I-Appliance BBS The Official Source for Internet Appliance Upgrades and Mods Click Here! BBS Main List | Sign In | Sign Up | Search | Help | Linux-Hacker.net Reply to Thread | Printer |

Home / I-Opener Areas / I-Opener Technical Stuff "The USB Problem" - SOLVED !!!! One more killed....

"The USB Problem" - SOLVED !!!! (modified 0 times) Georgie Profile Hello Everybody, I wish a happy day to You all. Yes, I seriously think I have managed to pinpoint and cure the USB transfer problem. But I am in a somewhat sadistic mood today, and will make you wait until tonight to find out all the details and read the whole saga...... Actually, I am at work now, and have no time to type the whole thing in. So, I apologize for the delay, but rest assured, it will appear here tonight. Cheers Georgie 08-28-2001 09:28:48

Yes, me too, I had the USB problem... (modified 0 times) Georgie Profile OK, here I am, and the whole story... Just as many other people on this list, originally I thought that ckbone's USB problems were isolated problems, related to some configuration error. I had all the latest drivers installed, Programmer's fixusb loaded from autoexec.bat, my machine ran fine, so why worry? But.... But then I began using my HP USB Cd-Writer. And the awful experience came with it. Whatever files I worte on CDs, they were corrupted!! DAMN! I HAD THE USB PROBLEM!!! Now I'm convinced, that every IO has it, with only very few exceptions (explanation coming up). So there I was, a week ago, quite disappointed. I reread the FixUSB thread, just to get the feeling, that no solution is about to come. Thus there was nothing else to do, but sit down, and think about it. This is what I was doing Sunday evening. 08-28-2001 22:50:17

The First Clues (modified 0 times) Georgie Profile First of all, just as Programmer, I tried to find some data pattern in the files that got corrupted. There is a small DOS program, "crcmp" - stands for CRC compare -, meant to test the integrity of huge amounts of files and data very quickly. I have used it for many years. Basically it reads all files in a directory tree, calculates 32bit CRCs, and writes the names, sizes and CRCs into a small data file. If the data file already exists, it compares the CRCs calculated on the fly against those in the database. It can test the integrity of gigabytes of data in minutes. So I used crcmp to find all the corrupted MP3s on the CD I had written. Well, every second file, on average turned out to be corrupted.... Next step was comparing pairs of original and corrupted files in a good hex editor. (Our good Programmer, poor Him, is trying to do the same stuff right now, but without having an IO to use, it's very very complicated, and slow...) Back to the files. So I had 10 pairs of files, original compared to corrupted side-by-side, and I found absolutely no suspicious data pattern (ie. prone to get corrupted easily)! No byte, hex, or bit pattern! It gets better: I wrote another CD, and once again checked data integrity with crcmp. Files having only one error on the first CD had none, or up to three errors on the second CD. Moreover, data corruption occurred at entirely different parts of the same or entirely different files, when comparing the two CDs. I had about 1 error in every 3MB of data transferred. Data corruption occurred at totally random points of data and/or time. There was only one identifiable pattern: error length. It was always 32 bytes. Always. 32 bytes of data taken at absolutely random intervals got replaced by 32 bytes of absolutely random garbage. There was no other problem with data transferred to the CD-Writer, only these randomly occurring 32-byte segments of garbage. All in all, it awfully looked like a timeout error, because of the fixed length of errors. So I scratched the possibility for a chip design error showing up at certain bit-patterns, and began looking for timing issues. 08-28-2001 23:16:14

Timing IS Everything! - the first, "quick and dirty" cure (modified 0 times) Georgie Profile I found no better method to test for timing than playing with PCI speed. I used a soft-FSB program for this. I have my machine running at 110MHz FSB, i.e. 110/3=36.7MHz PCI bus speed. This is the speed at which my USB circuit is working. I established a small test data set - 9 MP3 files + the crcmp data file, about 30MB in total -. I wrote this 30MB data set repeatedly on CD at different PCI speeds, and then checked the data for corruption. And BINGO!!! I had my first error-free data transmission! I tried the following speeds: FSB PCI Status (OK/Error) 110 36.7 Error 105 35 Error 100 33 Error 95 31.8 OK !!!! 83.3 27.7 OK !!!! 80 26.7 Error 75 25 Error As you can see, I managed to write 30MB of data on the CD without errors at 31.8 and 27.7 MHz PCI speeds. I have no data for 30MHz (60/2), because for this I would have had to open the IO, which I did not want to do. (Explanation: if your machine is using a "/3" FSB, and you try to switch to a "/2" FSB, or vice versa, the computer hangs tight. Only a power cycle will revive it. I warned you!) Anyway, I do think, that 30MHz would have been successful as well. One thing I have to emphasize here is that for these OK transmissions to occur I had to run Programmer's fixusb at startup. Once I disabled it in autoexec.bat, my error-free transmissions at these two PCI frequencies were history! I drew the following conclusions: it is a timing issue, Programmer's fixusb does help somehow, and that I could run my machine at 95 MHz at best, which wouldn't be too bad, but..... BUT!!!! It's not finished yet. I eventually decided to transfer more than 30MB. My test set was 150MB this time. I carried out numerous tests, and unfortunately, once I had one error in the files. One error - the same 32-byte random garbage error - in 150MB - much better than 1 in every 3MB, isn't it? Oh yes, it sounds very good, just not good enough... I knew, I was heading in the right direction, I just did not quite had the real solution yet. Back to thinking.... 08-29-2001 00:20:46

The fun part - software-only USB FIX !! (modified 0 times) Georgie Profile I opened up the Southbridge manual in Acrobat Reader, and began looking for speed and/or timing related PCI registers among the USB function registers. I found 8 suspicious values at 3 offsets total. I had a closer look at them in my IO. For reading/manipulating the PCI registers I have used John Fine's PCI for years now. You can find it at http://www.execpc.com/~geezer/johnfine/. A short command pci r,7,2,?100 gives the screen printout of all the interesting information in the USB-related PCI registers. So I was looking at the content of my USB register space - Bus:0, Device:7, Function:2 - when I noticed something very strange: offset D is called "Timer Value" in the VIA manual, and the manual gives a default value of "16" for it. Yet in my machine it said "20". Now, it is a nice 50% different in timing value!!! - The values are in hex 8-) -. I mean, it became clear immediately. This value is the value! Back to square one. I disabled fixusb - sorry, Programmer ;) -, went back to 110MHz FSB, and restarted the machine. And began playing with the timer value. The test set was 150MB. As I anticipated, the optimal timer value seems to be different in the case of different PCI bus speeds. So, I'll show you what I have tried so far, and we'll talk about the rest later: 110MHz FSB, 36,7MHz PCI speed: - transmission error-free with timer values: 16,12,f,c - transmission contained 1 error with timer value "8", many errors with "20" 75 MHz FSB, 25MHz PCI speed: - transmission error-free with timer value "8" - transmission contained errors with timer values: 20,16,12,c A few notes to add: it is interesting that there seems to exist a sweet spot for the timer value, which is somewhat different at different PCI speeds. The higher the PCI speed the higher the optimal value will be. I tried to establish a global value, which would work at any PCI speed - this is why I tested the two extremes of PCI speed -, but it seems that there is no such value. In any case, the further away I went from the sweet spot, the higher the number of errors was. Now where is this sweet spot?? Well. At 25MHz it seems to be 8, or even lower. I haven't examined it really thoroughly, but who on Earth would be interested in such a horrible underclocking of the machine anyway? As for 36.7MHz (110MHz FSB) it is around 10 (hex, as always!!!). I set it at 10 on my IO, and have burnt 2 CDs WITHOUT ERRORS since then. That is almost 1.5GB of data transmitted through my USB port without hiccup. It sure sounds good, doesn't it? ;))) So this is where I am now. My USB seems to be working flawlessly at last, and I am very happy to say this. My temporary solution is a bat file loading pci.exe to rewrite the timer value. I load it at Windows startup, from "RunServices". The reason for loading it from Windows is when the USB driver loads in Windows, it resets - most probably - the USB circuit and the timer value will acquire that stupid "20" hex value, with all its consequences. So I load the pci command after the USB drivers are loaded. Works like a charm! Just to let you experiment with it: the command to read the timer value out: pci rd,7,2,? Value will be given at the end of line. Rewriting the timer value to "10": pci rd,7,2,=10 This should be enough for today. It is getting very late, so I will have to leave you with my thoughts. But I would really appreciate some help in testing the fine details of all this, and it would be nice to know how that value "20" gets into that register. It may be a USB driver fault. (Read: correctable....) Finally, thank you for your patience, and I hope to be able to read a few success stories about your - hopefully now - working USB ports. Happy hacking, Everyone! Georgie 08-29-2001 01:50:27

RE: (modified 0 times) ckbone Profile Georgie.....Thank you for all this work and thanks to everyone else trying to straighten out these USB problems. Please answer a few questions: Should this patch work equally well on all I-O versions? Should it also work on modified boards running different multiplier speeds? Is this patch loaded up like Programmers patches......A line in Autoexec.bat that reads pci rd,7,2,=10.com You said all I-Os were affected with a few exceptions. What exceptions? Some people claim no USB problems....is this possible? Anyway thanks again....we'll all be thankful if this is a success. 08-29-2001 03:34:48

RE: (modified 0 times) laserfan Profile My question is even more basic: how do I identify & manipulate FSB. I know there is discussion of this around here somewhere but it would be nice to have an explanation and link in this thread. 08-29-2001 05:20:32

RE: (modified 0 times) laserfan Profile Oh, and I must say "thanks!" to Georgie for all your work. I must admit when I first saw your "teaser" post yesterday I thought our chains were being yanked, cuz I'd never seen your handle here before. Thanks again! 08-29-2001 05:24:39

RE: (modified 0 times) Programmer Profile Georgie: Thanks for the initiative to figure this out.. Here's another contribution from me.. ;) AFAIK, this setting is only specified in the BIOS and is not changed by anything else.. http://bethie.net/~programmer/georgie.com is a small tool to let you set the timers on BOTH USB ports.. usage: georgie [VAL] where VAL is a character. Using an ascii table http://asciitable.com find the value that you want to set the timer to, only the lower 5 bits are recognised by the program. Georgie, with his 110Mhz FSB and 37Mhz PCI Bus would use the 'sweet spot' that he found at 10h. On the ascii table 10h is a non-printable character that you have no good way of typing, in the next column over, 30h which after dropping the first 3 bits is 10h is the number '0', further, the letters 'P' and 'p' will also equate to 10h. So georgie could use 'georgie 0', 'georgie P', or 'georgie p'. most users out there will have a 33Mhz PCI bus and should use 'M' or 'N' once you find a value that suits your IO perfectly, you can put the instruction in your Autoexec.bat or, you can use http://bethie.net/~programmer/com2rom.zip to create an option rom that will set the setting for you in the bios. just enter your letter (or number) in the box labeled command string and it will be written to the option rom in a manner that it will be recognised that same as it is in DOS. Said option rom can be added to a bios using awardmod http://awardmod.sourceforge.net , or cbrom. from there you will need to flash it using awdflash. 08-29-2001 10:16:38

What about FSB... (modified 0 times) laserfan Profile Ok, I found "FSB 101" courtesy of ckbone in the iopener HOW-TO thread: http://www.linux-hacker.net/cgi-bin/UltraBoard/UltraBoard.pl?Action=ShowPost&Board=technical&Post=2142&Idle=0&Sort=0&Order=Descend&Page=0&Session= I will try to play with this stuff tonight. 08-29-2001 11:12:05

Answering questions (modified 0 times) Georgie Profile Thanks Everyone for joining the new thread! :))) First questions answered first... > Should this patch work equally well on all I-O versions? ckbone, To my knowledge there was no Southbridge revision change between the different iterations of the Eye-Opener. Meaning that I would definitely expect all IOs to have that idiot 20h timer value as default. So, in one word, YES, it has to work! > Should it also work on modified boards running different multiplier speeds? Yes, it is only related to the speed of your PCI bus, which is absolutely not related to your CPU multiplier. Just to clarify things, my IO is a nicely hacked V3b, with a K6-3+ 450 running at 550 Mhz (5x110). DC/DC hack installed. > Is this patch loaded up like Programmers patches.. A line in Autoexec.bat that reads pci rd,7,2,=10.com NOPE!!! Please, see explanation below. It is important! > You said all I-Os were affected with a few exceptions. What exceptions? Some people claim no USB problems....is this possible? Yes, I can imagine someone running the IO at one of the PCI speeds, which I identified as working "almost perfect" together with Programmer's fixusb. So, people running FSB at 60, 83.3 (only if FSB/PCI=3/1 !!!) or 95.3 MHz AND running fixusb might only experience one rare error from time to time. Someone running a Winchip at 180MHz (3x60) AND fixusb, for example... ......... Programmer, Thank you for your much valued contribution once again! A couple of notes here. At the same time I have to add, that again, I am at work, and cannot test anything until tonight. > georgie.com is a small tool to let you set the timers on BOTH USB ports Just a note. As far as I remember, you don't have to set the value for both ports. I mean, I think that you are talking about setting the timer value for Bus:0 Device:7 Function:3 as well, not only for Function:2. Actually, Function:2 deals with both USB ports that can be physically installed on the IO (ports 0 & 1). Yes, the chipset could handle two more USB ports directly (ports 2,3), for a total of 4, but there is no place on the mobo for them. So you can, but it is never going to get used actually.... > once you find a value that suits your IO perfectly, you can put the instruction in your > Autoexec.bat or, you can use com2rom.zip to create an option rom EVERYONE MUST READ IT. VERY IMPORTANT NOPE. Based on my experience you cannot execute it from autoexec.bat or put it in your BIOS and expect it to actually work. I have already written about this, but here it is once again. When Windows loads, something - most probably the USB driver - either resets USB or just writes a few default values in some of the registers. Either way, the result is that even if you set the timer value to your value of choice in autoexec.bat or in BIOS, it will get rewritten when Windows loads, and once in Windows, you will have your timer value set to "20". Conclusion: you have to set the timer value to your choice in Windows. Use whatever software to set the timer value, pci or georgie.com, or wpcre from H. Oda, but use it in Windows. And it will work! Just a quick note, prior to getting mad, first check for the actual value in your register using the pci program I wrote about. It works in a DOS window. Just don't forget: do it in Windows!!! 'Nuff said.... Cheers, Georgie 08-29-2001 12:13:15

About changing your FSB (modified 0 times) Georgie Profile > how do I identify & manipulate FSB. laserfan, Identifying your FSB is actually quite indirect. I mean without opening up your IO. The PLL chip of the IO - the chip which can be programmed to set the FSB -, is a write-only chip, so you cannot read the current value out of it. Funny, isn' it? Anyway, you can find many programs capable of giving you your current CPU speed. Most of the time it is quite easy to find your FSB based on this. The IO is able to produce only a limited amount of FSB values: 60, 66.7, 75, 80.3, 83.3, 95.3, 100, 105, 110,.. (not interesting, the chipset doesn't work correctly at or over 115MHz. Yes, I have tested it...). I may have left out something, but if say ctcm16r - from the german magazine, c't - gives you that you have a Winchip running at 200MHz, it is clear, that your FSB is 66,7. (The Winchip knows only integer multipliers, and it doesn't run at 100MHz FSB. >> It is running at 3x66.7). I have to admit, for this to work you have to know your CPU quite well. But by looking at the motherboard (SW2), it is clear at first sight. It is easy, if your IO motherboard is not modified - or if you have not modified SW2 on it yet-. There are two possibilities: up to version 3b, you had a Winchip, running at 3x66,7=200MHz. Version 4 and 5 had a RISE running at 2x100=200MHz. One nice program to find out much about your CPU is wcpuid from http://www.h-oda.com/ . Useful! As for manipulating your FSB, there are several softFSB programs out there. SoftFSB from H. Oda is not supported any more, but you can find it on many web sites. The only thing is that you need a data file for the PLL chip in the IO. It looks to be a Cypress W156A. Two other programs would be CPUCOOL or CPUFSB from http://www.podien.de/ . They have built in support for this PLL IC. One more thing: do select the correct PLL chip in the program, or you are risking your motherboard! Another warning: if you are using a FSB/2 PCI speed, don't try to switch to a FSB/3 type PCI speed, or vice versa, because it will just cause your machine to hang really tight.... Hope this helps. What the heck, it should! ;)) Georgie 08-29-2001 12:29:45

RE: (modified 0 times) Wild_Pencil Profile | Email Absolutely outstanding work, Georgie! You Rock! -WP 08-29-2001 14:49:03

RE: (modified 0 times) Programmer Profile Georgie: I must disagree with you, and I've just verified it on my daughters IO. On hers at least, if I set offset 0d from autoexec.bat (I used ckbone.com to do it since it sets it to 42 (!!!)) it is still set to 42 after windows loads (and my usb is running) It may be something that either the microsoft patch or the via filter driver or the via driver sets, I do not have those installed.. If someone (ckbone?) could check into that it'd be great.. Also there ARE at least 2 revs of the North and southbridges in use on different IOs CKbone has at least one of each.. (though that should not have any effect on this patch) PS.. I'd be grateful if someone could let me know the "true" sweet spot for a 33.3Mhz PCI bus 08-29-2001 15:21:12

Glug, glug. In over my head! (modified 0 times) laserfan Profile Sorry but I'm not following this very well. I did learn from WCPUID that my WinChip2 is running (mb completely unaltered by me in any way) at "200.45" MHz and my System Bus and Clock run at 66.82 with a 3 multiplier. I also tried "pci rd,7,2,?" and it returns a Val of 42 (is this Decimal or Hex?). Anyway I'm not sure what the heck to do next, but I'm game if someone can give me direction. My sense is "pci rd,7,2,=8" or less. Do I just experiment? I'm a little gunshy as SoftFSB crashed my IO pretty hard. 08-29-2001 19:38:29

Yes, it works from autoexec or BIOS!! (modified 0 times) Georgie Profile Oh well, what an a**hole...... I mean I am. Programmer, I must agree with you. You are perfectly right, the timer value will not get changed by Windows, or any Windows driver. So, I have to make a new announcement (should have just stayed quiet until testing it once more...) EVERYONE SHOULD READ THIS. IMPORTANT INFORMATION HERE I was wrong here. You can load the timer value patch from autoexec.bat or the BIOS, it will not get changed. I tried it with or without the VIA all-in-one and USB filter drivers, and no change. So please, just disregard my alarm above, and live a calm andlong life.... The only explanation I can come up with is that I was so sleepy at this point that I didn't realize that I had just rebooted the IO. There is no other way... Oh, please, stop laughing, will you? Briefly: best to add it to the BIOS, so that it will always be loaded, but you can also use it from autoexec.bat - and even from Windows, why not? ;))))) Thanks, and I apologize for messing up on this one.... Georgie 08-29-2001 19:55:03

RE: (modified 0 times) El_Kabong Profile | Email Georgie, dude thanks! You are the IO-Man of The Month. I ran PCI under WinME. (Start-Run-PCI.exe rd,7,2,?-enter)Returned a "20". Created a shortcut to PCI.exe in the C:\windows\Start Menu\Programs\StartUp dir. Selected Close on Exit in the properties. (Right click on PCI.exe) On the Cmd line I put PCI.exe rd,7,2,=10. Upon reboot I tested the PCI settings. I got a 10! So, it works under WinME. 08-29-2001 20:01:36

About the sweet spot for 33.3MHz (modified 0 times) Georgie Profile Programmer, Thank you very much for the clarification about the Southbridge versions. As for the sweet spot for 33.3MHz, I don't know it for certain, but expect it to be around c,d,or e. I would begin testing with e. I don't have the time right now, but will take a few hours one day, if no-one else will have reported back to us about this. laserfan, > "pci rd,7,2,?" and it returns a Val of 42 (is this Decimal or Hex?). PCI always talks in hex. That is 42 hex. It is waaaaaay off!! As I wrote above, in your case, which is 33.3MHz PCI speed, I would begin experimenting with "pci rd,7,2,=e". I expect it to work flawlessly at that speed, but it might need a little tweaking. I'm looking forward to your test results. Georgie 08-29-2001 20:12:13

Another note... (modified 0 times) Georgie Profile One more thing. Please, disable any other usb fixes possibly lurking in your autoexec.bat, this is the only USB patch that should be applied. The reason for me saying this is that laserfan says he has a default of "42", which makes me believe he is still running ckbone.com??? So, guys, this time no fixusb.com or ckbone.com, only georgie.com or pci.exe, OK?? ;)))))) Georgie 08-29-2001 20:35:11

RE: (modified 0 times) YouBecha Profile I am impressed! Hopefully this will solve my problems. Could you sum up the steps for me...I don't have the IO with me and I have had a few beers, so I got lost somewhere in the middle;) http://www.geocities.com/mr_bubba_zanetti/ 08-29-2001 21:16:05

RE: (modified 0 times) laserfan Profile Glad you gave the warning--I did still have ckbone.com in my AUTOEXEC.BAT. Taking this out, I started at E and went up, down, & all around and while some worked better than others I never got a clean transfer. Maybe someone else will find the holy grail but for me I'm dead tired and putting this away for tonight. YouBecha what I've been doing is I have pci.exe conveniently located at C:\, and I've been running from the MS-DOS Command window e.g. "pci rd,7,2,=E" and then xsferring a >1Mb JPG file (picture) out from my IO and then back in, and looking at it to see if it's hosed. Then I went up (to F & 10 etc.) and down (to D & C etc.) to see if it changed. No luck here (yet). Am I doing this right Georgie? 08-29-2001 21:38:00

RE: (modified 0 times) laserfan Profile Forgot to make clear: with every change I would send the file Up to my server and then bring it Back to my IO to look at it, taking care not to overwrite my original (good) file in the process by renaming the file before I bring it back down to the IO. 08-29-2001 21:41:21

33MHz (modified 0 times) Georgie Profile laserfan, Yes, you are doing it right. Two things. Could you please, check whether you really have the values you should have? I mean, prior to file transfer, read the register just to make sure you have what you think you have. What values have you tried so far? Could you give us all? Your IO is a Winchip-2 at 3x66,7MHz, isn't it? TIA Georgie 08-29-2001 22:26:53

Still 33MHz (modified 0 times) Georgie Profile laserfan, Something else. Your server. Does it have a mobo with VIA chipset???? Because you might need the same fix on the server as well. Anyway, the easiest way to do this would be to look at the file on the server. This way you are only testing the IO, not both. Would you please, try it this way once again. TIA Georgie 08-29-2001 22:32:37

RE: (modified 0 times) ckbone Profile laserfan....Regarding SoftFSB, I found Poiden's prog CPUFSB or CPUCool to work much better. I was able to O/C my Winchip2 machine to "75" (225mhz) and run stable. SoftFSB always crashes at that setting. As a bonus CPUFSB can automatically adjust to that setting at startup, and automatically go back to stock at shutdown.....which may stop any shutdown problems you may be having. In case anyone doesn't realize the benefits of an FSB overclock, what you get is an across-the-board increase in performance, very impressive. More noticeable to me than a multiplier increase to say 250mhz. Not every chip responds the same, sometimes your ram chip won't like the increased FSB...you just have to try. FWIW...There's a new VIA USB patch available...not much feed-back yet. I won't have time till next week to try this new stuff. To give it all a fair test, I'll try it on a clean install. It's hard to remember all the stuff I've tried on these three machines I have, and I don't like to apply patches over other patches. The results are never predictable for me. The ultimate USB fix will be one that loads and runs automatically...maybe wishful thinking. 08-30-2001 04:00:14

RE: (modified 0 times) Programmer Profile laserfan: IF you are unable to get a working solution just manipulating the usb timers, give the original fixusb a try as well http://bethie.net/~programmer/fixusb.com .. the settings it manipulates are on the northbridge and they are intended to boost the reliability of the PCI bus itself, they will not interfere with the functionality of the usb fix we are currently talking about.. 08-30-2001 06:52:32

RE: (modified 0 times) Georgie Profile laserfan, My note above your server having a VIA chipset applies - of course - only in the case if you have a USB NIC on your server as well. Now that I think of it, why would you have a USB NIC insetad of a PCI card, but one never knows... So if your server is not communicating through the USB port, please, disregard it. There are two reasons I can think of that might make the timer patch fail on your machine. One is transfer speed. I have only tried CD burning so far. It works flawlessly now, but it is only 600KB/s. When using a USB NIC, at least in theory, you may be able to burst more data through the port for short time periods, if it's a 10Mbit NIC. If it is 100Mbit, it means a lot more data... This would be the worse case. The other possibility is that I have a nice bunch of other PCI register enhancements on my machine here, which help performance tremendously, and actually may also help this USB patch. I haven't tried my USB patch without those other enhancements. It's easy to find out. I'm going to try it this evening. Nevertheless, I would like to be able to compare my PCI setup with yours, if you don't mind. It would mean executing pci.exe with some special commands, and sending the file created back to me. I'll send you the details in a few hours. TIA Georgie 08-30-2001 10:55:34

I want your PCI setup! (modified 0 times) Georgie Profile laserfan, A I wrote above, I would like you to help me to sort this problem out. I need your PCI configuration info for this. Would you mind, please, helping me. Thanks. I constructed a short command for you to execute. It will create a short text file named "laserfan.io" in the directory from which you execute it - i.e. the current directory-. Here is the command: pci r?100 r,7,1,?100 r,7,2,?100 > laserfan.io You should send "laserfan.io" to me at the following address: "my BBS ID" at chez dot com. ;)) Thanks Georgie 08-30-2001 11:16:03

RE: (modified 0 times) jbarr Profile Is the "Gorgie Fix" Windows-specific? I am (typically) running Midori Linux from the SanDisk. Of course, with Midori Linux, I rarely transfer lots of data FROM the i-Opener--primarily just web browsing and streaming MP3 audio. 08-30-2001 14:38:19

Final solution being prepared!!! (modified 0 times) Georgie Profile OK Guys, You will also need one of my other PCI patches for the USB patch to work. Once I disabled them, the errors came back... Telling, isn't it? I need some time to fine-tune the whole thing for you, but rest assured, you will have it tonight. I am preparing it right now. Keep faith! ;) Georgie 08-30-2001 18:39:22

Is it Windows-specific? (modified 0 times) Georgie Profile jbarr, No, the "Georgie-fix" is not Windows-specific. The program I actually use to set the PCI registers runs in DOS, but there are similar programs for all OSes. So if you point me to a Linux program to set PCI registers directly, I will give you all the info you need. It will work. Georgie 08-30-2001 18:44:06

RE: (modified 0 times) laserfan Profile It is the end of a long day for me, and I am dead tired, but here are a couple notes: My IO has been upgraded to WinChip2 from WinChipC6 and is otherwise unaltered. My clock runs at 66.82 and a multiplier of 3 (and I have no clue really what the HECK this is all about). My server has a PCI card, no VIA on the mobo, and my IO's USB-to-Ethernet device is an EA101 from Netgear. I don't recall right now, and the dongle itself has no clues, if it's a 100M or a 10M device. I have resisted doing anything at all w/FSB and Overclocking, first because my system works perfectly (it ain't broke, so I'm disinclined to fix it) and second because my 200MHz WinChip2 is cooled only by a hacked-off original IO heat sink, no fan at all. So I've not wanted to add to its burden. I will run the program for you Georgie before I go to bed--hope I'm not so tired as to not figure out your precise address!!! Thanks for your effort and your energy to go after this problem.... 08-30-2001 20:16:34

RE: (modified 0 times) Kcmjr Profile All this great guys, and I don't mean to complain, but it seems that all these slick utilities only work in DOS or Win95/98. I'm happy to wait for someone to come up with a BIOS patch to fix this, if it works, but I wonder if it will help me. I'm running Win2k and when I run the command "pci rd,7,2,?" I get a value of FFh (255 decimal), not 10h (16 decimal). Doesn't appear that the PCI util can change the value either. I tried it a number of times and always got FF when I rechecked it. Either way,I'm glad to see the problem has finally been isolated. I'd hate to have to revert to Win98 though, yuck!! Has anyone else besides me tried these various utils on Win2k and gotten different results? KCM 08-31-2001 23:42:53

Will it work in this or that OS? (modified 0 times) Georgie Profile Kcmjr and jbarr and others, Once I have the final solution it will be very easy to incorporate it into the BIOS, so that any OS can take advantage of the working USB. Read: it will work in any OS. Yes, I am still tuning and testing the fix, but it is about to be released. So stay tuned! Georgie 09-01-2001 08:48:33

Win2k (modified 0 times) Georgie Profile Kcmjr, Just a note about NT/Win2k. I am not at all surprised to hear that pci.exe doesn't work under the NT core, because it's HAL is supposed to isolate the hardware from programs runnning on the machine. So it is actually normal. You would need an NT utility, or the best is to put it into your BIOS. Georgie 09-01-2001 08:52:27

RE: (modified 0 times) Wild_Pencil Profile | Email Kcmjr, I am working on a CMOS-driven ROM that implements this fix. Along with an accompanying C program that should compile under DOS and Linux, I think this new ROM would address the non-WinDos users. I'll probably have something in Beta form sometime next week. -WP 09-01-2001 11:42:27

RE: (modified 0 times) Wild_Pencil Profile | Email Doh! I just re-read and saw what Kcmjr's problem really is about: Win2k is blocking him out. Since Kcmjr is always getting "FF", I'm certain that Win2k is blocking access to the raw PCI I/O ports. This is "protected" stuff that I don't think even Administrator can access directly. I think the same goes for the CMOS ports, now that I think about it. Hmmm.... If it's not a dual-bootable machine, and the Sandisk isn't bootable, I'm not sure how to address Win2k with my fix. Hmm.... By the way... can Win2k even do an Awdflash without blocking you out? -WP 09-01-2001 11:52:53

RE: (modified 0 times) laserfan Profile Programmer said (some time ago now, in the FixUSB thread--sorry for the duplicate post): >The only odd thing left I see with your system is the fact that irq routing for the >USB posts shows up as disabled. >ckbone: The winchip machine is the same as laserfans, IRQ routing is disabled... I don't know what this means. Where does one find this info about one's config? Another question I have is: can anyone suggest why I have two (count 'em) USB Root Hubs and 2 Host Controllers? Noticed in Device Manager, and then/now WCPUID shows I have 2 UHCI USB Controllers: Bus Dev Fnc VendorID DeviceID Revision SubSystem IRQ Device Type --- --- --- --------- --------- --------- --------- --- --------------------- #0 #0 #0 1106h 0501h 03h 00000000h - Host Bridge #0 #1 #0 1106h 8501h 00h 00000000h - PCI-PCI Bridge #0 #7 #0 1106h 0686h 1Bh 00001106h - ISA Bridge #0 #7 #1 1106h 0571h 06h 00000000h - IDE Controller #0 #7 #2 1106h 3038h 0Eh 12340925h 15 UHCI USB Controller #0 #7 #3 1106h 3038h 0Eh 12340925h 15 UHCI USB Controller #0 #7 #4 1106h 3057h 20h 00000000h - Host Bridge #1 #0 #0 1023h 8420h 5Ch 84001023h 15 VGA Compatible This strikes me as odd for a device that has only one USB port. Anyone understand why this is the case? Also, if "IRQ Routing" is disabled on my machine, isn't the fact that three devices have IRQ 15 a problem??? 09-02-2001 20:32:14

About those USB controllers... (modified 0 times) Georgie Profile laserfan, Nothing special about this. The SouthBridge contains two USB controllers, and each is able to drive two USB ports, for a total of 4 USB ports. Out of these four only one is installed on the IO motherboard, and there is place for a second one. In theory, one could install two more USB ports, but that would mean a *lot* of soldering... Geogie PS: Still finalizing the fix.... 09-02-2001 22:13:37

RE: (modified 0 times) ckbone Profile Laserfan....as far as the IRQ routing being disabled....On several of my I-Openers I disabled IRQ steering (same as routing I think). I did this to remove the USB controller from sharing it's IRQ with the trident chip. They normal1y share IRQ 15. Supposedly, USB works better if it has it's own IRQ......didn't help me though. TO do this you disable an IRQ for the VGA in the BIOS, you then assign IRQ 15 to Legacy ISA in the BIOS, then in Windows Device manager/system/pci bus/settings, you disable "use IRQ steering". (I think that's right). Anyway when you reboot the USB uses IRQ11, and the Trident chip is not assigned an IRQ. This made absolutely no difference for me......Maybe one of those tests I sent Programmer was done on a machine set up like that. 09-03-2001 05:25:11

RE: Brainpower (modified 0 times) laserfan Profile I just want to state publicly that I appreciate the brainpower that's being expended on this problem. Hopefully we will come to a happy conclusion. I have said before this IO is "my favorite computer", and I have also said that I have not heretofore considered this xsfer issue to be a big problem. However, I've realized recently that even as a fun-toy "surf and email" computer I have found myself downloading stuff off the Web which I subsequently want to off-load to my main PC, and omigosh the stored stuff got hosed in transfer and I didn't even know it! I haven't even tried using my USB CD-RW with the IO, but can REALLY imagine being aggravated given how damn long the CD-burn process is (i.e. even when it WORKS it's a time-consuming PIA). So I hope we can find a solution to this so we can trust our IOs to be reliable members of our network families. 09-03-2001 08:43:44

RE: (modified 0 times) Programmer Profile wild_pencil: how do you access the cmos memory?? and have you located a particular byte that can be used? 09-04-2001 06:53:13

RE: (modified 0 times) Wild_Pencil Profile | Email Programmer: I think I/O Ports 72 and 73 access the full 256 bytes of CMOS; while Ports 70 and 71 access the lower 128 bytes of CMOS. To find an acceptable byte, I'm planning to wipe the CMOS with a pattern, allow the BIOS to restore its defaults upon checksum error, and (hopefully) see what byte regions remain intact. -WP 09-04-2001 07:34:40

Status update (modified 0 times) Georgie Profile So.... After long hours (days??) of testing, I have what I would call a partial solution. Originally I thought I would only release a final solution, but I cannot donate all my spare time on this project right now - because this week I have none...-, but I would still like the lucky ones benefit from my work. Basically, I have a bunch of patches known to work error-free with a USB CF card reader and a USB CD-RW. The patches are somewhat complicated, because part of them is unfortunately PCI-speed specific. But still, if you have any of these USB devices "working" at one of the following PCI speeds, please, send me an email, and you will get the patch for testing and mostly enjoying! I have working patches for these PCI speeds (MHz): 36.7 (FSB=110) 33.3 (FSB=100 or 66,7) 31.8 (FSB=95) Once again. The patches work with my CF card reader/writer and my CD-RW drive. None of them work correctly with USB network adapters without TCP/IP - thanks to laserfan for testing! -, but please, feel free to try it! There is a fundamental difference between the patch needed for my CF writer and my CD writer, which is most certainly caused by the driver software. If you happen to have a USB device of which the driver needs the same settings as my CF card writer, you are lucky, because your device will work absolutely error-free at any PCI speeds! Please, in your message send me the following information: your hard drive model preferably, if not, at least its size and whether it is a UDMA, ATA-33 or ATA-66 capable drive. You can contact me at "my BBS ID" at chez dot com. Thanks. Georgie P.S.: No, I'm not done with it. I will have time next week. I will try to get to the guts of the problem and make your USB NICs work as well! 09-04-2001 21:35:52

RE: (modified 0 times) laserfan Profile Georgie wants to know: >your hard drive model preferably, if not, at least its size and >whether it is a UDMA, ATA-33 or ATA-66 capable drive I was puzzled about why he wanted to know this on my IO (I have a fast hard drive installed) but he returned to me some tweaks that have made my disk access "wildly faster" e.g. apps just leap open now. As he explained it to me, since the IO wasn't set-up to actually USE the IDE circuitry, it basically defaults to "slow" and so he has been tweaking these settings to get his CD-RW and HD to work together. We haven't found the "sweet spot" yet for my NIC, but in the meantime I am enjoying a palpably speedier i-opener here. Whoosh! 09-05-2001 05:04:06

Faster *is* better (modified 0 times) El_Kabong Profile | Email Faster hard drive access? Do tell! 09-05-2001 07:48:34

RE: (modified 0 times) ckbone Profile Yipes.....I tried the VIA IDE patch recently....could not get my new IBM drive to operate at a UDMA 33, although it is supposed to. A patch to fix that would be as valuable as a USB patch. This is getting interesting...... 09-05-2001 09:02:56

RE: (modified 0 times) Kcmjr Profile Wild_Pencil, I have not tried awdflash under Win2k, but I suspect that it won't work. I've grown accustomed to this since switching to NT a few years ago. In my job I exclusively support and engineer NT and Win2k based systems so I have no need for the other OS's at home. I suspect a software based solution will be a lot more work for NT/W2K so if someone does come up with a BIOS patch that would be best. I'll just sit here and patiently await the end result..... if it helps me great, if not I can work around the problem. Either way everyone's hard work is very much appreciated..... 09-08-2001 23:45:04

RE: (modified 0 times) Kcmjr Profile Wild_Pencil, I have not tried awdflash under Win2k, but I suspect that it won't work. I've grown accustomed to this since switching to NT a few years ago. In my job I exclusively support and engineer NT and Win2k based systems so I have no need for the other OS's at home. I suspect a software based solution will be a lot more work for NT/W2K so if someone does come up with a BIOS patch that would be best. I'll just sit here and patiently await the end result..... if it helps me great, if not I can work around the problem. Either way everyone's hard work is very much appreciated..... 09-08-2001 23:49:59

RE: (modified 1 times) Wild_Pencil Profile | Email Here's the code I am using for the CMOS-Driven Georgie USB Fix. It will be available through at least 3 channels: 1) Do it yourself -- use NASM or tailor this for COM2ROM, to build your own ROM. 2) (coming soon) Do it yourself via my BIOS-Analyzer Website: http://www.lalee.net/biosinfo/ 3) Wait for the new Full Version to reach BadFlash. ; ;CMOS Georgie-USB Fix 1.0 ;Written by L. Lee ;Released with Aloha to the Public Domain ; ; Usage in NASM (or COM2ROM): ; CALL WildPencil_GeorgieUSBFix ; %define CONST_GEORGIE_SLOT 129 ; %define CONST_GEORGIE_SLOT_CHECK 130 ;Check-Byte = $FF - [Georgie_Slot] WildPencil_GeorgieUSBFix: Push EAX Push EBX ;Retrieve the GEORGIE-Value from CMOS Mov BL,0Eh ;//Default GEORGIE-FIX is 0Eh ;//Retrieve the Checksum . . . Mov DX,072h Mov AL,CONST_GEORGIE_SLOT_CHECK Out DX,AL In AL,073h Mov AH,0FFh Sub AH,AL ;AH = Target Value . . . ;//Retrieve the Georgie-Value . . . Mov AL,CONST_GEORGIE_SLOT Out DX,AL In AL,073h Cmp AH,AL JNE USE_DEFAULT_GEORGIE Mov BL,AL ;Update GEORGIE with CMOS-Driven Value USE_DEFAULT_GEORGIE: ;------------------------------------------------ ;PCI CONFIGURATION SPACE . . . ;1000 0000 / 0000 0000 / 0000 0 ; 000 / 0000 0000 ; ^^== Reserved (0) ; ^^^^ ^^ == Register ; ^^^ == Function # ; ^^^^ ^ == Device # ; ^^^^ ^^^^ == PCI BUS # ; ^^^ ^^^^ / == Reserved ;^ == Enable Configuration Space ;------------------------------------------------ ; BUS0 DEV7 FN2 REG 0C-0F ;1000 0000 / 0000 0000 / 0011 1 ; 010 / 0000 1100 ;------------------------------------------------ Mov eax,080003A0Ch mov dx,00CF8h out dx,eax mov dx,00CFCh in eax, dx ;Modify and Update the DWORD . . . ;0F 0E 0D 0C Mov AH,BL ;xxxxxxxx xxxxxxxx !!!!!!!! xxxxxxxx Out DX,EAX ;Write to Port 0CFCh Pop EBX Pop EAX Ret And here's a Delphi Application that updates the CMOS Value. I don't have time to port it to C at the moment; but it shouldn't be hard for a programmer to figure out by reading through the code. program Georgie; Uses SysUtils; {$R *.RES} // // Georgie CMOS Updater 1.0 // Written by L. Lee // Released with Aloha to the Public Domain // Const CONST_GEORGIE_SLOT = 129; CONST_GEORGIE_SLOT_CHECK = 130; //Check-Byte = $FF - [Georgie_Slot] //As more CMOS Values get used, add other CONST_* Values here //to update them . . . var v : Integer; //Value used to set Port Byte. // Note: PortIn and PortOut will work in Windows 95, but will NOT work // under Windows NT. (* ------------------------------------------------------------ *) FUNCTION PortIn(IOport: WORD): BYTE; ASSEMBLER; REGISTER; ASM MOV DX,AX IN AL,DX END; (* ------------------------------------------------------------ *) PROCEDURE PortOut(IOport: WORD; Value: BYTE); ASSEMBLER; REGISTER; ASM XCHG DX,AX OUT DX,AL END; (* ------------------------------------------------------------ *) procedure Display_Usage; begin writeln('*** GEORGIE FIX-USB INITIALIZER ***'); writeln('Usage: '+ParamStr(0)+' <timer value>'); writeln('Specify Timer in HEX, from 00 through FF'); writeln; end; (* ------------------------------------------------------------ *) begin if (ParamCount <= 0) then begin Display_Usage; Exit; end; Try v := StrToInt('$'+ParamStr(1)); Except Try v := StrToInt(ParamStr(1)); Except Display_Usage; Exit; End; //Inner-Try End; //Outer-Try //Georgie-Value PortOut($72,CONST_GEORGIE_SLOT); PortOut($73,v); //Georgie-Checksum PortOut($72,CONST_GEORGIE_SLOT_CHECK); PortOut($73,($FF-v)); writeln('Georgie Timer Value updated to '+IntToHex(v,2)+' Hex'); end. 09-09-2001 13:08:19

RE: (modified 0 times) Kcmjr Profile I know this is off topic but I would sure would like to know why my posts are showing up twice????? Anyone have any ideas? Hate to annoy everyone all the time with duplicate posts....... 09-11-2001 15:24:15

RE: (modified 0 times) ckbone Profile I had a double-post problem when I tried Netscape6.0.....Crappy browser. 09-11-2001 17:10:45

RE: (modified 0 times) zonyl Profile This may be likely the reason Linux users have been having problems with USB NICs to work consitently. Many of us have observed that after a certain amount of data transfered via FTP, that the USB driver will timeout and cause the driver to bail. I was working with author of the pegasus driver (petkan@dce.bg) for quite a while to hunt down this problem but gave up as it was not able to be reproduced on his equipment. It appears that this problem was more deeply rooted and we were looking in the wrong place. The apparent problems in Linux have seemed to go away, but I think that was due to a more robust version of the USB core that handled timeout's better. It probably should be brought up to be included in the kernel though for this machine type. (although they sometimes do not like to include patches for broken hard/firmware) 09-15-2001 07:24:06

RE: (modified 0 times) El_Kabong Profile | Email So is this thread dead? Not much activity of late. 09-17-2001 22:54:38

RE: (modified 0 times) laserfan Profile Not dead, just slowed, I think, by other concerns, priorities, and current events. For my part I had broken my USB/Enet dongle, but expect to get at more testing later this week. 09-18-2001 04:59:24

RE: (modified 0 times) Wild_Pencil Profile | Email Not a whole lot to report on this fix.. the Beta-Testers haven't reported any problems with tbe BIOS implementation of this, so it looks like this issue is close to being closed. As for the Award BIOS upgrading website, I haven't had much time to invest in it lately. Tech companies in Silicon Valley are cutting back like mad, and I'm now among a huge pool of talented tech-workers applying and interviewing for the few remaining jobs. The terrorist attack just made things a whole lot worse for us, as most Silicon Valley companies are now in a panic and have issued hiring freezes. It's gonna get VERY ugly out here soon. How's the tech-industry outlook in the rest of the USA? -WP 09-19-2001 14:23:45

RE: (modified 0 times) 02U2 Profile The Aircraft industry not much better... 30,000 estimated to go soon. I've got over 15 years senority in my (Union) shop but the gators may be even nippin' at my rear end by end of next year too! I am often looking the Tech industry here in the Arse end as a interesting hobby. Electronic/computer parts discards from OEM and corporate shops have drastically dropped off. Trash dumpsters that used to be overfilled now are barely even 1/3 full (if that) on collection day...and most have lost their often unbelievable generosity. Looks like things are getting VERY tight! One more local small OEM looks like it's going belly up. A bunch of used desks are now setting out back along with some file cabinets... 09-19-2001 15:49:21

RE: (modified 0 times) Kcmjr Profile I work in the I.T. design and engineering department of a very large medical group. Although we've got a hiring freeze on we're not looking at any layoffs and don't expect to. Our freeze was due to budgetary issues from the first two quarters of this year not the current state of world affairs. We're actually hurting for people and will likely start hiring again but I don't expect the freeze to lift until January. Back to the subject, does anyone know if the USB can be upgraded to USB v2? I'm assuming that it's a hardware based upgrade rather than firmware or software so we would be out of luck here. Not that it would necessarily even help with this issue. Is the Via chipset even flashable? 09-20-2001 12:41:00

RE: (modified 0 times) ckbone Profile I don't know if this is a break-through but I have made some progress here. I just installed XP corporate final.....the first USB device I tried, a USB ZIP 250, now works fine both ways....I tried files 2-20mb and a variety of pictures, bitmap, jpeg 1-8mb. No corruption was found at a FSB of 70. This was not possible before using WIN98. I have been unable to get my LAN working, so no testing here is possible yet. As to why XP seems to be working with USB....no patches at all....well, we'll see. As I try other devices, I'll report back. 09-20-2001 18:26:57

RE: (modified 0 times) Mitchell Profile | Email I wonder how a stock v1 iopener with the stock winchip 200 and 128 megs of ram would run with winXP?? 09-20-2001 19:27:30

RE: (modified 0 times) ckbone Profile OK..I got my CF reader (Snapshot, Compusa, $20) working on XP. No drivers required...just plug it in. At an FSB of 66 and 70, the transfer of large photos was perfect without any corruption. At 75 however the file transfer crashed the system. I haven't tried out this 75 setting much...could be this setting is unstable at everything. At an FSB of 70 and a multiplier of 3.5, so far everything is working normal. I guess the big test will be to try a large cd burn....maybe this week end. More to come.... 09-21-2001 03:58:53

RE: (modified 0 times) Kcmjr Profile I installed a ZIP 250 on my IO which is running Win2k and have had no problems at all. All files transfer fine yet if I disconnect the ZIP and connect my network adapter the old USB problem returns. Go figure...... KCM 09-21-2001 16:38:49

RE: (modified 0 times) ckbone Profile Maybe I should start a new topic "USB Problem really solved". Not by me of course, but by Windows XP. I installed XP on a regular computer, EPOX VIA apollo pro 133a, PII 450. It has a pci HPNA card...everything installed perfectly. First thing I tried was the LAN...Guess what?....That pesky 1.1mb JPEG now transfered perfectly from the I-Opener to the EPOX. So far no problems with transfers. I'll do more testing.... The I-Opener I used is an AMDk62+450, 70FSB, 3.5 multiplier. I still am not connecting with the other computers on the LAN running WIN98. But I'm happy as a pig!!!! 09-22-2001 13:13:16

RE: (modified 0 times) ckbone Profile Any of you guys want me to run any tests on this setup...let me know. Maybe there is a setting XP uses, that could be applied to your fix patches. 09-22-2001 13:17:16

RE: (modified 0 times) ckbone Profile I can't prove it yet, but here's my theory: XP is using a new network protocol called "QoS Packet Scheduler". "Quality of Service Packet Scheduler. This component provides network traffic control, including rate-of-flow, and prioritization services". Which makes some sense in light of the network transfer problems we've been having. I think there is a way to apply this new protocol to other windows versions. XP in network settings gives you an option to apply this protocol to other computers on your network....it's not really explained this way, but I think that's what's happening. Maybe this QoS protocol will fix the LAN transfer problems on different systems, even without using XP. We'll see. FWIW...I just transferred about 250mb of programs out of the I-O to the EPOX using the LAN. None of the programs I've tried so far, have been corrupted. I still have to try all my USB drives, but this is a good start. 09-22-2001 15:30:10

RE: (modified 0 times) ckbone Profile He, he. Well...guess who has his Win98 I-Opener transferring files over his USB LAN? Look, this still needs more testing. The little file I have, modifies your network settings, requires two reboots, runs out of windows....easy install. Just double-click on it. I set mine up to fix my LAN only...I think you can also use it for internet connection sharing. Try it to see. I'm willing to send it to anyone who has helped out on this topic. Post your address. If it works for everyone, maybe we can make it available somewhere. It should work on all windows versions. ckbone 09-22-2001 16:24:51

RE: (modified 0 times) YouBecha Profile OK, hit me with the fix... mr_bubba_zanetti at yahoo dot com http://www.geocities.com/mr_bubba_zanetti/ 09-22-2001 16:32:01

RE: (modified 0 times) ckbone Profile More notes on this fix....I'm having problems getting this to work on my Winchip2 I-Opener (Win98). It's working on the AMD I-Opener with Win98 and XP. May be a network setting I'm overlooking. Uninstall NetBEUI before appling the patch. Be sure spx/ipx is installed before applying the patch. This needs more testing. 09-22-2001 19:00:29

RE: (modified 0 times) ckbone Profile Worked on this all morning....These are not final conclusions, and a different LAN system may perform better/differently.....but, it seems the success of a file transfer depends on the recieving computer...it must have XP installed. The I-Opener with XP, or with Win98 and this new patch applied, can successfully send files to another machine....but only if the recieving computer is running XP. Files into the I-Opener remain ok from any system. Of course I only have XP on one I-Opener so far....plus I'm not sure about the driver for this HPNA Lan, since it is using a Win2k driver. The normal computer uses a native XP driver for it's pci LAN card. This thing needs more testing, maybe by someone using a Cat5 LAN system, rather than an HPNA system...... 09-23-2001 08:19:26

RE: USB Problems (georgie can i contact you) (modified 0 times) Spine Profile | Email I have a webplayer and not an i-opener, but the webplayer suffers from USB problems as well. The only question I have is how to determine what bus and device the USB controller is on. For georgies example, it was "Bus:0, Device:7, Function:2" and using the pci.exe program the register values are displayed. Anyone want to give me a hand at trying to figure out the problem with the webplayers as well? ps. in order not to clog this message board, i can be emailed. Thanks for the help. 09-23-2001 16:19:29

RE: (modified 0 times) ckbone Profile Unless something else comes up, or I can find more info, here's the final result of the XP/I-Opener/USB LAN setup. XP fixed all my USB LAN I-Opener problems, but only if XP is on either the sending or recieving computer. I can send files from a Win98 I-O to any computer running XP on the LAN. I can send files from an XP I-O to any computer with any OS. As before files into the I-O are ok. I cannot send files from a Win98 I-O to any computer running Win98(same as before). The little file I have lets XP communicate with Win98 computers, but it doesn't solve I-Opener problems completely. If you only had one computer on your LAN with XP, any I-O on the LAN should be able to send files to it successfully. If you put XP on everything, all your USB problems should be solved. I tried all my USB drives....they all work now with two exceptions. I can't find drivers yet for my EXP Parallel port cdrom. My USB HP cdrw (8230) works as a cdrom using Win2k drivers. However, no recording software I tried works with the HP....not even on my EPOX with XP. There's more possibilities with this XP "fix", and I'll keep trying more stuff. The new protocal "QoS", does not install on Win98. Than may be the clue here...that may be what's making the I-Opener's USB work with XP. I'm not savy enough to figure all this out...but it's a start. 09-24-2001 16:55:19

RE: (modified 0 times) zyxw Profile | Email Hi, Did I miss it... is there a bios fix out in this thread somewhere to download??? 09-25-2001 09:42:51

RE: (modified 1 times) Wild_Pencil Profile | Email Only the Beta-Version of the BIOS was released thus far -- unless someone built and released a new ROM Image based on my code posted above. I'm still waiting for any problem reports or complaints on the Beta before I build the final commercial version for Badflash. (Had one complaint of the BIOS Beta using "F4" instead of the <ESC> key for the CMOS Setup-Menu; which was retracted when I explained this lets the original IOpener keyboard configure the BIOS..) Badflash will have chip-replacements as before; and I'm considering email/paypal distribution for those who don't want the chip. It's been slow going for the BIOS development, I know. I've got a few ideas on building a "Lojack BIOS" patch, once I get the floppy-booting stuff working. Unfortunately, I need to devote most of my time to resume writing and phone-screen interviews. CRM and Web-Portal gurus like me are a dime-a-dozen right now... -WP 09-25-2001 14:33:28

RE: (modified 0 times) zyxw Profile | Email Thanks, I'll start a thread and inquire if anyone is willing to provide a non-commercial bios version of the patch 09-26-2001 16:43:46

New version of the USB fix: now working. (modified 0 times) Georgie Profile I have managed to get to the guts of this problem finally. Please, see the new thread. Georgie 04-13-2002 21:46:14

Reply to Thread | Printer | All times are PST Powered by UltraBoard v1.62

I-Appliance BBS General - NEWS, Breaking NEWS - linux-hacker.net BBS on CD Questions - User Groups in Your Town - E-Bay, Yahoo Auction LINKS ONLY (MODERATED) - Jokes & Humor - usdtv box I-Opener Areas - I-Opener General Posts - I-Opener Technical Stuff - I-Opener Technical Info Reference (MODERATED) WebSurfer Areas - WebSurfer General Posts - WebSurfer Technical Stuff - WebSurfer Technical Info Reference (MODERATED) Other I-Appliances - Virgin Webplayer - WebPal - MailStation - 3Com Audrey - ThinkNIC / Oracle NIC - Gateway/AOL Connected Touch Pad - Philips/AOL-TV - ACER NT-150 STB - MSN Companion - Compaq IA - Emachines IA - IBM NetVista IA - Intel dot.station - MSNTV2 Network Computers - Fujitsu-Siemens Scovery Linux Net Computers - GallantWeb Server Consoles - Dreamcast - xbox hacking Web Pads - Epods One General Post - Panasonic CF-01 - Fujitsu Stylistic 1200 - Cybiko Thin Clients - Thin Clients DigitalVideoRecorders - DTV Recorders, Replay, TiVo, Etc... - Tivo Hacks Linux - Midori Linux General Board - ITX MotherBoards Single Board Computers - Earth Computers mARMalade Elite Member Area MISC Areas - The Swap Meet WE TAKE NO RESPONSIBILITY FOR ANY TRANSACTION HERE AND NO COMMERCIAL POST PERIOD !!! - Wireless Net - Cameras - Robotics - Routers/Gateways - Mattel JuiceBox