openldap-bugs

openldap-bugs@openldap.org

1 participants
21003 discussions

(ITS#8512) segfault for large entries
by steffen.hofmann＠fu-berlin.de 07 Oct '16

07 Oct '16

Full_Name: Steffen Hofmann Version: 2.4.44 OS: Debian 3.16.36-1+deb8u1 URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (87.77.77.78) When searching large entries (e.g. >65kb) slapd segfaults. from log: ber_flush2 failed errno=11 reason="Resource temporarily unavailable" from strace: 816254 write(9, "<deleted content>"..., 105957) = 67160 816254 write(9, "<deleted content>"..., 38797) = -1 EAGAIN (Resource temporarily unavailable) 816254 --- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_ACCERR, si_addr=0x7efd4f9438d0} --- from coredump: #0 0x00002aaec12318d0 in ?? () #1 0x000000000044d2d7 in slap_writewait_play (op=0x2aafb41041b0) at result.c:294 #2 send_ldap_ber (op=op@entry=0x2aafb41041b0, ber=ber@entry=0x2aafaeca3000) at result.c:367 #3 0x0000000000450550 in slap_send_search_entry (op=0x2aafb41041b0, rs=0x2aafaee34a70) at result.c:1430 #4 0x00000000004d135b in mdb_search (op=<oimimized out>, rs=0x2aafaee34a70) at search.c:1086 #5 0x00000000004a2f4d in overlay_op_walk (op=0x2aafb41041b0, rs=0x2aafaee34a70, which=op_search, oi=0x1907760, on=0x0) at backover.c:677 #6 0x00000000005a83e2 in slapi_op_func (op=0x2aafb41041b0, rs=0x2aafaee334c0) at slapi_overlay.c:650 #7 0x00000000004a2f0a in overlay_op_walk (op=op@entry=0x2aafb41041b0, rs=0x2aafaee34a70, which=op_search, oi=0x1907760, on=0x1907940) at backover.c:661 #8 0x00000000004a3087 in over_op_func (op=0x2aafb41041b0, rs=<optimized out>, which=which@entry=op_search) at backover.c:730 #9 0x00000000004a31aa in over_op_search (op=<optimized out>, rs=<optimized out>) at backover.c:757 #10 0x00000000004414aa in fe_op_search (op=0x2aafb41041b0, rs=0x2aafaee34a70) at search.c:402 #11 0x00000000004a2f4d in overlay_op_walk (op=op@entry=0x2aafb41041b0, rs=0x2aafaee34a70, which=op_search, oi=0x18a1430, on=0x0) at backover.c:677 #12 0x00000000004a3087 in over_op_func (op=0x2aafb41041b0, rs=<optimized out>C C which=which@entry=op_search) at backover.c:730 #13 0x00000000004a31aa in over_op_search (op=<optimized out>, rs=<optimized out>) at backover.c:757 #14 0x000000000044100c in do_search (op=0x2aafb41041b0, rs=0x2aafaee34a70) at search.c:247 #15 0x000000000043eca6 in connection_operation (ctx=ctx@entry=0x2aafaee34b90, arg_v=arg_v@entry=0x2aafb41041b0) at connection.c:1158 #16 0x000000000043fa4b in connection_read_thread (ctx=0x2aafaee34b90, argv=0x9) at connection.c:1294 #17 0x0000000000570293 in ldap_int_thread_pool_wrapper (xpool=0x184bd30) at tpool.c:696 #18 0x00002aaebcc360a4 in start_thread () from /lib/x86_64-linux-gnu/libpthread.so.0 #19 0x00002aaebcf3362d in clone () from /lib/x86_64-linux-gnu/libc.so.6 (gdb) up #1 0x000000000044d2d7 in slap_writewait_play (op=0x2aafb41041b0) at result.c:294 294 sc->sc_writewait( op, sc ); (gdb) list 289 { 290 slap_callback *sc = op->o_callback; 291 292 for ( ; sc; sc = sc->sc_next ) { 293 if ( sc->sc_writewait ) 294 sc->sc_writewait( op, sc ); 295 } 296 } 297 298 static long send_ldap_ber( I think, the change that cause the error was made with the new function slap_writewait_play in the version 2.4.40 in servers/slapd/result.c I hope it is helpful. Steffen

1 0

Re: (ITS#8511) Rename "mirrormode" to "multimaster", deprecate "mirrormode" parameter
by michael＠stroeder.com 06 Oct '16

06 Oct '16

quanah(a)openldap.org wrote: > In bconfig.c, the following should be reworked: > [..] > So that the parameter name is multimaster, and the config attr is > olcMultiMaster > > The term "mirrormode" is both confusing and misleading, as what it really > controls is whether or not this node is enabled for multimaster. +1 Ciao, Michael.

1 0

(ITS#8511) Rename "mirrormode" to "multimaster", deprecate "mirrormode" parameter
by quanah＠openldap.org 06 Oct '16

06 Oct '16

Full_Name: Quanah Gibson-Mount Version: 2.4.44 OS: N/A URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (47.208.148.26) In bconfig.c, the following should be reworked: { "mirrormode", "on|off", 2, 2, 0, ARG_DB|ARG_ON_OFF|ARG_MAGIC|CFG_MIRRORMODE, &config_generic, "( OLcfgDbAt:0.16 NAME 'olcMirrorMode' " So that the parameter name is multimaster, and the config attr is olcMultiMaster The term "mirrormode" is both confusing and misleading, as what it really controls is whether or not this node is enabled for multimaster.

1 0

Re: (ITS#8427) Incorrect value of tls_reqcert in syncrepl
by mkp37215＠gmail.com 05 Oct '16

05 Oct '16

Since more than four months have passed since last message here, I would like to ask if any progress has been made in this matter? Has anyone looked into my patch? Thank you

1 0

(ITS#8510) OpenLDAP Crash when using ppolicy
by poil＠quake.fr 03 Oct '16

03 Oct '16

Full_Name: Poil Version: 2.4.42+dfsg-2ubuntu3.1 OS: Ubuntu Xenial URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (62.240.254.57) Hi, I have an openldap with a master-master replication * Config directory is a HDB database * Main directory (people/group/policy) is a MDB database When using # ldapwhoami -x -H ldap://localhost -D uid=poil,ou=opople,dc=my,dc=test -W -e ppolicy LDAP crash When using # ldapwhoami -x -H ldap://localhost -D uid=poil,ou=people,dc=my,dc=test -W It's working No logs from crash Best regards

1 0

(ITS#8509) LMDB make targets for automake
by phantom+lmdb＠live.ie 02 Oct '16

02 Oct '16

Full_Name: Aido Version: 2.4.44 OS: Ubuntu Trusty URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (216.239.90.19) The current liblmdb Makefile does not make it easy to use liblmdb as a submodule to other projects that utilize automake e.g. automake uses a 'check' directive instead of 'test'. Adding the following to the Makefile would make liblmdb more automake friendly: EMPTY_AUTOMAKE_TARGETS = dvi pdf ps info html tags ctags distcheck .PHONY: $(EMPTY_AUTOMAKE_TARGETS) $(EMPTY_AUTOMAKE_TARGETS): check: test See: https://www.gnu.org/software/automake/manual/html_node/Third_002dParty-Make…

1 0

(ITS#8508) PATCH - ucgendat.c properly add title-case characters without upper-case equivalents (e.g. greek letters with iota subscript)
by z.figura12＠gmail.com 30 Sep '16

30 Sep '16

Full_Name: Zebediah Figura Version: trunk OS: (Kubuntu 16.04) URL: ftp://ftp.openldap.org/incoming/zebediah-figura-160930.patch Submission from: (NULL) (132.161.178.173) The file ucgendat.c, which is used to generate a list of (among other things) Unicode case correspondences, only generates correspondences for the title-case characters which have both a corresponding lower-case and upper-case forms (ǅ, ǈ, ǋ, ǲ). This is because read_cdata() checks whether there exists both a lower-case and an upper-case form, but Unicode does not provide an upper-case form for title-case characters (Greek characters with iota subscript) which do not have a distinct upper-case form. Thus these are mistakenly treated as upper-case characters. I have attached a patch which causes read_cdata() to treat a character as title-case if it has the property Lt (i.e. Letter, title) and has either an upper-case or lower-case form defined.

1 0

Re: (ITS#8505) LMDB vs. fork()
by h.b.furuseth＠usit.uio.no 29 Sep '16

29 Sep '16

I wrote: >On 29. sep. 2016 18:27, lmb(a)cloudflare.com wrote: >> If I understand you correctly, LMDB would be safe to use in these >> scenarios in an ideal world: >> >> 1. mdb_env_open(), fork(), exec() (without fd leaks I'd argue) >> 2. mdb_env_open(), pthread_create(), fork() >> 3. mdb_env_open(), fork() >> >> As far as I understand you are after 2 and 3, while I want 1. > > I'm after all three. ...if the user calls the cleanup function. But: Whoops, correction: If you mean using LMDB after fork(), (2) can't be safe unless the user makes it safe with pthread_atfork(). But at least it can clean up some of the stuff (3) also does, and he'll have the pthread_atfork option if he really needs it.

1 0

Re: (ITS#8505) LMDB vs. fork()
by h.b.furuseth＠usit.uio.no 29 Sep '16

29 Sep '16

On 29. sep. 2016 18:27, lmb(a)cloudflare.com wrote: > <h.b.furuseth(a)usit.uio.no> wrote: > >> Yes. Your concern seems to reduce to a doc bug and an >> otherwise-harmless file descriptor leak which it is too late to fix >> completely. So I ended up thinking of resource leaks in general. > > Why is it harmless in your opinion? Because I wrote that documentation, and now I know what it gets wrong:-) Only the lockfile descriptor is dangerous. Therefore it already had FD_CLOEXEC. The other FD leaks are just leaks, they can't break LMDB. >> It'd be pretty intrusive of a _library_ to forbid the user to >> fork() at all without exec(). > > From my point of view, that is what the LMDB doc says. Yes. And that doc is half wrong, so I'll fix it. And it's half right with pthread_exit(), so I want to fix LMDB about that. > If I understand you correctly, LMDB would be safe to use in these > scenarios in an ideal world: > > 1. mdb_env_open(), fork(), exec() (without fd leaks I'd argue) > 2. mdb_env_open(), pthread_create(), fork() > 3. mdb_env_open(), fork() > > As far as I understand you are after 2 and 3, while I want 1. I'm after all three. (1) is already safe. The mdb_env_get_fd() descriptor will leak since otherwise programs using it could break. But you can close() it yourself in the child. Well, or mdb_env_open() could take an MDB_CLOEXEC flag which says to set FD_CLOEXEC on that FD too. And since the child may need to do something to avoid leaks anyway, I figure we might as well provide a cleanup function do do it properly. So... > Case 2 > seems unlikely, given that forking a multi-threaded program is so hard > that is rarely makes sense [1]. It's very limited what you can do in the child, yes. But sometimes if you have do then you have to. E.g. if the exec()ed program may not exist, and the child has do some simple task as a fallback. Or if you're not free to rewrite the main program, only some module. Case 2 will at least have memory leaks, since free() in the child is unsafe. But we can get rid of some OS resources (FDs, memory map, maybe semaphores, need to check about that.) > Case 3 is simply a matter of reordering the calls. Only if the program logic allows it. Cumbersome e.g. if the LMDB database contains the config controlling the child. -- Hallvard

1 0

(ITS#8507) adding the syncprov overlay does not load the associated schema elements
by elecharny＠apache.org 29 Sep '16

29 Sep '16

Full_Name: Emmanuel L.charny Version: 2.4.44 OS: Linux CentOS 6 URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (176.154.3.51) Start an OpenLDAP server using the dynamic configuration (slapd.d) from a slapd.conf file that does not contain a synprov module : $ /opt/symas/bin/slaptest -f /opt/symas/etc/openldap/slapd.conf -F /opt/symas/etc/openldap/slapd.d $ /etc/init.d/solserver start -d 256 $ grep -ri "olcSpNoPresent" slapd.d $ ... Then inject the syncprov module in the server : dn: cn=module{0},cn=config changetype: modify add: olcModuleLoad olcModuleLoad: syncprov.la - The olcSpNoPresent attribute is not present : $ grep -ri "olcSpNoPresent" slapd.d $ ... You can stop and restart the server, the attribute is still not present. Going back to the slapd.conf file, add the module : ... # Uncomment the mululeloads as needed to enable additional # functionalityi when configured. NOTE: We package many # more modules options than those found below. moduleload back_mdb.la moduleload back_monitor.la moduleload syncprov.la # Added module ... and regenerate the slapd.d content : $ /etc/init.d/solserver stop $ rm -rf /opt/symas/etc/openldap/slapd.d/* $ /opt/symas/bin/slaptest -f /opt/symas/etc/openldap/slapd.conf -F /opt/symas/etc/openldap/slapd.d $ grep -ri "olcSpNoPresent" /opt/symas/etc/openldap/slapd.d slapd.d/cn=config/cn=schema.ldif:olcAttributeTypes: ( OLcfgOvAt:1.3 NAME 'olcSpNoPresent' DESC 'Omit Present slapd.d/cn=config/cn=schema.ldif: $ olcSpSessionlog $ olcSpNoPresent $ olcSpReloadHint ) ) $... This is problematic...

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

openldap-bugs