AtomBIOS
  • Order of the Butterfly
    Order of the Butterfly
    sailor
    Posts: 358 from 2019/5/9
    From: Central Bohemi...
    Does somebody know exactly, which are first Radeon cards with AtomBIOS?

    I did't find exact list or chip number, from which AtomBIOS is used.
    It is somewhere between R400 and R500.
    R500:
    Best graphics... Best video ... Peg2 died AGP cards...
    R400:
    More Radeon ... AMD video ... whats he building in there....

    It is not only academic question, I want select PCIe cards ( RV370 and newer ) but without AtomBIOS to test it with Efika + PCI-PCIe bridge.
    AmigaOS3: Amiga 1200
    AmigaOS4: Micro A1-C, AmigaOne XE, Pegasos II, Sam440ep, Sam440ep-flex, AmigaOneX1000
    MorphOS: Efika 5200b, Pegasos I, Pegasos II, Powerbook G4, Mac Mini, iMac G5, Powermac G5 Quad
  • »18.01.21 - 19:18
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12085 from 2003/5/22
    From: Germany
    > Does somebody know exactly, which are first Radeon cards with
    > AtomBIOS? I did't find exact list or chip number, from which
    > AtomBIOS is used. It is somewhere between R400 and R500.
    > R500: [...]

    Indeed, early AtomBIOS versions came somewhen during R400 series, but I also couldn't find out from which GPU/card on exactly. The reason I've been mentioning AtomBIOS to start only with R500 series here on MorphZone is that, as far as I understand, the AtomBIOS parser contained in MorphOS requires the newer AtomBIOS versions that started with the R500 series. I've now added a corresponding explanation to those comments to prevent future misleading.
  • »19.01.21 - 00:44
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    sailor
    Posts: 358 from 2019/5/9
    From: Central Bohemi...
    Please, are there somebody with linux and graphics card with chips:
    RV370: X600 variants (X300, X550, X1050 not in MOS compatible)
    RV380: X600 PRO/XT
    RV410: (X550, X700, X1050 not in MOS compatible)
    R423: X800 variants
    R430: X800, X800GT
    R480: X800 variants, X850 variants
    It can be any PC with Intel or AMD, not only PPC, with opensource driver (amdgpu, radeon).

    Can I ask you if you can try the "dmesg | grep ATOM" command? To check if there are ATOM BIOS or not. And please post here the result.

    Many thanks.
    AmigaOS3: Amiga 1200
    AmigaOS4: Micro A1-C, AmigaOne XE, Pegasos II, Sam440ep, Sam440ep-flex, AmigaOneX1000
    MorphOS: Efika 5200b, Pegasos I, Pegasos II, Powerbook G4, Mac Mini, iMac G5, Powermac G5 Quad
  • »19.01.21 - 07:36
    Profile
  • Priest of the Order of the Butterfly
    Priest of the Order of the Butterfly
    ernsteiswuerfel
    Posts: 545 from 2015/6/18
    From: Funeralopolis
    I have a X300 lying around and I think (?) a X600. I can check them on the weekend. AFAIK the X300 did not show up with AtomBIOS in the logs... But well, I schould rather check it.
    Talos II. [Gentoo Linux] | PMac G5 11,2. PMac G4 3,6. PBook G4 5,8. [MorphOS 3.18 / Gentoo Linux] | Vampire V4 SA [ApolloOS / Amiga OS 3.2.2]
  • »19.01.21 - 10:18
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    sailor
    Posts: 358 from 2019/5/9
    From: Central Bohemi...
    After searching in TechPowerUp BIOS collection I found, that all Rx4xx cards which are on database are with ATOMBios:

    RV410: X550 ATOMBios, X700 ATOMBios, (X1050-not in database )
    R423: X800 ATOMBios
    R430: X800 ATOMBios, X800GT ATOMBios (not all variants in database )
    R480: X800 ATOMBios, X850 ATOMBios (not all variants in database )

    For simplification we can expect, that all Rx4xx Radeon chips are with ATOMBios.

    For testing remains:
    RV370: X600 variants (X300, X550, X1050 not in MOS compatible)
    RV380: X600 PRO/XT

    Many thanks, ernsteiswuerfel!

    [ Edited by sailor 19.01.2021 - 14:15 ]
    AmigaOS3: Amiga 1200
    AmigaOS4: Micro A1-C, AmigaOne XE, Pegasos II, Sam440ep, Sam440ep-flex, AmigaOneX1000
    MorphOS: Efika 5200b, Pegasos I, Pegasos II, Powerbook G4, Mac Mini, iMac G5, Powermac G5 Quad
  • »19.01.21 - 14:12
    Profile
  • MorphOS Developer
    cyfm
    Posts: 537 from 2003/4/11
    From: Germany
    I'm not exactly sure what you are aiming at but even if a Radeon card has a so called AtomBIOS, it still has all the legacy BIOS functionality required by a X86 BIOS emulator.

    Also, even if R4xx cards have AtomBIOS support, it isn't supported by the current MorphOS driver anyway. It requires at least a R5xx chip equipped card to be supported.

    There are no R(V)3xx equipped cards that use any kind of AtomBIOS afaict.
  • »19.01.21 - 14:38
    Profile Visit Website
  • Order of the Butterfly
    Order of the Butterfly
    sailor
    Posts: 358 from 2019/5/9
    From: Central Bohemi...
    Quote:

    cyfm wrote:
    I'm not exactly sure what you are aiming at but even if a Radeon card has a so called AtomBIOS, it still has all the legacy BIOS functionality required by a X86 BIOS emulator.

    Also, even if R4xx cards have AtomBIOS support, it isn't supported by the current MorphOS driver anyway. It requires at least a R5xx chip equipped card to be supported.

    There are no R(V)3xx equipped cards that use any kind of AtomBIOS afaict.


    The aim is to test on Efika more "modern" cards then Radeon 9800.
    Here some results from testing.

    In short, MOS supported HD4350 cards doesn't work in Efika, and more, after a minute serial debug returns mesh of chars and efika don't boot to MOS.
    These card differs from 9xxx/R300 cards with three things:
    radeon chip: but HD4350 PCI works with Pegasos2 MOS
    PCI-PCIe bridge: PCI versions have PEX8112, PCI-e version with PEX8111. But X1900 G5 (PPC bios) with PEX8111 boots with Efika - no screen, card not recognized, but bridge chip is the same as I tested with HD4350, so bridge is probably no reason of Efika stuck.
    AtomBIOS: Radeon 9xxx have "classic" x86 bios and OF of Efika (and Pegasos) have emulator for it. And cards newer then R500 (and all R400 as we found) all have Atom bios. I don't find exactly with which chip ATOMBios was started.

    So it looks that main reason why Efika don't boot is ATOMBios. Yes, it is only my theory.
    Then I ask about rest of chips greater then Radeon 9xxx - i.e. now only RV370 and RV380 if they have Atombios or not. I am not sure, I don't find it on ATI related pages.

    If they cannot have ATOMBios, it makes sense to test them with Efika, as it probably most "modern" cards, which can be usable.
    AmigaOS3: Amiga 1200
    AmigaOS4: Micro A1-C, AmigaOne XE, Pegasos II, Sam440ep, Sam440ep-flex, AmigaOneX1000
    MorphOS: Efika 5200b, Pegasos I, Pegasos II, Powerbook G4, Mac Mini, iMac G5, Powermac G5 Quad
  • »19.01.21 - 17:37
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12085 from 2003/5/22
    From: Germany
    >> even if a Radeon card has a so called AtomBIOS, it still has all
    >> the legacy BIOS functionality required by a X86 BIOS emulator.

    > Radeon 9xxx have "classic" x86 bios and OF of Efika (and Pegasos)
    > have emulator for it. And cards newer then R500 (and all R400 as
    > we found) all have Atom bios. [...] So it looks that main reason why
    > Efika don't boot is ATOMBios. Yes, it is only my theory.

    I can see how this theory somehow contrasts with what cyfm just wrote ;-) But please go ahead, then we'll know for sure.
  • »19.01.21 - 20:52
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    sailor
    Posts: 358 from 2019/5/9
    From: Central Bohemi...
    Quote:

    Andreas_Wolf wrote:
    I can see how this theory somehow contrasts with what cyfm just wrote ;-) But please go ahead, then we'll know for sure.

    Many theories cotrast with something. They are theories ;-)

    From cyfm there is usable info "There are no R(V)3xx equipped cards that use any kind of AtomBIOS afaict." Unfortunatelly there is no link to some official source, but we can take id as a premise. I thought so too.

    Second info "Radeon card has a so called AtomBIOS, it still has all the legacy BIOS functionality required by a X86 BIOS emulator" is also true, but all functionality <> 100% code compatible.

    Without deeper knowledge of Efika firmware (or reverse engeneering of it) I have no other easy solution then search for card differencies. And ATOM bios is one the significant. And up to now fits in theory. ;-)

    Partial (up today) conclusion is, that make sense to test only one other card: Radeon X600 XT. No Atombios, with bridge, newer then R350.
    Unfortunatelly I have no one in my proximity, so I have to wait if some Efika user has, or wait for one.
    I also suppose, that this card cannot be faster then 9800PRO, but if it will be very close, it is better solution for me: 9800PRO 47W, X600XT 25W, Efika <= 3W.
    AmigaOS3: Amiga 1200
    AmigaOS4: Micro A1-C, AmigaOne XE, Pegasos II, Sam440ep, Sam440ep-flex, AmigaOneX1000
    MorphOS: Efika 5200b, Pegasos I, Pegasos II, Powerbook G4, Mac Mini, iMac G5, Powermac G5 Quad
  • »20.01.21 - 07:15
    Profile
  • Paladin of the Pegasos
    Paladin of the Pegasos
    Amigaharry2
    Posts: 1240 from 2010/1/6
    From: EU-Austria (Wien)
    The only person who knows about internas of EFIKA or Pegasos OpenFirmware, very well, is Nicolas Det from BPLAN. He has written PEG/EFIKA-relevant parts of OF and has also done some low-level drivers for MOS. I think it would be a good idea to ask him about that issue.
    Peg2, 3xPowerMac G5, 2xPowerbookG4, 2x MacMiniG4, Efika (again), A3000T and life is never boring.....
  • »20.01.21 - 08:14
    Profile
  • MorphOS Developer
    cyfm
    Posts: 537 from 2003/4/11
    From: Germany
    Quote:

    Amigaharry2 schrieb:
    The only person who knows about internas of EFIKA or Pegasos OpenFirmware, very well,


    So much for more theories but feel free to go ahead, whatever floats your boat ... :)
    Nicolas Det definitely was involved in the development of EFIKA's OF but the actual BIOS emulator used in both Efika & Pegasos has originally been written by Gerald Carda@bplan
  • »20.01.21 - 13:55
    Profile Visit Website
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12085 from 2003/5/22
    From: Germany
    > 9800PRO 47W, X600XT 25W, Efika <= 3W

    Don't forget the bridge card ;-)
  • »20.01.21 - 16:12
    Profile
  • Priest of the Order of the Butterfly
    Priest of the Order of the Butterfly
    ernsteiswuerfel
    Posts: 545 from 2015/6/18
    From: Funeralopolis
    @sailor: Checked the X300 I got lying around. It says:

    Code:
     $ lspci -v -s 02:00.0
    02:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] RV370 [Radeon X600/X600 SE] (prog-if 00 [VGA controller])
    Subsystem: Advanced Micro Devices, Inc. [AMD/ATI] RV370 [Radeon X600/X600 SE]
    Flags: bus master, fast devsel, latency 0, IRQ 40, NUMA node 0
    Memory at d8000000 (64-bit, prefetchable) [size=128M]
    Memory at fbde0000 (64-bit, non-prefetchable) [size=64K]
    I/O ports at c000 [size=256]
    Expansion ROM at 000c0000 [disabled] [size=128K]
    Capabilities: <access denied>
    Kernel driver in use: radeon
    Kernel modules: radeon

    And its' kernel dmesg:
    Code:
    kernel: [drm] radeon kernel modesetting enabled.
    kernel: radeon 0000:02:00.0: remove_conflicting_pci_framebuffers: bar 0: 0xd8000000 -> 0xdfffffff
    kernel: radeon 0000:02:00.0: remove_conflicting_pci_framebuffers: bar 2: 0xfbde0000 -> 0xfbdeffff
    kernel: [drm] initializing kernel modesetting (RV380 0x1002:0x5B62 0x1002:0x0F02 0x00).
    kernel: resource sanity check: requesting [mem 0x000c0000-0x000dffff], which spans more than PCI Bus 0000:00 [mem 0x000d0000-0x000dffff window]
    kernel: caller 0xffffffff9b3b0a04 mapping multiple BARs
    kernel: [drm] Generation 2 PCI interface, using max accessible memory
    kernel: radeon 0000:02:00.0: VRAM: 128M 0x00000000D8000000 - 0x00000000DFFFFFFF (128M used)
    kernel: radeon 0000:02:00.0: GTT: 512M 0x00000000B8000000 - 0x00000000D7FFFFFF
    kernel: [drm] Detected VRAM RAM=128M, BAR=128M
    kernel: [drm] RAM width 64bits DDR
    kernel: [TTM] Zone kernel: Available graphics memory: 16498482 KiB
    kernel: [TTM] Zone dma32: Available graphics memory: 2097152 KiB
    kernel: [TTM] Initializing pool allocator
    kernel: [drm] radeon kernel modesetting enabled.
    kernel: radeon 0000:02:00.0: remove_conflicting_pci_framebuffers: bar 0: 0xd8000000 -> 0xdfffffff
    kernel: radeon 0000:02:00.0: remove_conflicting_pci_framebuffers: bar 2: 0xfbde0000 -> 0xfbdeffff
    kernel: [drm] initializing kernel modesetting (RV380 0x1002:0x5B62 0x1002:0x0F02 0x00).
    kernel: resource sanity check: requesting [mem 0x000c0000-0x000dffff], which spans more than PCI Bus 0000:00 [mem 0x000d0000-0x000dffff window]
    kernel: caller 0xffffffff9b3b0a04 mapping multiple BARs
    kernel: [drm] Generation 2 PCI interface, using max accessible memory
    kernel: radeon 0000:02:00.0: VRAM: 128M 0x00000000D8000000 - 0x00000000DFFFFFFF (128M used)
    kernel: radeon 0000:02:00.0: GTT: 512M 0x00000000B8000000 - 0x00000000D7FFFFFF
    kernel: [drm] Detected VRAM RAM=128M, BAR=128M
    kernel: [drm] RAM width 64bits DDR
    kernel: [TTM] Zone kernel: Available graphics memory: 16498482 KiB
    kernel: [TTM] Zone dma32: Available graphics memory: 2097152 KiB
    kernel: [TTM] Initializing pool allocator
    kernel: [TTM] Initializing DMA pool allocator
    kernel: [drm] radeon: 128M of VRAM memory ready
    kernel: [drm] radeon: 512M of GTT memory ready.
    kernel: [drm] GART: num cpu pages 131072, num gpu pages 131072
    kernel: [drm] radeon: 1 quad pipes, 1 Z pipes initialized
    kernel: [drm] PCIE GART of 512M enabled (table at 0x00000000D8040000).
    kernel: radeon 0000:02:00.0: WB enabled
    kernel: radeon 0000:02:00.0: fence driver on ring 0 use gpu addr 0x00000000b8000000 and cpu addr 0x(____ptrval____)
    kernel: [drm] Supports vblank timestamp caching Rev 2 (21.10.2013).
    kernel: [drm] Driver supports precise vblank timestamp query.
    kernel: radeon 0000:02:00.0: radeon: MSI limited to 32-bit
    kernel: radeon 0000:02:00.0: radeon: using MSI.
    kernel: [drm] radeon: irq initialized.
    kernel: [drm] Loading R300 Microcode
    kernel: [drm] radeon: ring at 0x00000000B8001000
    kernel: [drm] ring test succeeded in 1 usecs
    kernel: [drm] ib test succeeded in 0 usecs
    kernel: [drm] Radeon Display Connectors
    kernel: [drm] Connector 0:
    kernel: [drm] DVI-I-1
    kernel: [drm] HPD1
    kernel: [drm] DDC: 0x60 0x60 0x60 0x60 0x60 0x60 0x60 0x60
    kernel: [drm] Encoders:
    kernel: [drm] CRT1: INTERNAL_DAC1
    kernel: [drm] DFP1: INTERNAL_TMDS1
    kernel: [drm] Connector 1:
    kernel: [drm] SVIDEO-1
    kernel: [drm] Encoders:
    kernel: [drm] TV1: INTERNAL_DAC2
    kernel: [drm] Cannot find any crtc or sizes
    kernel: [drm] Initialized radeon 2.50.0 20080528 for 0000:02:00.0 on minor 0
    Talos II. [Gentoo Linux] | PMac G5 11,2. PMac G4 3,6. PBook G4 5,8. [MorphOS 3.18 / Gentoo Linux] | Vampire V4 SA [ApolloOS / Amiga OS 3.2.2]
  • »27.01.21 - 12:26
    Profile
  • Order of the Butterfly
    Order of the Butterfly
    sailor
    Posts: 358 from 2019/5/9
    From: Central Bohemi...
    @ernsteiswuerfel:

    Many thanks! So X300 (RV380) have no ATOMBios.
    Cards with it have output like this:
    Code:
    # dmesg | grep ATOM
    [ 0.752011] ATOM BIOS: TURKS
    [ 2.434756] ATOM BIOS: Tahiti


    So I mean, that we can close this subject with this result:
    All Radeon chips upto R3xx don't have ATOMBios, and
    all Radeon chips from R4xx have ATOMBios.

    Many thanks for testing!
    AmigaOS3: Amiga 1200
    AmigaOS4: Micro A1-C, AmigaOne XE, Pegasos II, Sam440ep, Sam440ep-flex, AmigaOneX1000
    MorphOS: Efika 5200b, Pegasos I, Pegasos II, Powerbook G4, Mac Mini, iMac G5, Powermac G5 Quad
  • »27.01.21 - 17:41
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12085 from 2003/5/22
    From: Germany
    > X300 (RV380)

    RV370 (see also comment #3) :-)
  • »28.01.21 - 11:19
    Profile
  • Paladin of the Pegasos
    Paladin of the Pegasos
    Amigaharry2
    Posts: 1240 from 2010/1/6
    From: EU-Austria (Wien)
    Quote:

    cyfm schrieb:
    .......
    Nicolas Det definitely was involved in the development of EFIKA's OF but the actual BIOS emulator used in both Efika & Pegasos has originally been written by Gerald Carda@bplan



    Thanks for clarifycation. Whenever I needed anything about OF in the past from BPLAN, always Nicholas Det was refered to me. So I thought he was the "OF-man".
    Peg2, 3xPowerMac G5, 2xPowerbookG4, 2x MacMiniG4, Efika (again), A3000T and life is never boring.....
  • »28.01.21 - 11:34
    Profile