Skip to content

[RFC] Use GNULIB's compiler warning code#50

Closed
fidencio wants to merge 10 commits into
SSSD:masterfrom
fidencio:wip/manywarnings
Closed

[RFC] Use GNULIB's compiler warning code#50
fidencio wants to merge 10 commits into
SSSD:masterfrom
fidencio:wip/manywarnings

Conversation

@fidencio

@fidencio fidencio commented Oct 17, 2016

Copy link
Copy Markdown
Contributor

This patch series was sent to the sssd-devel and some discussions
already happened there0. I've decided to open the PR because there
are some few patches that can be pushed even if we decide to not use
the "many warnings" patches.

Let's keep track of those patches (and discussions related to them) in
the github, in this way we can avoid them to get lost. :-)

Best Regards,

Changes:

683f72d (Fabiano Fidêncio, 10 weeks ago)
BUILD: Make use of GNULIB's compiler warning code

As GNULIB has the 'manywarnings' module, which basically turns on every GCC
warning, let's make use of it. We can easily blacklist the warnings we
cannot cope with, but the main goal should be to have enabled every
possible GCC warning.

When new GCC warnings are created the 'manywarnings' file can be refreshed
from upstream GNULIB.

Signed-off-by: Fabiano Fidêncio [email protected]

f59828a (Fabiano Fidêncio, 5 days ago)
NSS: Fix "old-style-definition" warning caught by GCC

Signed-off-by: Fabiano Fidêncio [email protected]

f22aff7 (Fabiano Fidêncio, 5 days ago)
SIFP: Fix a "jump-misses-init" warning caught by GCC

Signed-off-by: Fabiano Fidêncio [email protected]

58609d3 (Fabiano Fidêncio, 5 days ago)
RESOLV: Fix a "-Werror=null-dereference" caught by GCC

Signed-off-by: Fabiano Fidêncio [email protected]

bd1d7fd (Fabiano Fidêncio, 5 days ago)
RESOLV: Simplify reply_weight_rearrange() a little bit

Signed-off-by: Fabiano Fidêncio [email protected]

@lslebodn

lslebodn commented Oct 19, 2016

Copy link
Copy Markdown
Contributor

It seems that resolv patches broke resolv-tests on fedora 22 and fedora 23.
The test has been running for 30 hours without any result. I've just cancelled it.

BTW CI did not passed on any supported distro.

@fidencio

fidencio commented Nov 9, 2016

Copy link
Copy Markdown
Contributor Author

I've updated this series and I'm still waiting for CI to finish.
A lot of warnings were just caught when building on RHEL6 machine (which passes gracefully now).

I'll post the CI results as soon as I have them.

@fidencio

fidencio commented Nov 9, 2016

Copy link
Copy Markdown
Contributor Author

http://sssd-ci.duckdns.org/logs/job/56/78/summary.html

@lslebodn

lslebodn commented Nov 9, 2016

Copy link
Copy Markdown
Contributor

I do not think that initialisation to NULL is the best solution for Wunitialized on RHLE6.

My experience is that such bugs occurs when there is some suspicious code in functions which are
inlined.

lslebodn
lslebodn reviewed Nov 9, 2016
View reviewed changes
Comment thread src/providers/krb5/krb5_child.c
prompter, kr, 0,
SSSD_KRB5_CHANGEPW_PRINCIPAL,
discard_const(SSSD_KRB5_CHANGEPW_PRINCIPAL),
kr->options);

@lslebodn lslebodn Nov 9, 2016

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Such patch was rejected by simo approximately 3 years ago.
Because it was fixed in krb5.

@fidencio fidencio Nov 9, 2016

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Okay, NACK to this one.

lslebodn
lslebodn reviewed Nov 9, 2016
View reviewed changes
Comment thread src/tests/sbus_codegen_tests.c Outdated
START_TEST(test_signals)
{
const struct sbus_signal_meta *signal;
const struct sbus_signal_meta *sm_signal;

@lslebodn lslebodn Nov 9, 2016

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What does the prefix "sm_" mean?

@fidencio fidencio Nov 9, 2016

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

"sm_" comes from sbus_meta.

@lslebodn lslebodn Nov 9, 2016

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It is not sbus_meta_signal but sbus_signal_meta But ssm_ does not increase readability either

What about a_signal or sig, signl

@fidencio fidencio Nov 9, 2016

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Okay, I'll follow one of your suggestions.

lslebodn
lslebodn reviewed Nov 9, 2016
View reviewed changes
Comment thread src/tools/sssctl/sssctl_cache.c
#define SSSCTL_CACHE_UPDATE {_("Cache entry last update time"), SYSDB_LAST_UPDATE, attr_time_fn}
#define SSSCTL_CACHE_EXPIRE {_("Cache entry expiration time"), SYSDB_CACHE_EXPIRE, attr_expire_fn}
#define SSSCTL_CACHE_IFP {_("Cached in InfoPipe"), SYSDB_IFP_CACHED, attr_yesno_fn}
#define SSSCTL_CACHE_NULL {NULL, NULL, NULL}

@lslebodn lslebodn Nov 9, 2016

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suffix "_fn" in name of function looks weird to me.
I would suggest either rename function (get_attr_name ...) to have more suitable name
or rename local variables which shadow functions.

@fidencio fidencio Nov 9, 2016

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Okay, added the get_ prefix to these functions.

lslebodn
lslebodn reviewed Nov 9, 2016
View reviewed changes
Comment thread src/providers/ad/ad_gpo.c
/* If INI_PARSE_IGNORE_NON_KVP is not defined, use 0 (no effect) */
#ifndef INI_PARSE_IGNORE_NON_KVP
#define INI_PARSE_IGNORE_NON_KVP 0
#warning INI_PARSE_IGNORE_NON_KVP not defined.

@lslebodn lslebodn Nov 9, 2016

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

NACK,
this warning is intentional. It was added because sssd need to be compiled even with libini_config-1.2.
Because required the latest libini_config-1.3 is only in fedora and is required for
https://fedorahosted.org/sssd/ticket/2751

@fidencio fidencio Nov 9, 2016

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Fair enough ...

@fidencio

fidencio commented Nov 9, 2016

Copy link
Copy Markdown
Contributor Author

@lslebodn, a way easier solution would be to disable -Werror when running CI on RHEL6 than fixing those issues only caught there. Is there a simple way to do that?

lslebodn
lslebodn reviewed Nov 9, 2016
View reviewed changes
Comment thread src/tools/tools_util.c
}

fprintf(stderr, "%s%s", error, nl);
fprintf(stderr, "%s", error);

@lslebodn lslebodn Nov 9, 2016

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ACK

lslebodn
lslebodn reviewed Nov 9, 2016
View reviewed changes
Comment thread src/tests/krb5_child-test.c
DEBUG(SSSDBG_FATAL_FAILURE,
"Failed to create dummy pam_data");
goto fail;
}

@lslebodn lslebodn Nov 9, 2016

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM,

@lslebodn

lslebodn commented Nov 9, 2016

Copy link
Copy Markdown
Contributor

On (09/11/16 07:54), fidencio wrote:

@lslebodn, a way easier solution would be to disable -Werror when running CI on RHEL6 than fixing those issues only caught there. Is there a simple way to do that?

I agree that it would be better to disable Werror on rhel6.

But your patch "BUILD: Make use of GNULIB's compiler warning code "
introduced Werror. And I am still not satisfied with the patch
because of increased time in CI (10%) caused by detection in manywarnings.

LS

@fidencio

fidencio commented Nov 9, 2016

Copy link
Copy Markdown
Contributor Author

Disabling Werror on RHEL6 will be just a matter of passing "--enable-werror=no" to configure. It should be really simple as long as we agree on doing that.

About the extra extra time, it was 5% last time you reviewed. 100 seconds for a process that we already have to wait for an hour or so. IMO should not block, at all, introducing more warnings to our code.

@lslebodn

lslebodn commented Nov 9, 2016

Copy link
Copy Markdown
Contributor

On (09/11/16 08:18), fidencio wrote:

Disabling Werror on RHEL6 will be just a matter of passing "--enable-werror=no" to configure. It should be really simple as long as we agree on doing that.

About the extra extra time, it was 5% last time you reviewed. 100 seconds for a process that we already have to wait for an hour or so. IMO should not block, at all, introducing more warnings to our code.

I am not against enabling new warnings.
More warnings can be enabled in Makefile.am directly.
The warnings which are not supported by older compilers (gcc on rhel6)
should be explicitely checked at configure time.

It is wasting of time to check evey time flags which
does not produce any warnings atm:
e.g.
-Wabi
-Warray-bounds
-Wcast-align
-Wcast-qual
-Wclobbered
-Wdisabled-optimization
-Wempty-body
-Werror-implicit-function-declaration
-Wformat-security
-Wignored-qualifiers
-Winit-self
-Winvalid-pch
-Wlarger-than=65536
-Wmissing-include-dirs
-Wmissing-parameter-type
-Wmultichar
-Wnested-externs
-Wold-style-declaration
-Woverlength-strings
-Woverride-init
-Wpacked
-Wpointer-arith
-Wshadow
-Wstack-protector
-Wstrict-prototypes
-Wsync-nand
-Wtype-limits
-Wundef
-funsafe-loop-optimizations
-Wunsafe-loop-optimizations
-Wunused-but-set-parameter
-Wwrite-strings
-Wall

Or warnings which will never be fixed:
e.g.
-Wsystem-headers
-Wformat-nonliteral
-Wpacked
-Wvla
-Wtraditional

LS

@fidencio
TESTS: Fix "-Werror=null-dereference" caught by GCC
1baeeab 
Signed-off-by: Fabiano Fidêncio <[email protected]>
Reviewed-by: Lukáš Slebodník <[email protected]>
@fidencio
TOOLS: Fix "-Wstack-protector" caught by GCC
e5ad8cf 
This warning only happens when building SSSD on RHEL6.

Signed-off-by: Fabiano Fidêncio <[email protected]>
Reviewed-by: Lukáš Slebodník <[email protected]>
@fidencio
SSSCTL: Fix "-Wshadow" warning caught by GCC
862d10c 
This warning only happens when building SSSD on RHEL6.

Signed-off-by: Fabiano Fidêncio <[email protected]>
@fidencio
SSSCTL: Use get_ prefix for the sssctl_attr_fn functions
87aa549 
As done for the attr_name_fn to avoid "-Wshadow", let's be consistent
and do the same for all the other sssctl_attr_fn functions.

Signed-off-by: Fabiano Fidêncio <[email protected]>
@fidencio
TESTS: Fix "-Wshadow" caught by GCC
145d5b1 
This warning only happens when building SSSD on RHEL6.

Signed-off-by: Fabiano Fidêncio <[email protected]>
@fidencio

fidencio commented Nov 9, 2016

Copy link
Copy Markdown
Contributor Author

Lukáš,

I'm afraid you didn't understand how the usage of the GNULIB's compiler warnings is supposed to be.
There are basically 3 essential files:

  • warnings.m4
  • manywarnings.m4
  • sssd-compile-warnings.m4

The first two you can just get from GNULIB's git repo and update in our repo whenever there's a update there. So, tweaking those files is something I could do, but it's wrong per si.
There's the third file then, sssd-compile-warnings.m4, where we disable the warnings we don't need, want or can't cope with. And in this file I've added all the warnings we don't need, want or can't cope with. In this very same file, we check whether a warning is or is not supported.

That's how it's supposed to work. If you think patching Makefile.am is better, okay, just reject this patch series, open a bug and we will mark as "patches welcome".

Again, I really don't see a valid point for rejecting this series because it introduces 100s more in a process that takes an hour or so.

So, what's your suggestion? Can we accept the series? Can you, at least, finish reviewing the patches that are not RHEL6 specific?

@fidencio

fidencio commented Nov 9, 2016

Copy link
Copy Markdown
Contributor Author

@lslebodn:
After some proper checking, seems that all "-Wunitialized" reports are false positives, so I've removed the patches from this series.
I've added "Reviewed-by: " to the patches you've already ACKed and dropped the ones you've already NACKed. For the ones you gave some suggestions, I followed them.

The build is failing in CentOS CI and I'll take a look on it later on. I've updated my branch based only on my local tests (and they're okay on F25). I haven't done any test on our CI because it seems to be broken due to lack of space.

Feel free to wait till I solve those issues before starting reviewing. Or feel free to NACK the series as soon as possible, so I don't spend time on it anymore.

@fidencio
CI: Disable _FORTIFY_SOURCE when building with -O0
dc7c6e7 
This patch is based on cabc051.

Our coverage build uses -O0 but the macro _FORTIFY_SOURCE requires to be
compiled with optimization.
be compiled w

Signed-off-by: Fabiano Fidêncio <[email protected]>
@fidencio
BUILD: Make use of GNULIB's compiler warning code
e2a7403 
As GNULIB has the 'manywarnings' module, which basically turns on every
GCC warning, let's make use of it. We can easily blacklist the warnings
we cannot cope with, but the main goal should be to have enabled every
possible GCC warning.

When new GCC warnings are created the 'manywarnings' file can be
refreshed from upstream GNULIB.

Compilation with -Werror is explicitly disabled for RHEL6 in our CI, as
a lot of warnings show up there due to old dependencies or even as
false-positives.

Signed-off-by: Fabiano Fidêncio <[email protected]>
@fidencio

fidencio commented Nov 10, 2016 •
edited
Loading

Copy link
Copy Markdown
Contributor Author

So, here are some stats about the configure time that you complained before:
So, I've ran the contrib/ci/run script on our F25 machine on Jenkins.

This is what we have without GNULIB patch:

-bash-4.3$ ./contrib/ci/run
install-deps:           success  00:00:05 ci-install-deps.log
autoreconf:             success  00:00:22 ci-autoreconf.log
DEBUG BUILD:                              ci-build-debug
configure:              success  00:00:16 ci-build-debug/ci-configure.log
make-tests:             success  00:02:55 ci-build-debug/ci-make-tests.log
make-check-valgrind:    success  00:01:33 ci-build-debug/ci-make-check-valgrind.log
SUCCESS

And this is what we have with GNULIB patch applied:

-bash-4.3$ ./contrib/ci/run 
install-deps:           success  00:00:05 ci-install-deps.log
autoreconf:             success  00:00:23 ci-autoreconf.log
DEBUG BUILD:                              ci-build-debug
configure:              success  00:00:22 ci-build-debug/ci-configure.log
make-tests:             success  00:03:56 ci-build-debug/ci-make-tests.log
make-check-valgrind:    success  00:01:33 ci-build-debug/ci-make-check-valgrind.log
SUCCESS

It adds 6 seconds more to the configure time. I really think that dropping those patches because of 6 seconds on the configure is far, but really really, far away from being reasonable.

So, can we have the patches reviewed and merged?

fidencio added a commit to fidencio/sssd that referenced this pull request Dec 12, 2016
@fidencio
WATCHDOG: Avoid non async-signal-safe from the signal_handler
8072323 
While debugging rhbz#1396912 some deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

The removal of those functions we lose some debug messages, which is the
least of problems now. Also instead of re-setting the watchdog timeout
in case of time-shift (please, see b8ceaeb for more info), let's just
re-set the timer's time.

Along with this commit, the function to teardown the watchdog has been
removed as it's not used anymore.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    SSSD#1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    SSSD#2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    SSSD#3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    SSSD#4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    SSSD#5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    SSSD#6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
@fidencio fidencio mentioned this pull request Dec 12, 2016
Closed
fidencio added a commit to fidencio/sssd that referenced this pull request Dec 13, 2016
@fidencio
WATCHDOG: Avoid non async-signal-safe from the signal_handler
f82f300 
While debugging rhbz#1396912 some deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

The removal of those functions we lose some debug messages, which is the
least of problems now. Also instead of re-setting the watchdog timeout
in case of time-shift (please, see b8ceaeb for more info), let's just
exit the let the monitor restart the process.

A proper fix for the clock screw situation should be implemented on
samba's side, by having tevent using monotonic (or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    SSSD#1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    SSSD#2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    SSSD#3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    SSSD#4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    SSSD#5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    SSSD#6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
fidencio added a commit to fidencio/sssd that referenced this pull request Dec 20, 2016
@fidencio
WATCHDOG: Avoid non async-signal-safe from the signal_handler
8079834 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and a single byte is written to this pipe (which
is an async-signal-safe operation) and the logic currently done by the
timer handler will be done inside the fd handler in a safe way.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    SSSD#1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    SSSD#2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    SSSD#3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    SSSD#4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    SSSD#5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    SSSD#6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
fidencio added a commit to fidencio/sssd that referenced this pull request Dec 21, 2016
@fidencio
WATCHDOG: Avoid non async-signal-safe from the signal_handler
b9b8c04 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and, in case of clock screw,  a single byte is
written to this pipe (which is an async-signal-safe operation) and the
logic currently done by the timer handler to reset the watchdog will be
done inside the fd handler in a safe way.

With this patch we ended up losing some debug messages as
orderly_shutdown() has been replaced by _exit(). Personally I don't
think is worth the trouble to try to log those messages properly in this
specific case.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    SSSD#1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    SSSD#2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    SSSD#3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    SSSD#4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    SSSD#5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    SSSD#6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
@fidencio

fidencio commented Jan 2, 2017

Copy link
Copy Markdown
Contributor Author

It's been almost 2 months this code has been around and there's anyone interested to review.
So, I'm closing the PR (as "rejected") as this patchset doesn't seem to be useful for the project.

@fidencio fidencio closed this Jan 2, 2017
fidencio added a commit to fidencio/sssd that referenced this pull request Jan 3, 2017
@fidencio
WATCHDOG: Avoid non async-signal-safe from the signal_handler
01584b4 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and, in case of clock screw,  a single byte is
written to this pipe (which is an async-signal-safe operation) and the
logic currently done by the timer handler to reset the watchdog will be
done inside the fd handler in a safe way.

With this patch we ended up losing some debug messages as
orderly_shutdown() has been replaced by _exit(). Personally I don't
think is worth the trouble to try to log those messages properly in this
specific case.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    SSSD#1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    SSSD#2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    SSSD#3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    SSSD#4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    SSSD#5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    SSSD#6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
fidencio added a commit to fidencio/sssd that referenced this pull request Jan 3, 2017
@fidencio
WATCHDOG: Avoid non async-signal-safe from the signal_handler
62b06d4 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and, in case of clock screw,  a single byte is
written to this pipe (which is an async-signal-safe operation) and the
logic currently done by the timer handler to reset the watchdog will be
done inside the fd handler in a safe way.

With this patch we ended up losing some debug messages as
orderly_shutdown() has been replaced by _exit(). Personally I don't
think is worth the trouble to try to log those messages properly in this
specific case.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    SSSD#1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    SSSD#2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    SSSD#3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    SSSD#4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    SSSD#5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    SSSD#6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
fidencio added a commit to fidencio/sssd that referenced this pull request Jan 3, 2017
@fidencio
WATCHDOG: Avoid non async-signal-safe from the signal_handler
37a1f69 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and, in case of clock screw,  a single byte is
written to this pipe (which is an async-signal-safe operation) and the
logic currently done by the timer handler to reset the watchdog will be
done inside the fd handler in a safe way.

With this patch we ended up losing some debug messages as
orderly_shutdown() has been replaced by kill(-getpgrp(), SIGTERM).
Personally I don't think is worth the trouble to try to log those messages
properly in this specific case.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    SSSD#1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    SSSD#2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    SSSD#3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    SSSD#4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    SSSD#5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    SSSD#6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
fidencio added a commit to fidencio/sssd that referenced this pull request Jan 5, 2017
@fidencio
WATCHDOG: Avoid non async-signal-safe from the signal_handler
b56d187 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and, in case of clock screw,  a single byte is
written to this pipe (which is an async-signal-safe operation) and the
logic currently done by the timer handler to reset the watchdog will be
done inside the fd handler in a safe way.

With this patch we ended up losing some debug messages as
orderly_shutdown() has been replaced by _exit(1). Personally I don't
think is worth the trouble to try to log those messages properly in
this specific case.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    SSSD#1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    SSSD#2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    SSSD#3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    SSSD#4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    SSSD#5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    SSSD#6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
fidencio added a commit to fidencio/sssd that referenced this pull request Jan 9, 2017
@fidencio
WATCHDOG: Avoid non async-signal-safe from the signal_handler
241c418 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and, in case of clock screw,  a single byte is
written to this pipe (which is an async-signal-safe operation) and the
logic currently done by the timer handler to reset the watchdog will be
done inside the fd handler in a safe way.

With this patch we ended up losing some debug messages as
orderly_shutdown() has been replaced by kill(-getpgrp(), SIGTERM).
Personally I don't think is worth the trouble to try to log those messages
properly in this specific case.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    SSSD#1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    SSSD#2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    SSSD#3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    SSSD#4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    SSSD#5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    SSSD#6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
fidencio added a commit to fidencio/sssd that referenced this pull request Jan 9, 2017
@fidencio
WATCHDOG: Avoid non async-signal-safe from the signal_handler
34fa8d0 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and, in case of clock screw,  a single byte is
written to this pipe (which is an async-signal-safe operation) and the
logic currently done by the timer handler to reset the watchdog will be
done inside the fd handler in a safe way.

With this patch we ended up losing some debug messages as
orderly_shutdown() has been replaced by kill(-getpgrp(), SIGTERM).
Personally I don't think is worth the trouble to try to log those messages
properly in this specific case.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    SSSD#1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    SSSD#2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    SSSD#3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    SSSD#4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    SSSD#5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    SSSD#6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
fidencio added a commit to fidencio/sssd that referenced this pull request Jan 10, 2017
@fidencio
WATCHDOG: Avoid non async-signal-safe from the signal_handler
8833b6b 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and, in case of clock screw,  a single byte is
written to this pipe (which is an async-signal-safe operation) and the
logic currently done by the timer handler to reset the watchdog will be
done inside the fd handler in a safe way.

With this patch we ended up losing some debug messages as
orderly_shutdown() has been replaced by kill(-getpgrp(), SIGTERM) (or
_exit(1) considering the cases where setting up the process group during
the server_setup() has failed).
Personally I don't think is worth the trouble to try to log those messages
properly in this specific case.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    SSSD#1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    SSSD#2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    SSSD#3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    SSSD#4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    SSSD#5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    SSSD#6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
jhrozek pushed a commit that referenced this pull request Jan 25, 2017
@fidencio Lukas Slebodnik
WATCHDOG: Avoid non async-signal-safe from the signal_handler
e6a5f8c 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and, in case of clock screw,  a single byte is
written to this pipe (which is an async-signal-safe operation) and the
logic currently done by the timer handler to reset the watchdog will be
done inside the fd handler in a safe way.

With this patch we ended up losing some debug messages as
orderly_shutdown() has been replaced by kill(-getpgrp(), SIGTERM) (or
_exit(1) considering the cases where setting up the process group during
the server_setup() has failed).
Personally I don't think is worth the trouble to try to log those messages
properly in this specific case.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    #1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    #2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    #3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    #4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    #5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    #6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    #7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    #8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    #9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    #10 <signal handler called>
    #11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    #12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    #13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    #14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    #15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    #16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    #17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    #18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    #19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    #20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    #21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    #22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    #23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    #24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    #25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    #26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    #27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    #28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    #29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    #30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    #31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    #32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    #33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    #34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    #35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    #36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    #37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    #38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    #39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    #40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    #41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    #42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    #43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    #44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    #45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    #46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    #47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    #48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    #49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    #50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    #51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    #52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    #53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    #54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    #55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    #56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    #57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    #58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    #59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    #60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
Reviewed-by: Pavel Březina <[email protected]>
Reviewed-by: Jakub Hrozek <[email protected]>
Reviewed-by: Simo Sorce <[email protected]>
jhrozek pushed a commit that referenced this pull request Jan 25, 2017
@fidencio Lukas Slebodnik
WATCHDOG: Avoid non async-signal-safe from the signal_handler
0606a71 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and, in case of clock screw,  a single byte is
written to this pipe (which is an async-signal-safe operation) and the
logic currently done by the timer handler to reset the watchdog will be
done inside the fd handler in a safe way.

With this patch we ended up losing some debug messages as
orderly_shutdown() has been replaced by kill(-getpgrp(), SIGTERM) (or
_exit(1) considering the cases where setting up the process group during
the server_setup() has failed).
Personally I don't think is worth the trouble to try to log those messages
properly in this specific case.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    #1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    #2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    #3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    #4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    #5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    #6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    #7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    #8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    #9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    #10 <signal handler called>
    #11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    #12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    #13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    #14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    #15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    #16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    #17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    #18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    #19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    #20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    #21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    #22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    #23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    #24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    #25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    #26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    #27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    #28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    #29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    #30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    #31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    #32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    #33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    #34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    #35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    #36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    #37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    #38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    #39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    #40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    #41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    #42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    #43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    #44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    #45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    #46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    #47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    #48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    #49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    #50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    #51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    #52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    #53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    #54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    #55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    #56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    #57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    #58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    #59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    #60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
Reviewed-by: Pavel Březina <[email protected]>
Reviewed-by: Jakub Hrozek <[email protected]>
Reviewed-by: Simo Sorce <[email protected]>
(cherry picked from commit e6a5f8c)
jhrozek pushed a commit to jhrozek/sssd that referenced this pull request Jul 20, 2017
@fidencio @jhrozek
WATCHDOG: Avoid non async-signal-safe from the signal_handler
6955a5c 
While debugging rhbz#1396912 a deadlock on sssd_be was noticed[0] and
it's been caused by the use of non async-signal-safe functions from the
signal_handler (please, see man 7 signal for more info about which are
the async-signal-safe functions that can be used).

In order to work this situation around a pipe has been added to the
watchdog_ctx structure and, in case of clock screw,  a single byte is
written to this pipe (which is an async-signal-safe operation) and the
logic currently done by the timer handler to reset the watchdog will be
done inside the fd handler in a safe way.

With this patch we ended up losing some debug messages as
orderly_shutdown() has been replaced by kill(-getpgrp(), SIGTERM) (or
_exit(1) considering the cases where setting up the process group during
the server_setup() has failed).
Personally I don't think is worth the trouble to try to log those messages
properly in this specific case.

It's really worth to mention that a proper fix the clock screw situation
should be implemented on samba's side, by having tevent using monotonic
(or boottime) clock.

[0]:
  [root@dusan ~]# pstack 17922
    #0  __lll_lock_wait_private () at ../nptl/sysdeps/unix/sysv/linux/x86_64/lowlevellock.S:95
    #1  0x00007fe707d04f93 in _L_lock_14932 () from /lib64/libc.so.6
    #2  0x00007fe707d02013 in __GI___libc_malloc (bytes=140630248638304, bytes@entry=15) at malloc.c:2891
    #3  0x00007fe707d0888a in __GI___strdup (s=0x7fe707dff4f7 "/etc/localtime") at strdup.c:42
    #4  0x00007fe707d31b61 in tzset_internal (always=<optimized out>, explicit=explicit@entry=1) at tzset.c:438
    #5  0x00007fe707d32523 in __tz_convert (timer=timer@entry=0x7ffcd5d2b090, use_localtime=use_localtime@entry=1, tp=tp@entry=0x7fe708041d40 <_tmbuf>) at tzset.c:621
    #6  0x00007fe707d30521 in __GI_localtime (t=t@entry=0x7ffcd5d2b090) at localtime.c:42
    SSSD#7  0x00007fe70886c7b0 in sss_vdebug_fn (file=<optimized out>, line=<optimized out>, function=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=16, flags=flags@entry=0, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n", ap=ap@entry=0x7ffcd5d2b130) at src/util/debug.c:248
    SSSD#8  0x00007fe70886c995 in sss_debug_fn (file=file@entry=0x7fe70bff263b "src/util/util_watchdog.c", line=line@entry=82, function=function@entry=0x7fe70bff27f0 <__FUNCTION__.9379> "watchdog_handler", level=level@entry=16, format=format@entry=0x7fe70bff2760 "Watchdog timer overflow, killing process!\n") at src/util/debug.c:284
    SSSD#9  0x00007fe70bfdb409 in watchdog_handler (sig=<optimized out>) at src/util/util_watchdog.c:81
    SSSD#10 <signal handler called>
    SSSD#11 0x00007fe707cff664 in _int_malloc (av=av@entry=0x7fe70803c760 <main_arena>, bytes=bytes@entry=151) at malloc.c:3494
    SSSD#12 0x00007fe707d01fbc in __GI___libc_malloc (bytes=bytes@entry=151) at malloc.c:2893
    SSSD#13 0x00007fe708450749 in __talloc_with_prefix (prefix_len=0, size=55, context=0x7fe718373210) at ../talloc.c:668
    SSSD#14 __talloc (size=55, context=0x7fe718373210) at ../talloc.c:708
    SSSD#15 _talloc_named_const (name=0x7fe70bb7015d "../common/ldb_pack.c:425", size=55, context=0x7fe718373210) at ../talloc.c:865
    SSSD#16 talloc_named_const (context=<optimized out>, size=size@entry=55, name=name@entry=0x7fe70bb7015d "../common/ldb_pack.c:425") at ../talloc.c:1606
    SSSD#17 0x00007fe70bb61803 in ldb_unpack_data_only_attr_list (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=0x7fe7184aa1e0, list=list@entry=0x0, list_size=list_size@entry=0, nb_elements_in_db=nb_elements_in_db@entry=0x0) at ../common/ldb_pack.c:425
    SSSD#18 0x00007fe70bb61a7d in ldb_unpack_data (ldb=ldb@entry=0x7fe70e4d52c0, data=data@entry=0x7ffcd5d2b990, message=<optimized out>) at ../common/ldb_pack.c:470
    SSSD#19 0x00007fe6fdc29b46 in ltdb_parse_data_unpack (key=..., data=..., private_data=0x7ffcd5d2ba70) at ../ldb_tdb/ldb_search.c:249
    SSSD#20 0x00007fe70a5e0a24 in tdb_parse_data (tdb=tdb@entry=0x7fe70e4eaa10, key=..., offset=15619748, len=414772, parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/io.c:637
    SSSD#21 0x00007fe70a5dc1fc in tdb_parse_record (tdb=0x7fe70e4eaa10, key=..., parser=parser@entry=0x7fe6fdc29b10 <ltdb_parse_data_unpack>, private_data=private_data@entry=0x7ffcd5d2ba70) at ../common/tdb.c:253
    SSSD#22 0x00007fe6fdc29e7b in ltdb_search_dn1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, msg=msg@entry=0x7fe7184aa1e0) at ../ldb_tdb/ldb_search.c:287
    SSSD#23 0x00007fe6fdc2acbb in ltdb_dn_list_load (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183c4940, list=list@entry=0x7fe7183c3a30) at ../ldb_tdb/ldb_index.c:181
    SSSD#24 0x00007fe6fdc2bbbb in ltdb_index_add1 (module=module@entry=0x7fe70e4eab50, dn=dn@entry=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", v_idx=v_idx@entry=0, el=<optimized out>, el=<optimized out>) at ../ldb_tdb/ldb_index.c:1134
    SSSD#25 0x00007fe6fdc2c62c in ltdb_index_add_el (el=0x7fe7184aa3e0, dn=0x7fe7183bf3e0 "[email protected],cn=users,cn=DOMAIN.COM,cn=sysdb", module=0x7fe70e4eab50) at ../ldb_tdb/ldb_index.c:1180
    SSSD#26 ltdb_index_add_element (module=module@entry=0x7fe70e4eab50, dn=<optimized out>, el=el@entry=0x7fe7184aa3e0) at ../ldb_tdb/ldb_index.c:1290
    SSSD#27 0x00007fe6fdc290bb in ltdb_modify_internal (module=module@entry=0x7fe70e4eab50, msg=0x7fe7183bf0c0, req=req@entry=0x7fe7183bdc10) at ../ldb_tdb/ldb_tdb.c:903
    SSSD#28 0x00007fe6fdc2958a in ltdb_modify (ctx=0x7fe7183c2950, ctx=0x7fe7183c2950) at ../ldb_tdb/ldb_tdb.c:998
    SSSD#29 ltdb_callback (ev=<optimized out>, te=<optimized out>, t=..., private_data=<optimized out>) at ../ldb_tdb/ldb_tdb.c:1380
    SSSD#30 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4d2890) at ../tevent_timed.c:341
    SSSD#31 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4d2890, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#32 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4d2890, location=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent_standard.c:114
    SSSD#33 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4d2890, location=location@entry=0x7fe70bb72ec5 "../common/ldb.c:631") at ../tevent.c:533
    SSSD#34 0x00007fe70bb6bc4f in ldb_wait (handle=0x7fe7183c4530, type=<optimized out>) at ../common/ldb.c:631
    SSSD#35 0x00007fe70bb6c793 in ldb_autotransaction_request (ldb=0x7fe70e4d52c0, req=0x7fe7183bdc10) at ../common/ldb.c:573
    SSSD#36 0x00007fe70bb6d263 in ldb_modify (ldb=ldb@entry=0x7fe70e4d52c0, message=<optimized out>) at ../common/ldb.c:1655
    SSSD#37 0x00007fe70bfa2ab5 in sysdb_set_cache_entry_attr (ldb=0x7fe70e4d52c0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bf680, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1159
    SSSD#38 0x00007fe70bfa304d in sysdb_rep_ts_entry_attr (sysdb=0x7fe70e4eadd0, attrs=0x7fe7183bf680, entry_dn=0x7fe7183c4760) at src/db/sysdb_ops.c:1218
    SSSD#39 sysdb_set_ts_entry_attr (sysdb=sysdb@entry=0x7fe70e4eadd0, entry_dn=entry_dn@entry=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1248
    SSSD#40 0x00007fe70bfa4aa9 in sysdb_set_entry_attr (sysdb=0x7fe70e4eadd0, entry_dn=0x7fe7183c4760, attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1199
    SSSD#41 0x00007fe70bfa4b5f in sysdb_set_user_attr (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", attrs=attrs@entry=0x7fe7183bb840, mod_op=mod_op@entry=2) at src/db/sysdb_ops.c:1285
    SSSD#42 0x00007fe70bfa58c3 in sysdb_add_user (domain=domain@entry=0x7fe70e4d62f0, name=name@entry=0x7fe7183c01f0 "[email protected]", uid=uid@entry=1415408147, gid=<optimized out>, gid@entry=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:1928
    SSSD#43 0x00007fe70bfab271 in sysdb_store_new_user (now=1481105315, cache_timeout=5400, attrs=0x7fe7183bb840, orig_dn=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", shell=0x0, homedir=0x0, gecos=0x7fe710465d00 "Test User7045", gid=1415400513, uid=1415408147, name=0x7fe7183c01f0 "[email protected]", domain=0x7fe70e4d62f0) at src/db/sysdb_ops.c:2549
    SSSD#44 sysdb_store_user (domain=domain@entry=0x7fe70e4d62f0, name=0x7fe7183c01f0 "[email protected]", pwd=pwd@entry=0x0, uid=1415408147, gid=1415400513, gecos=gecos@entry=0x7fe710465d00 "Test User7045", homedir=homedir@entry=0x0, shell=shell@entry=0x0, orig_dn=orig_dn@entry=0x7fe710465940 "CN=Test User7045,OU=Sales,DC=DOMAIN,DC=COM", attrs=attrs@entry=0x7fe7183bb840, remove_attrs=0x7fe7183c08a0, cache_timeout=cache_timeout@entry=5400, now=now@entry=1481105315) at src/db/sysdb_ops.c:2499
    SSSD#45 0x00007fe6fba0d9f9 in sdap_save_user (memctx=memctx@entry=0x7fe70e544ee0, opts=opts@entry=0x7fe70e518400, dom=dom@entry=0x7fe70e4d62f0, attrs=<optimized out>, _usn_value=_usn_value@entry=0x7ffcd5d2c260, now=now@entry=1481105315) at src/providers/ldap/sdap_async_users.c:509
    SSSD#46 0x00007fe6fba0df9a in sdap_save_users (memctx=memctx@entry=0x7fe70e544e40, sysdb=0x7fe70e4eadd0, dom=0x7fe70e4d62f0, opts=0x7fe70e518400, users=<optimized out>, num_users=10006, _usn_value=_usn_value@entry=0x7fe70e544e60) at src/providers/ldap/sdap_async_users.c:572
    SSSD#47 0x00007fe6fba0e460 in sdap_get_users_done (subreq=<optimized out>) at src/providers/ldap/sdap_async_users.c:938
    SSSD#48 0x00007fe6fba0c9d5 in sdap_search_user_process (subreq=0x0) at src/providers/ldap/sdap_async_users.c:814
    SSSD#49 0x00007fe6fba07379 in generic_ext_search_handler (subreq=0x0, opts=<optimized out>) at src/providers/ldap/sdap_async.c:1689
    SSSD#50 0x00007fe6fba0991b in sdap_get_generic_op_finished (op=<optimized out>, reply=<optimized out>, error=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:1621
    SSSD#51 0x00007fe6fba083cd in sdap_process_message (ev=<optimized out>, sh=<optimized out>, msg=0x7fe70e5f9ce0) at src/providers/ldap/sdap_async.c:353
    SSSD#52 sdap_process_result (ev=<optimized out>, pvt=<optimized out>) at src/providers/ldap/sdap_async.c:197
    SSSD#53 0x00007fe708664b4f in tevent_common_loop_timer_delay (ev=ev@entry=0x7fe70e4cbc30) at ../tevent_timed.c:341
    SSSD#54 0x00007fe708665b5a in epoll_event_loop_once (ev=0x7fe70e4cbc30, location=<optimized out>) at ../tevent_epoll.c:911
    SSSD#55 0x00007fe708664257 in std_event_loop_once (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:114
    SSSD#56 0x00007fe70866040d in _tevent_loop_once (ev=ev@entry=0x7fe70e4cbc30, location=location@entry=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:533
    SSSD#57 0x00007fe7086605ab in tevent_common_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent.c:637
    SSSD#58 0x00007fe7086641f7 in std_event_loop_wait (ev=0x7fe70e4cbc30, location=0x7fe70bfee8e7 "src/util/server.c:702") at ../tevent_standard.c:140
    SSSD#59 0x00007fe70bfd1993 in server_loop (main_ctx=0x7fe70e4cd080) at src/util/server.c:702
    SSSD#60 0x00007fe70c84cb82 in main (argc=8, argv=<optimized out>) at src/providers/data_provider_be.c:587

Resolves:
https://fedorahosted.org/sssd/ticket/3266

Signed-off-by: Fabiano Fidêncio <[email protected]>
Reviewed-by: Pavel Březina <[email protected]>
Reviewed-by: Jakub Hrozek <[email protected]>
Reviewed-by: Simo Sorce <[email protected]>
(cherry picked from commit e6a5f8c)
(cherry picked from commit 0606a71)
@fidencio fidencio mentioned this pull request Aug 22, 2017
Closed
@fidencio fidencio mentioned this pull request Sep 8, 2017
Closed
@fidencio

fidencio commented Sep 9, 2017

Copy link
Copy Markdown
Contributor Author

So, I've re-worked those patches a little bit and here is the time difference when running reconfing with the patches:

real	0m26.047s
user	0m21.318s
sys	0m4.635s

And now without:

real	0m25.565s
user	0m20.696s
sys	0m4.433s

This patch set is rebased on top of PR #377.

I really would appreciate if someone could review and give their opinion.

The reason this PR was blocked is because this time difference has been considered a "performance issue".

@jhrozek, could you take a look on this?

@fidencio

fidencio commented Sep 9, 2017

Copy link
Copy Markdown
Contributor Author

Ah, somehow seems that I cannot reopen the PR. So, I'm opening a new one.

@fidencio fidencio mentioned this pull request Sep 9, 2017
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

1 more reviewer
@lslebodn lslebodn lslebodn left review comments
Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

Rejected

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@fidencio @lslebodn