I think I have an understanding of what needs to be done now. Set the env variables to load the umem libs on the idmap daemon, then restart the idmap daemon and wait for it to happen again.
Then I can dump the core and run the ::findleaks dohickie. I was dumping a core of the idmap daemon without the proper preloaded libraries, so the ::findleaks function would not work, as pldd shows: pldd /tmp/idmapd-core.326 core '/tmp/idmapd-core.326' of 326: /usr/lib/idmapd /usr/lib/libc/libc_hwcap1.so.1 /usr/lib/libidmap.so.1 /lib/libnsl.so.1 /lib/libavl.so.1 /lib/libscf.so.1 /lib/libuutil.so.1 /lib/libresolv.so.2 /lib/libsocket.so.1 /usr/lib/libldap.so.5 /usr/lib/libnspr4.so /lib/librt.so.1 /lib/libdl.so.1 /lib/libgcc_s.so.1 /lib/libsecdb.so.1 /usr/lib/libsldap.so.1 /lib/libuuid.so.1 /usr/lib/libadutils.so.1 /lib/libmp.so.2 /lib/libmd.so.1 /lib/libgen.so.1 /usr/lib/libsasl.so.1 /usr/lib/libplc4.so /usr/lib/libnss3.so /usr/lib/libsoftokn3.so /usr/lib/libplds4.so /usr/lib/libssl3.so /lib/libdlpi.so.1 /lib/libinetutil.so.1 /lib/libdladm.so.1 /lib/libdevinfo.so.1 /lib/librcm.so.1 /lib/libnvpair.so.1 /lib/libkstat.so.1 /lib/libsec.so.1 /usr/lib/sasl/crammd5.so.1 /usr/lib/sasl/gssapi.so.1 /usr/lib/libgss.so.1 /usr/lib/sasl/plain.so.1 /usr/lib/sasl/digestmd5.so.1 /usr/lib/libpkcs11.so.1 /lib/libcryptoutil.so.1 /usr/lib/security/pkcs11_softtoken.so.1 /usr/lib/libsoftcrypto.so.1 /usr/lib/libsoftcrypto/libsoftcrypto_hwcap1.so.1 /usr/lib/gss/mech_krb5.so.1 -----Original Message----- From: John Keiffer Sent: Tuesday, March 03, 2009 1:44 PM To: [email protected]; Nicolas Williams Cc: [email protected]; John Rogers Subject: RE: [cifs-discuss] idmapd leaking? Any way to tell if it really is a dupe of this defect? http://bugs.opensolaris.org/view_bug.do?bug_id=6652090 -----Original Message----- From: [email protected] [mailto:[email protected]] Sent: Tuesday, March 03, 2009 1:15 PM To: Nicolas Williams Cc: John Keiffer; [email protected] Subject: Re: [cifs-discuss] idmapd leaking? Nicolas Williams wrote: > On Tue, Mar 03, 2009 at 12:48:59PM -0800, John Keiffer wrote: >> Looks like we could not run this command: >> >> r...@leopard-6:/volumes# gcore -c /tmp/idmapd-core `pgrep idmapd` >> gcore: invalid content string '/tmp/idmapd-core' >> usage: gcore [ -pgF ] [ -o filename ] [ -c content ] pid ... >> r...@leopard-6:/volumes# > > Sorry, that was a typo. I meant: > > # gcore -o /tmp/idmapd-core `pgrep idmapd` > I wonder if this isn't just an instance of: 6652090 libidmap.so.1`idmap_get_mappings leaks memory -Mark _______________________________________________ cifs-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/cifs-discuss
