[Nut-upsdev] Patch for the bestfortress driver 0.02

[Arnaud Quette]

Hi Jørgen, Oliver and Stuart,

I'm not forgetting you...
while looking again at bestfortress, I realized I completely missed
Jørgen's patch!

2010/4/19 J. Thomsen <list at jth.net>:
> I could not make the bestfortress driver 0.02 work for my Best Fortress LI675VA,
> so I have made a patch for the driver.
> The reason was, that all responses are preceded and terminated by CR LF, but the driver would discard everything after the first CR
> LF.
>
> The following patch discards everything up to the first LF.
> It also adds a debugging feature of displaying the accepted data  (level 5 i.e. -D -D -D -D -D -D)
>
> /Jørgen Thomsen
>
> --- nut-2.4.3/drivers/bestfortress.c    2010-02-11 22:43:23.000000000 +0100
> +++ nut-2.4.3j/drivers/bestfortress.c   2010-04-19 12:17:07.929852889 +0200
> @@ -34,7 +34,7 @@
>  #endif
>
>  #define DRIVER_NAME             "Best Fortress UPS driver"
> -#define DRIVER_VERSION  "0.02"
> +#define DRIVER_VERSION  "0.03"
>
>  /* driver description structure */
>  upsdrv_info_t   upsdrv_info = {
> @@ -203,20 +203,26 @@
>        int loadva;
>        int len;
>        int retry;
> +       char ch;
>
>        int checksum_ok, is_online=1, is_off, low_batt, trimming, boosting;
>
>        for (retry = 0; retry < 5; ++retry) {
>                upsflushin (0, 0, "\r ");
>                upssend ("f\r");
> +               while (ser_get_char(upsfd, &ch, 0, UPSDELAY) > 0 && ch != '\n'); // response starts with \r\n
> +               temp[2] = 0;
>                do {
> -                       if (upsrecv (temp+2, sizeof temp - 2, ENDCHAR, IGNCHARS) <= 0) {
> +                       if ((len = upsrecv (temp+2, sizeof temp - 2, ENDCHAR, IGNCHARS)) <= 0) {
>                                upsflushin (0, 0, "\r ");
>                                upssend ("f\r");
> +                               while (ser_get_char(upsfd, &ch, 0, UPSDELAY) > 0 && ch != '\n'); // response starts with \r\n
>                        }
>                } while (temp[2] == 0);
>
> -               /*syslog (LOG_DAEMON | LOG_NOTICE,"ups: got '%s'\n", p);*/
> +               /* syslog (LOG_DAEMON | LOG_NOTICE,"ups: got %d chars '%s'\n", len, temp + 2); */
> +
> +               upsdebugx(5, "data from UPS: len=%d '%s'", len, &temp[2]);
>                /* status example:
>                   000000000001000000000000012201210000001200014500000280600000990025000000000301BE
>                   000000000001000000000000012401230000001200014800000280600000990025000000000301B7


I've still not had time to actually get back on the protocol and implementation.
All apologies for that: budget restrictions and time constraints are
killing me :-(

That said, I've updated Jørgen's patch to apply to the current
trunk... just to see if things are a bit better.
@Jørgen, Oliver and Stuart: would you be kind enough to test the attached patch?
I'm not much confident considering the change, and Oliver's specific
issue (with a 660 LI)...

I've still a small room to address some final things for 2.6.4.
So if possible within this timeframe, I'd like to check if we can
solve your issues, and improve the situation...

Thanks for your help.

cheers,
Arnaud
-- 
Linux / Unix Expert R&D - Eaton - http://powerquality.eaton.com
Network UPS Tools (NUT) Project Leader - http://www.networkupstools.org/
Debian Developer - http://www.debian.org
Free Software Developer - http://arnaud.quette.free.fr/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: bestfortress-JTomsen-AQ.diff
Type: application/octet-stream
Size: 1730 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/nut-upsdev/attachments/20120525/43fd77cc/attachment.obj>