From owner-freebsd-questions Thu Aug 14 19:01:54 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id TAA29458 for questions-outgoing; Thu, 14 Aug 1997 19:01:54 -0700 (PDT) Received: from www2.shoppersnet.com (shoppersnet.com [204.156.152.112]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id TAA29451 for ; Thu, 14 Aug 1997 19:01:47 -0700 (PDT) Received: (from hlew@localhost) by www2.shoppersnet.com (8.8.5/8.8.5) id TAA15273; Thu, 14 Aug 1997 19:07:02 -0700 (PDT) Date: Thu, 14 Aug 1997 19:07:02 -0700 (PDT) From: Howard Lew To: questions@freebsd.org Subject: Re: Cyrix cpus Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-questions@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Okay, I had a chance to check out the new Cyrix 6x86MX processor with FreeBSD... Sorry, I only got ahold of the PR166 model (new 133MHz rev, not the 150MHz one), so here is some info about my experiences. I don't have a system for "make worlds" so I won't be able to test that for you guys. 1) Is this chip fast? Feels much faster than an K5-PR166. Coretest disk I/O also went up from about 10.8MB/sec to almost 13MB/sec on the Quantum BigFoot 4Gig EIDE. 2) The chip requires a very new Bios update on the motherboard or it may not even bootup. (Panics when booting FreeBSD 3.0/2.2.2 on an older MB i.e. VXpro). 3) You can now use the 586 compile settings for the kernel, so unlike the predecessor 6x86 or 6x86L, the 6x86MX seems to be a "Pentium instruction set" clone. 3) On a motherboard with an older bios (before the 6x86MX's arrival), comptest.exe in dos says it is a 486. On this new (very new) motherboard (TXpro) comptest says it is a Pentium running at 196MHz. AMI Bios v1.4 I believe... 7/15/97 Whether it is a true Pentium clone I don't know, but perhaps without the proper bios it won't be configured right and it may require 486 compile settings (???). 4) I have tried the new chip with FreeBSD 3.0-June something snap and it works great. Also with FreeBSD 2.2.2 RELENG and it also works fine. 5) Probe Message: shell: {14} dmesg Copyright (c) 1992-1997 FreeBSD Inc. Copyright (c) 1982, 1986, 1989, 1991, 1993 The Regents of the University of California. All rights reserved. FreeBSD 2.2-970713-RELENG #0: Thu Aug 14 15:38:37 PDT 1997 howard@shell.shoppersnet.com:/usr/src/sys/compile/POWEROS Calibrating clock(s) ... i586 clock: 133305791 Hz, i8254 clock: 1193169 Hz CLK_USE_I8254_CALIBRATION not specified - using default frequency CLK_USE_I586_CALIBRATION not specified - using old calibration method CPU: Cyrix 6x86MX (133.31-MHz 586-class CPU) Origin = "CyrixInstead" Id = 0x600 Device ID = 0x0 Stepping=0 Revision=0 real memory = 33554432 (32768K bytes) avail memory = 30949376 (30224K bytes) pcibus_setup(1): mode 1 addr port (0x0cf8) is 0x8000004c pcibus_setup(1a): mode1res=0x80000000 (0x80000000) pcibus_check: device 0 is there (id=153110b9) Probing for devices on PCI bus 0: configuration mode 1 allows 32 devices. chip0 rev 177 on pci0:0 chip1 rev 178 on pci0:2 vga0 rev 0 on pci0:4 mapreg[10] type=0 addr=ef800000 size=800000. fxp0 rev 1 int a irq 11 on pci0:5 mapreg[10] type=0 addr=ef7ff000 size=1000. mapreg[14] type=1 addr=0000ef80 size=0020. mapreg[18] type=0 addr=ff900000 size=100000. reg16: virtual=0xf44a1000 physical=0xef7ff000 size=0x1000 fxp0: Ethernet address 00:a0:c9:39:19:42 pci0:11: ACER Labs, device=0x5229, class=storage (ide) int a irq ?? [no driver assigned] map(20): io(fff0) pci0: uses 9441280 bytes of memory from ef7ff000 upto ff9fffff. pci0: uses 32 bytes of I/O space from ef80 upto ef9f. 6) Memory speeds... This was done with 32MB 10ns 100MHz sdram with fastest BIOS settings. Your results may vary depending on the the cpu clock. This new motherboard has 512K onboard and is a TXpro (ALI Alladin IV+ Chipset), so I don't think it is too bad for a 133MHz processor (66x2). Setting npx flags as 0x7 also helps improve memory speed. shell: {14} dd if=/dev/zero of=/dev/null bs=1m count=1000 1000+0 records in 1000+0 records out 1048576000 bytes transferred in 13.358422 secs (78495499 bytes/sec) shell: {15} dd if=/dev/zero of=/dev/null bs=1m count=1000 1000+0 records in 1000+0 records out 1048576000 bytes transferred in 13.506459 secs (77635152 bytes/sec) shell: {16} dd if=/dev/zero of=/dev/null bs=1m count=1000 1000+0 records in 1000+0 records out 1048576000 bytes transferred in 12.914553 secs (81193364 bytes/sec) 7) My Conclusion: Not bad for a chip that's only a little over $100.