3 ms·

Will RISC-V end up with the same (or even worse) platform fragmentation as ARM? Because of absence of any common platform standard we have phones that are only

by storus 6mo ago

Will RISC-V end up with the same (or even worse) platform fragmentation as ARM? Because of absence of any common platform standard we have phones that are only good for landfill once their support lifetime is up, drivers never getting upstreamed to Linux kernel (or upstreaming not even possible due to completely quixotic platforms and boot protocols each manufacturer creates). RISC-V allows even higher fragmentation in the portions of instruction sets each CPU supports, e.g. one manufacturer might decide MUL/DIV are not needed for their CPU etc. ("M" extension).

hajile 6mo ago
RVA23 is the standard target for compilers now. If you support newer stuff, it’ll take a while before software catches up (just like SVE in ARM or AVX in x86). If you try to make your own extensions, the standard compiler flags won’t be supporting it and it’ll probably be limited to your own software. If it’s actually good, you’ll have to get everyone on board with a shared, open design, then get it added to a future RVA standard.
- storus 6mo ago
  Thanks, that however addresses only a part of the problem. ARM is also suffering from no boot/initialization standard where each manufacturer does it their own way instead of what PC had with BIOS or UEFI, making ARM devices incompatible with each other. I believe the same holds with RISC-V.
  - hajile 6mo ago
    There has been concerted effort to start working on these kinds of standards, but it takes time to develop and reach a consensus. Some stuff like BRS (Boot and Runtime Services Specification)and SBI (Supervisor Binary Interface) already exist.
  - Findecanor 6mo ago
    There is a RISC-V Server Platform Spec [0] on the way supposed to standardise SBI, UEFI and ACPI for server chips, and it is expected to be ratified next month. (I have not read it myself yet) [0]: https://github.com/riscv-non-isa/riscv-server-platform https://github.com/riscv-non-isa/riscv-server-platform
- MobiusHorizons 6mo ago
  Compiling the code is not the issue. The hard part is the system integration. Most notably the boot process and peripherals. It's not actually hard to compile code for any given ARM or x86 target. Even much less open ecosystems like IBM mainframes have free and open source compilers (eg GCC). The ISA is just how computation happens. But you have to boot the system, and get data in and out for the system to be actually useful, and pretty much all of that contains vendor specific quirks. Its really only the x86 world where that got so standardized across manufacturers, and that was mostly because people were initially trying to make compatible clones of the IBM PC.
indolering 6mo ago
The answer is unequivocally yes: RISC-V is designed to be customizable and a vendor can put whatever they like into a given CPU. That being said, profiles and platform specs are designed to limit fragmentation. The modular design and core essential ISA also makes fat binaries much more straight-forward to implement than other ISAs.
- hajile 6mo ago
  You can choose to develop proprietary extensions, but who’s going to use them? A great case study is the companies that implemented the pre-release vector standard in their chips. The final version is different in a few key ways. Despite substantial similarities to the ratified version, very few people are coding SIMD for those chips. If a proprietary extension does something actually useful to everyone, it’ll either be turned into an open standard or a new open standard will be created to replace it. In either case, it isn’t an issue. The only place I see proprietary extensions surviving is in the embedded space where they already do this kind of stuff, but even that seems to be the exception with the RISCV chips I’ve seen. Using standard compilers and tooling instead of a crappy custom toolchain (probably built on an old version of Eclipse) is just nicer (And cheaper for chip makers).
  - LeFantome 6mo ago
    Yes, extensions are perfect for embedded. But not just there. Extensions allow you to address specific customer needs, evolve specific use cases, and experiment. AI is another perfect fit. And the hyperscaler market is another one where the hardware and software may come from the same party and be designed to work together. Compatibility with the standard is great for toolchains and off-the-shelf software but there is no need for a hyperscaler or AI specific extension to be implemented by anybody else. If something more universally useful is discovered by one party, it can be added to a future standard profile.
bsder 6mo ago
> Will RISC-V end up with the same (or even worse) platform fragmentation as ARM? Sadly, yes. RISC-V vendors are repeating literally every single mistake that the ARM ecosystem made and then making even dumber ones.
- snvzz 6mo ago
  Please elaborate.
pjmlp 6mo ago
Just like everything else outside PC thanks to clones becoming a thing. One reason UNIX became widely adopted, besides being freely available versus the other OSes, was that allowed companies to abstract their hardware differences, offering some market differentiation, while keeping some common ground. Those phones common ground is called Android with Java/Kotlin/C/C++ userspace, folks should stop seeing them as GNU/Linux.
tliltocatl 6mo ago
PC/x86 was an extreme outlier, sadly, and it was because of Microsoft/Intel business model. The architecture details was historically mostly decided on by Wintel, yet the system integration was done by many vendors, whose best interest was to stay as compatible as possible. Its unlikely that another platform would be able to reach this state, the PC architecturing was subsidized from the M$ software monopoly that nobody would have wanted to suffer thru again.
- ajxs 6mo ago
  > Its unlikely that another platform would be able to reach this state... Is this really true? The computer ecosystem is more open now than ever. The original PC BIOS (which PC-compatible manufacturers needed to implement) was never an open, documented standard. It was a proprietary, closed system made by IBM. It's pretty fair to say that IBM didn't anticipate a PC/x86 ecosystem developing around their product. They even sued companies who made their own compatible BIOSes (like Corona). Intel didn't really have much to do with the success of the product at that point in time either, much less Microsoft. In contrast, every widely-used modern system for hardware abstraction (UEFI/ACPI/DeviceTree/OpenSBI/etc) are open, royalty-free standards that anyone can use. Their implementation in ARM is newer, and inconsistent, but that's only because of how hugely diverse the ARM ecosystem is.
  - tliltocatl 6mo ago
    The early PC era was a mess, and that's not the period I'm talking about. IBM was clearly not up to the task and Intel didn't care much yet, but Microsoft certainty did a lot for compatibility from the start (i. e. DOS abstracted away a lot of BIOS routines, so it would be easy to port MS-DOS to a non-IBM x86). But after IBM revealed MCA to show just exactly how much do they care about compatibility and platform openness, Intel realized they are missing out and cleaned up the MCA/EISA/VLB mess with PCI. Then Microsoft and Intel jointly released APM 1992 (which was clearly not enough), and then ACPI in 1996 (which is a total dumpster fire, but a sufficiently functional dumpster fire). I. e. ACPI and UEFI are exactly the product of the monopoly. M$/Intel profited from the abundance of cheapo white boxes, so it was in their best interest to come up with a standard even DELL can implement. The fact that AMD is going to implement the ACPI too wasn't much bother for Intel - they were so dominant that they could afford not to care. On the other hand, ARM sells the cores to SoC vendors (and doesn't care much what becomes of it), SoC vendors ducktape the ARM cores to a bunch of Synopsys peripherals and sell the resulting SoCs to smartphone and car makers (and doesn't care much for the product). System integrators throw Android on top and sell it to the customers. Then Google, who get all the cream via Play, hides all the mess behind a thousand layers of Java abstractions. DeviceTree is an offshot of Sun's OpenFirmware (and it leaves out all the hard stuff - OpenFirmware had Forth, DeviceTree expects the kernel to support every single brand of fan switch). OpenSBI is a disaster. I'm sorry, but what kind of bright mind came up with the idea of hiding damn *timer* behind a privilege switch? Timers were enough of a pain point on x86 already, then it settled on userspace-accesable RDTSC. RISC-V SBI? Reproducing x86 one stupid decision at a time.
LeFantome 6mo ago
> platform fragmentation RISC-V is addressing this issue quite directly. For things like desktops, laptops, SBCs and servers we have the RVA23 profile which defines quite specifically what features a chip must support to ensure code portability. On top of this, there are platform specifications. For example, the server spec is about to finalize next month. It extends RVA23 which things like UEFI, SBI, and ACPI to ensure that your can take something like a Linux distro and easily install it on any RISC-V server, like you can in the world of x86-64. > we have phones that are only good for landfill once their support lifetime is up RISC-V will probably not solve that problem in general. First, the ISA cannot really demand that your phone avoid a Broadcom wireless chip that requires proprietary firmware for example. Also, the phone vendor can still lock down the devices to prevent running arbitrary code. Thankfully, the RISC-V world is developing a culture of openness. If a company wants to create a fully “open” phone, they are quite likely to adopt RISC-V. And, because of RISC-V, even the SoC itself could be fully Open Source. But your typical Android phone is not going to get more Open just because they contain a RISC-V CPU.

by storus 6mo ago

hajile 6mo ago

RVA23 is the standard target for compilers now. If you support newer stuff, it’ll take a while before software catches up (just like SVE in ARM or AVX in x86). If you try to make your own extensions, the standard compiler flags won’t be supporting it and it’ll probably be limited to your own software. If it’s actually good, you’ll have to get everyone on board with a shared, open design, then get it added to a future RVA standard.

storus 6mo ago
Thanks, that however addresses only a part of the problem. ARM is also suffering from no boot/initialization standard where each manufacturer does it their own way instead of what PC had with BIOS or UEFI, making ARM devices incompatible with each other. I believe the same holds with RISC-V.
- hajile 6mo ago
  There has been concerted effort to start working on these kinds of standards, but it takes time to develop and reach a consensus. Some stuff like BRS (Boot and Runtime Services Specification)and SBI (Supervisor Binary Interface) already exist.
- Findecanor 6mo ago
  There is a RISC-V Server Platform Spec [0] on the way supposed to standardise SBI, UEFI and ACPI for server chips, and it is expected to be ratified next month. (I have not read it myself yet) [0]: https://github.com/riscv-non-isa/riscv-server-platform https://github.com/riscv-non-isa/riscv-server-platform
MobiusHorizons 6mo ago
Compiling the code is not the issue. The hard part is the system integration. Most notably the boot process and peripherals. It's not actually hard to compile code for any given ARM or x86 target. Even much less open ecosystems like IBM mainframes have free and open source compilers (eg GCC). The ISA is just how computation happens. But you have to boot the system, and get data in and out for the system to be actually useful, and pretty much all of that contains vendor specific quirks. Its really only the x86 world where that got so standardized across manufacturers, and that was mostly because people were initially trying to make compatible clones of the IBM PC.

indolering 6mo ago

The answer is unequivocally yes: RISC-V is designed to be customizable and a vendor can put whatever they like into a given CPU. That being said, profiles and platform specs are designed to limit fragmentation. The modular design and core essential ISA also makes fat binaries much more straight-forward to implement than other ISAs.

hajile 6mo ago
You can choose to develop proprietary extensions, but who’s going to use them? A great case study is the companies that implemented the pre-release vector standard in their chips. The final version is different in a few key ways. Despite substantial similarities to the ratified version, very few people are coding SIMD for those chips. If a proprietary extension does something actually useful to everyone, it’ll either be turned into an open standard or a new open standard will be created to replace it. In either case, it isn’t an issue. The only place I see proprietary extensions surviving is in the embedded space where they already do this kind of stuff, but even that seems to be the exception with the RISCV chips I’ve seen. Using standard compilers and tooling instead of a crappy custom toolchain (probably built on an old version of Eclipse) is just nicer (And cheaper for chip makers).
- LeFantome 6mo ago
  Yes, extensions are perfect for embedded. But not just there. Extensions allow you to address specific customer needs, evolve specific use cases, and experiment. AI is another perfect fit. And the hyperscaler market is another one where the hardware and software may come from the same party and be designed to work together. Compatibility with the standard is great for toolchains and off-the-shelf software but there is no need for a hyperscaler or AI specific extension to be implemented by anybody else. If something more universally useful is discovered by one party, it can be added to a future standard profile.

bsder 6mo ago

> Will RISC-V end up with the same (or even worse) platform fragmentation as ARM? Sadly, yes. RISC-V vendors are repeating literally every single mistake that the ARM ecosystem made and then making even dumber ones.

snvzz 6mo ago
Please elaborate.

pjmlp 6mo ago

Just like everything else outside PC thanks to clones becoming a thing. One reason UNIX became widely adopted, besides being freely available versus the other OSes, was that allowed companies to abstract their hardware differences, offering some market differentiation, while keeping some common ground. Those phones common ground is called Android with Java/Kotlin/C/C++ userspace, folks should stop seeing them as GNU/Linux.

tliltocatl 6mo ago

PC/x86 was an extreme outlier, sadly, and it was because of Microsoft/Intel business model. The architecture details was historically mostly decided on by Wintel, yet the system integration was done by many vendors, whose best interest was to stay as compatible as possible. Its unlikely that another platform would be able to reach this state, the PC architecturing was subsidized from the M$ software monopoly that nobody would have wanted to suffer thru again.

ajxs 6mo ago
> Its unlikely that another platform would be able to reach this state... Is this really true? The computer ecosystem is more open now than ever. The original PC BIOS (which PC-compatible manufacturers needed to implement) was never an open, documented standard. It was a proprietary, closed system made by IBM. It's pretty fair to say that IBM didn't anticipate a PC/x86 ecosystem developing around their product. They even sued companies who made their own compatible BIOSes (like Corona). Intel didn't really have much to do with the success of the product at that point in time either, much less Microsoft. In contrast, every widely-used modern system for hardware abstraction (UEFI/ACPI/DeviceTree/OpenSBI/etc) are open, royalty-free standards that anyone can use. Their implementation in ARM is newer, and inconsistent, but that's only because of how hugely diverse the ARM ecosystem is.
- tliltocatl 6mo ago
  The early PC era was a mess, and that's not the period I'm talking about. IBM was clearly not up to the task and Intel didn't care much yet, but Microsoft certainty did a lot for compatibility from the start (i. e. DOS abstracted away a lot of BIOS routines, so it would be easy to port MS-DOS to a non-IBM x86). But after IBM revealed MCA to show just exactly how much do they care about compatibility and platform openness, Intel realized they are missing out and cleaned up the MCA/EISA/VLB mess with PCI. Then Microsoft and Intel jointly released APM 1992 (which was clearly not enough), and then ACPI in 1996 (which is a total dumpster fire, but a sufficiently functional dumpster fire). I. e. ACPI and UEFI are exactly the product of the monopoly. M$/Intel profited from the abundance of cheapo white boxes, so it was in their best interest to come up with a standard even DELL can implement. The fact that AMD is going to implement the ACPI too wasn't much bother for Intel - they were so dominant that they could afford not to care. On the other hand, ARM sells the cores to SoC vendors (and doesn't care much what becomes of it), SoC vendors ducktape the ARM cores to a bunch of Synopsys peripherals and sell the resulting SoCs to smartphone and car makers (and doesn't care much for the product). System integrators throw Android on top and sell it to the customers. Then Google, who get all the cream via Play, hides all the mess behind a thousand layers of Java abstractions. DeviceTree is an offshot of Sun's OpenFirmware (and it leaves out all the hard stuff - OpenFirmware had Forth, DeviceTree expects the kernel to support every single brand of fan switch). OpenSBI is a disaster. I'm sorry, but what kind of bright mind came up with the idea of hiding damn *timer* behind a privilege switch? Timers were enough of a pain point on x86 already, then it settled on userspace-accesable RDTSC. RISC-V SBI? Reproducing x86 one stupid decision at a time.

LeFantome 6mo ago

> platform fragmentation RISC-V is addressing this issue quite directly. For things like desktops, laptops, SBCs and servers we have the RVA23 profile which defines quite specifically what features a chip must support to ensure code portability. On top of this, there are platform specifications. For example, the server spec is about to finalize next month. It extends RVA23 which things like UEFI, SBI, and ACPI to ensure that your can take something like a Linux distro and easily install it on any RISC-V server, like you can in the world of x86-64. > we have phones that are only good for landfill once their support lifetime is up RISC-V will probably not solve that problem in general. First, the ISA cannot really demand that your phone avoid a Broadcom wireless chip that requires proprietary firmware for example. Also, the phone vendor can still lock down the devices to prevent running arbitrary code. Thankfully, the RISC-V world is developing a culture of openness. If a company wants to create a fully “open” phone, they are quite likely to adopt RISC-V. And, because of RISC-V, even the SoC itself could be fully Open Source. But your typical Android phone is not going to get more Open just because they contain a RISC-V CPU.