| 02:23.52 | *** join/#openjtag davygravy (n=davygrav@h69-128-156-59.mdtnwi.dsl.dynamic.tds.net) |
| 05:57.55 | *** join/#openjtag pleemans (n=peter@dD577D009.access.telenet.be) |
| 06:15.56 | *** join/#openjtag drath_ (i=vmaster@p5B07C3DB.dip.t-dialin.net) |
| 07:57.33 | *** join/#openjtag rwhitby (n=rwhitby@nslu2-linux/rwhitby) |
| 13:32.32 | *** join/#openjtag prpplague (n=dave@mail.americanmicrosystems.com) |
| 13:57.02 | *** part/#openjtag rwhitby (n=rwhitby@nslu2-linux/rwhitby) |
| 14:08.15 | *** join/#openjtag prpplague^2 (n=dave@mail.americanmicrosystems.com) |
| 16:53.13 | *** join/#openjtag pleemans (n=peter@d51A5E76A.access.telenet.be) |
| 17:47.17 | *** join/#openjtag Tom_itx (n=Tl@ip68-102-124-72.ks.ok.cox.net) |
| 18:00.45 | *** part/#openjtag Tom_itx (n=Tl@ip68-102-124-72.ks.ok.cox.net) |
| 18:22.06 | *** join/#openjtag davygravy_ (n=davygrav@h69-128-156-59.mdtnwi.dsl.dynamic.tds.net) |
| 19:54.42 | *** join/#openjtag ink_ (n=ink@82-35-232-51.cable.ubr01.camd.blueyonder.co.uk) |
| 19:56.09 | ink_ | Hi, what type of devices are discussed in here? |
| 19:56.46 | ink_ | I'm hoping someone can discuss JTAG on x86 devices |
| 19:57.41 | prpplague | ink_: although alot of x86 devices have jtag, it is very seldom used since most x86 devices have removable media |
| 19:57.55 | bjdooks | i think quite a lot of people here use jtag for arm |
| 19:58.07 | bjdooks | prpplague: it is possibly there's jtag for at-least boundary scan |
| 19:58.19 | ink_ | yes, I have noted that my AMD K6-2 device has JTAG |
| 19:58.24 | prpplague | bjdooks: yea, all of the geodes have jtag |
| 19:58.51 | prpplague | ink_: the real question is, "what do you wish to accomplish by using the JTAG interface?" |
| 19:58.52 | ink_ | But the flash chip is a tiny TSOP I'm not willing to remove |
| 19:59.20 | ink_ | And I sadly lost the serial connection I was using to update the flash image |
| 19:59.41 | ink_ | rendering it (a Cobalt Qube) quite useless |
| 20:00.51 | ink_ | I was going to build a Wiggler clone, but found one premade online for a little over a tenner |
| 20:01.39 | bjdooks | yeah, there's plenty of them around |
| 20:01.58 | ink_ | it's a pity its such an old device, i cant find any datasheets for anything other than the AMD CPU and the AMD flash chip |
| 20:02.51 | ink_ | I was hoping i could find some specs for the southbridge (an Ali M1543). |
| 20:03.13 | ink_ | as i'm guessing i'll need to twiddle some pins on that to program the flash |
| 20:05.45 | bjdooks | hmm, the m1543 is old |
| 20:06.16 | ink_ | The flash chip itself is an Am29F080B (JEDEC compatible if that helps) |
| 20:06.54 | ink_ | I'm guessing there is no support at all for an old CMOS chip like it in and of the open source JTAG projects |
| 20:07.50 | ink_ | But I imagine (?) that thats the least of my problems |
| 20:16.11 | ink_ | Should I expect to be able to identify a device (the CMOS flash chip) on the chain if I find I can get the K6-2's JTAG pins to respond at all ? |
| 20:18.11 | ink_ | I'm assuming I'll need to add a device definition file to allow it to be addressed at some later point |
| 20:24.18 | bjdooks | ink: devices should respond to IDCODE |
| 20:34.33 | drath | ink_: you'll only be able to identify devices that are part of the chain - the Am29F080B doesn't have a JTAG interface, and thus wont be part of the JTAG chain |
| 20:35.25 | drath | ink_: what you need is knowledge of the JTAG boundary scan of a device that has a JTAG interface and connects to all the data, address and control pins of the flash chip |
| 20:36.20 | drath | ink_: flashing would then work by applying the necessary bit pattern to the flash pins via the output registers of the boundary scan chain |
| 20:38.59 | ink_ | ah, well as the southbridge chip seems to be the only nearby chip and i've no idea if it has jtag support, even worse i'd have little ot no chance of getting a bsdl? file for it which i believe i would definately need |
| 20:39.26 | drath | ink_: yeah, the bsdl file describes the boundary scan chain |
| 20:40.23 | drath | ink_: i've seen people claim that some x86s support debug via the jtag interface, which could help you, but there's absolutely no documentation available |
| 20:40.48 | ink_ | bugger is the word then. Ali (in Taiwan) say that NVidia now own the rights to the M1541/1543 chipset |
| 20:41.11 | ink_ | and they're notoriously unhelpful. |
| 20:42.08 | bjdooks | are cobalt still around? |
| 20:42.19 | ink_ | no, Sun bought them. |
| 20:42.42 | ink_ | I spoke to a chap called Tim Hockin who worked on the BIOS development |
| 20:43.03 | ink_ | but he claims there were never schematics for the boards |
| 20:43.29 | ink_ | And always programmed the flash chips in system after careful testing. |
| 20:44.11 | ink_ | he thought the jtag headers (if any) were almost certainly unpopulated. |
| 20:45.34 | ink_ | I think I'm convincing myself that I'm going to have to replace the TSOP with a pre-programmed chip |
| 20:45.48 | ink_ | But I was rather hoping to tinker with JTAG if possible |
| 21:03.04 | ink_ | Well, anyone know of another cost effective way of writing to a 5v CMOS flash chip perhaps? |
| 21:33.48 | *** join/#openjtag prpplague (n=dave@mail.americanmicrosystems.com) |
| 22:06.56 | *** join/#openjtag bjdooks (n=ben@trinity.fluff.org) [NETSPLIT VICTIM] |
| 22:23.14 | ink_ | i found the datasheets finally |
| 22:24.04 | ink_ | no jtag. only smbus and an xd-bus ? that appears to connect to the ROM |
| 23:53.08 | *** join/#openjtag rwhitby (n=rwhitby@nslu2-linux/rwhitby) |
| 23:53.40 | *** part/#openjtag rwhitby (n=rwhitby@nslu2-linux/rwhitby) |