Fix GCC lazy init race condition and add test #2562

c1728p9 · 2016-08-26T21:07:38Z

Fix lazy initialization of objects in GCC along with adding a test for it.

geky · 2016-08-26T21:41:16Z

hal/common/retarget.cpp

+        singleton_unlock();
+        return 0;
+    }
+    MBED_ASSERT(0 == *(uint8_t*)guard_object);


Is this assert helpful? Why not change the above to a simple boolean check:

if (*(uint8_t*)guard_object) {

This will catch if the guard object is uninitialized or clobbered.

Ah ok, so it's an assert against the toolchain

geky · 2016-08-26T21:47:53Z

Small nits, but looks good 👍

c1728p9 · 2016-08-26T22:30:43Z

/morph test

c1728p9 · 2016-08-26T22:31:15Z

@mbed-bot: TEST

HOST_OSES=ALL
BUILD_TOOLCHAINS=GCC_ARM
TARGETS=K64F,NRF51_DK,NRF51_MICROBIT,NUCLEO_F411RE

mbed-bot · 2016-08-26T22:33:02Z

[Build 848]
FAILURE: Something went wrong when building and testing.

mbed-bot · 2016-08-27T00:31:52Z

Result: FAILURE

Your command has finished executing! Here's what you wrote!

/morph test

Output

mbed Build Number: 735

Test failed!

pan- · 2016-08-30T09:19:43Z

hal/common/retarget.cpp

@@ -713,6 +713,39 @@ extern "C" void __env_unlock( struct _reent *_r )
 {
    __rtos_env_unlock(_r);
 }
+
+extern "C" int __cxa_guard_acquire (uint64_t *guard_object_p)


The signature is a bit different for ARM aeabi, an int* is used as parameter.
Same thing for __cxa_guard_release and __cxa_guard_abort.

More details here: http://infocenter.arm.com/help/topic/com.arm.doc.ihi0043d/IHI0043D_rtabi.pdf

Good catch @pan- . It is also described on ARM pages, that they use int* as a parameter, even for ARM6 compiler.

@c1728p9 Where did you take these prototypes from ? (uint64_t *guard_object_p) ?

@0xc0170 Russ take them from https://mentorembedded.github.io/cxx-abi/abi.html#once-ctor .

I updated this to more closely match the run-time ABI as specified by ARM.

0xc0170 · 2016-08-30T10:37:20Z

nit:
The commit message:

Implement the functions __cxa_guard_acquire, __cxa_guard_release and
__cxa_guard_abort so any objects lazily initialized in GCC are done so
in a thread safe manner.

I assume the aim of the fix is to guard lazily initialised objects (as they are done at any time), but this fix is regarding locks for local static variable initialisation. (might be more clear to a reader).

+1 for providing locks for this to have it thread safe.

Test SingletonPtr initailization along the with lazy initailization that is built into the C++ language itself.

Implement the functions __cxa_guard_acquire, __cxa_guard_release and __cxa_guard_abort so lazily initialized function-local static objects are done so in a thread safe manner in GCC.

c1728p9 · 2016-08-30T16:26:19Z

Rebased pr to master, updated functions to match run-time ABI from ARM and updated commit message to make it more clear.

c1728p9 · 2016-08-30T16:26:23Z

/morph test

mbed-bot · 2016-08-30T18:08:34Z

Result: FAILURE

Your command has finished executing! Here's what you wrote!

/morph test

Output

mbed Build Number: 752

Test failed!

c1728p9 · 2016-08-30T19:15:02Z

/morph test

0xc0170 · 2016-08-31T11:03:15Z

Seems like this job stalled CI. I'll try to terminate it or ask someone do it

mbed-bot · 2016-08-31T13:48:34Z

Result: FAILURE

Your command has finished executing! Here's what you wrote!

/morph test

Output

mbed Build Number: 755

Test failed!

c1728p9 · 2016-08-31T22:00:46Z

/morph test

mbed-bot · 2016-08-31T22:04:00Z

Result: ABORTED

Your command has finished executing! Here's what you wrote!

/morph test

Output

mbed Build Number: 765

Build Prep failed!

bridadan · 2016-08-31T22:32:07Z

/morph test

mbed-bot · 2016-09-01T00:02:54Z

Result: FAILURE

Your command has finished executing! Here's what you wrote!

/morph test

Output

mbed Build Number: 766

Test failed!

c1728p9 · 2016-09-01T02:05:25Z

/morph test

mbed-bot · 2016-09-01T05:14:59Z

Result: FAILURE

Your command has finished executing! Here's what you wrote!

/morph test

Output

mbed Build Number: 770

Test failed!

sg- · 2016-09-08T23:22:42Z

@mbed-bot: TEST

HOST_OSES=ALL
BUILD_TOOLCHAINS=ALL
TARGETS=ALL

sg- · 2016-09-08T23:22:47Z

/morph test

mbed-bot · 2016-09-09T00:47:54Z

Result: SUCCESS

Your command has finished executing! Here's what you wrote!

/morph test

Output

mbed Build Number: 788

All builds and test passed!

mbed-bot · 2016-09-09T01:33:14Z

[Build ${MBED_BUILD_ID}]
FAILURE: Something went wrong when building and testing.

bridadan · 2016-09-09T02:12:18Z

@mbed-bot: TEST

HOST_OSES=ALL
BUILD_TOOLCHAINS=ALL
TARGETS=ALL

mbed-bot · 2016-09-09T15:37:29Z

[Build 882]
FAILURE: Something went wrong when building and testing.

Release mbed-os-5.1.4 Changes: New Targets: 2504: [Disco_F769NI] adding new target [#2504] 2654: DELTA_DFBM_NQ620 platform porting [#2654] 2615: [MTM_MTCONNECT04S] Added support for MTM_MTCONNECT04S [#2615] 2548: Nucleof303ze [#2548] Fixes: 2678: Fixing NCS36510 compile on Linux #2678 2657: [MAX326xx] Removed echoing of characters and carriage return. #2657 2651: Use lp_timer to count time in the deepsleep tests #2651 2645: NUCLEO_F446ZE - Enable mbed5 release version #2645 2643: Fix thread self termination #2643 2634: Updated USBHost for library changes #2634 2633: Updated USBDevice to use Callback #2633 2630: Test names not dependent on disk location of root #2630 2624: CFSTORE Bugfix for realloc() moving KV area and cfstore_file_t data structures not being updated correctly #2624 2623: DISCO_L476VG - Add Serial Flow Control pins + add SERIAL_FC macro #2623 2617: STM32F2xx - Enable Serial Flow Control #2617 2613: Correctly providing directories to build_apis #2613 2607: Fix uvisor memory tracing #2607 2604: Tools - Fix fill section size variation #2604 2601: Adding ON Semiconductor copyright notice to source and header files. #2601 2597: [HAL] Fixed "intrinsic is deprecated" warnings #2597 2596: [HAL] Improve memory tracer #2596 2594: Fix TCPServer constructor #2594 2593: Add app config command line switch for test and make #2593 2589: [NUC472] Fix heap configuration error with armcc #2589 2588: Timing tests drift refactor #2588 2587: add PTEx pins as option for SPI on Hexiwear - for SD Card Interface #2587 2584: Set size of callback irq array to IrqCnt #2584 2583: github issue and PR templates #2583 2582: [GCC_CR] fix runtime hang for baremetal build #2582 2580: lwip - Add check for previously-bound socket #2580 2579: lwip - Fix handling of max sockets in socket_accept #2579 2578: Fix double free in NanostackInterface #2578 2576: Add smoke test that builds example programs with mbed-cli #2576 2575: tools-config! - Allow an empty or mal-formed config to be passed to the config system #2575 2562: Fix GCC lazy init race condition and add test #2562 2559: [utest]: Allow the linker to remove any part of utest if not used #2559 2545: Added define guards for SEQUENTIAL_FLASH_JOURNAL_MAX_LOGGED_BLOBS so #2545 2538: STM32F4xx - Add support of ADC internal channels (Temp, VRef, VBat) #2538 2521: [NUCLEO_F207ZG] Add MBED5 capability #2521 2514: Updated FlexCan and SAI SDK drivers #2514 2487: Runtime dynamic memory tracing #2487 2442: Malloc heap info #2442 2419: [STM32F1] Add asynchronous serial #2419 2393: [tools] Prevent trace-backs from incomplete args #2393 2245: Refactor export subsystem #2245 2130: stm32 : reduce number of device.h files #2130

Changes: New Targets: #2504: [Disco_F769NI] adding new target [ARMmbed/mbed-os#2504] #2654: DELTA_DFBM_NQ620 platform porting [ARMmbed/mbed-os#2654] #2615: [MTM_MTCONNECT04S] Added support for MTM_MTCONNECT04S [ARMmbed/mbed-os#2615] #2548: Nucleof303ze [ARMmbed/mbed-os#2548] Fixes: #2657: [MAX326xx] Removed echoing of characters and carriage return. [ARMmbed/mbed-os#2657] #2651: Use lp_timer to count time in the deepsleep tests [ARMmbed/mbed-os#2651] #2643: Fix thread self termination [ARMmbed/mbed-os#2643] #2623: DISCO_L476VG - Add Serial Flow Control pins + add SERIAL_FC macro [ARMmbed/mbed-os#2623] #2617: STM32F2xx - Enable Serial Flow Control [ARMmbed/mbed-os#2617] #2601: Adding ON Semiconductor copyright notice to source and header files. [ARMmbed/mbed-os#2601] #2597: [HAL] Fixed "intrinsic is deprecated" warnings [ARMmbed/mbed-os#2597] #2589: [NUC472] Fix heap configuration error with armcc [ARMmbed/mbed-os#2589] #2587: add PTEx pins as option for SPI on Hexiwear - for SD Card Interface [ARMmbed/mbed-os#2587] #2584: Set size of callback irq array to IrqCnt [ARMmbed/mbed-os#2584] #2582: [GCC_CR] fix runtime hang for baremetal build [ARMmbed/mbed-os#2582] #2562: Fix GCC lazy init race condition and add test [ARMmbed/mbed-os#2562] #2538: STM32F4xx - Add support of ADC internal channels (Temp, VRef, VBat) [ARMmbed/mbed-os#2538] #2514: Updated FlexCan and SAI SDK drivers [ARMmbed/mbed-os#2514] #2442: Malloc heap info [ARMmbed/mbed-os#2442] #2419: [STM32F1] Add asynchronous serial [ARMmbed/mbed-os#2419] #2130: stm32 : reduce number of device.h files [ARMmbed/mbed-os#2130] #2678: Fixing NCS36510 compile on Linux [ARMmbed/mbed-os#2678] #2607: Fix uvisor memory tracing [ARMmbed/mbed-os#2607] #2596: [HAL] Improve memory tracer [ARMmbed/mbed-os#2596] #2487: Runtime dynamic memory tracing [ARMmbed/mbed-os#2487]

c1728p9 mentioned this pull request Aug 26, 2016

[utest]: Allow the linker to remove any part of utest if not used #2559

Merged

sg- added needs: review needs: CI labels Aug 26, 2016

geky reviewed Aug 26, 2016
View reviewed changes

c1728p9 force-pushed the init_race_condition branch from fc83e0d to ecd4f11 Compare August 26, 2016 21:59

pan- reviewed Aug 30, 2016
View reviewed changes

c1728p9 added 2 commits August 30, 2016 11:23

Add a test to check for init race conditions

69431da

Test SingletonPtr initailization along the with lazy initailization that is built into the C++ language itself.

Fix GCC locks for lazy object initailization

ef45ef8

Implement the functions __cxa_guard_acquire, __cxa_guard_release and __cxa_guard_abort so lazily initialized function-local static objects are done so in a thread safe manner in GCC.

c1728p9 force-pushed the init_race_condition branch from ecd4f11 to ef45ef8 Compare August 30, 2016 16:23

sg- removed the needs: review label Sep 8, 2016

sg- added ready for merge and removed needs: CI labels Sep 9, 2016

sg- merged commit 0128dd2 into ARMmbed:master Sep 9, 2016

sg- added release-version: 5.1.4 and removed ready for merge labels Sep 9, 2016

bridadan mentioned this pull request Nov 21, 2016

RZ_A1H Tests Failing #3273

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix GCC lazy init race condition and add test #2562

Fix GCC lazy init race condition and add test #2562

c1728p9 commented Aug 26, 2016

geky Aug 26, 2016 •

edited

Loading

c1728p9 Aug 26, 2016

geky Aug 26, 2016 •

edited

Loading

geky commented Aug 26, 2016

c1728p9 commented Aug 26, 2016

c1728p9 commented Aug 26, 2016

mbed-bot commented Aug 26, 2016

mbed-bot commented Aug 27, 2016

pan- Aug 30, 2016

0xc0170 Aug 30, 2016

0xc0170 Aug 30, 2016

pan- Aug 30, 2016

c1728p9 Aug 30, 2016

0xc0170 commented Aug 30, 2016

c1728p9 commented Aug 30, 2016

c1728p9 commented Aug 30, 2016

mbed-bot commented Aug 30, 2016

c1728p9 commented Aug 30, 2016

0xc0170 commented Aug 31, 2016

mbed-bot commented Aug 31, 2016

c1728p9 commented Aug 31, 2016

mbed-bot commented Aug 31, 2016

bridadan commented Aug 31, 2016

mbed-bot commented Sep 1, 2016

c1728p9 commented Sep 1, 2016

mbed-bot commented Sep 1, 2016

sg- commented Sep 8, 2016

sg- commented Sep 8, 2016

mbed-bot commented Sep 9, 2016

mbed-bot commented Sep 9, 2016

bridadan commented Sep 9, 2016

mbed-bot commented Sep 9, 2016

Fix GCC lazy init race condition and add test #2562

Fix GCC lazy init race condition and add test #2562

Conversation

c1728p9 commented Aug 26, 2016

geky Aug 26, 2016 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

geky Aug 26, 2016 • edited Loading

Choose a reason for hiding this comment

geky commented Aug 26, 2016

c1728p9 commented Aug 26, 2016

c1728p9 commented Aug 26, 2016

mbed-bot commented Aug 26, 2016

mbed-bot commented Aug 27, 2016

Output

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

0xc0170 commented Aug 30, 2016

c1728p9 commented Aug 30, 2016

c1728p9 commented Aug 30, 2016

mbed-bot commented Aug 30, 2016

Output

c1728p9 commented Aug 30, 2016

0xc0170 commented Aug 31, 2016

mbed-bot commented Aug 31, 2016

Output

c1728p9 commented Aug 31, 2016

mbed-bot commented Aug 31, 2016

Output

bridadan commented Aug 31, 2016

mbed-bot commented Sep 1, 2016

Output

c1728p9 commented Sep 1, 2016

mbed-bot commented Sep 1, 2016

Output

sg- commented Sep 8, 2016

sg- commented Sep 8, 2016

mbed-bot commented Sep 9, 2016

Output

mbed-bot commented Sep 9, 2016

bridadan commented Sep 9, 2016

mbed-bot commented Sep 9, 2016

geky Aug 26, 2016 •

edited

Loading

geky Aug 26, 2016 •

edited

Loading