Announcement

Collapse
No announcement yet.

8-pin NOR flash and bad flash update for MSI B350

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    8-pin NOR flash and bad flash update for MSI B350

    Hi. Do all of those 8-pin SPI NOR flash chips found in motherboards always have the same pinouts?

    I have an MSI B350 Tomahawk with a Ryzen processor. The BIOS was version 1.20. We couldn't get the RAM to run anywhere near what it was supposed to run at (3200 MHz). The CPU, motherboard, and RAM came as a packaged deal from Newegg. The newest BIOS is 1.90 I belive and says it fixes memory compatibility issues. We were just enabling XMP and it wouldn't POST. We tried updating the BIOS through the UEFIs BIOS update program.

    It wouldn't post at all after that. Just a black screen. I guess this is a common issue with these boards and these processors.

    Anyway, my spi programmer shows the MISO and MOSI pins switched compared to what the data sheet says. Would that matter? I would like to think all these 25xx flashes share a standard pinouts. There's only one BIOS on this board.

    This board has the 2x6 JSPI1 header and I have the pinouts for it. I can dump it and reprogram it in circuit. Just not sure if that really mattered or not.

    Thanks.
    Last edited by Spork Schivago; 12-09-2017, 12:53 AM. Reason: Fixed autocorrect making ttp
    -- Law of Expanding Memory: Applications Will Also Expand Until RAM Is Full

    #2
    Re: 8-pin NOR flash and bad flash update for MSI B350

    You are not alone.
    https://www.reddit.com/r/Amd/comment...i6&sh=17aa60c4

    Maybe this will help.
    https://forum-en.msi.com/index.php?topic=283908.0

    Comment


      #3
      Re: 8-pin NOR flash and bad flash update for MSI B350

      Originally posted by gabiz_ro View Post
      Yes, this is what I meant by a common problem with the Ryzen + AM4 + MSI combination. It appears not to be just the B350. This could be why certain MSI boards have the JSPI1 jumper next to the BIOS. I don't have a lot of experience with MSI boards, so maybe they've been doing this from the git go, but the J stands for Jumper, the SPI stands for Serial Peripheral Interface, the 1 is the first jumper of this type on this board (and the only). There are pinouts available. Some older boards have a 2x5 header, mine has the 2x6. I don't have the proper jumper cables to hook to it though, so I have to unsolder the chip.

      Originally posted by gabiz_ro View Post
      This is what I tried, kinda. I could not get the chip to be detected by my clip and the CH341A programmer I have, so I had to unsolder it. I flashed 1.90 directly to the chip, and resoldered. The CPU and DRAM lights come directly on, but that's it.

      Notice how the chip is a MX25U12873F but in the thread, they tell flashrom it's the MX25U12835F. I use a program called CH341A Programmer, version 1.31. It identifies the chip as a MX25U12835F as well. The sizes match just fine. My programmer uses 3.3v and I haven't had any issues with the chip frying, despite the datasheet saying the eeprom is 1.8v. The program I use lists the chip as an MX25U12835F, 1.8v. Perhaps the program can somehow change the voltage on my programmer?

      Anyway, here's where it gets a bit weird. I pull the NOR. I try flashing 1.90v, verify fails. I try over and over again, verify always fails. I compare a dump of what I burned and the original firmware. It always fails in the same spot. There was NVRAM data in that area. It was like the NOR wasn't properly erasing certain blocks.

      I did not try soldering it back on until I got the verify to match. I had to erase the NOR flash first, and then I was able to successfully burn 1.90 without having issues with the verify, but it still don't POST. I made a mistake and forgot to backup the NOR first, before attempting to flash 1.90 on there. I figured because it was bad flash, it didn't matter. This was wrong. Maybe that information needs to be there? Or maybe I should attempt to flash the original firmware version back on?

      I dunno. I had removing that chip. A good bit of work (unhooking everything, removing the board, unsoldering the chip isn't too bad, soldering it back on is a bit of a pain, putting it all back together, etc).

      I'll try flashing it again on board. Maybe my clip wasn't properly seated....
      -- Law of Expanding Memory: Applications Will Also Expand Until RAM Is Full

      Comment


        #4
        Re: 8-pin NOR flash and bad flash update for MSI B350

        Oh, the original question remains though. Does having the Master Input Slave Output and the Master Output Slave Input switched for programming?

        With my clip, the way I had it setup, according to this programmer, the NOR flash's MISO and MOSI will be reversed. For whatever reason, I was able to burn and successfully verify the burn, so maybe when programming, it doesn't matter so much? Maybe it only matters if there's another SPI chip in slave mode? What do you guys think?
        -- Law of Expanding Memory: Applications Will Also Expand Until RAM Is Full

        Comment


          #5
          Re: 8-pin NOR flash and bad flash update for MSI B350

          miso/mosi are just clever names for tx/rx or in/out
          if you had them wrong, you would not be able to id the chip.

          Comment


            #6
            Re: 8-pin NOR flash and bad flash update for MSI B350

            The BIOS/UEFI update from manufacturers isn't usually the whole dump, just the update.
            What size is the chip and what size is the file from MSI ?
            Did i read right, you didn't backup the original ?

            Comment


              #7
              Re: 8-pin NOR flash and bad flash update for MSI B350

              Yes, because it was a bad update, I did not backup the original. That was a mistake in my part. But I have success. The 16,384KB. The chip is 128Mb / 16MB, so to me, I looked at that as being a full dump from MSI.

              I redownloaded the original from the manufacturer again, just to make sure. I think the problem was the soldering technique I was using and wasn't getting proper connection on some. When I used the DMM to check continuity, I think it would add some pressure to the pin and make the contact.

              I ended up tinning the pads on the board, like I did before, then I'd push down on each pin of the chip. But this time, I went back through and touched the iron to the pad and the pin and added a dab of solder and just drew it to cover the whole pin and pad. As soon as I did this, I powered it up and it POSTed!
              -- Law of Expanding Memory: Applications Will Also Expand Until RAM Is Full

              Comment


                #8
                Re: 8-pin NOR flash and bad flash update for MSI B350

                That's good, I tried using one of those clips in the past but don't any more due to reliability.

                Comment


                  #9
                  Re: 8-pin NOR flash and bad flash update for MSI B350

                  Ah shit. The DMI information was not stored in the BIOS. The UUID info is filled with all FF's. I believe that was there before, however, most of the other fields had originally said to be filled out by O.E.M.

                  I can use an editor to fill in the values for things like serial, but I don't know what the UUID was. Maybe it's calculated from the serial number? I think Serial and UUID where the only two fields populated before.
                  -- Law of Expanding Memory: Applications Will Also Expand Until RAM Is Full

                  Comment


                    #10
                    Re: 8-pin NOR flash and bad flash update for MSI B350

                    Originally posted by diif View Post
                    That's good, I tried using one of those clips in the past but don't any more due to reliability.
                    I noticed there's a new version of the program I use. It's written by a guy over in one of the Asian countries. Because I have the paid version, he was supposed to e-mail when a new version came out. I'm running 1.31. The latest is 1.34. I was looking at the change logs for the newer versions that I don't have, and I see that one says it fixes a bug with write and something with not properly erasing something. This could be why I had to blank the chip first....
                    -- Law of Expanding Memory: Applications Will Also Expand Until RAM Is Full

                    Comment


                      #11
                      Re: 8-pin NOR flash and bad flash update for MSI B350

                      are you using a CH341 ?

                      Comment


                        #12
                        Originally posted by stj View Post
                        are you using a CH341 ?
                        Yes.
                        -- Law of Expanding Memory: Applications Will Also Expand Until RAM Is Full

                        Comment


                          #13
                          Re: 8-pin NOR flash and bad flash update for MSI B350

                          it's now supported by flashrom.
                          http://www.flashrom.org/Flashrom

                          there is also some russian stuff.
                          https://github.com/nofeletru/UsbAsp-flash

                          Comment


                            #14
                            Re: 8-pin NOR flash and bad flash update for MSI B350

                            Originally posted by stj View Post
                            it's now supported by flashrom.
                            http://www.flashrom.org/Flashrom

                            there is also some russian stuff.
                            https://github.com/nofeletru/UsbAsp-flash
                            I like how it's open source, but I like GUIs right now. On top of that, even though it might support my CH341A (or the clone of a CH341A), I don't see anything on the page you linked too about I2C support.

                            What I like best is with flashrom, people could support new chips to the database if they confirm they work or not. With the program I use (you linked me to it originally), the database is in a proprietary format, but does support a lot of chips.

                            flashroom still has a good bit of work that needs to be done I think. It doesn't list the MX25U12873F at all, just like the program I use, unfortunately, but flashrom does list the MX25U12835F, which my chip was falsely identified as with the program I use. However, flashrom does NOT list it as compatible.

                            I wish I had known before I fixed the PC. I would have tried and we could have updated the wiki.

                            I also wish I had the latest version of the program I use. I see there was a bug fix. I had a weird problem writing BIOSes to the MX25U12873F. Every verified failed, for whatever reason. I had to manually erase the chip, then it worked. For whatever reason, always, in the same area (or from what I could tell), it wasn't properly overwriting the bytes. I had it configured to write FF's first, and then the new data. But it always failed. It could have been that chip in particular and maybe that's why there's so many bad flashes (using the BIOS's M-FLASH utility or the Windows flasher). I dunno.

                            I don't understand what they mean by
                            Code:
                            Partial Windows support is available (no internal programmer support at the moment,
                            hence no "BIOS flashing").
                            The no internal programmer support is what is messing me up. flashrom is software, not hardware. I use my CH341A as the programmer. Why would this not work on Windows?

                            Still, flashrom seems promising. It'd be nice to have a program that works in Linux and Windows, and is open-source, stable, free.

                            **EDIT: Does the no internal programmer just mean if I'm booted into Windows and want to flash the BIOS on the machine I'm booted into, I cannot use flashrom to do it, but if I'm booted into Linux, more than likely, I can just run the program and flash the BIOS, without any extra hardware? Like a universal flashing program? Kinda like the Windows BIOS updater programs?
                            Last edited by Spork Schivago; 12-11-2017, 06:41 PM.
                            -- Law of Expanding Memory: Applications Will Also Expand Until RAM Is Full

                            Comment


                              #15
                              Re: 8-pin NOR flash and bad flash update for MSI B350

                              offtopic
                              On CH341 I encountered many times problems writing some chips (random chips mostly from TV)
                              As a rule,after I backup original dump from chip,always do erase and blank check,result is no more problems.

                              Comment


                                #16
                                Re: 8-pin NOR flash and bad flash update for MSI B350

                                Originally posted by Spork Schivago View Post
                                I like how it's open source, but I like GUIs right now. On top of that, even though it might support my CH341A (or the clone of a CH341A), I don't see anything on the page you linked too about I2C support.
                                25 series are not i2c, they are microwire AKA SPI

                                Comment


                                  #17
                                  Re: 8-pin NOR flash and bad flash update for MSI B350

                                  Originally posted by gabiz_ro View Post
                                  offtopic
                                  On CH341 I encountered many times problems writing some chips (random chips mostly from TV)
                                  As a rule,after I backup original dump from chip,always do erase and blank check,result is no more problems.
                                  What software are you using? I wonder for me, if it's the programmer, the software, or the physical chip I was programming. What made me think for me it was the programming software, I saw this on the software page's changelog:

                                  Code:
                                  ...
                                  1.34 version 2017-7-8
                                  Repair S25FL256 read and write BUG
                                  ...
                                  1.33 version 2017-6-15
                                  Repair the previous version write \ erase bug will be reported
                                  ...
                                  That Repair the previous version write \ erase bug was the one I was thinking that was causing the problem. Google translate doesn't do the best job of converting Chinese to English, so we have to kinda try to figure out exactly what the sentence means.

                                  The mention about S25FL256 I don't think applies, but maybe the bug affected more than just the S25FL256? I wish I could have had a chance to try writing with the new software or tried that flashrom program, just to rule out the programmer and software.

                                  Thanks for sharing!
                                  -- Law of Expanding Memory: Applications Will Also Expand Until RAM Is Full

                                  Comment


                                    #18
                                    Re: 8-pin NOR flash and bad flash update for MSI B350

                                    Originally posted by stj View Post
                                    25 series are not i2c, they are microwire AKA SPI
                                    Yes, you're right, but my programmer supports I2C, which are 24xx series EEPROMs, I do believe. And although I run into a lot of the 25xx series, I have ran across a few 24xx series stuff. You follow?

                                    Thanks.
                                    Last edited by Spork Schivago; 12-12-2017, 11:36 AM.
                                    -- Law of Expanding Memory: Applications Will Also Expand Until RAM Is Full

                                    Comment


                                      #19
                                      Re: 8-pin NOR flash and bad flash update for MSI B350

                                      there is a Linux programmer on sourceforge to do i2c on the ch341

                                      Comment


                                        #20
                                        Re: 8-pin NOR flash and bad flash update for MSI B350

                                        Yes, but there's also the current program that I use, that does both of them. So instead of using two programs with probably no graphical user interface, I get to use one program with a graphical user interface.

                                        It'd be nice if both programs merged. That'd be kinda cool.
                                        -- Law of Expanding Memory: Applications Will Also Expand Until RAM Is Full

                                        Comment

                                        Working...
                                        X