platform/core/security/security-manager.git
4 months agoRelease 1.9.18 accepted/tizen_unified_x_asan accepted/tizen/8.0/unified/20250206.155754 accepted/tizen/9.0/unified/20250206.165047 accepted/tizen/unified/20250207.071323 accepted/tizen/unified/x/20250212.043832 accepted/tizen/unified/x/asan/20250211.003437
Tomasz Swierczek [Wed, 5 Feb 2025 07:18:20 +0000 (08:18 +0100)]
Release 1.9.18

* Change the way we pre-load symbol for syscall by calling it

Change-Id: Iee6cf0a127c50c5dfb9b44e88fad0f86857c1334

4 months agoChange the way we pre-load symbol for syscall by calling it
Tomasz Swierczek [Wed, 5 Feb 2025 07:05:06 +0000 (08:05 +0100)]
Change the way we pre-load symbol for syscall by calling it

Apparently using -1 as invalid argument can fail with SIGILL
on some configurations (happened on VD side).

Change-Id: I26d1d0411c637a67888a8d1e73892171004d05a3

4 months agoRelease 1.9.17 25/318425/4 accepted/tizen/8.0/unified/20250117.145609 accepted/tizen/9.0/unified/20250117.022504 accepted/tizen/unified/20250117.011625 accepted/tizen/unified/x/20250117.020910
Tomasz Swierczek [Thu, 16 Jan 2025 12:37:01 +0000 (13:37 +0100)]
Release 1.9.17

* Fix possible issue with lazy-symbol binding

Change-Id: I813fc8f6d1e61d01937664e09a1e8ee2546391ab

4 months agoFix possible issue with lazy-symbol binding 24/318424/5
Tomasz Swierczek [Thu, 16 Jan 2025 12:30:17 +0000 (13:30 +0100)]
Fix possible issue with lazy-symbol binding

It seems possible that lazy binding of the SO symbols loaded by linker can cause problems
when the SIGSETXID arrives in some thread at the precise moment that a symbol in a thread
is being lazy-initialized. By doing calls below we're ensuring that symbols used inside
our handler are already loaded - lazy symbols (or any not-loaded) should not be called inside
a handler.

Also, changed the handler to be a regular function for easier debugging in kernel team.

Change-Id: I6a293374c38f47b18af86166d8b150af55469739

5 months agoRelease 1.9.16 63/317863/1 accepted/tizen/8.0/unified/20250114.155529 accepted/tizen/9.0/unified/20250114.041604 accepted/tizen/unified/20250113.133507 accepted/tizen/unified/x/20250114.041839
Tomasz Swierczek [Thu, 9 Jan 2025 05:53:52 +0000 (06:53 +0100)]
Release 1.9.16

* Add security_manager_is_app_from_pid() function

Change-Id: Iea0b4651a91b39cf4db20f82db2b722a4113d82a

5 months agoAdd security_manager_is_app_from_pid() function 23/317823/1
Tomasz Swierczek [Wed, 8 Jan 2025 07:15:41 +0000 (08:15 +0100)]
Add security_manager_is_app_from_pid() function

Change-Id: I53ed6e507a945e1829e851621e521417bef831b5

5 months agoRelease 1.9.15 88/317388/1 accepted/tizen/8.0/unified/20250107.161525 accepted/tizen/9.0/unified/20250104.070051 accepted/tizen/unified/20250104.061052 accepted/tizen/unified/x/20250104.065706 accepted/tizen/unified/x/asan/20250113.002112
Krzysztof Jackiewicz [Tue, 31 Dec 2024 13:04:49 +0000 (14:04 +0100)]
Release 1.9.15

* Mark thread as dead in case of ESRCH
* Disable logging during thread sync
* Remove log from get_alive_threads()
* Put abort() in place to facilitate debugging

Change-Id: I7e56ad62b4338ddc0149bf322d155f939ecd478f

5 months agoMark thread as dead in case of ESRCH 87/317387/1
Krzysztof Jackiewicz [Tue, 31 Dec 2024 12:33:27 +0000 (13:33 +0100)]
Mark thread as dead in case of ESRCH

This may speed up the thread sync a bit

Change-Id: Ib6a9c8f11a719615227efd24da18c65b91d69bf0

5 months agoDisable logging during thread sync 85/317385/3
Krzysztof Jackiewicz [Tue, 31 Dec 2024 10:58:40 +0000 (11:58 +0100)]
Disable logging during thread sync

Change-Id: I78e3154a6f5ae4697a461a0cbcce4affca2f50dd

5 months agoRemove log from get_alive_threads() 84/317384/2
Krzysztof Jackiewicz [Tue, 31 Dec 2024 10:30:50 +0000 (11:30 +0100)]
Remove log from get_alive_threads()

No allocations allowed during thread sync

Change-Id: I3e539c4572ccfba96ba798383c1eea7d0260cd44

5 months agoPut abort() in place to facilitate debugging 82/317382/5
Krzysztof Jackiewicz [Tue, 31 Dec 2024 09:36:55 +0000 (10:36 +0100)]
Put abort() in place to facilitate debugging

This way the backtrace will show us the point of failure

Change-Id: I97f70db007661b91baa757442b01c0cdc3270a5b

5 months agopackaging: asan: skip Werror=stringop-overread 89/317089/1 accepted/tizen/unified/20241226.121340 accepted/tizen/unified/x/20241227.014404
Dongkyun Son [Tue, 5 Nov 2024 08:28:46 +0000 (17:28 +0900)]
packaging: asan: skip Werror=stringop-overread

Change-Id: I10972a6d7bb01cc2d7e017b4d2a06189e33f674f
Signed-off-by: Dongkyun Son <dongkyun.s@samsung.com>
5 months agoRelease 1.9.14 16/316916/4 accepted/tizen/8.0/unified/20241220.161204 accepted/tizen/9.0/unified/20241220.020357 accepted/tizen/unified/20241220.010614 accepted/tizen/unified/x/20241220.014648
Tomasz Swierczek [Thu, 19 Dec 2024 08:11:52 +0000 (09:11 +0100)]
Release 1.9.14

* Change runtime Smack check in dev_wos cases to compile-time
* Add argument check to security_manager_set_identity()

Change-Id: I6ff748258a07d98c32d31f43c2b9e30a65f2dc9a

5 months agoChange runtime Smack check in dev_wos cases to compile-time 13/316913/4
Tomasz Swierczek [Thu, 19 Dec 2024 08:01:08 +0000 (09:01 +0100)]
Change runtime Smack check in dev_wos cases to compile-time

Previous patch didn't work for PontusM smack-enabled environment.

Now this patch ensures that WITHOUT dev_wos flag, there is definitely
no difference in logic of security-manager, even in the absence of Smack
(ie. during image creation).

Change-Id: If37d9ceae8bcb5727e82f7310acff0358b820240

5 months agoAdd argument check to security_manager_set_identity() 57/316857/1
Jan Wojtkowski [Wed, 18 Dec 2024 11:30:04 +0000 (12:30 +0100)]
Add argument check to security_manager_set_identity()

Change-Id: Ibac8f33ae7eaef9bff7b67fdea53bc6f82591fbb

6 months agoRelease 1.9.13 79/316579/2 accepted/tizen/unified/20241216.010927 accepted/tizen/unified/x/20241218.032709
Tomasz Swierczek [Fri, 13 Dec 2024 12:53:28 +0000 (13:53 +0100)]
Release 1.9.13

* Change process_type elements with additional prefix

Change-Id: I5eb61bbc15a5aa8b8cc66ef046e2fb4f280d6605

6 months agoChange process_type elements with additional prefix 76/316576/2
Tomasz Swierczek [Fri, 13 Dec 2024 12:24:23 +0000 (13:24 +0100)]
Change process_type elements with additional prefix

askuser-notification couldn't be built because of ambiguity
in the APIs/enums.

Also, added missing header.

Change-Id: I1dbc4ef6cac4cdf74db200b85b790424d3b03ee4

6 months agoRelease 1.9.12 99/316299/4
Tomasz Swierczek [Thu, 12 Dec 2024 09:56:11 +0000 (10:56 +0100)]
Release 1.9.12

* Make security_manager_set_identity() and security_manager_self_is_app() work in dev_wos env
* Add function security_manager_set_identity()
* Add function security_manager_get_app_owner_uid()
* Silence SVACE warning
* Add function security_manager_self_is_app()
* Add security-manager permissive mode build option - dev_wos

Change-Id: I3c225668a16f500fd875a951534b4754e73bb4a2

6 months agoMake security_manager_set_identity() and security_manager_self_is_app() work in dev_w... 96/316296/5
Tomasz Swierczek [Thu, 12 Dec 2024 09:50:39 +0000 (10:50 +0100)]
Make security_manager_set_identity() and security_manager_self_is_app() work in dev_wos env

Change-Id: I42437b6db64b5beb20c06339ae8f782e15562bf4

6 months agoAdd function security_manager_set_identity() 86/315486/14
Jan Wojtkowski [Fri, 29 Nov 2024 10:17:32 +0000 (11:17 +0100)]
Add function security_manager_set_identity()

Change-Id: Iec954c9a41cbf0547de685590685fdbf2b03e5d6

6 months agoAdd function security_manager_get_app_owner_uid() 94/316294/6
Tomasz Swierczek [Thu, 12 Dec 2024 09:32:58 +0000 (10:32 +0100)]
Add function security_manager_get_app_owner_uid()

Change-Id: Ie8281381b9c2ce509aec7ef3f5a0dbdcf7e58a1f

6 months agoSilence SVACE warning 33/315433/7
Tomasz Swierczek [Tue, 19 Nov 2024 09:51:05 +0000 (10:51 +0100)]
Silence SVACE warning

Change-Id: Ib9baacdbdd47f197b39c530a2b5bff75e0fc944e

6 months agoAdd function security_manager_self_is_app() 05/315405/7
Tomasz Swierczek [Mon, 25 Nov 2024 14:27:57 +0000 (15:27 +0100)]
Add function security_manager_self_is_app()

Change-Id: I03c4b947295359964578d1eb9c7e78ea2b9223ff

6 months agoAdd security-manager permissive mode build option - dev_wos 34/316134/4
Tomasz Swierczek [Mon, 15 Apr 2024 14:01:29 +0000 (16:01 +0200)]
Add security-manager permissive mode build option - dev_wos

Usage: gbs <your favourite flags> --define "dev_wos 1".

If not explicitly specified, the option is disabled (switched to 0).

This option is intended only for development, not for release builds.

With this option, security-manager enters "permissive mode"
and doesn't operate on Smack or anything that may be related:
* at launching:
** doesn't change/modify/configure mount namespaces
** doesn't change/modify/configure capabilities
** doesn't change/modify/set Smack labels of app candidate threads
** does set group IDs based on privileges in the Cynara DB
* at app installation:
** doesn't change any Smack labels of files/folders

Other operations (ie. registering apps in the DB) should be performed
as usual.

rules-loader service is only checking the DB and making sure it has
all proper schema applied (DB versioning, etc.) but otherwise,
the one-shot service is not loading any Smack rules.

With this option, the application identification functions that rely
on Smack being the application identifier, that is:

security_manager_identify_app_from_socket
security_manager_identify_app_from_pid

...will return same string default_app_no_Smack_mode as the app pkg ID.

The API function:

security_manager_identify_app_from_cynara_client

...will normally parse Smack label, as its expected that cynara helpers
will be returning the string:

User::Pkg::default_app_no_Smack_mode

...as cynara client.

The API function:

security_manager_app_has_privilege

...will always return success & access granted.

This mode is intended to be used by for bring-up development
of Tizen working in container.

Change-Id: I24c15bb51f87faab3ac14b9dae5e296d682bd768

6 months agoRelease 1.9.11 34/316034/3 accepted/tizen/8.0/unified/20241210.170135 accepted/tizen/9.0/unified/20241210.020008 accepted/tizen/unified/20241210.041709
Krzysztof Jackiewicz [Mon, 9 Dec 2024 10:40:27 +0000 (11:40 +0100)]
Release 1.9.11

* Remove remaining allocations
* Properly handle disappearing threads during sync

Change-Id: I9f45d44385d0a14805301707fe5e34602f571f35

6 months agoRemove remaining allocations 36/316036/2
Krzysztof Jackiewicz [Mon, 9 Dec 2024 11:43:46 +0000 (12:43 +0100)]
Remove remaining allocations

Change-Id: Ibbfe3b014f32b40065db35081110806f5545d83f

6 months agoProperly handle disappearing threads during sync 33/316033/1
Krzysztof Jackiewicz [Mon, 9 Dec 2024 10:21:25 +0000 (11:21 +0100)]
Properly handle disappearing threads during sync

Plus few minor fixes

Change-Id: I0928f0ce811e26b25e3e3f566410f2ba95054416

6 months agoRelease 1.9.10 73/315973/1 accepted/tizen/unified/20241209.090229
Krzysztof Jackiewicz [Fri, 6 Dec 2024 19:37:04 +0000 (20:37 +0100)]
Release 1.9.10

* Don't use allocations during signaling
* Initialize all resource_pid_t members
* Add negative test for generateProcessLabel

Change-Id: Iab1dbdf5dc17ba082c14db2fb65440221278673e

6 months agoDon't use allocations during signaling 89/315889/3
Krzysztof Jackiewicz [Thu, 5 Dec 2024 16:26:59 +0000 (17:26 +0100)]
Don't use allocations during signaling

This is to prevent deadlock in malloc. It may happen if a supplementary
thread is interrupted after it took an internal glibc lock in malloc()
and it waits in a signal handler for the main thread. The main thread
won't be able to allocate memory due to lock taken by supplementary
thread.

Change-Id: I218075c2c2d6befa8fafb141e0507e64b5b47406

6 months agoInitialize all resource_pid_t members 59/319659/2
Krzysztof Jackiewicz [Fri, 25 Oct 2024 10:22:54 +0000 (12:22 +0200)]
Initialize all resource_pid_t members

Change-Id: I819693bc3cf895c28204c07dc2ee66a50efa7aeb

6 months agoAdd negative test for generateProcessLabel 63/319763/3
Krzysztof Jackiewicz [Thu, 31 Oct 2024 14:57:13 +0000 (15:57 +0100)]
Add negative test for generateProcessLabel

Change-Id: I8864eaae630302489d64b39013b433a267d3ed29

7 months agoRelease 1.9.9 99/319499/3 accepted/tizen/8.0/unified/20241111.162216 accepted/tizen/9.0/unified/20241109.041958 accepted/tizen/unified/20241030.154513
Tomasz Swierczek [Thu, 24 Oct 2024 09:59:25 +0000 (11:59 +0200)]
Release 1.9.9

* Remove unused code
* Fix double free in label monitor

Change-Id: I68dda88fefdee0b54f8791a153d63f7f1f98c1db

7 months agoRemove unused code 71/319471/5
Krzysztof Jackiewicz [Wed, 23 Oct 2024 13:15:08 +0000 (15:15 +0200)]
Remove unused code

Change-Id: Ib0423186dc3183ecedb519600922995d438bf2b8

7 months agoFix double free in label monitor
Krzysztof Jackiewicz [Thu, 24 Oct 2024 08:03:39 +0000 (10:03 +0200)]
Fix double free in label monitor

The pointer passed to initialize_inotify() is already managed by a unique_ptr.
In case of initialize_inotify() failure the pointer was being passed to
security_manager_app_labels_monitor_finish() where it was wrapped in a
unique_ptr again.

Add a helper non-throwing function operating on raw app_labels_monitor pointer
for properly closing it. Use it in security_manager_app_labels_monitor_finish()
and security_manager_app_labels_monitor_init().

Change-Id: I6f3b5883fde53ba3ded7764f3121f59ace053a10

7 months agoRelease 1.9.8 01/319401/7
Tomasz Swierczek [Tue, 22 Oct 2024 18:31:28 +0000 (20:31 +0200)]
Release 1.9.8

* Add check if HIGHEST priority is only for PREPARE_APP call
* Add check for CAP_MAC_ADMIN inside prepare_app2 call
* Added warning logs for events priority
* Fix logic in managing EAGAIN on sending signal in client (preapre_app2)
* Introduce separate thread for highest-priority events
* Make objects used in service_impl.cpp (mostly) thread safe
* Add highest priority & use it for SecurityModuleCall::PREPARE_APP
* Rename service-thread.h to service-thread-dispatcher.h
* Fix typo

Change-Id: Iea3507f1401b1b651ec2919626ddd1b863da9e44

7 months agoAdd check if HIGHEST priority is only for PREPARE_APP call 11/319411/3
Tomasz Swierczek [Tue, 22 Oct 2024 18:17:17 +0000 (20:17 +0200)]
Add check if HIGHEST priority is only for PREPARE_APP call

In case client was malicious, we could get events with highest
priority that are not PREPARE_APP calls, which could lead to
ie. multiple DB transactions being done in parallel threads
which the logic (service_impl.cpp) doesn't support as of now.

This technically eliminates the need for HIGHEST priority at all,
but it can be used in the future (if ie. a thread poll will be needed)
so it was not removed.

Change-Id: I30441b041610ef770cd0a23eb0d52216c7c1a7f3

7 months agoAdd check for CAP_MAC_ADMIN inside prepare_app2 call 00/319400/5
Tomasz Swierczek [Tue, 22 Oct 2024 14:52:08 +0000 (16:52 +0200)]
Add check for CAP_MAC_ADMIN inside prepare_app2 call

Apparently in some weird scenarios, the wrt-loader or other launcher
that attempts at running apps, doesn't have the caps required.

Added check just prints proper error logs.

TODO: libsmack lacks function to get relabel-self list of labels
which probably should also be checked in the future (this is why
the check for CAP_MAC_ADMIN doesn't exit when there's no capability).

Change-Id: I5eeacb5ecb84883f6a4b2097887b82708297e98f

7 months agoAdded warning logs for events priority 02/319302/2
Tomasz Swierczek [Mon, 21 Oct 2024 08:21:12 +0000 (10:21 +0200)]
Added warning logs for events priority

This will ease debugging in case when many requests are being handled.

Change-Id: I85ee91d868c1e335e569fa67153a0871ce569eb9

7 months agoFix logic in managing EAGAIN on sending signal in client (prepare_app2) 17/319217/3
Tomasz Swierczek [Thu, 17 Oct 2024 14:19:37 +0000 (16:19 +0200)]
Fix logic in managing EAGAIN on sending signal in client (prepare_app2)

Previous buggy version could lead to abort in case
2nd attempt at sending a signal went okay.

Change-Id: Ia0b0001f4ed602f9f7e5d40922998b5912a2414a

7 months agoIntroduce separate thread for highest-priority events 16/319216/2
Tomasz Swierczek [Thu, 17 Oct 2024 13:49:54 +0000 (15:49 +0200)]
Introduce separate thread for highest-priority events

So far, only PREPARE_APP command on server side is treated as such.

Change-Id: Ia1dde64d0eaf0758f9f8c6c5e68d3e7a289eb3bb

7 months agoMake objects used in service_impl.cpp (mostly) thread safe 15/319215/1
Tomasz Swierczek [Thu, 17 Oct 2024 09:05:14 +0000 (11:05 +0200)]
Make objects used in service_impl.cpp (mostly) thread safe

So that its easier to introduce multiple-threads into the daemon.

**Disclaimer**
--------------
(1)
privilege_db is thread safe as long as only one thread is managing RW transactions
on the DB (each API of the DB takes & releases internal mutex).

This is enough to introduce a new separate thread just for prepare_app2 functionality
that only READS data and 2nd thread (current worker) operating on everything else,
including the RW operations.

To make a thread pool operate on many RW requests, transactions would need to be
handled in different way, so that they do not intertwine & the process
makes just one transaction at a time on the shared connection to the DB.

(2)
NSMountLogic is used only on policy update, so will be used only from the
standard worker thread (not the prepare_app2 thread). If thread pool will have
to be implemented in security-manager (or more than 1 thread to handle updates),
this would have to get reviewed.

(3)
Updates of permissible file with Smack labels of apps were left untouched to be done
on one thread (worker thread).

Change-Id: Idb228c52104058dce0c2ce43322613de09b8a06f

7 months agoAdd highest priority & use it for SecurityModuleCall::PREPARE_APP 14/319214/1
Tomasz Swierczek [Thu, 17 Oct 2024 07:49:52 +0000 (09:49 +0200)]
Add highest priority & use it for SecurityModuleCall::PREPARE_APP

So far without its separate thread, just to have separate queues
for the top-priority events.

Change-Id: I6d52e51fd8188a18f6b9650413ee6926139bdcd9

7 months agoRename service-thread.h to service-thread-dispatcher.h 13/319213/1
Tomasz Swierczek [Thu, 17 Oct 2024 07:32:51 +0000 (09:32 +0200)]
Rename service-thread.h to service-thread-dispatcher.h

This file will no longer implement only one service thread.
Changed the name not to be misleading.

Change-Id: Ibbf90daaf8a399108e5caabf4627ad8646156464

7 months agoFix typo 70/319170/1
Tomasz Swierczek [Wed, 16 Oct 2024 09:39:43 +0000 (11:39 +0200)]
Fix typo

Change-Id: If73a6d618e446851689cbbe77e9ceee8a7fc292c

8 months agoRelease 1.9.7 13/318813/1 accepted/tizen/8.0/unified/20241009.160104
Tomasz Swierczek [Tue, 8 Oct 2024 13:53:30 +0000 (15:53 +0200)]
Release 1.9.7

* Add robot privileges

Change-Id: I9a9ee8878025e28485fa3cef4129f6c0c0737a16

8 months agoAdd robot privileges 12/318812/1
Tomasz Swierczek [Tue, 8 Oct 2024 13:51:57 +0000 (15:51 +0200)]
Add robot privileges

http://tizen.org/privilege/robot.depthsensor
http://tizen.org/privilege/robot.map.get
http://tizen.org/privilege/robot.map.set
http://tizen.org/privilege/robot.movingpart
http://tizen.org/privilege/robot.navigation.get
http://tizen.org/privilege/robot.navigation.set
http://tizen.org/privilege/robot.property.read
http://tizen.org/privilege/robot.property.write
http://tizen.org/privilege/robot.speaker
http://tizen.org/privilege/robot.wheel

Change-Id: I525b92c01cc604f05f22c2ef6f942d891ebd16b9

8 months agoRelease: 1.9.6 05/318605/2 accepted/tizen_unified_toolchain accepted/tizen/9.0/unified/20241030.233141 accepted/tizen/unified/20241008.173927 accepted/tizen/unified/toolchain/20241022.122517 accepted/tizen/unified/toolchain/20241022.122941 accepted/tizen/unified/x/20241009.055408 accepted/tizen/unified/x/asan/20241014.000118 tizen_9.0_m2_release
Krzysztof Jackiewicz [Thu, 3 Oct 2024 08:06:55 +0000 (10:06 +0200)]
Release: 1.9.6

* Revert "Fix tgkill retry logic"
* Set minimal cynara-admin version
* Optimization: batch cynara admin checks into one API call

Change-Id: I755213257aa8001b19bc67d106807349db06dc3f

8 months agoRevert "Fix tgkill retry logic" 18/318518/3
Krzysztof Jackiewicz [Tue, 1 Oct 2024 10:45:24 +0000 (12:45 +0200)]
Revert "Fix tgkill retry logic"

This reverts commit f4e8020905f10fbcb12a06060aab132d2e447f3f.

Apparently there are some failures/crashes in other modules only if the
reverted commit is used. Reverting temporarily. To be investigated
later.

Change-Id: I9bd9e6b7acff8868609e2a91823ac1e73f87ef22

8 months agoSet minimal cynara-admin version 07/318607/1
Krzysztof Jackiewicz [Thu, 3 Oct 2024 08:18:22 +0000 (10:18 +0200)]
Set minimal cynara-admin version

This is necessary as we started using the new cynara-admin API.

Change-Id: I4c660d22510326539c6e114fcbcf6d05307c558e

8 months agoOptimization: batch cynara admin checks into one API call 58/317558/7
Krzysztof Malysa [Wed, 11 Sep 2024 16:02:06 +0000 (18:02 +0200)]
Optimization: batch cynara admin checks into one API call

Although the optimization's focus is on minimizing latency under load it shows some improvements overall.

Before (on emulator):
  additionalThreads 32 nConcurrentApps 1 samples 94
    prepareAppCandidate [us]:              min 145 max 755 avg 470 median 459 stddev 72
    prepareApp [us]:                       min 25 424 max 76 408 avg 40 300 median 42 693 stddev 7 330
    prepareAppCandidate + prepareApp [us]: min 25 868 max 76 857 avg 40 770 median 43 166 stddev 7 337
  nConcurrentApps 1 samples 135
    prepareAppCandidate [us]:              min 146 max 1 102 avg 476 median 457 stddev 88
    prepareApp [us]:                       min 13 769 max 26 616 avg 21 278 median 20 438 stddev 1 968
    prepareAppCandidate + prepareApp [us]: min 13 916 max 27 718 avg 21 754 median 20 893 stddev 1 993
  nConcurrentApps 2 samples 216
    prepareAppCandidate [us]:              min 317 max 1 285 avg 522 median 500 stddev 114
    prepareApp [us]:                       min 12 992 max 34 919 avg 24 143 median 23 894 stddev 4 035
    prepareAppCandidate + prepareApp [us]: min 13 598 max 35 250 avg 24 666 median 24 419 stddev 4 028
  nConcurrentApps 4 samples 384
    prepareAppCandidate [us]:              min 85 max 4 969 avg 678 median 600 stddev 552
    prepareApp [us]:                       min 12 682 max 49 079 avg 25 782 median 24 161 stddev 6 644
    prepareAppCandidate + prepareApp [us]: min 12 824 max 49 732 avg 26 461 median 24 774 stddev 6 609
  nConcurrentApps 8 samples 672
    prepareAppCandidate [us]:              min 77 max 4 609 avg 512 median 435 stddev 397
    prepareApp [us]:                       min 12 260 max 59 107 avg 25 890 median 24 256 stddev 7 760
    prepareAppCandidate + prepareApp [us]: min 12 378 max 59 575 avg 26 403 median 24 723 stddev 7 795
  nConcurrentApps 16 samples 1168
    prepareAppCandidate [us]:              min 79 max 4 619 avg 317 median 220 stddev 355
    prepareApp [us]:                       min 12 656 max 57 372 avg 26 940 median 26 660 stddev 7 674
    prepareAppCandidate + prepareApp [us]: min 12 806 max 57 963 avg 27 257 median 26 958 stddev 7 725
  nConcurrentApps 32 samples 1824
    prepareAppCandidate [us]:              min 76 max 4 945 avg 184 median 136 stddev 287
    prepareApp [us]:                       min 12 589 max 87 912 avg 37 803 median 36 480 stddev 15 334
    prepareAppCandidate + prepareApp [us]: min 12 757 max 88 015 avg 37 987 median 36 606 stddev 15 379

After (on emulator):
  additionalThreads 32 nConcurrentApps 1 samples 100
    prepareAppCandidate [us]:              min 197 max 655 avg 444 median 434 stddev 62
    prepareApp [us]:                       min 30 372 max 44 675 avg 38 708 median 41 516 stddev 4 686
    prepareAppCandidate + prepareApp [us]: min 30 795 max 45 328 avg 39 152 median 41 924 stddev 4 684
  nConcurrentApps 1 samples 165
    prepareAppCandidate [us]:              min 83 max 803 avg 221 median 132 stddev 159
    prepareApp [us]:                       min 12 224 max 23 058 avg 15 407 median 13 172 stddev 3 727
    prepareAppCandidate + prepareApp [us]: min 12 319 max 23 638 avg 15 628 median 13 294 stddev 3 879
  nConcurrentApps 2 samples 356
    prepareAppCandidate [us]:              min 66 max 1 470 avg 104 median 90 stddev 111
    prepareApp [us]:                       min 11 949 max 20 065 avg 13 307 median 13 160 stddev 912
    prepareAppCandidate + prepareApp [us]: min 12 025 max 20 144 avg 13 412 median 13 261 stddev 907
  nConcurrentApps 4 samples 552
    prepareAppCandidate [us]:              min 72 max 696 avg 110 median 94 stddev 58
    prepareApp [us]:                       min 12 115 max 18 504 avg 14 116 median 14 030 stddev 1 305
    prepareAppCandidate + prepareApp [us]: min 12 251 max 18 756 avg 14 226 median 14 128 stddev 1 316
  nConcurrentApps 8 samples 984
    prepareAppCandidate [us]:              min 71 max 2 196 avg 135 median 103 stddev 155
    prepareApp [us]:                       min 11 948 max 25 692 avg 16 838 median 16 695 stddev 2 939
    prepareAppCandidate + prepareApp [us]: min 12 041 max 26 019 avg 16 974 median 16 814 stddev 2 952
  nConcurrentApps 16 samples 1184
    prepareAppCandidate [us]:              min 75 max 5 523 avg 401 median 227 stddev 584
    prepareApp [us]:                       min 12 035 max 54 686 avg 25 184 median 24 884 stddev 7 269
    prepareAppCandidate + prepareApp [us]: min 12 175 max 56 117 avg 25 585 median 25 202 stddev 7 398
  nConcurrentApps 32 samples 1760
    prepareAppCandidate [us]:              min 72 max 4 791 avg 262 median 157 stddev 451
    prepareApp [us]:                       min 12 654 max 87 433 avg 35 770 median 34 678 stddev 14 002
    prepareAppCandidate + prepareApp [us]: min 12 777 max 88 638 avg 36 033 median 34 875 stddev 14 108

Change-Id: I0f039cf9c5023b78830816f776f108afdc8e2134

8 months agoRelease 1.9.5 70/318270/2
Tomasz Swierczek [Thu, 26 Sep 2024 07:02:56 +0000 (09:02 +0200)]
Release 1.9.5

* Change usleep time from 2 ms to 10 ms
* Add warning log for actually receiving a request from socket
* Fix potential memory leak

Change-Id: I4e43bf5ee16bf83214802b1fe3cc898758b584e7

8 months agoChange usleep time from 2 ms to 10 ms 69/318269/2
Tomasz Swierczek [Thu, 26 Sep 2024 06:46:09 +0000 (08:46 +0200)]
Change usleep time from 2 ms to 10 ms

Apparently, the longer the sleep, less CPU time is used in heavy-load scenarios.

Change-Id: I79b89303eb14a74e699b7ba2e2980dd499dcb781

8 months agoAdd warning log for actually receiving a request from socket 82/318082/2
Tomasz Swierczek [Tue, 24 Sep 2024 10:00:55 +0000 (12:00 +0200)]
Add warning log for actually receiving a request from socket

Needed by VD to investigate few issues with performance on slower boards.

Change-Id: I3d0ba83303b4897212afe1dd100eef26a160f27a

9 months agoFix potential memory leak 57/317557/1
Krzysztof Malysa [Wed, 11 Sep 2024 15:51:32 +0000 (17:51 +0200)]
Fix potential memory leak

Change-Id: I6a14537d3e510046b37603aa99e75f8f96b3a845

9 months agoRelease 1.9.4 54/317454/2 accepted/tizen_unified_dev accepted/tizen/8.0/unified/20240913.151459 accepted/tizen/unified/20240911.161942 accepted/tizen/unified/dev/20240913.055518 accepted/tizen/unified/toolchain/20241004.101746 accepted/tizen/unified/x/20240912.013758
Dariusz Michaluk [Tue, 10 Sep 2024 11:25:10 +0000 (13:25 +0200)]
Release 1.9.4

* Add core privilege: isu.admin
* Add unit tests for increased coverage
* Fix tgkill retry logic

Change-Id: I5a1d33a07134d9d747912b0c26c9c5ffd08a4313

9 months agoAdd core privilege: isu.admin 45/317445/1
Mateusz Moscicki [Tue, 10 Sep 2024 08:30:23 +0000 (10:30 +0200)]
Add core privilege: isu.admin

Change-Id: I71d23fe21f9b82226d5604e5a173608068715c19

9 months agoAdd unit tests for increased coverage 48/316448/2
Tomasz Swierczek [Wed, 21 Aug 2024 10:28:05 +0000 (12:28 +0200)]
Add unit tests for increased coverage

Change-Id: Ifbb5ad446b4aa4c2d350f630ea713cd2d7bbbb9d

10 months agoFix tgkill retry logic 75/314875/2
Krzysztof Jackiewicz [Mon, 22 Jul 2024 09:56:11 +0000 (11:56 +0200)]
Fix tgkill retry logic

If second tgkill attempt after EAGAIN succeeds, the abort() would still
be called. This commit fixes it.

Change-Id: I507c32188924fbb38f521c5d7fd1c1897c7ce534

11 months agoRelease 1.9.3 17/314417/1 accepted/tizen/unified/20240712.160116 accepted/tizen/unified/dev/20240715.080544 accepted/tizen/unified/toolchain/20240812.132959 accepted/tizen/unified/x/20240715.042510 accepted/tizen/unified/x/asan/20240813.231250
Tomasz Swierczek [Thu, 11 Jul 2024 09:57:03 +0000 (11:57 +0200)]
Release 1.9.3

* Add more warning logs for performance analysis on release builds
* Remove pkg existance check in labelPaths function
* Remove unused function

Change-Id: If9efe93e4ffb4e2019724c258632ff6d60086dba

11 months agoAdd more warning logs for performance analysis on release builds 86/314186/2
Tomasz Swierczek [Mon, 8 Jul 2024 07:40:50 +0000 (09:40 +0200)]
Add more warning logs for performance analysis on release builds

Change-Id: Ie83e2768d4da2208e1e4666b2c1969b2c38e2a6a

11 months agoRemove pkg existance check in labelPaths function 87/314187/2
Tomasz Swierczek [Mon, 8 Jul 2024 07:42:44 +0000 (09:42 +0200)]
Remove pkg existance check in labelPaths function

The function is called 2 out of 3 times in places
where the package was already added (appInstall, appUpdate),
so no check is needed.

Added the check manually before function call in registering paths API,
where in fact we need to return invalid param for backward compatibility.

Change-Id: I7fb8b967409972ef1f0734f40320682b6bf6cc1b

12 months agoRemove unused function 33/312533/1
Krzysztof Jackiewicz [Tue, 11 Jun 2024 08:21:44 +0000 (10:21 +0200)]
Remove unused function

Change-Id: I07191ea6a8a7632c7f0446fe5724eea678dae0f2

12 months agoRelease 1.9.2 accepted/tizen/8.0/unified/20240614.162128 accepted/tizen/unified/20240603.032414 accepted/tizen/unified/dev/20240620.005646 accepted/tizen/unified/toolchain/20240624.121447 accepted/tizen/unified/x/20240604.013000 accepted/tizen/unified/x/asan/20240625.092453
Tomasz Swierczek [Fri, 31 May 2024 04:38:47 +0000 (06:38 +0200)]
Release 1.9.2

* Add core privilege: keymanager.extended

Change-Id: Ib758362fe042a0d25f599a70304be926c8637897

12 months agoAdd core privilege: keymanager.extended 15/311315/1
Dariusz Michaluk [Fri, 17 May 2024 11:08:34 +0000 (13:08 +0200)]
Add core privilege: keymanager.extended

- keymanager.extended (platform level): app can use extended CKMC API

Change-Id: I1eee03db7511a7bbf18daf5dcc979857f258a4cf

13 months agoRelease 1.9.1 68/309868/1 accepted/tizen/8.0/unified/20240419.170006 accepted/tizen/unified/20240423.164619 accepted/tizen/unified/x/20240425.051117
Tomasz Swierczek [Wed, 17 Apr 2024 12:46:48 +0000 (14:46 +0200)]
Release 1.9.1

* Add mitigation for DDOS attack
* Adjust coverage script after lcov upgrade

Change-Id: Ib852206bcbeb276db97152c7b463916f62eff483

13 months agoAdd mitigation for DDOS attack 59/309859/2
Tomasz Swierczek [Wed, 17 Apr 2024 11:48:46 +0000 (13:48 +0200)]
Add mitigation for DDOS attack

Adversary can send too big message to the daemon,
making it fail on allocation. Few places were required
to be enhanced with catch for std::bad_alloc & proper
graceful closing of connection instead of terminating
the process.

Change-Id: Id05dd3ee3d323a8d47ba93e33fae9d9bc6bb255d

14 months agoAdjust coverage script after lcov upgrade 90/309190/1
Dariusz Michaluk [Fri, 5 Apr 2024 10:07:43 +0000 (12:07 +0200)]
Adjust coverage script after lcov upgrade

Change-Id: I5e02bfd884ea1fa8b6f96796bcaea143ac6eb10c

15 months agoRelease 1.9.0 44/306544/1 accepted/tizen/8.0/unified/20240227.172259 accepted/tizen/unified/20240223.160530 accepted/tizen/unified/toolchain/20240311.065531 accepted/tizen/unified/x/20240226.075419
Tomasz Swierczek [Thu, 22 Feb 2024 16:47:27 +0000 (17:47 +0100)]
Release 1.9.0

* Add core privilege: wifi.secureinfo

Change-Id: I33704a6c60d798d788fd89dcfef7fe9c33a94871

15 months agoAdd core privilege: wifi.secureinfo 42/306342/1
Krzysztof Malysa [Tue, 20 Feb 2024 10:35:14 +0000 (11:35 +0100)]
Add core privilege: wifi.secureinfo

- wifi.secureinfo (platform level): app can read Wi-Fi security information

Change-Id: Iad23ef25c8c2aa65ecf27e47cc1ebbf59c645292

16 months agoRelease 1.8.14 accepted/tizen_unified_riscv accepted/tizen/8.0/unified/20240118.170034 accepted/tizen/unified/20240117.163258 accepted/tizen/unified/riscv/20240118.054545
Tomasz Swierczek [Wed, 17 Jan 2024 08:53:05 +0000 (09:53 +0100)]
Release 1.8.14

* Removed unnecessary bloat & fix readme
* Modified error log in case of failed relabel_self setup
* Add unit tests for permissible file update, initialization and removal
* Implement a mechanism for permissible file repair
* Implement permissible file integrity verification

Change-Id: I1155d61b5343f8981facd2a3fb5b5a9d7f71a144

16 months agoRemoved unnecessary bloat & fix readme
Tomasz Swierczek [Mon, 15 Jan 2024 14:39:30 +0000 (15:39 +0100)]
Removed unnecessary bloat & fix readme

Change-Id: Iafa766759f035a2b9f5d4232d0b16eac46781f4e

18 months agoModified error log in case of failed relabel_self setup
Tomasz Swierczek [Fri, 27 Oct 2023 14:05:59 +0000 (16:05 +0200)]
Modified error log in case of failed relabel_self setup

Error in this place sometimes happens on VD images, but without
harsh consequences, as if it happened in places where the label
list is still empty. Added number of labels to be printed.

Change-Id: I564b819dd53ae359b908ee111573cdde950343ce

18 months agoAdd unit tests for permissible file update, initialization and removal
Filip Skrzeczkowski [Wed, 4 Oct 2023 09:12:23 +0000 (11:12 +0200)]
Add unit tests for permissible file update, initialization and removal

Change-Id: Id05b5deaa8d4fd39c53478c555524f4be4ddac2b

18 months agoImplement a mechanism for permissible file repair
Filip Skrzeczkowski [Thu, 28 Sep 2023 11:19:50 +0000 (13:19 +0200)]
Implement a mechanism for permissible file repair

Permissible files that have been marked as invalid due to their hash
not matching can now be repaired by fetching the relevant data directly
form PrivilegeDb. This works for both the service and the client library.
As the client lacks the privileges to do it on their own, in that case
a request is sent to the Security Manager daemon.

Change-Id: I43fee3d39c015a7fbba0e674c29c904ac26e91bf

18 months agoImplement permissible file integrity verification
Filip Skrzeczkowski [Wed, 23 Aug 2023 13:01:00 +0000 (15:01 +0200)]
Implement permissible file integrity verification

Permissible files have a SHA-1 hash attached at the beginning.
Upon opening it is compared with a new hash calculated from the
file content in order to verify if the file's integrity is intact.
An error is thrown should the hashes differ. Files with no hash are
still supported but reading/updating the rules stored in them
will cause them to automatically switch to the new system.

Change-Id: I5ec379b58cc78e63bcde084ada43273237d61beb

18 months agoRelease 1.8.13 accepted/tizen/8.0/unified/20231212.161348 accepted/tizen/unified/20231130.180925 accepted/tizen/unified/riscv/20231226.211352
Tomasz Swierczek [Wed, 29 Nov 2023 10:06:34 +0000 (11:06 +0100)]
Release 1.8.13

* Rework security_manager_sync_threads_internal()

Change-Id: I03550d31dd5d8b1230dddc0f8c0b04e022beafce

18 months agoRework security_manager_sync_threads_internal() 66/301866/12
Tomasz Swierczek [Fri, 24 Nov 2023 12:47:07 +0000 (13:47 +0100)]
Rework security_manager_sync_threads_internal()

Add "stop-the-world" implemenation in signal handler.

Why this change may be needed?
------------------------------

On slower platforms, some dotnet-based app candidate processes
tend to have 10-15-... threads. Sending signals that change
Smack labels & drop caps to each of them can be delayed - for apps
that don't have a priority to get launched, this can take up
to 10-15 seconds, in pathological cases (observed in real life,
on low-end boards). In such situation, there's non-zero time window
when some threads have changed security attributes and some
haven't  - access control issues occur in app candidate's logic,
leading to serious errors & abort()s in the launched app.

With this modification, threads change the Smack labels & capabilities
when none of them are performing any of their original app candidate
logic and the only thread NOT in signal handler is the one
with security-manager client library.

This is expected to limit Smack/permission issues mentioned earlier.

Details, disclaimers:
---------------------

This implementation CANNOT fix an issue where app candidate process
created a resource (ie. a file) BEFORE call to security_manager_prepare_app()
and then, wants to access it (for read) AFTER call to the function.
This is because, before the transition, the process has either User
or System::Privileged Smack label, to which apps do not have R Smack rule
(such issues can still occur and can't be dealt with at security-manger level).

This implementation can be up to 3x slower on low-end boards,
as we have 3 waiting points here, instead of one:
    1) waiting for signals to arrive in all threads (place from old implementation)
    2) waiting for signals to be able to start changing caps/Smack labels
    3) waiting for signals to sync after attributes are changed to start work

Tests performed on Tizen emulator show about 30-40% slowdown
of prepare app vs. previous implementation.

This implementation assumes that the size of set of all TIDs
that can appear in a process (already dead + currently alive) will not exceed 1000.
A global array of std::atomic's is used to keep track of each thread's
state in the execution of signal handler.

This implementation uses sleep() and std::atomic to implement
waiting for the barrier inside a signal handler (home-made spinlock,
basically). While this can be done more elegantly, ie. with std::atomic's wait()
functionality, this C++ feature doesn't seem to be supported in current Tizen
toolchain (c++2a supported, not full c++20).

An alternative way for synchronization would be to use high-profile system
resources like fd's or semaphores, however, because the VD H/W platform
where the issue of thread dis-synchronization happens is very slow, interfacing
with kernel was limited to minimum, as adding any context switch to kernel
may slow down even more.

Change-Id: Ic7037acaeb4e3eaab03284ae63216e7ab4d6d862

18 months agoRelease 1.8.12 42/301842/1 accepted/tizen/unified/20231127.165002
Tomasz Swierczek [Fri, 24 Nov 2023 10:20:27 +0000 (11:20 +0100)]
Release 1.8.12

* Add even more logs around IPC

Change-Id: I15012a4e14844397b3115ff8f21a6fb57256d524

18 months agoAdd even more logs around IPC 41/301841/1
Tomasz Swierczek [Fri, 24 Nov 2023 10:19:43 +0000 (11:19 +0100)]
Add even more logs around IPC

Needed by VD division for diagnostic purposes on release images.

Change-Id: Id45f577c50234a80152d646b6dc11f12b6633690

18 months agoRelease 1.8.11 82/301482/1 accepted/tizen/8.0/unified/20231117.165421 accepted/tizen/unified/20231122.172547
Tomasz Swierczek [Thu, 16 Nov 2023 10:49:00 +0000 (11:49 +0100)]
Release 1.8.11

* Add better logs around IPC

Change-Id: Ib5aa13e91250469b64c301ebfc17e7f5e7f2b2d1

18 months agoAdd better logs around IPC 81/301481/2
Tomasz Swierczek [Thu, 16 Nov 2023 10:30:00 +0000 (11:30 +0100)]
Add better logs around IPC

Needed by VD division to debug socket clogging problem in the kernel.

Change-Id: Ic1764b8e9342f93911dcf26f6c03fca028a7f72f

19 months agoRelease 1.8.10 47/300147/2 accepted/tizen/8.0/unified/20231027.023627 accepted/tizen/unified/20231026.032433
Tomasz Swierczek [Tue, 17 Oct 2023 10:33:22 +0000 (12:33 +0200)]
Release 1.8.10

* Add mitigation for long processing of signals in VD's environments

Change-Id: I499a9b16bd2ba9552ac8972c7efba8e26b1906b2

19 months agoAdd mitigation for long processing of signals in VD's environments 45/300145/3
Tomasz Swierczek [Tue, 17 Oct 2023 09:46:40 +0000 (11:46 +0200)]
Add mitigation for long processing of signals in VD's environments

* waiting for signals - timeout increased to 20s from 10s
* added 2 ms timeout if tgkill() will return EAGAIN (happend in actual product image in testing once)

Change-Id: I26ce63e66fc36536136a3eab7903efb07fb35437

21 months agoRelease 1.8.9 85/298085/1 accepted/tizen/8.0/unified/20231005.093706 accepted/tizen/unified/20230907.055346 tizen_8.0_m2_release
Tomasz Swierczek [Thu, 31 Aug 2023 06:25:16 +0000 (08:25 +0200)]
Release 1.8.9

* Create DB file backup in case it was discovered to be corrupted
* Fix app-defined privileges certificate verification
* Migrate to openssl3

Change-Id: I5ebe3bbf8e081d0fbb2efe9f3ae2813db2cb158f

21 months agoCreate DB file backup in case it was discovered to be corrupted 49/297649/16
Tomasz Swierczek [Tue, 22 Aug 2023 07:54:48 +0000 (09:54 +0200)]
Create DB file backup in case it was discovered to be corrupted

This way, the DB can be later analyzed in detail.
Unit tests were modified to check if the backup
is properly created.

This change was requested by VD division.

Change-Id: I151113f793a2d87ea5060931ffeef72f511558cf

21 months agoFix app-defined privileges certificate verification 79/297979/1
Dariusz Michaluk [Tue, 29 Aug 2023 09:30:50 +0000 (11:30 +0200)]
Fix app-defined privileges certificate verification

Since OpenSSL 3.0 a number of additional checks introduced to X509_verify_cert()
according to requirements in RFC 5280 in case X509_V_FLAG_X509_STRICT is set.
For example, the basicConstraints of CA certificates must be marked critical.

This checks cause verification failure for not well-formed certificates.
Having backward compatibility in mind, it's better to disable strict checking
than replacing all broken certificates in applications.

Change-Id: I8c6cb50be7c36ff7ef3d5c221dd4f066c2626be6

21 months agoMigrate to openssl3 95/295695/2
Dariusz Michaluk [Wed, 12 Jul 2023 10:13:04 +0000 (12:13 +0200)]
Migrate to openssl3

Change-Id: Ibab2eed97c6011e303433e48322b3f00fbb48a62

22 months agoRelease 1.8.8 78/296778/1 accepted/tizen/7.0/unified/20230808.014650 accepted/tizen/unified/20230808.011405
Tomasz Swierczek [Fri, 4 Aug 2023 06:26:37 +0000 (08:26 +0200)]
Release 1.8.8

* Fix error checking for open

Change-Id: I8780a3601482f84681f37009aa7dd0fd6ad7f4c8

22 months agoFix error checking for open 55/296755/2
Tomasz Swierczek [Fri, 4 Aug 2023 04:39:49 +0000 (06:39 +0200)]
Fix error checking for open

stdin should not get closed, but anyway, this should get fixed.

Change-Id: Icfeb0e8be0893ec14640f579d7818670d84ef918

22 months agoRelease 1.8.7 81/296181/1 accepted/tizen/7.0/unified/20230726.165535 accepted/tizen/unified/20230725.013002
Krzysztof Jackiewicz [Fri, 21 Jul 2023 08:56:28 +0000 (10:56 +0200)]
Release 1.8.7

* Improve threads' privilege synchronisation

Change-Id: I178bee9a70016bf29ac62de67576b3580ecbdf9b

22 months agoImprove threads' privilege synchronisation 96/296096/3
Tomasz Swierczek [Thu, 13 Jul 2023 14:55:50 +0000 (16:55 +0200)]
Improve threads' privilege synchronisation

* Drop the caps after the threads have been listed for a second time
  (after the sync). This is to avoid errors during accessing /proc for
  newly spawned threads as a unprivileged process.
* Check if newly spawned threads have correct labels.
* Retry the privileges sync twice for all remaining privileged threads.
* Retry listing of /proc/self/task/ in case of failure.
* Use set instead of vector for easier tid checks.
* Omit main thread from the list.

Change-Id: I21e7e5dd3d5efb70fe51a1597bd7bc4ccf1099e8

23 months agoRelease 1.8.6 accepted/tizen/7.0/unified/20230713.041539 accepted/tizen/unified/20230713.014423
Tomasz Swierczek [Tue, 11 Jul 2023 17:14:43 +0000 (19:14 +0200)]
Release 1.8.6

* Revert "Change the signal used by security-manager prepare_app implementation"

Change-Id: I26f7f7e384a4b0ca127acad3399d94e2c3a3fce1

23 months agoRevert "Change the signal used by security-manager prepare_app implementation"
Tomasz Swierczek [Tue, 11 Jul 2023 17:14:07 +0000 (19:14 +0200)]
Revert "Change the signal used by security-manager prepare_app implementation"

This reverts commit 4e26238c9e62a442042145a42207c8b669bf939d.

TCT tests seem to fail when signal is changed to SIGRTMIN+2
from SIGSETXID used previously.

Change-Id: Ib7df235addc30d6fb4b5d503a6f43d012218b55d

23 months agoRelease 1.8.5 99/295499/5
Tomasz Swierczek [Mon, 10 Jul 2023 05:49:31 +0000 (07:49 +0200)]
Release 1.8.5

* Increase waiting time for threads to receive signal
* Change the signal used by security-manager prepare_app implementation
* Fix an ugly bug in printing debug information about threads

Change-Id: I29507a05612dab8fcff9629bd743885b02c19ad5

23 months agoIncrease waiting time for threads to receive signal 98/295498/5
Tomasz Swierczek [Mon, 10 Jul 2023 05:47:24 +0000 (07:47 +0200)]
Increase waiting time for threads to receive signal

As discussed with VD engineers, thats now increased to 10 seconds.
Also, additional log with information about waiting (warning log)
was added.

Change-Id: Iaf76777af2adb2ce65f98a71349a99e1a0704cd5

23 months agoChange the signal used by security-manager prepare_app implementation 45/295445/7
Tomasz Swierczek [Fri, 7 Jul 2023 08:10:20 +0000 (10:10 +0200)]
Change the signal used by security-manager prepare_app implementation

SIGSETXID was used previously to mimic the exact behaviour of glibc.

However, in testing on VD side using SIGSETXID can cause interference
with regular glibc usage, which also leads to abort().

To give more details: if threads are so delayed that their signal handling
is done ie. > 2 seconds after the sending was done, the signal arrives
in the thread when security-manager already did change the signal handler
back to original glibc. Then, glibc, checks whether it sent the signal,
it recognizes it didn't, and aborts/quits.

Changing signal number will not cause glibc to abort/fail, and if execution
gets back to security-manager main thread, IT will fail, printing
all necessary information about threads for debugging, before calling abort().

Changed signal number to SIGRTMIN+2, as advised.

Change-Id: I9b755504898daee71997ee35023c3d647e332b84

23 months agoFix an ugly bug in printing debug information about threads 38/295438/6
Tomasz Swierczek [Fri, 7 Jul 2023 07:28:06 +0000 (09:28 +0200)]
Fix an ugly bug in printing debug information about threads

The for range was calcualted based on wrong variable,
resulting in not all threads information being printed.

Also, changed error handling a bit.

Change-Id: I45fb88c889fb158ba63e0eb55f8d9c813fb40f9c

23 months agoRelease 1.8.4 78/295378/1 accepted/tizen/unified/20230710.013127 accepted/tizen/unified/dev/20230726.115819
Tomasz Swierczek [Thu, 6 Jul 2023 07:29:14 +0000 (09:29 +0200)]
Release 1.8.4

* Fix static analysis issues
* Fix build for new toolchain
* Add additional debug information in prepare_app

Change-Id: I5ff8f36c31f7e5527a816842433a56c1ec2d1a00

23 months agoFix static analysis issues 77/295377/1
Tomasz Swierczek [Thu, 6 Jul 2023 07:26:13 +0000 (09:26 +0200)]
Fix static analysis issues

Change-Id: I90839d874efd4e8b53a7c30870182fbf83c8c248

23 months agoFix build for new toolchain 71/295371/1
Tomasz Swierczek [Thu, 6 Jul 2023 07:05:04 +0000 (09:05 +0200)]
Fix build for new toolchain

Change-Id: I156cf22c4a752057490a6d6d7fbe040358fc933d