When STM32F746-DISCO board was being used in (unsupported) USBHost mode,
the communication was unreliable. Our investigation revealed that the
problem lied in redundant IN tokens that the host generated even though
it shouldn't. This could lead to endless high-frequency NAKs being
received from device, which caused watchdog reset as USBHost spent all
time in interrupt handlers.
In our application the clocks frequencies are:
* HCLK = 48 MHz
* APB1 = 6 MHz
* APB2 = 12 MHz
We have captured the raw USB High-Speed traffic using OpenVizsla.
Without this change, when USB MSD device connected to the system
responded to IN with NAK, there were excessive IN tokens generated about
667 ns after the NAK. With this commit the IN tokens are generated no
sooner than 10 us after the NAK.
The high frequency of the IN/NAK pairs is not the biggest problem.
The biggest problem is that the USB Host did continue to send the IN token
after DATA and ACK packets were received from device - *without* any request
from upper layer (USB MSD).
The USB MSD devices won't have extra data available on Bulk IN endpoint
after the expected data was received by Host. In such case IN/NAK cycle
time is only houndreds of nanoseconds, the MCU has no time for anything else.
The problem manifested not only on Bulk endpoints, but also during
Control transfers. Example correct scenario (when this fix is applied):
* SETUP stage
* SETUP [host -> address 0 endpoint 0]
* DATA0 [80 06 00 01 00 00 08 00] [CRC16: EB 94]
* ACK
* DATA stage
* IN
* NAK
... the IN/NAK repeated multiple time until device was ready
* IN
* DATA1 [12 01 10 02 00 00 00 40] [CRC16: 55 41]
* ACK
* STATUS stage
* OUT
* DATA1 ZLP
* ACK
Without this commit, in DATA stage, after the ACK was received, the host
did send extra IN to which device responded with STALL. On bus it was:
* DATA stage
...
* IN
* DATA1 [12 01 10 02 00 00 00 40] [CRC16: 55 41]
* IN
* STALL
* IN
* STALL
* STATUS stage
* OUT
* DATA1 ZLP
* STALL
In the fault case the next SETUP was sent only after 510 ms, which
indicates timeout in upper layer.
With this commit the next SETUP is sent 120 us after the STATUS stage ACK.
LP TICKER mbed-os wrapper needs to be disabled as it introduces too much latencies.
LP TICKER wrapper has been disabled and we need to managed the HW constraints at low level:
- main HW constraint is that once the comparator has been programmed once,
driver cannot program it again before CMPOK HW flag is set, which takes about 3 30us cycles.
To make it even more complex, the driver also needs to cope with "LP ticker workaround"
See commit:
LP ticker workaround
There is an errata in LPTIM specification that explains that CMP Flag
condition is not an exact match (COUNTER = MATCH) but rather a
comparison (COUNTER >= MATCH).
Also the disable interrupt is more complete now:
- always check sleep manager status and restore it
- remove irq_handler as comparator is always programed and might get called
eventually when LP TICK is restarted
- reset delayed_prog
Also in set_interrupt, make sure interrupt does not fire early.
If needed, we decide to slightly delay the tick to cope with the HW limitation to
make sure it will fire as soon as HW is capable.
Functions are called under critical section as they may be called from
the IRQ handler now, not only from driver layer.
There is an errata in LPTIM specification that explains that CMP Flag
condition is not an exact match (COUNTER = MATCH) but rather a
comparison (COUNTER >= MATCH).
As a consequence the interrupt is firing early than expected when
programing a timestamp after the 0xFFFF wrap-around.
In order to
work-around this issue, we implement the below work-around.
In case timestamp is after the work-around, let's decide to program the
CMP value to 0xFFFF, which is the wrap-around value. There would anyway be
a wake-up at the time of wrap-around to let the OS update the system time.
When the wrap-around interrupt happen, OS will check the current time and
program again the timestamp to the proper value.
Add a pulse when using hardware chip select for SPI transmissions.
CS is at low level when a transmission is on-going.
Be careful, this is not compatible with all modes. It will work only
if PHA is 0, ie spi mode is 0 or 2. See stm32xx reference manual,
chapter "NSS pulse mode" for more details.
Fix#10671
Signed-off-by: Vincent Veron <vincent.veron@st.com>
Fix WATCHDOG_STATUS_INVALID_ARGUMENT for timeout values from 1 ms to
407 ms (fix inability to set prescaler bits to zero).
Fix timeout <-> IWDG registers conversions.
Fix read & write access to IWDG_PR and IWDG_RLR registers.
Fix LSI RC frequency setting.
Limit MAX_TIMEOUT_MS to UINT32_MAX.
int_szyk
Martin Kojtal
Laurent Meunier
Seth Itow
Szymon Szantula
Ireneusz Gaicki
jeromecoutant
Seppo Takalo
Kevin Bracey
Yuan Cao
Tymoteusz Bloch
Arto Kinnunen
itziar
Tymoteusz Bloch
Alexandre Bourdiol
Malavika Sajikumar
Vincent Veron
Filip Jagodzinski
toyowata
Teijo Kinnunen
Filip Jagodzinski
Steven Cartmell