This is the mail archive of the ecos-patches@sources.redhat.com mailing list for the eCos project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

RE: RFC - fix to at49xxxx crash during erase/program when interrupt happens


> No, I meant to say that the code executed by your threads
> would still reside in flash and hence no threads can run
> while the flash is being programmed. 

I see. 

The application could keep a list of "flash safe" and "flash
unsafe" threads, implement at49xxxx_enable/disable() and override
the default functionality to suspend "flash unsafe" threads.

I'd still like to have interrupts running all the time though.

> It is very common,
> and generally desirable, to have the interrupt handlers
> do as little work as possible, and to delegate as much as 
> possible to a thread they notify.

IMHO, the at91 serial drivers should do more than it does
today on the interrupt: the drivers should store the 
byte incoming byte in a ring buffer 
and do the rest on a DSR or high priority thread.

So far I've had little luck with >38400 baud serial ports.

Øyvind


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]