On 24/11/2014 22:27, Michael Black wrote:
Hi Mike,
> The string length logic seems a bit inconsistent on hamlib.
> It looks to me like all rigs that use hamlib serial com should be having
> problems.
>
> I checked some other rigs and it appears they should have similar problems
> with truncated responses.
>
> If you read_string() with a correct expected length for the response you get
> back length-1. Maybe this doesn't cause problems if most rigs return a 0x0d
> or such on the end of their responses.
> If you ask for more bytes than you expect you get a timeout and an extra
> byte.
>
> It's fairly obvious that you don't want to hit a timeout on every
> command...so the "expected response length" should be the right argument.
> The Omni VII has some commands which are binary and exact length is returned
> plus the implementation for several commands was wrong and I'm fixing those.
> I've got basic operation working now and am working on split mode.
>
> In read_string() it does this incorrect logic which ends up return one less
> byte than requested
> while (total_count < rxmax-1) {
> Presumably to make room for this
> rxbuffer[total_count] = '\000';
>
> I changed that and commented out the insertion of the terminator
> while (total_count < rxmax) {
> // rxbuffer[total_count] = '\000';
> // maybe this should be rxbuffer[total_count+1]='\000' but I'm leary of this
> since I'm sure many calling functions may not expect to have to allow for an
> extra byte
> Either string_read should add the terminating zero and the calling functions
> ensure they have length+1 bytes available....
> Or...the calling functions should be responsible for ensuring a terminating
> zero.
>
>
> With that change it behaves for the Omni VII with the corrected command set
> that I've implemented getting back the correct length for all commands.
>
> And I think the extra byte on timeout is fixed with this in iofunc.c by
> adding a rd_count > 0 check instead of always incrementing total_count
> rd_count = port_read(p, &rxbuffer[total_count], 1);
> if (rd_count > 0) {
> ++total_count;
> }
These changes are very familiar. What version of hamlib are you basing
your changes on?
> Mike W9MDB
>
73
Bill
G4WJS.
------------------------------------------------------------------------------
Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
from Actuate! Instantly Supercharge Your Business Reports and Dashboards
with Interactivity, Sharing, Native Excel Exports, App Integration & more
Get technology previously reserved for billion-dollar corporations, FREE
http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
_______________________________________________
wsjt-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/wsjt-devel
