Re: A3000 problems
Frank Wille
On 26.12.10 20:44:03 you wrote:
> My A1200 refuses to start again and temporarily A3000 is my test
> machine.
>
> Some observed problems:
>
> 1.Turbo board is not identified properly (as CyberStorm PPC). Again,
> amigappc_identify() is at fault - this CSPPC's serial number starts
> with 'DD'.
Yes, my CSPPC is also identified as "PowerUp". I don't know why Adam
Ciarcinski, who wrote the initial amigappc port code, decided that only 'E'
serial numbers are CSPPC, but I also didn't care much.
> 2. Kernel does not boot with ZorRAM RAM expansion installed.
Maybe because the kernel and/or the loader are loaded into Z3-RAM? I'm not
even sure if PPC code works reliably in any other than the CSPPC onboard
RAM.
What experience did you have with WarpOS or PowerUp?
> Do we support Zorro III RAM expansions at all in amigappc port?
When the ZorRAM memory was recognized by AmigaOS it will be in the memlist
passed by the bootloader. But currently initppc() will only recognize the
memory segment where the kernel was loaded (which is usually the onboard
RAM).
> 3.
> Kernel does not boot with Deneb USB card installed, even when flash ROM
> is disabled and Poseidon USB stack unloaded.
I have a Deneb in my A3000 too and there was never any problem with it.
Neither when booting amigappc nor amiga.
The card is just ignored by the kernel. Where exactly is the problem? Does
the bootloader fail?
> 4. Serial port does
> transmit only in one way (same symptoms as earlier). I have machdep.c
> rev 1.47 with recent Frank's changes.
Did it ever work for you with my last patch?
The symptoms I had were that the system freezes, as soon as the first
character is received. I tested my serial connection with tip(1) and a
proper entry in /etc/remote.
I didn't try to boot with a serial console yet.
--
Frank Wille
--
Posted automagically by a mail2news gateway at muc.de e.V.
Please direct questions, flames, donations, etc. to news-...@muc.de
Radosław Kujawa
On Dec 27, 2010, at 7:40 PM, Frank Wille wrote:
>> 1.Turbo board is not identified properly (as CyberStorm PPC). Again,
>> amigappc_identify() is at fault - this CSPPC's serial number starts
>> with 'DD'.
>
> Yes, my CSPPC is also identified as "PowerUp". I don't know why Adam
> Ciarcinski, who wrote the initial amigappc port code, decided that only 'E'
> serial numbers are CSPPC, but I also didn't care much.
Maybe there's better way to differentiate between these two than checking serial number? We could assume that we're dealing with CSPPC if it's 604 and BPPC if it's 603. But this solution is as hackish as current.
>> 2. Kernel does not boot with ZorRAM RAM expansion installed.
>
> Maybe because the kernel and/or the loader are loaded into Z3-RAM? I'm not
> even sure if PPC code works reliably in any other than the CSPPC onboard
> RAM.
Yup, that's the case - gobsdppc prints that kernel is loaded at 0x5000 0000. Loader should follow priority information from AmigaOS (CSPPC RAM is prioritized to be allocated first). Using firstseg option to gobsdppc is a workaround to this problem - kernel is loaded and executed successfully at 0x0800 0000 (and works as expected).
> What experience did you have with WarpOS or PowerUp?
I don't remember ever running out of CSPPC RAM (128MB) while working with OS 3.x. I bought ZorRAM before I acquired CSPPC, when I had only 4MB of onboard Fast RAM and A3640. Now it's mostly unused, because no AmigaOS 3.x program needs more than 128MB of RAM...
MorphOS does not support Z3 RAM at all. Current version of AmigaOS 4.0 does not work with Z3 RAM, but upcoming 4.1 is rumored to support it. If additional work was needed, then probably it's not straightforward.
>> Do we support Zorro III RAM expansions at all in amigappc port?
>
> When the ZorRAM memory was recognized by AmigaOS it will be in the memlist
> passed by the bootloader. But currently initppc() will only recognize the
> memory segment where the kernel was loaded (which is usually the onboard
> RAM).
Looks like something bad happens if kernel is loaded into Z3 RAM. Lack of BAT setup is the first problem here?
>> 3.
>> Kernel does not boot with Deneb USB card installed, even when flash ROM
>> is disabled and Poseidon USB stack unloaded.
>
> I have a Deneb in my A3000 too and there was never any problem with it.
> Neither when booting amigappc nor amiga.
Actually, today I removed CV64/3D from this A3000 and kernel booted with Deneb. No idea what difference it makes...
> The card is just ignored by the kernel. Where exactly is the problem? Does
> the bootloader fail?
Kernel is loaded into RAM, then usual "flying colors", then black screen. Console is not initialized, no output on serial port.
>> 4. Serial port does
>> transmit only in one way (same symptoms as earlier). I have machdep.c
>> rev 1.47 with recent Frank's changes.
>
> Did it ever work for you with my last patch?
Hmm, I'm not sure, but it worked fine on A1200 after applying your patch, I'll need to check again - I didn't test it extensively.
> The symptoms I had were that the system freezes, as soon as the first
> character is received. I tested my serial connection with tip(1) and a
> proper entry in /etc/remote.
I think that's not the case here, but I can't test it properly now, as I don't have a full userspace installed (and can't do it on A3000 due to unsupported FastATA controller, no compatible SCSI disk, etc. etc.), so... I need to revive my A1200 ;).
--
Best regards,
Radosław Kujawa
Frank Wille
>> Yes, my CSPPC is also identified as "PowerUp". I don't know why Adam
>> Ciarcinski, who wrote the initial amigappc port code, decided that
>> only 'E' serial numbers are CSPPC, but I also didn't care much.
>
> Maybe there's better way to differentiate between these two than
> checking serial number?
I don't even know where the difference between a "PowerUp" and a "CSPPC"
board is. AFAIR there were early developer boards. Were those just called
"PowerUp"?
> We could assume that we're dealing with CSPPC
> if it's 604 and BPPC if it's 603. But this solution is as hackish as
> current.
Better would be: it's a BPPC when it is installed in an A1200. CSPPC
otherwise.
Some developer boards had a 603e. Were A2000-BPPCs ever released?
> Yup, that's the case - gobsdppc prints that kernel is loaded at 0x5000
> 0000. Loader should follow priority information from AmigaOS (CSPPC RAM
> is prioritized to be allocated first). Using firstseg option to
> gobsdppc is a workaround to this problem - kernel is loaded and
> executed successfully at 0x0800 0000 (and works as expected).
The FIRSTSEG/S option is misleading. In reality it is implemented as "use
segment with highest priority" instead of "use biggest segment".
>>> Do we support Zorro III RAM expansions at all in amigappc port?
>> [...]
> Looks like something bad happens if kernel is loaded into Z3 RAM. Lack
> of BAT setup is the first problem here?
The kernel will be automatically relocated for the Z3 RAM, like a BPPC
kernel. The only problem could be that I'm always using a 128MB BAT for
CSPPC, while your Z3 RAM is 256MB.
>>> Kernel does not boot with Deneb USB card installed, even when flash
>>> ROM is disabled and Poseidon USB stack unloaded.
>> [...]
> Actually, today I removed CV64/3D from this A3000 and kernel booted
> with Deneb. No idea what difference it makes...
Maybe too much memory space used by CV64/3D + ZorRAM + Deneb? For CSPPC a
256MB BAT will be reserved at 0x40000000 for Z3 cards. But when the CV64/3D
is already using all the 256MB, the Deneb will not be BAT-mapped. Without
the CV64/3D it returns into the mapped 0x40000000-space.
--
Frank Wille
Radosław Kujawa
> Radoslaw Kujawa wrote:
>
>>> Yes, my CSPPC is also identified as "PowerUp". I don't know why Adam
>>> Ciarcinski, who wrote the initial amigappc port code, decided that
>>> only 'E' serial numbers are CSPPC, but I also didn't care much.
>>
>> Maybe there's better way to differentiate between these two than
>> checking serial number?
>
> I don't even know where the difference between a "PowerUp" and a "CSPPC"
> board is. AFAIR there were early developer boards. Were those just called
> "PowerUp"?
Yes, dev boards for CyberStorm Mk-II were called "PowerUP": http://bboah.com/download_photos/ppcproto_5_big.jpg
>
>> We could assume that we're dealing with CSPPC
>> if it's 604 and BPPC if it's 603. But this solution is as hackish as
>> current.
>
> Better would be: it's a BPPC when it is installed in an A1200. CSPPC
> otherwise.
> Some developer boards had a 603e. Were A2000-BPPCs ever released?
Some guy stuck a dev board onto Blizzard 2060: http://www.youtube.com/watch?v=foWfXhmi1wY . So it's possible to have a PPC in A2000 ;). Dunno if that behaves more like an BPPC or like CSPPC. But I never heard of official PPC card for A2000.
>
>> Yup, that's the case - gobsdppc prints that kernel is loaded at 0x5000
>> 0000. Loader should follow priority information from AmigaOS (CSPPC RAM
>> is prioritized to be allocated first). Using firstseg option to
>> gobsdppc is a workaround to this problem - kernel is loaded and
>> executed successfully at 0x0800 0000 (and works as expected).
>
> The FIRSTSEG/S option is misleading. In reality it is implemented as "use
> segment with highest priority" instead of "use biggest segment".
Fine then, but shouldn't it be the default in amigappc port? You'd always want to load kernel into CSPPC/BPPC RAM, which always have highest priority.
>
>>>> Do we support Zorro III RAM expansions at all in amigappc port?
>>> [...]
>> Looks like something bad happens if kernel is loaded into Z3 RAM. Lack
>> of BAT setup is the first problem here?
>
> The kernel will be automatically relocated for the Z3 RAM, like a BPPC
> kernel. The only problem could be that I'm always using a 128MB BAT for
> CSPPC, while your Z3 RAM is 256MB.
Ah, that's not really true - my ZorRAM is a 128MB model. Though, for some reason it allocates 256MB of Z3 space. ZorRAM is currently produced in two versions - 128MB and 256MB. I suspect that CPLD core is same for both versions. What happens if actual size of RAM is different than allocated Z3 space size? It should not be a problem, because memlist contains valid values passed from AmigaOS?
>>>> Kernel does not boot with Deneb USB card installed, even when flash
>>>> ROM is disabled and Poseidon USB stack unloaded.
>>> [...]
>> Actually, today I removed CV64/3D from this A3000 and kernel booted
>> with Deneb. No idea what difference it makes...
>
> Maybe too much memory space used by CV64/3D + ZorRAM + Deneb? For CSPPC a
> 256MB BAT will be reserved at 0x40000000 for Z3 cards. But when the CV64/3D
> is already using all the 256MB, the Deneb will not be BAT-mapped. Without
> the CV64/3D it returns into the mapped 0x40000000-space.
Makes sense. ShowConfig looks like this:
PROCESSOR: CPU 68040/68040fpu
CUSTOM CHIPS: ECS PAL Agnus (id=$0020), ECS Denise (id=$00FC)
VERS: Kickstart version 40.68, Exec version 40.10, Disk version 40.42
RAM: Node type $A, Attributes $505 (FAST), at $8000000-$FEFFFFF (127.0 meg)
Node type $A, Attributes $505 (FAST), at $7C00000-$7FFFFFF (4.0 meg)
Node type $A, Attributes $405 (FAST), at $50000000-$57FFFFFF (128.0 meg)
Node type $A, Attributes $703 (CHIP), at $4000-$1FFFFF (~2.0 meg)
BOARDS:
Board (unidentified): Prod=3643/16($E3B/$10)
(@$40000000, size 16meg, subsize same)
Board (unidentified): Prod=3643/19($E3B/$13) (@$E90000 64K)
RAM (unidentified): Prod=3643/32($E3B/$20)
(@$50000000, size 256meg, subsize autosized Mem)
Board (unidentified): Prod=8512/67($2140/$43)
(@$60000000, size 256meg, subsize same)
Board + ROM (HD?) (unidentified): Prod=2206/62($89E/$3E)
(@$70000000, size 16meg, subsize same)
Board + ROM (HD?) (unidentified): Prod=8512/100($2140/$64) (@$F01060 128K)
First two boards is Deneb, then ZorRAM, CV64/3D, FastATA 4000 and CSPPC. Is it possible to BAT-map some additional space to support configuration like this? I see that BAT_BL_256M is max possible block size for older PPC CPUs...
--
Best regards,
Radosław Kujawa
Frank Wille
> Yes, dev boards for CyberStorm Mk-II were called "PowerUP":
> http://bboah.com/download_photos/ppcproto_5_big.jpg
Ok. I wonder how much of those exist.
>> Some developer boards had a 603e. Were A2000-BPPCs ever released?
>
> Some guy stuck a dev board onto Blizzard 2060:
> http://www.youtube.com/watch?v=foWfXhmi1wY . So it's possible to have a
> PPC in A2000 ;). Dunno if that behaves more like an BPPC or like CSPPC.
> But I never heard of official PPC card for A2000.
I guess we can ignore that. :)
I'm open for suggestions how to detect CSPPC, BPPC (and maybe even PowerUp)
cards reliably.
>> The FIRSTSEG/S option is misleading. In reality it is implemented as
>> "use segment with highest priority" instead of "use biggest segment".
>
> Fine then, but shouldn't it be the default in amigappc port? You'd
> always want to load kernel into CSPPC/BPPC RAM, which always have
> highest priority.
Yes, that's true. I changed the default behaviour now, removed FIRSTSEG/S
and introduced LARGESTSEG/S instead.
The updated gobsdppc can be found under:
ftp://ftp.netbsd.org/pub/NetBSD/misc/phx/amigappc/
I can also upload the full gobsdppc source somewhere, when anybody is
interested. But it may only be compilable with vbcc at the moment.
>> The kernel will be automatically relocated for the Z3 RAM, like a BPPC
>> kernel. The only problem could be that I'm always using a 128MB BAT
>> for CSPPC, while your Z3 RAM is 256MB.
>
> Ah, that's not really true - my ZorRAM is a 128MB model. Though, for
> some reason it allocates 256MB of Z3 space.
Probably just to make the next board aligned to a 256MB boundary. Even the
Deneb wastes 256MB in your configuration.
> ZorRAM is currently
> produced in two versions - 128MB and 256MB. I suspect that CPLD core is
> same for both versions. What happens if actual size of RAM is different
> than allocated Z3 space size? It should not be a problem, because
> memlist contains valid values passed from AmigaOS?
Yes. We are relying on the memlist.
I just tried to run WarpOS and PowerUp programs in Chip-RAM, which didn't
work because Chip-RAM is not mapped as an executable segment. So I'm still
unsure what happens when trying to execute PPC code outside of the onboard
RAM.
So I have no explanation for what happened to you ATM.
> Makes sense. ShowConfig looks like this:
> [...]
I have never seen so much allocated Z3 space before!
> First two boards is Deneb, then ZorRAM, CV64/3D, FastATA 4000 and
> CSPPC. Is it possible to BAT-map some additional space to support
> configuration like this? I see that BAT_BL_256M is max possible block
> size for older PPC CPUs...
Yes. And we have just four IBAT and four DBAT registers. But the kernel
supports more than four data BATs in the battable. In the dsitrap function
of powerpc/trap_subr.S you can see that it will randomly use the DBAT2 and
DBAT3 to load an entry from battable to fix the current dsitrap.
So, yes, we could add more 256MB BATs for your needs. Maybe you want to try
it yourself in initppc() - look for amigappc_batinit().
--
Frank Wille
Radosław Kujawa
On Jan 1, 2011, at 1:05 PM, Frank Wille wrote:
> Radoslaw Kujawa wrote:
>
>> Yes, dev boards for CyberStorm Mk-II were called "PowerUP":
>> http://bboah.com/download_photos/ppcproto_5_big.jpg
>
> Ok. I wonder how much of those exist.
powerup.amigaworld.de estimates that 19 cards were made.
> I'm open for suggestions how to detect CSPPC, BPPC (and maybe even PowerUp)
> cards reliably.
First question is: what do we need this information for?
- We should use it to set up BAT mappings, not base it on Amiga model like we do it now (it's possible to put CSPPC into A1200 if you have Micronik Zorro 3 busboard, which has CPU Fast Slot).
- We can use it to inform the user, but that's not critical.
- Anything else?
My suggestions:
- Check for Zorro product IDs of BPPC/CSPPC in gobsdppc, pass this information to kernel.
- We can search for BlizzardPPC ROM tag. BPPC flash lives somewhere within ROM extension area (starting 0xf00000). This method should be 100% accurate, as every BPPC has this flash ROM and will not work without one.
- Rely on serial numbers, like we do it now. At least we can be sure that numbers staring with H or I are BPPC. We could assume that other numbers are handled like CSPPC.
>> Fine then, but shouldn't it be the default in amigappc port? You'd
>> always want to load kernel into CSPPC/BPPC RAM, which always have
>> highest priority.
>
> Yes, that's true. I changed the default behaviour now, removed FIRSTSEG/S
> and introduced LARGESTSEG/S instead.
Great, it works!
>
> I can also upload the full gobsdppc source somewhere, when anybody is
> interested. But it may only be compilable with vbcc at the moment.
I am interested, as I stated some days ago when I asked for gobsdppc source ;).
Why not commit it to sys/arch/amigappc/stand/ ?
> I just tried to run WarpOS and PowerUp programs in Chip-RAM, which didn't
> work because Chip-RAM is not mapped as an executable segment. So I'm still
> unsure what happens when trying to execute PPC code outside of the onboard
> RAM.
I did some testing today - filled RAM: with large files and exhaused CSPPC and mainboard memory, which forced newly started programs to use Z3 RAM.
PowerUP seems to handle this situation well, I ran some PPC programs this way. I was unable to fill CSPPC RAM completely, so ppc.library managed to fit into CSPPC RAM. Started PPC tasks were loaded and executed in Z3 RAM (I checked this with SysInfo, tasks were indeed running somewhere within 0x5000 0000).
WarpOS did not like Z3 RAM - no messages, no exceptions, just hung when I tried to execute a WOS program.
>
>> First two boards is Deneb, then ZorRAM, CV64/3D, FastATA 4000 and
>> CSPPC. Is it possible to BAT-map some additional space to support
>> configuration like this? I see that BAT_BL_256M is max possible block
>> size for older PPC CPUs...
>
> Yes. And we have just four IBAT and four DBAT registers. But the kernel
> supports more than four data BATs in the battable. In the dsitrap function
> of powerpc/trap_subr.S you can see that it will randomly use the DBAT2 and
> DBAT3 to load an entry from battable to fix the current dsitrap.
So number of BATs is limited only by battable size (512) ?
> So, yes, we could add more 256MB BATs for your needs. Maybe you want to try
> it yourself in initppc() - look for amigappc_batinit().
I changed initppc() like this:
amigappc_batinit(0x00000000, BAT_BL_16M, BAT_I|BAT_G,
(startkernel & 0xf8000000), BAT_BL_128M, 0,
0xfff00000, BAT_BL_512K, 0,
0x40000000, BAT_BL_256M, BAT_I|BAT_G, 0,
0x50000000, BAT_BL_256M, BAT_I|BAT_G, 0,
0x60000000, BAT_BL_256M, BAT_I|BAT_G, 0,
0x70000000, BAT_BL_256M, BAT_I|BAT_G,
~0);
Still, kernel did not boot with CV64/3D inserted, but works if I remove it. Should I modify anything else?
Any idea how to debug this before console is initialized?
Ignatios Souvatzis
> First question is: what do we need this information for?
> - We can use it to inform the user, but that's not critical.
That's why I was wondering about the fuss, but your other point:
> - We should use it to set up BAT mappings, not base it on Amiga
> model like we do it now (it's possible to put CSPPC into A1200 if
> you have Micronik Zorro 3 busboard, which has CPU Fast Slot).
seems certainly valid.
> - Anything else?
>
> My suggestions:
> - Check for Zorro product IDs of BPPC/CSPPC in gobsdppc, pass
> this information to kernel.
The kernel does have the zorro list (although it doesn't parse
it early enough for BAT setup). Let me rephrase: all NetBSD/Amiga boot
loaders are required to pass a cooked AutoConfig list to the kernel.
A quick search of the list could be done; it is unlikely to be longer
than, say, 8 entries for real cards. Some cards might have more than
one Zorro card on them, and cpu cards define some, but the number is
small enough to search in a loop.
-is
Frank Wille
>> I can also upload the full gobsdppc source somewhere, when anybody is
>> interested. But it may only be compilable with vbcc at the moment.
>
> I am interested, as I stated some days ago when I asked for gobsdppc
> source ;).
The current source archive can be downloaded from:
http://sun.hasenbraten.de/~frank/NetBSD/gobsdppc.lha
> Why not commit it to sys/arch/amigappc/stand/ ?
Because it cannot be compiled under NetBSD. And it is just a temporary
solution. When amigappc ever reaches a usable state we should use/enhance
the port-amiga bootblock code.
> PowerUP seems to handle this situation well, I ran some PPC programs
> this way. I was unable to fill CSPPC RAM completely, so ppc.library
> managed to fit into CSPPC RAM. Started PPC tasks were loaded and
> executed in Z3 RAM (I checked this with SysInfo, tasks were indeed
> running somewhere within 0x5000 0000).
Ok, so it would theoretically work. But I think that it is practically
useless to run PPC code in memory which is so slow as Z3 RAM.
> So number of BATs is limited only by battable size (512) ?
Data-BATs, yes. ISI traps are not handled by replacing BATs. So the kernel
and maybe other firmware code it needs, should be covered by IBAT0 and
IBAT1.
> amigappc_batinit(0x00000000, BAT_BL_16M, BAT_I|BAT_G,
> (startkernel & 0xf8000000), BAT_BL_128M, 0,
> 0xfff00000, BAT_BL_512K, 0,
> 0x40000000, BAT_BL_256M, BAT_I|BAT_G, 0,
> 0x50000000, BAT_BL_256M, BAT_I|BAT_G, 0,
> 0x60000000, BAT_BL_256M, BAT_I|BAT_G, 0,
> 0x70000000, BAT_BL_256M, BAT_I|BAT_G,
> ~0);
>
> Still, kernel did not boot with CV64/3D inserted, but works if I remove
> it. Should I modify anything else?
This should have worked. I have no more spontaneous ideas. You have to debug
it.
Did you try to swap position with the Deneb, so the CV64/3D appears at
0x40000000?
> Any idea how to debug this before console is initialized?
Yes, write data into Chip-RAM (e.g. 0x1f0000), which you can inspect after a
reboot.
--
Frank Wille
Frank Wille
>> - We should use it to set up BAT mappings, not base it on Amiga
>> model like we do it now (it's possible to put CSPPC into A1200 if
>> you have Micronik Zorro 3 busboard, which has CPU Fast Slot).
>
> seems certainly valid.
>
>> - Anything else?
>>
>> My suggestions:
>> - Check for Zorro product IDs of BPPC/CSPPC in gobsdppc, pass
>> this information to kernel.
>
> The kernel does have the zorro list (although it doesn't parse
> it early enough for BAT setup).
True. set_boot_args() is called before initppc(). So we can use the cfdev
list to find a Cyberstorm or Blizzard entry.
AFAIK the rom.manid for all is 8512, the rom.prodid is 100 for CSPPC, 110
for BPPC (?) and probably 25 for PowerUp (because it is mounted on a
CSMk2)?
--
Frank Wille
Ignatios Souvatzis
> AFAIK the rom.manid for all is 8512, the rom.prodid is 100 for CSPPC, 110
> for BPPC (?) and probably 25 for PowerUp (because it is mounted on a
> CSMk2)?
The former two: yes, the latter one I don't know.
Radosław Kujawa
On Jan 2, 2011, at 4:38 PM, Frank Wille wrote:
> Because it cannot be compiled under NetBSD. And it is just a temporary
> solution. When amigappc ever reaches a usable state we should use/enhance
> the port-amiga bootblock code.
Certainly, but without it we can't even boot now. Besides, it's useful to be able to boot from AmigaOS (think about installation procedure).
Can loadbsd (which resides in sys/arch/amiga/stand...) be compiled under NetBSD now? I think not ;).
> Ok, so it would theoretically work. But I think that it is practically
> useless to run PPC code in memory which is so slow as Z3 RAM.
That's true. It might be useful for swap space (provided someone will write something similar to Linux's z2ram block device driver).
>> amigappc_batinit(0x00000000, BAT_BL_16M, BAT_I|BAT_G,
>> (startkernel & 0xf8000000), BAT_BL_128M, 0,
>> 0xfff00000, BAT_BL_512K, 0,
>> 0x40000000, BAT_BL_256M, BAT_I|BAT_G, 0,
>> 0x50000000, BAT_BL_256M, BAT_I|BAT_G, 0,
>> 0x60000000, BAT_BL_256M, BAT_I|BAT_G, 0,
>> 0x70000000, BAT_BL_256M, BAT_I|BAT_G,
>> ~0);
>>
> This should have worked. I have no more spontaneous ideas. You have to debug
> it.
>
> Did you try to swap position with the Deneb, so the CV64/3D appears at
> 0x40000000?
I removed it, and tried different positions of the cards. I tried examining addresses from 0x4000 0000 - 0x5000 0000 range in DDB and it works ok. But if I try to read from 0x5000 0000 - 0x8000 0000 I'll always get DSI read trap. If I move the same cards below 0x5000 0000 then examining their space works fine. I think that's the source of all problems I've been experiencing here. If I understand correctly, DSI read trap happen when kernel is trying to access unavailable address.
On Jan 2, 2011, at 5:14 PM, Ignatios Souvatzis wrote:
> On Sun, Jan 02, 2011 at 04:44:36PM +0100, Frank Wille wrote:
>
>> AFAIK the rom.manid for all is 8512, the rom.prodid is 100 for CSPPC, 110
>> for BPPC (?) and probably 25 for PowerUp (because it is mounted on a
>> CSMk2)?
>
> The former two: yes, the latter one I don't know.
Ok, so this is on my todo-list, right after busclock stuff.
--
Best regards,
Radosław Kujawa