Okay, I know you guy probably thought you got rid of me, but I'm still around. I'm currently trudging through everything I can find for docs on both the 440zx-66 and 440bx chipsets, and looking what's currently there for code and what exactly needs to be implemented. I still have the 440zx mobo (bios mod never worked out), and in the morning I should be able to get a 440bx, one that hopefully has a removable bios chip. I'm curious, is there any way to test the code through software, without a flash? Also, as far as POST cards go, any one thats better than the rest? Or just the cheapest thing I can find will work?
Corey Osgood
----- Original Message ---- From: Richard Smith smithbone@gmail.com To: Uwe Hermann uwe@hermann-uwe.de Cc: LinuxBIOS linuxbios@linuxbios.org Sent: Sunday, November 5, 2006 2:06:40 PM Subject: Re: [LinuxBIOS] 440bx stuff
Uwe Hermann wrote:
This has been done. It's now i82371eb.
Why not 82371eb (that's the official name AFAIK)? Is there some requirement that directories have to start with a letter?
Its a C thing. You can't start a variable with a number. That allows you to match the directory name to the struct statements in the code.
I think I'll have a look at this in the next few days. But I might use the 855pm code (derived from e7501) as a basis, that looks even better.
the e7501 datasheet has a date on it thats almost a full year earlier than the 855pm. I think the register names and operation may be closer to the 440bx. But feel free to use whatever. Its mostly about framework.
* Corey bario2004@yahoo.com [061107 10:43]:
Okay, I know you guy probably thought you got rid of me, but I'm still around.
Glad you are. We never intended to get rid of (potential) contributors. ;-)
I'm currently trudging through everything I can find for docs on both the 440zx-66 and 440bx chipsets, and looking what's currently there for code and what exactly needs to be implemented. I still have the 440zx mobo (bios mod never worked out), and in the morning I should be able to get a 440bx, one that hopefully has a removable bios chip.
great.
I'm curious, is there any way to test the code through software, without a flash?
Not for this part of the code. Software emulations dont emulate the chipset well enough to find out whether your code does the right thing. Unless you are using SimNow to simulate an AMD system. Dont know if something like this exists for Intel CPUs though.
Also, as far as POST cards go, any one thats better than the rest? Or just the cheapest thing I can find will work?
Get yourself a serial null modem cable attached to a second box... its far more useful than a post card. getting serial output is not too hard.
Stefan Reinauer stepan@coresystems.de writes:
I'm curious, is there any way to test the code through software, without a flash?
Not for this part of the code. Software emulations dont emulate the chipset well enough to find out whether your code does the right thing. Unless you are using SimNow to simulate an AMD system. Dont know if something like this exists for Intel CPUs though.
I would say that this is a shame. We really need a top-notch full system simulator. But doing complete and correct simulation of chipsets is hard, error prone and quite boring. Therefor emulators like QEMU take the short path, and put in a "phony" BIOS and more or less do not implement a northbridge at all.
What we need is a full-system simulator with the following characteristics;
* correct, * deterministic, * visible device state, * visible cpu state, * scriptable, * and pluggable (for profiling and coverage tools)
Note that I have not put "fast" on the list. I think that is secondary for our use case.
~j
* Johan Rydberg jrydberg@gnu.org [061107 14:28]:
I would say that this is a shame. We really need a top-notch full system simulator. But doing complete and correct simulation of chipsets is hard, error prone and quite boring. Therefor emulators like QEMU take the short path, and put in a "phony" BIOS and more or less do not implement a northbridge at all.
I know SimNow can theoretically do it, and it can be enhanced by Plugins for CPU, northbridge, southbridge, other bridges, other pci devices
But I doubt AMD is going to add Intel CPU plugins to it ;-)
Can Simics to that stuff?
What we need is a full-system simulator with the following characteristics;
- correct,
- deterministic,
- visible device state,
- visible cpu state,
- scriptable,
- and pluggable (for profiling and coverage tools)
Note that I have not put "fast" on the list. I think that is secondary for our use case.
For this simulator to be done, a whole lot of understanding of the hardware components is required, maybe including their errata. Adding support for a new component for such a simulator _might_ be more work than writing a bios for the same chipset.
But no doubt this would be great to have.
Stefan
Stefan Reinauer stepan@coresystems.de writes:
But I doubt AMD is going to add Intel CPU plugins to it ;-)
Can Simics to that stuff?
Yes, it can. Simics is more or less a set of connected components, that is glued together by a set of scripts that builds the system configuration. I know LinuxBIOS has been run on some of the models.
What we need is a full-system simulator with the following characteristics; [...]
For this simulator to be done, a whole lot of understanding of the hardware components is required, maybe including their errata. Adding support for a new component for such a simulator _might_ be more work than writing a bios for the same chipset.
Yeah. I didn't say _we_ should implement a simulator just to run LinuxBIOS on it. Other projects should gain from a simulator, but LinuxBIOS might benefit the most from having one. Lets just hope someone identifies the need outside this project.
~j
Johan Rydberg wrote:
For this simulator to be done, a whole lot of understanding of the hardware components is required, maybe including their errata. Adding support for a new component for such a simulator _might_ be more work than writing a bios for the same chipset.
Yeah. I didn't say _we_ should implement a simulator just to run LinuxBIOS on it. Other projects should gain from a simulator, but LinuxBIOS might benefit the most from having one. Lets just hope someone identifies the need outside this project.
The only people who can do this are the people who have the hardware description info (the VHDL or whatever). I'm quite sure Intel has sme sort of tool like SimNow internally but they just aren't talking or its only released under NDA.
Stefan Reinauer wrote:
there for code and what exactly needs to be implemented. I still have the 440zx mobo (bios mod never worked out), and in the morning I should be able to get a 440bx, one that hopefully has a removable bios chip.
Yeah! A developer to the rescue.
Also, as far as POST cards go, any one thats better than the rest? Or just the cheapest thing I can find will work?
Get yourself a serial null modem cable attached to a second box... its far more useful than a post card. getting serial output is not too hard.
+1 And depending on your superIO it might already "Just Work"
Stefan Reinauer wrote:
there for code and what exactly needs to be implemented. I still have the 440zx mobo (bios mod never worked out), and in the morning I should be able to get a 440bx, one that hopefully has a removable bios chip.
Yeah! A developer to the rescue. Go! man Go!
Also, as far as POST cards go, any one thats better than the rest? Or just the cheapest thing I can find will work?
Get yourself a serial null modem cable attached to a second box... its far more useful than a post card. getting serial output is not too hard.
And depending on your superIO it might already "Just Work"
btw, the digitallogic/smartcore-p3 in v1 is a 440-bx based board.
ron
ron minnich wrote:
btw, the digitallogic/smartcore-p3 in v1 is a 440-bx based board.
If it is of any help, I could try this on my i440bx board.
Just point me to the tar-ball.
Thanks!
-- Al
-
Al Boldi -
Corey -
Johan Rydberg -
Richard Smith -
ron minnich -
Stefan Reinauer