Experimental Physics and
Industrial Control System

Len Lawrence <[email protected]> · Wed, 5 Mar 1997 15:36:10 +0000 (GMT)

Sorry folks for introducing a VxWorks query, but after two weeks of
frustration trying to make the EPICS log server work I have returned
to simple file i/o (fopen, fprintf, etc) on an mv167 IOC running
VxWorks 5.2.  It would be useful to be able to examine the output on
the fly but no file ever appears until fclose is called.  fflush does
not do the business.  In general our applications never exit - they
simply run and run until a reboot, so there is never a time when
fclose should be called.  One clumsy attempt at a solution was to
fopen the host file with "w+" access and to close it periodically and
reopen it with "a+" access.  Now this works fine under Unix, and
should do on the mv167 according to the VxWorks manual, but in fact
VxWorks complains that the file already exists, which it must.

Does anybody know what is going on?  Is our VxWorks broken or what?

Len

==============================================================================
Len Lawrence                       Tel.   : +44 (0)131 668 8255 (direct)
Royal Observatory                           +44 (0)131 668 8100 (switchboard)
Blackford Hill                     [email protected]
Edinburgh EH9 3HJ                  aka [email protected]
UK                                 WWW    : http://www.roe.ac.uk/lclwww/

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System