From owner-freebsd-hardware Mon Sep 24 7:12: 8 2001 Delivered-To: freebsd-hardware@freebsd.org Received: from hal9000.net.au (cu764.adelaide.adsl.on.net [150.101.237.2]) by hub.freebsd.org (Postfix) with ESMTP id 0697A37B40E; Mon, 24 Sep 2001 07:11:54 -0700 (PDT) Received: (from bruins@localhost) by hal9000.net.au (8.11.1/8.11.1) id f8OEBnP05151; Mon, 24 Sep 2001 22:11:49 +0800 (WST) (envelope-from bruins) Date: Mon, 24 Sep 2001 22:11:49 +0800 From: Mike Bruins To: freebsd-questions@freebsd.org, freebsd-hardware@freebsd.org Cc: Mike Bruins Subject: health chip driver - vt82c686 - finishing off Message-ID: <20010924221149.A4971@wall.hal9000.net.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i Sender: owner-freebsd-hardware@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org Hello, I have put together most of a device driver for the health monitoring chip via technologies vt82c686 function 4 (PCI 3057:1106). All the kernel loadable module, device probe/attach code, temperature/fan/voltage conversion routines are there. I can read the device registers for PCI Vendor/Device, chip enable, class, subclass etc. Basically what I am not able to do is read the values hardware offset registers. The call to bus_alloc_resource fails (returns null) and I can't work out why. If anyone could please help with the following questions: 1. It looks like bus_alloc_resource() gets resources from it's parent device driver. Does the vt82c686 function 4, talk I2C to it's parent to get resources? If so, do I need a device driver for this parent? Anyone know it's PCI vendor/device chip code? 2. Perhaps I am going down the wrong path. Ideas, example code references appreciated. I have looked at many examples of code, but can't work out what is relevent/required for the 686a/b, in terms of resources and how it communicates. 3. The bit of my code that is actually failing is below. Like I said, I am pretty much sure the code to this point is correct and working as I can print any of the registers and they have sane values. I have tried every conbination I can think of for hwm_io_base_reg and the type of resource I am asking for. I have read the 686a.pdf file many times, tried all I can think of. According to the 686a.pdf the data size is 0x80 bytes. via->mem_rid = hwm_io_base_reg; via->res_mem = bus_alloc_resource(dev, SYS_RES_MEMORY, &(via->mem_rid), 0ul, ~0ul, 0x80, RF_ACTIVE); if (!via->res_mem) { device_printf(dev, "could not map memory\n"); goto bad; } via->memt = rman_get_bustag(via->res_mem); via->memh = rman_get_bushandle(via->res_mem); Ultimately I need the memt and memh values, so I can make calls like: val = bus_space_read_1(via->memt, via->memh, regno); My system: Motherboard: Gigabyte 6vxd7 CPU : dual 866 OS : 4.3-RELEASE free-bsd I have altered chip_match() so no driver is loaded at boot time for 0x30571106. chip0@pci0:0:0: class=0x060000 card=0x00000000 chip=0x06911106 rev=0xc4 hdr=0x00 pcib1@pci0:1:0: class=0x060400 card=0x00000080 chip=0x85981106 rev=0x00 hdr=0x01 isab0@pci0:7:0: class=0x060100 card=0x06861106 chip=0x06861106 rev=0x22 hdr=0x00 uhci0@pci0:7:2: class=0x0c0300 card=0x12340925 chip=0x30381106 rev=0x10 hdr=0x00 uhci1@pci0:7:3: class=0x0c0300 card=0x12340925 chip=0x30381106 rev=0x10 hdr=0x00 none0@pci0:7:4: class=0x0c0500 card=0x00000000 chip=0x30571106 rev=0x30 hdr=0x00 sym0@pci0:9:0: class=0x010000 card=0x00000000 chip=0x00201000 rev=0x01 hdr=0x00 sym1@pci0:9:1: class=0x010000 card=0x00000000 chip=0x00201000 rev=0x01 hdr=0x00 rl0@pci0:10:0: class=0x020000 card=0x0027a0a0 chip=0x813910ec rev=0x10 hdr=0x00 pcm0@pci0:13:0: class=0x040100 card=0x13711274 chip=0x13711274 rev=0x08 hdr=0x00 none1@pci0:14:0: class=0x040100 card=0x20601458 chip=0x13711274 rev=0x07 hdr=0x00 none2@pci1:0:0: class=0x030000 card=0x00000000 chip=0x002c10de rev=0x15 hdr=0x00 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hardware" in the body of the message