worker_map.c (720): wildchar rule '/jkmanager/*=jkstatus' source 'JkMount' was added [Thu Mar 19 12:29:50.109 2009] [7590:3] [debug] jk_uri_worker_map.c (729): exact rule '/status/=jkstatus' source 'JkMount' was added [Thu Mar 19 12:29:50.109 2009] [7590:3] [debug] jk_uri_worker_map.c (171): uri map dump after map open: index=0 file='(null)' reject_unsafe=0 reload=60 modified=0 checked=0 [Thu Mar 19 12:29:50.109 2009] [7590:3] [debug] jk_uri_worker_map.c (176): generation 0: size=0 nosize=0 capacity=0 [Thu Mar 19 12:29:50.109 2009] [7590:3] [debug] jk_uri_worker_map.c (176): generation 1: size=3 nosize=0 capacity=4 [Thu Mar 19 12:29:50.109 2009] [7590:3] [debug] jk_uri_worker_map.c (186): NEXT (1) map #0: uri=/xylos-soa/* worker=xylos-soa-850 context=/xylos-soa/* source=JkMount type=Wildchar len=12 [Thu Mar 19 12:29:50.109 2009] [7590:3] [debug] jk_uri_worker_map.c (186): NEXT (1) map #1: uri=/jkmanager/* worker=jkstatus context=/jkmanager/* source=JkMount type=Wildchar len=12 [Thu Mar 19 12:29:50.109 2009] [7590:3] [debug] jk_uri_worker_map.c (186): NEXT (1) map #2: uri=/status/ worker=jkstatus context=/status/ source=JkMount type=Exact len=8 [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_worker.c (116): found a worker xylos-soa-850 [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_worker.c (293): Found worker type 'ajp13' [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_uri_worker_map.c (512): Checking extension for worker 0: xylos-soa-850 of type ajp13 (2) [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_worker.c (116): found a worker jkstatus [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_worker.c (293): Found worker type 'status' [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_uri_worker_map.c (512): Checking extension for worker 1: jkstatus of type status (6) [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_worker.c (116): found a worker jkstatus [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_worker.c (293): Found worker type 'status' [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_uri_worker_map.c (512): Checking extension for worker 2: jkstatus of type status (6) [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_uri_worker_map.c (171): uri map dump after extension stripping: index=0 file='(null)' reject_unsafe=0 reload=60 modified=0 checked=0 [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_uri_worker_map.c (176): generation 0: size=0 nosize=0 capacity=0 [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_uri_worker_map.c (176): generation 1: size=3 nosize=0 capacity=4 [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_uri_worker_map.c (186): NEXT (1) map #0: uri=/xylos-soa/* worker=xylos-soa-850 context=/xylos-soa/* source=JkMount type=Wildchar len=12 [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_uri_worker_map.c (186): NEXT (1) map #1: uri=/jkmanager/* worker=jkstatus context=/jkmanager/* source=JkMount type=Wildchar len=12 [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_uri_worker_map.c (186): NEXT (1) map #2: uri=/status/ worker=jkstatus context=/status/ source=JkMount type=Exact len=8 [Thu Mar 19 12:29:50.191 2009] [7590:3] [debug] jk_uri_worker_map.c (482): Switching uri worker map from index 0 to index 1 [Thu Mar 19 12:29:53.099 2009] [7590:3] [debug] jk_util.c (458): Pre-processed log time stamp format is '[%a %b %d %H:%M:%S.000 %Y] ' [Thu Mar 19 12:29:53.099 2009] [7590:3] [debug] jk_uri_worker_map.c (770): rule map size is 3 [Thu Mar 19 12:29:53.099 2009] [7590:3] [debug] jk_uri_worker_map.c (720): wildchar rule '/xylos-soa/*=xylos-soa-850' source 'JkMount' was added [Thu Mar 19 12:29:53.099 2009] [7590:3] [debug] jk_uri_worker_map.c (720): wildchar rule '/jkmanager/*=jkstatus' source 'JkMount' was added [Thu Mar 19 12:29:53.099 2009] [7590:3] [debug] jk_uri_worker_map.c (729): exact rule '/status/=jkstatus' source 'JkMount' was added [Thu Mar 19 12:29:53.099 2009] [7590:3] [debug] jk_uri_worker_map.c (171): uri map dump after map open: index=0 file='(null)' reject_unsafe=0 reload=60 modified=0 checked=0 [Thu Mar 19 12:29:53.099 2009] [7590:3] [debug] jk_uri_worker_map.c (176): generation 0: size=0 nosize=0 capacity=0 [Thu Mar 19 12:29:53.099 2009] [7590:3] [debug] jk_uri_worker_map.c (176): generation 1: size=3 nosize=0 capacity=4 [Thu Mar 19 12:29:53.099 2009] [7590:3] [debug] jk_uri_worker_map.c (186): NEXT (1) map #0: uri=/xylos-soa/* worker=xylos-soa-850 context=/xylos-soa/* source=JkMount type=Wildchar len=12 [Thu Mar 19 12:29:53.099 2009] [7590:3] [debug] jk_uri_worker_map.c (186): NEXT (1) map #1: uri=/jkmanager/* worker=jkstatus context=/jkmanager/* source=JkMount type=Wildchar len=12 [Thu Mar 19 12:29:53.099 2009] [7590:3] [debug] jk_uri_worker_map.c (186): NEXT (1) map #2: uri=/status/ worker=jkstatus context=/status/ source=JkMount type=Exact len=8 > 5) Is it possible on iSeries to get a stack of the relevant thread from the > core, i.e. not only the top level method "allocator_free". Nope > 6) What is strange: JK doesn't use APR. Are you sure, that core is actually > related to JK, or did you do some more upgrades like e.g. upgrading APR or > httpd? > > Regards, --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org For additional commands, e-mail: dev-h...@tomcat.apache.org
