On Thu, Nov 10, 2005 at 04:13:02PM -0800, Chen, Kenneth W wrote: > Robin Holt wrote on Thursday, November 10, 2005 2:39 PM > > I am going to attach the dispersal analysis the modified objdump that > > Jack has produced. > > > 5c: 8 S cmp.eq p6,p7=5,r17 > > 60: 8 [MII] nop.m 0x0 > > 66: 8 shr.u r18=r22,36;; > > Hmm, I think the dispersal analysis software is buggy. At least for > the above instance. Immediate form of shr is pseudo-op'ed to extr. > It should count the above bundle with 2 cycles instead of one cycle. Very possible. The tool is very old & non-supported. We have used it several times and had good luck. It seems to point out ares of code that stall unexpectedlt. Eliminating the stalls usually improved performance. However, I would be surprised if there were no bugs in the code. I would like to see a standard tool that gives dispersal analysis. There was a discussion on the IA64 list ~ month ago but so far, no takers. > > - Ken > > - > To unsubscribe from this list: send the line "unsubscribe linux-ia64" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- Thanks Jack Steiner (steiner@sgi.com) 651-683-5302 Principal Engineer SGI - Silicon Graphics, Inc. - To unsubscribe from this list: send the line "unsubscribe linux-ia64" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.htmlReceived on Fri Nov 11 11:25:08 2005
This archive was generated by hypermail 2.1.8 : 2005-11-11 11:25:20 EST