On Thu, 23 Apr 2009 09:58:18 -0400 "Farley, Peter x23353" <peter.far...@broadridge.com> wrote:
:>> -----Original Message----- :>> From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On :>> Behalf Of Kirk Wolf :>> Sent: Wednesday, April 22, 2009 8:44 PM :>> To: IBM-MAIN@bama.ua.edu :>> Subject: Re: METAL C: CodeGen defeciency? :><Snipped> :>> Interesting that they didn't use XC, isn't it? If you spend any time :>> looking at XLC generated code you will see lots of interesting stuff. :>> Of particular interest is some of the loop unwinding optimizations. :>> After a while you start to wonder if you should ever write in :>> assembler again, since only IBM knows what instruction heuristics :>> slide through the pipelines the best. :>It seems to me to be counterproductive of IBM to "hide" (or simply fail :>to publish) knowledge or advice on effective assembler programming to :>take advantage of (or at least not to fall afoul of) the pipeline nature :>of today's z processors. :>Obviously a thorough reading of the z/Arch PoOP will give an experienced :>assembler programmer a few "don't do this, it flushes the pipeline" :>cautions, like the following that I recently ran across while :>researching something else entirely: :>In the notes under the Branch on Condition instruction in Chapter 7: :>"4. Execution of BCR 15,0 (that is, an instruction :>with a value of 07F0 hex) may result in significant :>performance degradation. To ensure optimum :>performance, the program should avoid use of :>BCR 15,0 except in cases when the serialization :>or checkpoint-synchronization function is actually :>required." As there is no reason to code a BCR 15,0 this provides little value. CNOP / NOPR generate BCR 0,0. :>But let's face it -- how many of us have the time to be "thorough" about :>reading the PoOP? :>So, does anyone who regularly reads the IBM Systems Journal or :>Redbooks/Redpapers or other IBM pubs know of anything that IBM may have :>published that consolidates and exposes such knowledge? I, for one, :>would be very happy to reduce my ignorance with a little such light :>reading. -- Binyamin Dissen <bdis...@dissensoftware.com> http://www.dissensoftware.com Director, Dissen Software, Bar & Grill - Israel Should you use the mailblocks package and expect a response from me, you should preauthorize the dissensoftware.com domain. I very rarely bother responding to challenge/response systems, especially those from irresponsible companies. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
