Experimental Physics and
Industrial Control System

[email protected] (Jeff Hill) · Thu, 11 Feb 1999 16:54:06 -0700

Geoff,

My initial message on the proper consistent use of EPICS C type "dbr_long_t" with 
type DBR_LONG is still valid. Ehhh, However ...

After a closer inspection of this particular change I now see that you
are making  a valid fix for a 64 bit specific bug which has existed at least
since 3.12.

Perhaps the standard tests have not been run on the alpha architecture
(and this data type has not been used with a put before on the alpha)?
Could you verify that "ca_test <channel name>" reproduces this bug before
the fix is applied?

I have committed your change (and several similar fixes) into CVS at APS.

Sorry about the confusion.

Now if there was only an undo for that mail send button...

Jeff

On Thursday, February 11, 1999 4:27 PM, Jeff Hill [SMTP:[email protected]] wrote:
> 
> 
> On Thursday, February 11, 1999 1:36 PM, Geoff Savage [SMTP:[email protected]] wrote:
> > 
> > I moved to the libCom directory and built libCom.a.  Then I went to the ca
> > directory and made one minor change.
> > 
> > <d0ola> diff access.c access.c.orig
> > 1915c1915
> > < *(dbr_long_t *)pdest = htonl (*(dbr_long_t *)pvalue);
> > ---
> > > *(long *)pdest = htonl (*(dbr_long_t *)pvalue);
> > <d0ola>
> > 

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System