Path: csiph.com!weretis.net!feeder9.news.weretis.net!panix!.POSTED.spitfire.i.gajendra.net!not-for-mail From: cross@spitfire.i.gajendra.net (Dan Cross) Newsgroups: comp.arch Subject: Re: Performance of xchg reg,reg Date: Sun, 18 Oct 2026 17:47:51 -0000 (UTC) Organization: PANIX Public Access Internet and UNIX, NYC Message-ID: <11b30o7$iqv$1@reader2.panix.com> References: <2026Oct13.104347@mips.complang.tuwien.ac.at> <1792257076-5857@newsgrouper.org> <11b0bd3$b5b$1@reader2.panix.com> Injection-Date: Sun, 18 Oct 2026 17:47:51 -0000 (UTC) Injection-Info: reader2.panix.com; posting-host="spitfire.i.gajendra.net:166.84.136.80"; logging-data="19295"; mail-complaints-to="abuse@panix.com" X-Newsreader: trn 4.0-test77 (Sep 1, 2010) Originator: cross@spitfire.i.gajendra.net (Dan Cross) Xref: csiph.com comp.arch:118695 In article , EricP wrote: >On 2026-Oct-17 13:31, Dan Cross wrote: >[snip] >Intel Volume 3 System Programmers guide 10.1.2 Bus Locking >10.1.2.1 Automatic Locking is explicit:... >[snip] Yes, which in my view, lends more support to requiring that the sentence in question in the description of the `LOCK` prefix be read in the context of the overall paragraph, which is clearly talking about cases where the destination is a memory operand. Sadly, however, that doesn't make the sentence itself any less ambiguous, and once you have ambiguity in a technical document like the SDM, you've got a problem. Intel can fix this easily by adding another phrase to that last sentence, ", when the destination is a memory operand." (Or words to that effect.) I'm speculating, but they probably haven't because it seems rather obvious that that is the intent, and no one has brought it up as a problem. - Dan C.