To: J3 07-323r1 From: Bill Long Subject: UTI 138 (p337): Collective subroutines described in wrong place Date: 2007 November 13 References: J3/07-007r3 Discussion Yes, 13.1 should have a better definition of a collective subroutine. An edit is supplied and a corresponding edit for the definition in 2.1.22. We need to keep the rule that the same statement is invoked on all images of the team since this is not the case for SYNC TEAM. Yes, we should have all the rules for implicit team synchronization in 8.5.3, so we supply an edit for that section and reference it from 13.1. The new words cover OPEN and CLOSE, too. An edit adds a reference to 8.5.3 for OPEN. CLOSE already has such a reference. Edits to 07-007r3 [12:6] In 2.1.22 collective subroutine, replace the definition by "intrinsic subroutine that is invoked on a team of images to perform a calculation for data on those images and assign the value of the result on all of them." [201:9-11]. In 8.5.3 SYNC TEAM statement, para. 2, replace the final sentence by "If the statement contains an , it specifies the team and shall satisfy the conditions required of an in a SYNC TEAM statement; otherwise, the team is the connect team for the unit in a CLOSE statement or the set of all images for a CALL to a collective subroutine." [222:8] In 9.5.6.19 TEAM= specified in the OPEN statement, at the end of para. 2, after "synchronization" add "(8.5.3)". [337:15-17] In 13.1 Classes of intrinsic procedure, replace the first three sentences of para. 4 by "A collective subroutine is one that is invoked on a team of images to perform a calculation for data on those images and assign the value of the result on all of them." [337:18] In 13.1 Classes of intrinsic procedure, in the final sentence of para. 4, after "synchronization" add "(8.5.3)".