openldap-bugs

openldap-bugs@openldap.org

1 participants
21000 discussions

[Issue 7980] LMDB added support for user context in compare function
by openldap-its＠openldap.org 02 Jun '21

02 Jun '21

https://bugs.openldap.org/show_bug.cgi?id=7980 Howard Chu <hyc(a)openldap.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Resolution|--- |INVALID Status|UNCONFIRMED |RESOLVED --- Comment #1 from Howard Chu <hyc(a)openldap.org> --- Not needed. Pass additional state using the MDB_val *a pointer. -- You are receiving this mail because: You are on the CC list for the issue.

1 0

[Issue 7853] Improve safety of mdb_env_close
by openldap-its＠openldap.org 02 Jun '21

02 Jun '21

https://bugs.openldap.org/show_bug.cgi?id=7853 --- Comment #4 from Howard Chu <hyc(a)openldap.org> --- (In reply to dw(a)hmmz.org from comment #0) > Full_Name: David Wilson > Version: LMDB 0.9.11 > OS: > URL: > https://raw.githubusercontent.com/dw/py-lmdb/master/misc/readers_mrb_env. > patch > Submission from: (NULL) (178.238.153.20) > > > Currently if a user (wilfully or accidentally, say, through composition of > third party libs) opens the same LMDB environment twice within a process, and > maintains active read transactions at the time one mdb_env_close() is called, > all reader slots will be deallocated in all environments due to the logic > around line 4253 that unconditionally clears reader slots based on PID. > > I'd like to avoid this in py-lmdb, and seems there are a few alternatives to > fix it: > 4) Modify lock.mdb to include MDB_env* address within the process, and update > mdb_env_close() to invalidate only readers associated with the environment > being closed. I dislike using the MDB_env* as an opaque publicly visible > cookie, but perhaps storing this field might be reused for other things in > future. I was looking at this approach, which initially seems harmless. But unfortunately it does nothing to address the fact the fcntl lock on the lock file will still go away on the first env_close() call. If there are no other processes with the env open, then the next process that opens the env will re-init the lock file, because it doesn't know about the existing process. At that point the existing process is hosed anyway. So the question is, if a process has made the mistake of opening the same env twice, what is the anticipated lifetime of each env instance? If the code was written according to doc guidelines, then both env instances should only be closed after all processing is done and just before program exit. In that case it doesn't really matter what happens to the reader table or the fcntl lock. But if the two instances have drastically different lifetimes, then all bets are off. -- You are receiving this mail because: You are on the CC list for the issue.

1 0

[Issue 7842] mdb readers/writer exclusion protocol, as implemented, is racy.
by openldap-its＠openldap.org 02 Jun '21

02 Jun '21

https://bugs.openldap.org/show_bug.cgi?id=7842 Howard Chu <hyc(a)openldap.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution|--- |DUPLICATE --- Comment #3 from Howard Chu <hyc(a)openldap.org> --- *** This issue has been marked as a duplicate of issue 7969 *** -- You are receiving this mail because: You are on the CC list for the issue.

1 0

[Issue 7969] LMDB: Globally shared fields of meta-data are not 'volatile'.
by openldap-its＠openldap.org 02 Jun '21

02 Jun '21

https://bugs.openldap.org/show_bug.cgi?id=7969 Howard Chu <hyc(a)openldap.org> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |rsbx(a)acm.org --- Comment #18 from Howard Chu <hyc(a)openldap.org> --- *** Issue 7842 has been marked as a duplicate of this issue. *** -- You are receiving this mail because: You are on the CC list for the issue.

1 0

[Issue 7969] LMDB: Globally shared fields of meta-data are not 'volatile'.
by openldap-its＠openldap.org 02 Jun '21

02 Jun '21

https://bugs.openldap.org/show_bug.cgi?id=7969 Howard Chu <hyc(a)openldap.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution|--- |FIXED --- Comment #17 from Howard Chu <hyc(a)openldap.org> --- Patch committed, fix released in LMDB 0.9.15 -- You are receiving this mail because: You are on the CC list for the issue.

1 0

[Issue 7668] LMDB enhancement, corruption detection
by openldap-its＠openldap.org 02 Jun '21

02 Jun '21

https://bugs.openldap.org/show_bug.cgi?id=7668 Howard Chu <hyc(a)openldap.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |RESOLVED Resolution|--- |TEST --- Comment #2 from Howard Chu <hyc(a)openldap.org> --- MDB_PREVSNAPSHOT flag was added for LMDB 1.0 to allow opening env with the previous instead of current meta page. mdb_env_set_checksum() was added for LMDB 1.0 to allow specifying per-page checksums. -- You are receiving this mail because: You are on the CC list for the issue.

1 0

[Issue 7165] Killing slap tools breaks the running mdb
by openldap-its＠openldap.org 02 Jun '21

02 Jun '21

https://bugs.openldap.org/show_bug.cgi?id=7165 --- Comment #5 from Howard Chu <hyc(a)openldap.org> --- Most of this bug appears to be obsolete. (In reply to Hallvard Furuseth from comment #0) > Full_Name: Hallvard B Furuseth > Version: master, d861910f55cb1beb5e443caa7e961ed760074352 > OS: Linux x86_64 > URL: > Submission from: (NULL) (195.1.106.125) > Submitted by: hallvard > > > Aborting an MDB process can break MDB if another MDB process is > running, because then locks/reader table info are not reset. Using mdb_stat -rr will purge stale readers. I suppose we could tweak back-mdb to also do a reader purge if it ever gets a MDB_READERS_FULL error from mdb_txn_begin. > > The problems below go away when the last slap process terminates > so the lockfile can be reset. Sometimes kill -KILL is needed. > > > ==== lock.conf ==== > include servers/slapd/schema/core.schema > database mdb > suffix o=hi > directory lock.dir > > > (A) If a process holding a shared mutex dies, other processes' ops > on the mutex can hang/fail: Robust mutex support was added in a2ac10107e2fb845c4a38a339239063ec4407d84 in 2014. > > rm -rf lock.dir; mkdir lock.dir > servers/slapd/slapd -f lock.conf -h ldapi://ldapi -d0 & > MDB_KILL_R=true servers/slapd/slapd -Tcat -f lock.conf > ldapsearch -xLLH ldapi://ldapi -b o=hi l=x 1.1 > ^C (kills ldapsearch which is waiting for results) > kill %% (fails to kill slapd) > kill -KILL %% > > The Posix fix, robust mutexes, is outlined below. With _WIN32, > check return code WAIT_ABANDONED. __APPLE__ Posix semaphores: > Don't know. SysV semaphores have SEM_UNDO to cancel the process' > effect if the process dies, but the peers do not seem to be > informed that this happened. > > if ((rc = pthread_mutexattr_init( &mattr )) || > (rc = pthread_mutexattr_setpshared(&mattr, PTHREAD_PROCESS_SHARED)) || > (rc = pthread_mutexattr_setrobust( &mattr, PTHREAD_MUTEX_ROBUST )) || > (rc = pthread_mutex_init(&mutex, &mattr ))) > return rc; > ... > switch (pthread_mutex_lock(&mutex)) { > case 0: > break; > case EOWNERDEAD: > Repair the possibly half-updated data protected by <mutex>; > if (successful repair) { > pthread_mutex_consistent(&mutex); > break; > } > /* With Posix mutexes, make future use of <mutex> return failure: */ > pthread_mutex_unlock(&mutex); > /* FALLTHRU */ > default: > return MDB_PANIC; > } > ... > pthread_mutex_unlock(&mutex); > > > BTW, the above also happens with MDB_KILL_W=true slapcat followed > by ldapadd, dunno why slapcat uses a txn without MDB_TXN_RDONLY. slapcat already opens the env RDONLY. In 3e47e825fd61e1bff001c01f40d70263c89ffabd in 2012 was patched to also create the txn RDONLY. > In this case, raise(SIGINT) is sufficient. With MDB_KILL_R, that > does not kill slapd and we need SIGKILL - it catches SIGINT I > guess. Don't know if that difference is intentional. > > > (B1) Repeated kill -KILL slap<tool> while slapd is running can use > up the reader table since it does not clear its table slots, > making the db unusable. > > (B2) slapcat report no error after it fails to read the db due to > full reader table, it exits with success status. Unable to reproduce, dunno when this was fixed but it returns an error code now. > > (B3) After one "kill -KILL slap<tool>" while slapd is running, I > imagine the stale slot which does not go away can prevent freelist > reuse so the db grows quickly. But I have not seen that. Depends on timing I suppose; it's possible that there was no active read txn at the moment, in which case the stale reader slot has no effect on page reuse. > > bash$ rm -rf lock.dir; mkdir lock.dir > bash$ servers/slapd/slapd -f lock.conf -h ldapi://ldapi -d0 & > bash$ for (( i=0; i<130; i++)) { > MDB_KILL_UR=true servers/slapd/slapd -Tcat -f lock.conf > } > /dev/null > bash$ servers/slapd/slapd -Tcat -f lock.conf -d-1 2>cat.out > (success result, no output, no errors in the log) > bash$ ldapsearch -xLLH ldapi://ldapi -b o=hi l=x 1.1 > version: 1 > > [R..RLOCKED..R1] Other (e.g., implementation specific) error (80) > Additional information: internal error > bash$ > > Maybe it's easier to have a per-process reader table after all, > and a shared table with just the oldest txn per process. > > Or, each transaction chould have an exclusive file region lock. > A writer could try to grab the oldest txn's region lock if that > belongs to another process, and clear the txn it if successful. > A reader, before giving up, would just search for another pid's > txn whose lock it can grab. The current code uses an fcntl lock to detect process liveness. > > Except since the system reuses pids, the ID would either have to > be more exclusive (e.g. pid + mdb_env_open() time) or mdb_env_open > must clear away reader table entries with its own pid. > > Speaking of which, I don't see the use of the repeated getpid() > calls. Should be enough to store the pid in the MDB_env. fcntl > record locks do not survive fork(), so an MDB_env doesn't either. > And the thread ID is unsed, and must be reset with memset, not > 'mr_tid = 0' since pthread_t can be a struct. > The above is no longer true. > > (C) Mentioned earlier: Startup can fail due to a race condition, > when another process starts MDB at the same time and then aborts: > - Process 1 starts MDB, gets write lock in mdb_env_setup_locks(). > - Process 2 starts MDB, awaits read lock since write lock is taken. > - Process 1 dies. > - Process 2 gets the read lock and proceeds, but with an uninitialized > lockfile since it expects that process 1 initialized it. Probably still true, but already doc'd in Caveats. > > Fix: An extra exclusive lock around the current lock setup code. > > kill slapd process, if any > rm -rf lock.dir; mkdir lock.dir > MDB_RACE=true servers/slapd/slapd -Tcat -f lock.conf & sleep 1; \ > servers/slapd/slapd -Tcat -f lock.conf > > In addition, the above dumps core in mdb_env_close: > for (i=0; i<env->me_txns->mti_numreaders; i++) > where env->me_txns == (MDB_txninfo *) 0xffffffffffffffff. > > > (D) And of course ^Z when holding a mutex can block other processes' > threads, but I suppose BDB has the same problem. > > DB_SUSPEND=true servers/slapd/slapd -Tadd -f lock.conf < /dev/null > servers/slapd/slapd -Tadd -f lock.conf < /dev/null > (hanging...) ^C > kill %% > > The fix I can see for now is "don't do that" in the doc. MDB > could in addition use pthread_mutex_timedlock() so slapd at least > can keep running and can be killed without kill -KILL. Final fix > = a lock-free library, when a useful portable one emerges. At this point the only potential action I see is to add a check for MDB_READERS_FULL as noted at the beginning of this reply. -- You are receiving this mail because: You are on the CC list for the issue.

1 0

[Issue 9565] New: radlib.h: No such file or directory - passwd slapd-module
by openldap-its＠openldap.org 29 May '21

29 May '21

https://bugs.openldap.org/show_bug.cgi?id=9565 Issue ID: 9565 Summary: radlib.h: No such file or directory - passwd slapd-module Product: OpenLDAP Version: 2.5.4 Hardware: All OS: Linux Status: UNCONFIRMED Severity: normal Priority: --- Component: contrib Assignee: bugs(a)openldap.org Reporter: rguichardspam(a)Gmail.com Target Milestone: --- Hi, while compile passwd slapd-modules, I've got error on Radius compilation. gcc -DOPENLDAP_FD_SETSIZE=4096 -O2 -g -DSLAP_SCHEMA_EXPOSE -g -O2 -I/usr/kerberos/include -I../../../include -I../../../include -I../../../servers/slapd -c radius.c -fPIC -DPIC -o .libs/radius.o radius.c:27:20: fatal error: radlib.h: No such file or directory #include <radlib.h> I've searched which package could provide the radlib file but I didn't found it. Please confirm which package should be used .. or link to the source.. I'm on CentOS -- You are receiving this mail because: You are on the CC list for the issue.

1 2

[Issue 9567] New: Double free error after attempt to insert a duplicate entry
by openldap-its＠openldap.org 29 May '21

29 May '21

https://bugs.openldap.org/show_bug.cgi?id=9567 Issue ID: 9567 Summary: Double free error after attempt to insert a duplicate entry Product: LMDB Version: 0.9.17 Hardware: All OS: All Status: UNCONFIRMED Severity: normal Priority: --- Component: liblmdb Assignee: bugs(a)openldap.org Reporter: wietse(a)porcupine.org Target Milestone: --- This problem was originally discovered by Adi Prasaja on Ubuntu with lmdb-0.9.24, and reproduced by me on Fedora with lmdb-0.9.25, as well as multiple source repo branches from github.com/openldap. The problem was introduced with lmdb version 0.9.17 and still exists in the 'head' version 0.9.70 as retrieved from github.com on May 28, 2021. Quick demo (any Postfix version with LMDB support): create a key-value store from a file containing one (key, value) per line. $ cat /tmp/aa aa 1 aa 2 $ postmap lmdb:/tmp/aa postmap: warning: lmdb:/tmp/aa: duplicate entry: "aa" <== Postfix message free(): double free detected in tcache 2 <== libc message Aborted (core dumped) Commenting out the free(env->me_txn0) call mdb_env_close0() will prevent the double free() error. Of course, that results in a memory leak when the input contains no duplicate line. The remainder of this message show the steps to reproduce this with lmdb version 0.9.70 (from github 'head') on Fedora 32 with postfix-3.7-20210424. These steps have been verified to also work for Postfix 3.2. The steps assume that the LMDB library and header files are installed in /usr/local, by using the default lmdb Makefile. $ cat >makemakefiles-lmdb-local <<'EOF' #!/bin/sh export OPT="" export CCARGS="-DNO_NIS" export AUXLIBS="" # Add LMDB export CCARGS="$CCARGS -DHAS_LMDB" export AUXLIBS_LMDB="-L/usr/local/lib -Wl,-rpath,/usr/local/lib -llmdb" unset shlib_directory make -f Makefile.in makefiles shared=no dynamicmaps=no CCARGS="-I/usr/local/include $CCARGS" AUXLIBS="$AUXLIBS $AUXLIBS_LMDB" EOF $ make tidy [some output] $ sh makemakefiles-lmdb-local $ make -j8 [lots of output] $ cat >/tmp/aa << EOF aa 1 aa 2 EOF $ valgrind --tool=memcheck bin/postmap lmdb:/tmp/aa ==340318== Memcheck, a memory error detector ==340318== Copyright (C) 2002-2017, and GNU GPL'd, by Julian Seward et al. ==340318== Using Valgrind-3.16.1 and LibVEX; rerun with -h for copyright info ==340318== Command: bin/postmap lmdb:/tmp/aa ==340318== postmap: warning: lmdb:/tmp/aa: duplicate entry: "aa" ==340318== Invalid free() / delete / delete[] / realloc() ==340318== at 0x483B9F5: free (vg_replace_malloc.c:538) ==340318== by 0x4858CFD: mdb_env_close0 (in /usr/local/lib/liblmdb.so) ==340318== by 0x4859B0C: mdb_env_close (in /usr/local/lib/liblmdb.so) ==340318== by 0x13CBC3: slmdb_close (slmdb.c:780) ==340318== by 0x13B297: dict_lmdb_close (dict_lmdb.c:475) ==340318== by 0x113DFF: mkmap_close (mkmap_open.c:211) ==340318== by 0x10F080: postmap (postmap.c:557) ==340318== by 0x1108E4: main (postmap.c:1140) ==340318== Address 0x7320c30 is 0 bytes inside a block of size 258 free'd ==340318== at 0x483B9F5: free (vg_replace_malloc.c:538) ==340318== by 0x48561DE: mdb_txn_commit (mdb.c:4130) ==340318== by 0x13CB7D: slmdb_close (slmdb.c:771) ==340318== by 0x13B297: dict_lmdb_close (dict_lmdb.c:475) ==340318== by 0x113DFF: mkmap_close (mkmap_open.c:211) ==340318== by 0x10F080: postmap (postmap.c:557) ==340318== by 0x1108E4: main (postmap.c:1140) ==340318== Block was alloc'd at ==340318== at 0x483CAE9: calloc (vg_replace_malloc.c:760) ==340318== by 0x48599F2: mdb_env_open (in /usr/local/lib/liblmdb.so) ==340318== by 0x13CD91: slmdb_open (slmdb.c:851) ==340318== by 0x13B6C6: dict_lmdb_open (dict_lmdb.c:612) ==340318== by 0x113F50: mkmap_open (mkmap_open.c:283) ==340318== by 0x10EC02: postmap (postmap.c:438) ==340318== by 0x1108E4: main (postmap.c:1140) ==340318== ==340318== ==340318== HEAP SUMMARY: ==340318== in use at exit: 27,608 bytes in 634 blocks ==340318== total heap usage: 1,430 allocs, 797 frees, 3,356,293 bytes allocated ==340318== ==340318== LEAK SUMMARY: ==340318== definitely lost: 0 bytes in 0 blocks ==340318== indirectly lost: 0 bytes in 0 blocks ==340318== possibly lost: 27,582 bytes in 633 blocks ==340318== still reachable: 26 bytes in 1 blocks ==340318== suppressed: 0 bytes in 0 blocks ==340318== Rerun with --leak-check=full to see details of leaked memory ==340318== ==340318== For lists of detected and suppressed errors, rerun with: -s ==340318== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 0 from 0) Similar errors happen with lmdb-0.9.17. The problem does not exist in lmdb-0.9.16 or earlier versions. -- You are receiving this mail because: You are on the CC list for the issue.

1 3

[Issue 9564] New: Race condition with freeing the spilled pages from transaction
by openldap-its＠openldap.org 27 May '21

27 May '21

https://bugs.openldap.org/show_bug.cgi?id=9564 Issue ID: 9564 Summary: Race condition with freeing the spilled pages from transaction Product: LMDB Version: 0.9.29 Hardware: Other OS: Mac OS Status: UNCONFIRMED Severity: normal Priority: --- Component: liblmdb Assignee: bugs(a)openldap.org Reporter: kriszyp(a)gmail.com Target Milestone: --- Created attachment 825 --> https://bugs.openldap.org/attachment.cgi?id=825&action=edit Free the spilled pages and dirty overflows before unlocking the write mutex The spilled pages (a transaction's mt_spill_pgs) is freed *after* a write transaction's mutex is unlocked (in mdb.master3). This can result in a race condition where a second transaction can start and subsequently assign a new mt_spill_pgs list to the transaction structure that it reuses. If this occurs after the first transaction unlocks the mutex, but before it performs the free operation on mt_spill_pgs, then the first transaction will end up calling a free on the same spilled page list as the second transaction, resulting in a double free (and crash). It would seem to be an extremely unlikely scenario to actually happen, since the free call is literally the next operation after the mutex is unlocked, and the second transaction would need to make it all the way to the point of saving the freelist before a page spill list is likely to be allocated. Consequently, this probably has rarely happened. However, one of our users (see https://github.com/DoctorEvidence/lmdb-store/issues/48 for the discussion) has noticed this occurring, and it seems that it may be particularly likely to happen on MacOS on the new M1 silicon. Perhaps there is some peculiarity to how the threads are more likely to yield execution after a mutex unlock, I am not sure. I was able to reproduce the issue by intentionally manipulating the timing (sleeping before the free) to verify that the race condition is technically feasible, and apparently this can happen "in the wild" on MacOS, at least with an M1. It is also worth noting that this is due to (or a regression from) the fix for ITS#9155 (https://github.com/LMDB/lmdb/commit/cb256f409bb53efeda4ac69ee8165a0b4fc1a277) where the free call was moved outside the conditional (for having a parent) that had previously never been executed after the mutex is unlocked, but now is called after the unlock. Anyway, the solution is relatively simple, the free call simply has to be moved above the unlock. In my patch, I also moved the free call for mt_dirty_ovs. I am not sure what OVERFLOW_NOTYET/mt_dirty_ovs is for, but presumably it should be handled the same. This could alternately be solved by saving the reference to these lists before unlocking, and freeing after unlocking, which would slightly decrease the amount of processing within the mutex guarded code. Let me know if you prefer a patch that does that. -- You are receiving this mail because: You are on the CC list for the issue.

1 1

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