| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
| |
NSS 3.11 introduces a new library (libfreebl3.so) that is loaded as part of NSS initialization. With Fedora DS 1.0, we moved NSS initialization to occur after the setuid from root to the runtime uid so that the files created during NSS init would have the correct ownership. However, the bin/slapd/server directory is set to 0700 meaning no execute permission for the runtime uid. The OS requires this directory to be 711 to allow the slapd process to load in the shared libraries needed by NSS. We use 711 to disallow reading in this directory because if slapd crashes shortly after startup, a core file may go in this directory which may contain secret information.
|
|
|
|
| |
2) Allow moz objdir name to be passed in
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bug Description: 64bit issues with normalize_path(), make_dn(), and
add_aci_v()
Reviewed by: Nathan and Noriko (Thanks!)
Fix Description: 1) use sizeof(char *) as the elemsize to pass to
slapi_ch_calloc(). 2) create a variable const char *NULLSTR = 0 to pass
as the last argument to the varargs functions.
Platforms tested: RHEL4 64
Flag Day: no
Doc impact: no
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
RHEL4_x86_64_gcc3_OPT.OBJ
The makefiles were assuming that if the string 86 was found in uname -m
it should use x86 for the arch. However, it should only do this if
uname -m is not x86_64. Also, for RHEL/Linux, we do not have to add the
NS64TAG just before the _OPT (or _DBG) since it's already in the arch.
Other platforms continue to use the other naming convention. The
default naming convention for the mozilla components is
Linux2.6_x86_64_glibc_PTH_64_OPT.OBJ. I considered using Wan-Teh's
instructions about source builds, but that would cause many changes to
be made to our makefiles, so I just changed the way we calculate the
mozilla OBJDIR name from the regular OBJDIR name. These changes are
pretty much the same for adminutil, setuputil, adminserver, and ldapserver.
For ldapserver, I had to change nsarch (adminserver changed nsarch in a
different way). I also got rid of the 32 bit specific setup stuff. I
assume this was due to 32 bit NES admin server, so we can get rid of
this for the future, because we will have all native 64 bit apps. I also
went ahead and rolled in the gcc4 changes since they are safe for gcc3
as well.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bug Description: Password Modify LDAPv3 extended operation erroneously
forces the client to supply old password
Reviewed by: Pete & Nathan (Thanks!)
Fix Description: If the BIND operation was successful, the CONN_DN field
is always set to the proper DN. This is even the case during a SASL or
client cert DN if the authentication was successful AND the given
identity could be mapped to a real user in the directory. Also, the
authmethod will be something other than NULL or none. So, if the old
password was not given, that is ok if there is a non-anonymous bind DN
and a real authmethod. The rest of the operation passes through the usual access control.
Platforms tested: Fedora Core 4
Flag Day: no
Doc impact: no
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bug Description: Most of our integer based config parameters do not do a good job
validating input. For example, alphabetic characters will be accepted as a value
if "0" is a valid setting for a given parameter since atoi returns "0" on error. We
also don't properly check for range errors for many parameters.
Reviewed by: Rich, Pete, and Noriko (Thanks!)
Files: ldapserver/ldap/servers/slapd/libglobs.c
Branch: HEAD
Fix Description: This adds better input validation for a number of integer based config
settings. I replaced our usage of atoi() with strtol() so we can recognize
conversion errors. I also added more descriptive error strings where they were
missing. I revised the fix due to feedback from Noriko. She recommended that
we check for ERANGE areas for all calls to strtol insead of only checking when
LONG_MAX is a valid setting for a config parameter.
Flag Day: no
Doc impact: no
QA impact: should be covered by regular nightly and manual testing
New Tests integrated into TET: none
|
|
|
|
| |
_instead of_ libfreebl3, not _in addition to_. So, set the list of platform specific freebl libs and add them to the list of libs to package.
|
|
|
|
|
| |
and more on Solaris and HP-UX.
Use setup.inf from admin server, and make sure the admin component is in there.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bug Description: DSMLGW code uses non-standard sun.misc.Base64Encoder
Reviewed by: The team (Thanks!)
Fix Description: Added jakarta-commons-codec.jar to the DS build and
packaging. We will need to add this file to the bundle of dsmlgw jar
files on /share/builds/components.
Platforms tested: Fedora Core 4
Flag Day: no
Doc impact: no
QA impact: should be covered by regular nightly and manual testing
New Tests integrated into TET: none
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bug Description: If you try to add one of the new password policy attributes to a fine-grained
policy, you get an objectclass violation error.
Reviewed by: Rich, Noriko (thanks!)
Files: ldapserver/ldap/schema/00core.ldif
Branch: HEAD
Fix Description: This adds new password policy config attributes need to the schema and updates
the "passwordPolicy" objectclass needs to allow them. I've also updated our master schema document
to keep track of the OIDs.
Flag Day: no
Doc impact: no
QA impact: should be covered by regular nightly and manual testing
New Tests integrated into TET: none
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bug Description: Chain BIND requests with chain on update + global pw policy
Reviewed by: Nathan, Pete (Thanks!)
Fix Description: When using global password policy, we need to chain the
BIND request back to a master so that it can update the pw policy op
attrs in the user's entry and replicate them to all other servers. The
call to config_get_pw_is_global_policy() is expensive (acquires a lock)
so we delay it as long as possible.
The reason we have to use config_get_pw_is_global_policy() is because the entry distribution plugin interface is poor - we have no way to register an init or start function to get the config, and no way to register a statechange callback to be notified of changes to the global password policy.
Platforms tested: Fedora Core 4
Flag Day: no
Doc impact: no
|
|
|
|
|
|
|
|
|
|
|
|
| |
Bug Description: crash in PTA plugin when bind returned controls
Reviewed by: Pete (Thanks!)
Fix Description: Fix by Ulf. We are passing the ***LDAPControl when it
needs the **LDAPControl.
Platforms tested: Fedora Core 4
Flag Day: no
Doc impact: no
QA impact: should be covered by regular nightly and manual testing
New Tests integrated into TET: none
|
|
|
|
|
| |
Fixed the broken build (64-bit).
LDFLAG (set in the tools/Makefile) was not passed to rsearch and ldclt.
|
|
|
|
|
|
| |
Fixed broken build on Solaris.
1) removed redeclared getsubopt in port.h
2) defined _XOPEN_SOURCE 500 for Linux
|
| |
|
| |
|
|
|
|
| |
Checking in Ulf's proposal; should not get connid from pid if the op is internal. (plus the value is just for debug print which is disabled now)
|
| |
|
|
|
|
|
|
|
| |
backup; Restore tries to restore all the fles from the backupdir
1) if the specified backup directory exists, rename it to <dir>.bak
2) support relative dir for the backup/restore dir, e.g., "bak2db bak/mybak"
|
| |
|
|
|
|
| |
integrated ldclt from DSRK into the DS source tree.
|
|
|
|
|
|
|
| |
If the specified backup dir does not exist, does not contain necessary files
(including the directory is not accessible), is not a directory, or is
identical to the path to the db dir, issues an error and stops restoring before
wiping out the database.
|
| |
|
| |
|
|
|
|
| |
Integrating rsearch and infadd source codes from DSRK into the Directory Server build tree.
|
| |
|
| |
|
|
|
|
| |
greater than or equal to 1
|
| |
|
| |
|
|
|
|
| |
It used to be taken to the "convert to quoted value" section even though the rdn did not have a quote.
|
| |
|
|
|
|
|
|
|
|
|
| |
Bug Description: The dsgw cookie directory needs to be writable by the admin server uid
Reviewed by: Nathan (Thanks!)
Fix Description: DS Gateway authentication breaks because the admin server uid cannot write to the bin/slapd/authck directory. This fix makes sure that directory is owned by the correct uid. I've also put a similar fix into the ds spec file %post section to fix this when upgrading from fds10 to fds101.
Platforms tested: Fedora Core 4
Flag Day: no
Doc impact: no
|
|
|
|
|
|
|
|
|
| |
Bug Description: The BUILD_BOMB is on by default
Reviewed by: Nathan (Thanks!)
Fix Description: Move the pumpkin and build bomb stuff into the internal makefiles. If you want the build bomb on, you have to set BUILD_PUMPKIN=1 on the make command line. The diff also contains some of the versioning for 1.0 to 1.0.1.
Platforms tested: Fedora Core 4
Flag Day: no
Doc impact: no
|
|
|
|
|
| |
If missing entry_id is found, skip it and continue reading the primary db to be
reindexed.
|
| |
|
|
|
|
| |
When reading a rotation info file, compare the contents with the files in the logs dir, if mismatch is found, update the rotation info file. Once the files arestored in the rotation info file, they will be the target of cleanup at the nextlog ration.
|
| |
|
|
|
|
| |
need to set localuser before creating/opening error log.
|
|
|
|
|
| |
Modified to change the owner to the "localuser" if the error log file is not
owned by the user.
|
|
|
|
|
|
|
|
|
|
| |
the contents. I'm not sure why we haven't caught this earlier, but
I believe it has something to do with the patch to make ds build on
Fedora Core 4 with gcc4. To do that, we turn off the -fwriteable-strings
argument to gcc. I suppose with it on, it moves those strings to
some sort of writeable memory location. With it off, constant strings
are definitely in the data section. There was one place in views that
used a constant string, and a couple of places in the windows sync code.
|
|
|
|
|
|
| |
NSS/SSL init. For example, import needs to hash passwords, export of encrypted attrs needs encryption.
2) Only create, configure (for SSL) and bind TCP ports if running in regular or referral mode. Before, the code short circuited if doing import, export, etc. before getting to the port stuff. But since 1) above, the code needs to take care only to do network related stuff if in network mode.
|
| |
|
|
|
|
|
| |
related files are owned by the correct user, but make that happen before
the detach so we can ask for the pin on the terminal.
|
| |
|
|
|
|
| |
Fixed to get the adminid from the right place: adminpw
|
|
|
|
|
|
| |
was due to the security CGI not being able to read the ds key/cert db files. They were owned by root instead of the server uid because they were being created by NSS_Initialize which was being called before the server did the setuid.
The fix is to move the NSS/SSL initialization code to just after the setuid call.
|
|
|
|
|
| |
1) setup was not using the correct value for the config admin ID during reconfig. The fix is to get the correct uid from the admpw file.
2) It is often difficult to figure out the correct id/url for the admin server when running it for the first time. The fix is for setup to print out the correct command to use so the user can copy/paste it into the terminal. The commands are also saved to setup.log in case they close the terminal window.
|
|
|
|
| |
tab. The solution is to chown the alias directory to the admin server uid so that the security CGI can create the key/cert dbs in that directory.
|