On Sat, 25 Jun 2011 22:28:07 +0300, Binyamin Dissen <[email protected]> wrote:
>On Fri, 24 Jun 2011 10:48:14 -0500 Walt Farrell <[email protected]> wrote: > >:>On Fri, 24 Jun 2011 15:32:19 +0300, Binyamin Dissen >:><[email protected]> wrote: > >:>>After the GETSHARED I do a > >:>> IARV64 REQUEST=SHAREMEMOBJ, >:>> USERTKN=ORIGIN, >:>> RANGLIST=@RANGLIST, >:>> ALETVALUE=0, >:>> COND=NO, >:>> MF=(E,OIARV64,COMPLETE) > >:>>setting the USERTKN to the base address of the block. > >:>I'm not sure you're setting your USERTKN field (ORIGIN) as the book >:>suggests. It says the token should be something unique to the >:>caller/request, and I don't see that the address of the block is necessarily >:>unique in the way the book suggests. > >The address of the 64 bit shared block is quite unique. True, but do you ever have more than one requestor getting the same block? If so, I think they need different USERTKN values on the SHAREMEMOBJ call or the first one that deletes its interest in the block will also delete the other user's interest. And then the second delete will fail. And since you're seeing a delete fail... -- Walt ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
