[STM32] Sleep code refactor #2192
Conversation
In order to enable sleep API factorization, HAL_Suspend/ResumeTick functions have been implemented in hal_tick.c for each platform.
In order to enable sleep API factorization, HAL_Suspend/ResumeTick functions have been implemented in hal_tick.c for each platform.
In order to enable sleep API factorization, HAL_Suspend/ResumeTick functions have been implemented in hal_tick.c for each platform.
In order to enable sleep API factorization, HAL_Suspend/ResumeTick functions have been implemented in hal_tick.c for each platform.
In order to enable sleep API factorization, HAL_Suspend/ResumeTick functions have been implemented in hal_tick.c for each platform.
In order to enable sleep API factorization, HAL_Suspend/ResumeTick functions have been implemented in hal_tick.c for each platform.
|
Build issue due to late rebase. |
In order to enable sleep API factorization, HAL_Suspend/ResumeTick functions have been implemented in hal_tick.c for each platform.
In order to enable sleep API factorization, HAL_Suspend/ResumeTick functions have been implemented in hal_tick.c for each platform.
erwango
force-pushed
the
sleep_code_refactor
branch
from
fbddb16 to
f712041
Compare
|
Non regression results (No regression found vs master status (which has some issues already)) |
|
Why hal_tick.c is per target and not for a family, for instance Other than that, looks good. +1 |
For now, hal_tick.c contains differences accross boards in some families. For instance, some targets dont have native 32 bit counters, hence we use addtional IT vs 32bit counter enabled boards (NUCLEO_F030R8 for instance). |
|
This is generally related to STM32 and BLUEPILL_F103C8 should be updated accordingly. I can either make fork and PR to this branch or wait for this PR to be merged to upstream master and then fork & PR to mbedmicro/mbed. How do you prefer this to be done? |
@zgoda Send a separate PR once this one lands. Thanks |
|
@mbed-bot: TEST HOST_OSES=windows |
|
[Build 669] |
|
Test results are fine |
Rework of sleep.c file across all STM32 families.
Aim was to correct sleep bug in DISCO_F469NI, not entering sleep mode due to use of wrong timer.
In the same time I took opportunity to refactor sleep code taking advantage of hal_tick.c functions.
STM32F7xxx code benefits from __INLINE, which is not compiling on other families for ARM and uARM toolchains (not compatible with __weak as per keil documentation).