5 ms·
TinyBIOS – A minimalist open-source BIOS project for fun
- jamesy0ung 4y agoI’m quite sure the PC Engines alix2 uses TinyBIOS
- spaam 4y agomaybe you thinking about https://en.wikipedia.org/wiki/Coreboot https://en.wikipedia.org/wiki/Coreboot?
- st_goliath 4y agoMost of the ALIX boards use something called TinyBIOS[1][2]. Browsing through the source tree from the pcengines site, it doesn't seem to share any relation with this project here. The current APU boards use Coreboot[3] [1] https://www.pcengines.ch/alix.htm https://www.pcengines.ch/alix.htm [2] https://www.pcengines.ch/tinybios.htm https://www.pcengines.ch/tinybios.htm [3] https://www.pcengines.ch/apu2.htm https://www.pcengines.ch/apu2.htm
- pwk4m1 4y agoAuthor here, there's indeed no relation at all between the projects apart from my unfortunate choice of name for my project.
- rkagerer 4y agoThe demo image is the clearest boot logging I've ever seen, while still being succinct. I wish all BIOS's presented themselves so straighforwardly.
- LoganDark 4y agoYou're absolutely right, I'm impressed. The log is so... polite.
- adrian_b 4y agoThe main task for any BIOS is to initialize the memory controller, which for modern DDR4 or DDR5 memory is a very complex and undocumented task, so it must be done usually by a binary firmware blob provided by Intel or AMD. The usual operating systems might be able to initialize the peripherals, but they expect an already working memory and a memory map provided by the BIOS. After a quick browsing through the sources, it seems that this project has not progressed so far yet, even if it probes for working memory and it may be able to use the cache memory as RAM on some CPU models (as all BIOSes must do, before the memory controller is initialized). Under an emulator like QEMU, a BIOS can work even if it is not able to initialize a real memory controller, but on real bare-metal hardware that has to work.
- monocasa 4y agoThese days in x86 SoCs, DRAM is normally initialized by other cores smaller like the ME and PSP before the main core which would run the bios come up. For instance, apparently oxide's code doesn't have to bring up DRAM despite being the first code to run on the main cores.
- mkopec 4y agoOn AMD Zen and later that is true, but on Intel the BIOS is still responsible for DRAM init (done via Intel's FSP)
- buildbot 4y agoThe fact that DDR works at all is kinda black magic, every boot the memory controller does a training routine to understand the analog aspects of the specific ram and board in order to actually make it function at all. https://www.systemverilog.io/ddr4-initialization-and-calibration https://www.systemverilog.io/ddr4-initialization-and-calibra... Edit: Reading the article is really cool, I just searched it as a reference but it is very in depth!
- packetlost 4y agoA lot of wifi/other wireless analog devices go through analogous training procedures. Things like timing and delays are pretty important to these types of systems and you don't have a clock edge to synchronize against like you do in wired protocols like SPI (though I imagine DDR initialization does have pins used for this or similar).
- mjovanovic0 4y agoSide question: what IDE/editor ASM developers use today?
- monocasa 4y agoWhatever they're comfortable with. There isn't really much IDE support for ASM, so it doesn't really matter what you pick as long as you can do dirty things with the build system. I tend to just use vscode.
- snvzz 4y agoI favor KDE's kate as editor, nasm as assembler, gnu make as build system.
- gps0 4y agoNot an ASM developer, but I know SASM (https://dman95.github.io/SASM/english.html https://dman95.github.io/SASM/english.html) exists.
- snvzz 4y agoChecked sources just to see what assembler was used. It uses NASM, my go-to assembler. A great choice.
- 13of40 4y agoA86 is a great minimalist assembler for 16-bit code. In the humble words of its author: "A86 (with its 32-bit version A386) is the finest assembler available, at any cost under any terms, for the Intel 86-family of microprocessors (IBM-PC, compatibles, and not-so-compatibles)." Actually, come to think of it, I'm not sure if a86.exe is a 16-bit binary or 32, so it might not run on modern Windows.
- snvzz 4y agoA86 might as well be perfect, and I'd still not use it. It is not Open Source.
- Karrot_Kream 4y agoDoes anyone have any recs to learn about what goes into a basic BIOS? The TinyBIOS source is super easy to read but I'm looking more for something like xv6 which is both fun to read and also pedagogic (a tall order I'm asking for free, I know ;)
- hakfoo 4y agoThere are at least three "real" 8086 BIOSes-- "real" as in they can work on actual hardware rather than just emulators-- available in source form, which might be worth examining. https://www.phatcode.net/downloads.php?id=101 https://www.phatcode.net/downloads.php?id=101 -- this is a respin of code that apparently originated in the 1980s in sketchy Taiwanese clones, so I think some of the docs are reverse engineered http://glabios.org/ http://glabios.org/ is an all-new-still-under-development project, but it works well enough for my hobby usage. https://github.com/skiselev/8088_bios https://github.com/skiselev/8088_bios These are more targeted towards some specific hardware projects (Xi8088/Micro8088) Obviously, being for a simple 8086 mrans they don't have some of the complex chipset config used in a modern PC, and even stuff like "hard disc support" is outsourced (for that, you might consider stuff like the XT-IDE Universal BIOS)
- snvzz 4y agoGlabios is GPL3, although unfortunately based upon MASM (non-free and quite old-fashioned assembler). The other two are sketchy and no license (i.e. unusable), respectively.
- 752963e64 4y ago