In a message dated 11/21/2007 5:44:16 P.M. Central Standard Time, [EMAIL PROTECTED] writes: >Acquiring a lock does not turn on a super bit. It turns on a bit in the locks held string. Right. I mis-remembered what I found by instruction stepping through GETMAIN many years ago, which was that a request for certain system subpools which require the requester's being authorized caused three tests, the success of any one of which would allow the request to be honored. The tests were (1) any super bit's being on, (2) any lock held bit's being on, and (3) the CPU's system mask has I/O interrupts masked off (aka "physical disablement"). I was building a front end to the I/O FLIH at the time and wanted to get a piece of ECSA while disabled. I discovered that having interrupts disabled was not honored everywhere in the GETMAIN instruction path, so I turned on a super bit and then got my storage OK. >I would recommend against writing any code which turns super bits on or off. I did this because I saw it done in IBM's I/O FLIH (IIRC), so I knew I could get away with it. Maybe next time I'll remember to get a disabled spin lock instead. It's also not a good idea to build FLIH front ends, in general, and unless your employer wants one built. Bill Fairchild Franklin, TN
**************************************Check out AOL's list of 2007's hottest products. (http://money.aol.com/special/hot-products-2007?NCID=aoltop00030000000001) ---------------------------------------------------------------------- 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
