https://yarchive.net/risks/sgi_irix.html https://yarchive.net/risks/sgi_irix.html PERFORMANCE UPDATE "Indy: an Indigo without the 'go'". -- Mark Hughes "X an

https://yarchive.net/risks/sgi_irix.html https://yarchive.net/risks/sgi_irix.html PERFORMANCE UPDATE "Indy: an Indigo without the 'go'". -- Mark Hughes "X and Motif are the reasons that UNIX deserves to die." -- Larry Kaplan >The performance story is just as bad. I was tempted to write simply, "Try to do some real work on a 16 megabyte Indy. Case closed.", but I'll include some details. >In May, I listed some unacceptable Motif performance measurements. Just before 5.1 MR, someone reran my tests and discovered that the performance had gotten even worse. Some effort was expended to tune the software so that instead of being intolerable, it was back to merely unacceptable performance. >We no longer report benchmark results on our standard system. The benchmarks are not done with the DSO libraries; they are all compiled non-DSO so that the performance in 5.1 has not declined too much. >Before I upgraded from 4.0.5 to the MR version of 5.1, I ran some timings of some everyday activities to see what would happen. These timings were all made with the wall clock, so they represent precisely what our users will see. I run a 32 megabyte R4000 Elan. Test 4.0.5 5.1 % change ---- ----- --- -------- C compile of a 25 sec 35 sec 40% small application C++ compile of a 68 sec 105 sec 54% small application Showcase startup, 13 sec 18 sec 38% May report file Start a shell <2 sec ~3 sec ~50% Jot 2 MB file <2 sec ~3 sec ~50% >What's most frightening about the 5.1 performance is that nobody knows exactly where it went. If you start asking around, you get plenty of finger-pointing and theories, but few facts. In the May report, I proposed a "5% theory", which states that each little thing we add (Motif, internationalization, drag-and-drop, DSOs, multiple fonts, and so on) costs roughly 5% of the machine. After 15 or 20 of these, most of the performance is gone. >Bloating by itself causes problems. There's heavy paging, there's so much code and it's so scattered that the cache may as well not be there. The window manager and X and Toto are so tangled that many minor operations like moving the mouse or deleting a file wake up all the processes on the machine, causing additional paging, and perhaps graphics context swaps. >But bloat isn't the whole story. Rocky Rhodes recently ran a small application on an Indy, and noticed that when he held the mouse button down and slid it back and forth across the menu bar, the (small) pop-up menus got as much as 25 seconds behind. He submitted a bug, which was dismissed as paging due to lack of memory. But Rocky was running with 160 megabytes of memory, so there was no paging. The problem turned out to be Motif code modified for the SGI look that is even more sluggish than regular Motif. Perhaps the problem is simply due to the huge number of context swaps necessary for all the daemons we're shipping. >The complexity of our system software has surpassed the ability of average SGI programmers to understand it. And perhaps not just average programmers. Get a room full of 10 of our best software people, and you'll get 10 different opinions of what's causing the lousy performance and bloat. What's wrong is that the software has simply become too complicated for anyone to understand.