Article 121445 of comp.os.vms: peressini@espace.dcl.com wrote: > > In article <3326E79A.50AD@mvpsi.com.nospam> John Vottero wrote: > > > I have it working under OpenVMS/VAX V6.2/DEC C V5.3. I could NOT get it > > to work under OpenVMS/Alpha V7.0/DEC C V5.5. The qio to do the write > > always failed with %SYSTEM-F-BADPARAM. I'm not sure if it's an > > Alpha/VAX issue or a V6.2/V7.0 issue. If anyone knows why GKDRIVER will > > return SYSTEM-F-BADPARAM please let me know (all of the parameters look > > fine to me!). Also, does anyone know what happened to the documentation > > for GK driver? I can't find it in the V7.1 Bookreader documentation. I > > have the V5.4 documentation but I'm looking for the lastest. > > I got CDWRITE13_VMS to work on a VAXstation 3100 M38 running > OpenVMS/VAX V6.0 and VAX C V3.2 using an HP 4020i with 1.20 firmware. > I had to change some things to get it to work. I was getting the > exact same error message at the write_6 step. I knew from past > performance tests on this machine that IOs greater than 31 blocks > (15872 bytes) were slower than expected. So I tried changing the > BLOCKS_PER_WRITE value in the code to 7 (14336 bytes) from 25 (51200 > bytes) and now the program works. I also made the container file > a multiple of 28 blocks instead of 100. I think the SCSI hardware > in my VAXstation 3100 M38 is old or broken somehow. It dosen't seem > likely that ALPHA systems would have the same problem, but it is an > easy think to try. > -- > I found out what the problem was with the Alpha and it is now fixed. The SCSI drivers for Alpha do more checking than previous (i.e. VAX) versions. The WRITE_6 routine was specifying SCSI timeouts of 3600 seconds and the SCSI standard says that the max is 3000. Lowering the timeouts fixed the problem.