Re: Stream device problem with redirection in @mismatch

[Dirk Zimoch <dirk.zimoch@psi.ch>]

On 10.04.2015 18:50, Michael Johnson wrote:

> In Stream, I do not set any terminators (redundant?).

You can set terminators in the startup script using asynOctetSet*Eos or 
in the StreamDevice protocol. The protocol defined terminators overwrite 
those in the startup script (StreamDevice restores them when it is done 
with the protocol).

> Program heap: free of bad pointer 98assertion "(uread(BSP_UART_COM1,
> LSR) & THRE) != 0" failed: file
> "../../../../../../../../__rtems-4.9.2/c/src/lib/libbsp/__i386/pc386/../../i386/shared/__comm/uart.c",
> line 585, function: BSP_uart_termios_write_com1

This seems to be outside StreamDevice. But if you send me a stack dump I 
may able to find out from where it gets called.

> (after changing "print" to "printf" to get it to compile)

Maybe this gets called wrongly. I renamed the StreamBuffer method from 
"printf" to "print" because I found out that on some platforms (RTEMS) 
printf is actually a macro (for fprintf(stdout, ...))

I really need to publish the current StreamDevice version. :-)

BTW: assertions should not call functions or macros. They should only 
test but never modify values, because switching them off results in no 
code at all, thus code like this is "illegal":

assert((uread(BSP_UART_COM1, LSR) & THRE) != 0);

The legal variant would be something like this:

int ureadResult = uread(BSP_UART_COM1, LSR);
assert ((ureadResult & THRE) != 0);

Dirk