LDAP: Adding SIGCHLD callback This patch adds SIGCHLD callback for ldap_child. So if timeout is reached and ldap_child is terminated by handler we have debug message about it. Resolves: https://fedorahosted.org/sssd/ticket/3106 Reviewed-by: Pavel Březina <pbrezina@redhat.com> Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>

LDAP: Adding SIGTERM signal before SIGKILL We add better termination of ldap_child. If ldap_child reaches the timeout for termination parent sents SIGTERM signal. Child has 2 seconds for removing temporary file and exit. If it is not sufficient there is SIGKILL send to the child. Resolves: https://fedorahosted.org/sssd/ticket/3106 Reviewed-by: Pavel Březina <pbrezina@redhat.com> Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>

Do not leak fds in case of failures setting up a child process Resolves: https://fedorahosted.org/sssd/ticket/3006 The handling of open pipes in failure cases was suboptimal. Moreover, the faulty logic was copied all over the place. This patch introduces helper macros to: - initialize the pipe endpoints to -1 - close an open pipe fd and set it to -1 afterwards - close both ends unless already closed These macros are used in the child handling code. The patch also uses child_io_destructor in the p11_child code for safer fd handling. Reviewed-by: Petr Cech <pcech@redhat.com>

UTIL: exit() the forked process if exec()-ing a child process fails When exec() fails, we should not attempt to continue, but just kill the forked process. The patch adds this logic to the exec_child() and exec_child_ex() functions to avoid code duplication Resolves: https://fedorahosted.org/sssd/ticket/3016 Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>

sssd: incorrect checks on length values during packet decoding https://fedorahosted.org/sssd/ticket/1697 It is safer to isolate the checked (unknown/untrusted) value on the left hand side in the conditions to avoid overflows/underflows. Reviewed-by: Petr Cech <pcech@redhat.com>

UTIL: Unify the fd_nonblocking implementation The responder and child_common modules each had their own implementation. Unify it instead and add a unit test. Reviewed-by: Pavel Březina <pbrezina@redhat.com>

LDAP: Use child_io_destructor instead of child_cleanup in a custom desctructor ldap_child was the only child process that used child_cleanup instead of the common child_io_destructor. Unify the implementation to use the common function instead. Reviewed-by: Pavel Březina <pbrezina@redhat.com>

UTIL: Allow dup-ing child pipe to a different FD Related to: https://fedorahosted.org/sssd/ticket/2544 Adds a new function exec_child_ex and moves setting the extra_argv[] to exec_child_ex() along with specifying the input and output fds. Reviewed-by: Sumit Bose <sbose@redhat.com>

Add extra_args to exec_child() Related: https://fedorahosted.org/sssd/ticket/2503 Currently all child processes use the same arguments, the construction of argv[] is even hardcoded in exec_child(). Add an extra_args[] array that extends the common set of argvs so that we can have child-specific arguments. Also adds a unit test. Reviewed-by: Sumit Bose <sbose@redhat.com>

UTIL: Remove more code duplication setting up child processes All our child processes duplicated the same code that set up the debugging all around. Instead of adding yet another copy for the selinux_child, add a common utility function. Reviewed-by: Michal Židek <mzidek@redhat.com>

LDAP: Drop privileges after kinit in ldap_child After ldap_child initializes privileges using root-owned keytab, it drops privileges to the SSSD user, minimizing the amount of code that runs as root. Reviewed-by: Michal Židek <mzidek@redhat.com>

Update DEBUG* invocations to use new levels Use a script to update DEBUG* macro invocations, which use literal numbers for levels, to use bitmask macros instead: grep -rl --include '*.[hc]' DEBUG . | while read f; do mv "$f"{,.orig} perl -e 'use strict; use File::Slurp; my @map=qw" SSSDBG_FATAL_FAILURE SSSDBG_CRIT_FAILURE SSSDBG_OP_FAILURE SSSDBG_MINOR_FAILURE SSSDBG_CONF_SETTINGS SSSDBG_FUNC_DATA SSSDBG_TRACE_FUNC SSSDBG_TRACE_LIBS SSSDBG_TRACE_INTERNAL SSSDBG_TRACE_ALL "; my $text=read_file(\*STDIN); my $repl; $text=~s/ ^ ( .* \b (DEBUG|DEBUG_PAM_DATA|DEBUG_GR_MEM) \s* \(\s* )( [0-9] )( \s*, ) ( \s* ) ( .* ) $ / $repl = $1.$map[$3].$4.$5.$6, length($repl) <= 80 ? $repl : $1.$map[$3].$4."\n".(" " x length($1)).$6 /xmge; print $text; ' < "$f.orig" > "$f" rm "$f.orig" done Reviewed-by: Jakub Hrozek <jhrozek@redhat.com> Reviewed-by: Stephen Gallagher <sgallagh@redhat.com> Reviewed-by: Simo Sorce <simo@redhat.com>

Make DEBUG macro invocations variadic Use a script to update DEBUG macro invocations to use it as a variadic macro, supplying format string and its arguments directly, instead of wrapping them in parens. This script was used to update the code: grep -rwl --include '*.[hc]' DEBUG . | while read f; do mv "$f"{,.orig} perl -e \ 'use strict; use File::Slurp; my $text=read_file(\*STDIN); $text=~s#(\bDEBUG\s*\([^(]+)\((.*?)\)\s*\)\s*;#$1$2);#gs; print $text;' < "$f.orig" > "$f" rm "$f.orig" done Reviewed-by: Jakub Hrozek <jhrozek@redhat.com> Reviewed-by: Stephen Gallagher <sgallagh@redhat.com> Reviewed-by: Simo Sorce <simo@redhat.com>

Fix formating of variables with type: size_t

SIGCHLD handler: do not call callback when pvt data where freed https://fedorahosted.org/sssd/ticket/1992

Remove unneeded parameter of setup_child and namespace it setup_child() was accepting a parameter it didn't use. Also the function name was too generic, so I added a sdap prefix.

More generalized function open_debug_file_ex() Function open_debug_file_ex() set flag FD_CLOEXEC to opened file according to the value of third parameter. Removed duplicity of unsetting FD_CLOEXEC after calling function open_debug_file_ex()

Collect krb5 trace on high debug levels If the debug level contains SSSDBG_TRACE_ALL, then the logs would also include tracing information from libkrb5. https://fedorahosted.org/sssd/ticket/1539

Remove unused variable

remove left over principal selection https://fedorahosted.org/sssd/ticket/1303 Domain start up was taking too long when there are many principals in a kerberos keytab. We were looking up in the keytab two times. The first time we try to select a proper principal and remember it. The second call happens almost right after the first one and it is just a check if the principal exists in the keytab, without any output information other than success/failure. It is probably a left over from https://fedorahosted.org/sssd/ticket/781. This patch removes the second call.

execv, excvp and exec_child never return EOK * So don't need to handle that case

Move child_common routines to util

Multiline macro cleanup This is mostly a cosmetic patch. The purpose of wrapping a multi-line macro in a do { } while(0) is to make the macro usable as a regular statement, not a compound statement. When the while(0) is terminated with a semicolon, the do { } while(0); block becomes a compound statement again.

Modify principal selection for keytab authentication Currently we construct the principal as host/fqdn@REALM. The problem with this is that this principal doesn't have to be in the keytab. In that case the provider fails to start. It is better to scan the keytab and find the most suitable principal to use. Only in case no suitable principal is found the backend should fail to start. The second issue solved by this patch is that the realm we are authenticating the machine to can be in general different from the realm our users are part of (in case of cross Kerberos trust). The patch adds new configuration option SDAP_SASL_REALM. https://fedorahosted.org/sssd/ticket/781

Make ldap_child report kerberos return code to parent

Validate keytab at startup In addition to validating the keytab everytime a TGT is requested, we also validate the keytab on back end startup to give early warning that the keytab is not usable. Fixes: #556

GSSAPI ticket expiry time is returned from ldap_child and stored in sdap_handle for future reference.

Fix realm_str dereference Fixes: #508

Add ldap_krb5_ticket_lifetime option

Properly set up SIGCHLD handlers Instead of having all-purpose SIGCHLD handlers that try to catch every occurrence, we instead create a per-PID handler. This will allow us to specify callbacks to occur when certain children exit.

Improve safe alignment buffer handling macros Make the counter optional so that alignment safe macros can be used also where there is no counter to update. Change arguments names so that they are not deceiving (ptr normlly identify a pointer) Turn the memcpy substitute into an inline function so that passing a pointer to rp and checking for it doesn't make the compiler spit lots of warnings.

Rename server/ directory to src/ Also update BUILD.txt