Yes, indeed, that is wrong.  Thanks, and thanks for the properly 
formatted patch.  Applied and pushed.

-corey

On 04/04/2016 09:06 AM, Janne Huttunen wrote:
> Don't set the 'write_enabled' field explicitly to one when pushing
> the telnet initialization sequence to the buffer. It will be done
> by the set_write_enable() call a bit later at the end of the routine.
>
> If the field is explicitly set to one, the set_write_enable() call
> won't notify the lower layer and the corresponding file descriptor
> will never be enabled for writing. This causes any input coming from
> the SoL connection just to be buffed indefinitely and never sent to
> the telnet server.
>
> Signed-off-by: Janne Huttunen <[email protected]>
> ---
>   lanserv/sol.c | 1 -
>   1 file changed, 1 deletion(-)
>
> diff --git a/lanserv/sol.c b/lanserv/sol.c
> index 01c4126..2a6afb3 100644
> --- a/lanserv/sol.c
> +++ b/lanserv/sol.c
> @@ -1062,7 +1062,6 @@ sol_port_init(ipmi_sol_t *sol)
>   
>       memcpy(sd->inbuf, sol_telnet_initseq, len);
>       sd->inlen = len;
> -     sd->write_enabled = 1;
>       }
>   
>       err = sd->sys->add_io_hnd(sd->sys, sd->fd, sol_data_ready, sol, 
> &sd->fd_id);


------------------------------------------------------------------------------
_______________________________________________
Openipmi-developer mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/openipmi-developer

Reply via email to