search

bluekitchen / btstack

Dual-mode Bluetooth stack, with small memory footprint.
http://bluekitchen-gmbh.com
Other
1.74k stars 618 forks source link

why VMIN = 1，is this useless? #501

Closed yangn0 closed 1 year ago

yangn0 commented 1 year ago

in the platform/posix/btstack_uart_posix.c

line 309 , open() set the O_NONBLOCK. In non-blocking mode, VMIN/VTIME have no effect (POSIX specifications).

line 330 , why VMIN = 1?

Is this useless?

yangn0 commented 1 year ago

https://stackoverflow.com/questions/20154157/termios-vmin-vtime-and-blocking-non-blocking-read-operations

mringwal commented 1 year ago

Hi. To be honest, I don't remember. Maybe I've used it without non-blocking first. Then, there might have been a reason (some USB-UART adapter, some USB-CDC device..) that didn't work with it and setting non-blocking made it work. Feel free to browse the version history.

Why did you stumble upon our uart driver?

mringwal commented 1 year ago

In any case, it's not a bug :)

yangn0 commented 1 year ago

Sorry for replying so late. thanks for you reply :)

I am porting btstack to a POSIX system. I have a problem about POSIX compliant.(https://devel.rtems.org/ticket/4926) I want to make it clear once and for all.

mringwal commented 1 year ago

The idea is that the run loop blocks with select() or similar until the fd becomes readable. Then, it calls read with a large buffer and expects that the read call will return immediately to not block.

For write, we don't want it to block either. Given that there's hardware flow control which might prevent the write call from immediate success, the non-blocking option must stay. If we have non-blocking, then the VMIN setting is probably not needed.