English Amiga Board


Go Back   English Amiga Board > Support > support.Hardware

 
 
Thread Tools
Old 06 October 2015, 13:17   #1
anotherlife
Registered User

 
Join Date: Oct 2015
Location: Sweden
Posts: 19
Amiga 1200 Mediator Radeon mem problem! HELP!!

Hi! I'm new here and I hope this is the right place for my problem

Ok! My setup is Amiga 1200, Apollo 060/80Mhz, Mediator SX, Voodoo 3, Radeon 9250, Soundblaster and ethernet card.

I have tried everything to get the Radeon ram to work as "fast ram" (slow i know) I can only see 128 mb of 256 mb in the wb list. I guess the radeon have two mem banks. Then when I try to fill the ram It works without problem as it use the Apollo ram first but then when it enters the radeon ram the whole Amiga freeze.

I have all the latest drivers from elbox and I have tried only using the Radeon card. Same problem apears. I have in env: RadeonMem and RadeonMemOS (Yes)

I once got the message "Allocated memory doesen't match memorymask"

The radeon mem is between §80000000 - 87fffffff I have tried addmem without any luck..

My english is not that good so I hope you understand

Thanks /Andreas
anotherlife is offline  
Old 08 October 2015, 08:24   #2
matthey
Banned
 
Join Date: Jan 2010
Location: Kansas
Posts: 1,284
Quote:
Originally Posted by anotherlife View Post
I have tried everything to get the Radeon ram to work as "fast ram" (slow i know) I can only see 128 mb of 256 mb in the wb list. I guess the radeon have two mem banks. Then when I try to fill the ram It works without problem as it use the Apollo ram first but then when it enters the radeon ram the whole Amiga freeze.
I don't have a 1200 Mediator but Elbox's Mediator drivers have bugs. My Voodoo 4 in a 3000(T)/4000(T) Mediator only adds half the graphics memory to the system although I'm pretty sure that all of it is available as P96 memory. Some versions of the Elbox drivers have been completely broken although the latest versions seem to be better.

Quote:
Originally Posted by anotherlife View Post
I have all the latest drivers from elbox and I have tried only using the Radeon card. Same problem apears. I have in env: RadeonMem and RadeonMemOS (Yes)

I once got the message "Allocated memory doesen't match memorymask"

The radeon mem is between §80000000 - 87fffffff I have tried addmem without any luck..
You might make sure the MMU env variable is turned on. It is possible that your MMU tables are not setup correctly but I don't think you can use another 68060.library with the Apollo accelerator so there isn't much you can do. You could contact Elbox for technical support or maybe someone will come along to help who has a closer Mediator setup. Good luck.
matthey is offline  
Old 08 October 2015, 10:40   #3
anotherlife
Registered User

 
Join Date: Oct 2015
Location: Sweden
Posts: 19
Quote:
Originally Posted by matthey View Post
I don't have a 1200 Mediator but Elbox's Mediator drivers have bugs. My Voodoo 4 in a 3000(T)/4000(T) Mediator only adds half the graphics memory to the system although I'm pretty sure that all of it is available as P96 memory. Some versions of the Elbox drivers have been completely broken although the latest versions seem to be better.



You might make sure the MMU env variable is turned on. It is possible that your MMU tables are not setup correctly but I don't think you can use another 68060.library with the Apollo accelerator so there isn't much you can do. You could contact Elbox for technical support or maybe someone will come along to help who has a closer Mediator setup. Good luck.
Different 060 libs works with apollo. Maybe ther is more of them out there. I have mmu set to yes. Not working with NO. We have tried this with blizzard 030 card and the memory works. So I guess it have something to do with 060 "mmu"

You need to be reg at elbox.. Otherwise they won't answer :@
anotherlife is offline  
Old 08 October 2015, 11:56   #4
Toni Wilen
WinUAE developer
 
Join Date: Aug 2001
Location: Hämeenlinna/Finland
Age: 43
Posts: 22,023
Memory outside of first 2G is not really supported and not safe, some memory handling functions use bit 31 as an error bit and some programs may use signed variables when comparing addresses.

Memory mask error probably comes from filesystem (pfs3?), make sure all partitions have mask FFFFFFFF (it probably is 7FFFFFFF). This change won't cause any problems with PIO (non-DMA) HD controllers.

Attach also showconfig output.
Toni Wilen is online now  
Old 08 October 2015, 12:21   #5
anotherlife
Registered User

 
Join Date: Oct 2015
Location: Sweden
Posts: 19
So i can change it without any problem. The amiga will boot up as normal? ;-) how do i upload a pic?
anotherlife is offline  
Old 08 October 2015, 12:28   #6
anotherlife
Registered User

 
Join Date: Oct 2015
Location: Sweden
Posts: 19
Quote:
Originally Posted by anotherlife View Post
So i can change it without any problem. The amiga will boot up as normal? ;-) how do i upload a pic?
I tried to change the mask in hdtoolboox but it wont save my changes.
anotherlife is offline  
Old 08 October 2015, 12:33   #7
anotherlife
Registered User

 
Join Date: Oct 2015
Location: Sweden
Posts: 19
Quote:
Originally Posted by anotherlife View Post
I tried to change the mask in hdtoolboox but it wont save my changes.
Ah.. I had to press enter ;-)
anotherlife is offline  
Old 08 October 2015, 12:43   #8
Toni Wilen
WinUAE developer
 
Join Date: Aug 2001
Location: Hämeenlinna/Finland
Age: 43
Posts: 22,023
Quote:
Originally Posted by anotherlife View Post
So i can change it without any problem. The amiga will boot up as normal? ;-) how do i upload a pic?
Yes and yes. Unless something unexpected happens

No need to upload pics, you can redirect showconfig output to a text file (tools/showconfig >file.txt)
Toni Wilen is online now  
Old 08 October 2015, 12:56   #9
anotherlife
Registered User

 
Join Date: Oct 2015
Location: Sweden
Posts: 19
PROCESSOR: CPU 68060/68060fpu/68060mmu
CUSTOM CHIPS: AA PAL Alice (id=$0023), AA Lisa (id=$00F8)
VERS: Kickstart version 45.65535, Exec version 45.20, Disk version 45.5
RAM: Node type $A, Attributes $505 (FAST), at $1080420-$4FFFFFF (~63.5 meg)
Node type $A, Attributes $2005 (FAST), at $84D00000-$84EFFFFF (2.0 meg)
Node type $A, Attributes $2005 (FAST), at $88000000-$8FFFFFFF (128.0 meg)
Node type $A, Attributes $703 (CHIP), at $4000-$1FFFFF (~2.0 meg)
BOARDS:
Board + ROM (HD?) (unidentified): Prod=8738/35($2222/$23) (@$EA0000 128K)
Board (Index Information Ltd): Prod=2206/40($89E/$28) (@$EC0000 128K)
Board (Index Information Ltd): Prod=2206/168($89E/$A8) (@$200000 8meg)
anotherlife is offline  
Old 08 October 2015, 14:37   #10
thomas
Registered User
thomas's Avatar
 
Join Date: Jan 2002
Location: Germany
Posts: 5,717
Quote:
Originally Posted by anotherlife View Post
how do i upload a pic?

Click on "Go Advanced", then on "Manage Attachments".
thomas is offline  
Old 08 October 2015, 16:59   #11
anotherlife
Registered User

 
Join Date: Oct 2015
Location: Sweden
Posts: 19
Quote:
Originally Posted by thomas View Post
Click on "Go Advanced", then on "Manage Attachments".
Ok thanks. But i have already post my showconfig as text :-)
anotherlife is offline  
Old 08 October 2015, 20:10   #12
Toni Wilen
WinUAE developer
 
Join Date: Aug 2001
Location: Hämeenlinna/Finland
Age: 43
Posts: 22,023
Ok, there really is RAM at >=0x80000000 addresses. Unfortunately I can't help more until emulated PCI video device exists

What does Mediator pci info tool output? (I don't remember exact name..)
Toni Wilen is online now  
Old 08 October 2015, 21:41   #13
anotherlife
Registered User

 
Join Date: Oct 2015
Location: Sweden
Posts: 19
I forgot to say that I can add mem from the radeon card with blizzard 030. The problem start with 060. But with my Voodoo I can add mem to wb.

Here is my pciinfo:




6.DH0:> pciinfo

Pci Device 1:
VendorID: $10ec
VendorName: Realtek Semiconductor Corp.
DeviceID: $8139
DeviceName: RTL8139 RTL8139 Fast Ethernet NIC
ClassCode: $20000
RevisionID: $10
InterruptPin: 1
SubsystemVendorID: $10ec
SubsystemID: $8139
IOSpace0: $ed6000 - $ed60ff
MemSpace1: $80000000 - $800000ff

Pci Device 2:
VendorID: $1274
VendorName: Ensoniq
DeviceID: $5000
DeviceName: ES1371 AudioPCI
ClassCode: $40100
RevisionID: $1
InterruptPin: 2
SubsystemVendorID: $4942
SubsystemID: $4c4c
IOSpace0: $ed6400 - $ed643f

Pci Device 3:
VendorID: $1002
VendorName: Advanced Micro Devices, Inc.
DeviceID: $5960
DeviceName: RV280 Radeon 9200 Pro
ClassCode: $30000
RevisionID: $1
InterruptPin: 4
SubsystemVendorID: $1002
SubsystemID: $2002
MemSpace0: $88000000 - $8fffffff
IOSpace1: $ed6800 - $ed68ff
MemSpace2: $80010000 - $8001ffff
RomSpace: $80020000 - $8003ffff

Pci Device 4:
VendorID: $1002
VendorName: Advanced Micro Devices, Inc.
DeviceID: $5940
DeviceName: RV280 Radeon 9200 Pro Secondary
ClassCode: $38000
RevisionID: $1
InterruptPin: 0
SubsystemVendorID: $1002
SubsystemID: $2003
MemSpace0: $90000000 - $97ffffff
MemSpace1: $80040000 - $8004ffff

Pci Device 5:
VendorID: $121a
VendorName: 3dfx Interactive Inc
DeviceID: $5
DeviceName: Voodoo3 All Voodoo3 chips, 3000
ClassCode: $30000
RevisionID: $1
InterruptPin: 2
SubsystemVendorID: $121a
SubsystemID: $36
MemSpace0: $82000000 - $83ffffff
MemSpace1: $84000000 - $85ffffff
IOSpace2: $ed6c00 - $ed6cff
RomSpace: $80050000 - $8005ffff

Last edited by anotherlife; 08 October 2015 at 21:57.
anotherlife is offline  
Old 08 October 2015, 22:09   #14
Toni Wilen
WinUAE developer
 
Join Date: Aug 2001
Location: Hämeenlinna/Finland
Age: 43
Posts: 22,023
I checked emulation code and remembered that 1200 Mediators only have 4M/8M banked view to PCI side address space. 4000 Mediators have much larger window (256M or 512M).

I think VRAM as system RAM is only possible with 4000 Mediators. 1200 Mediator PCI address space window is far too small.
Toni Wilen is online now  
Old 08 October 2015, 22:16   #15
anotherlife
Registered User

 
Join Date: Oct 2015
Location: Sweden
Posts: 19
Quote:
Originally Posted by Toni Wilen View Post
I checked emulation code and remembered that 1200 Mediators only have 4M/8M banked view to PCI side address space. 4000 Mediators have much larger window (256M or 512M).

I think VRAM as system RAM is only possible with 4000 Mediators. 1200 Mediator PCI address space window is far too small.
According to elbox it shall work with the latest pci.library. And it works perfect with 030 card.
anotherlife is offline  
Old 08 October 2015, 22:18   #16
Toni Wilen
WinUAE developer
 
Join Date: Aug 2001
Location: Hämeenlinna/Finland
Age: 43
Posts: 22,023
Ah, ok, so it must be a MMU hack, sounds like DOS EMS memory
I guess it require exactly correct MMU libraries etc..
Toni Wilen is online now  
Old 08 October 2015, 22:21   #17
anotherlife
Registered User

 
Join Date: Oct 2015
Location: Sweden
Posts: 19
Quote:
Originally Posted by Toni Wilen View Post
Ah, ok, so it must be a MMU hack, sounds like DOS EMS memory
I guess it require exactly correct MMU libraries etc..
Ok. So i have to try different mmu libs?
anotherlife is offline  
Old 08 October 2015, 23:49   #18
matthey
Banned
 
Join Date: Jan 2010
Location: Kansas
Posts: 1,284
Quote:
Originally Posted by Toni Wilen View Post
Ok, there really is RAM at >=0x80000000 addresses. Unfortunately I can't help more until emulated PCI video device exists
Wow! I would have never suspected these upper/negative addresses to be used. It should be alright to use these addresses if the pci.library and all drivers are carefully written but these addresses would probably not be safe to add as AmigaOS system memory. My Mediator maps the PCI memory to a different address for system use.

Voodoo 4
Pci Device 1:
VendorID: $121a
VendorName: 3dfx Interactive Inc
DeviceID: $9
DeviceName: Voodoo5 AGP X2, or PCI. Chip: 2x VSA100 with 32Mo
ClassCode: $30000
RevisionID: $1
InterruptPin: 2
SubsystemVendorID: $121a
SubsystemID: $5
MemSpace0: $60000000 - $67ffffff
MemSpace1: $68000000 - $6fffffff
IOSpace2: $40c06000 - $40c060ff
RomSpace: $70000000 - $7000ffff

Only 16 MB of the 32 MB of Voodoo 4 memory gets added to the system at $7000000 - $7fffffff (mh_Lower: $7000020, mh_Upper: $8000000). The mh_Upper for 16MB is right at the limit of what is safe for system use and may have something to do with my missing 16MB of memory. My system seems stable although the mapping is much different that for anotherlife.

Quote:
Originally Posted by Toni Wilen View Post
Ah, ok, so it must be a MMU hack, sounds like DOS EMS memory
Maybe. The 1200 support could be complicated with a memory window and MMU mapping.

Quote:
Originally Posted by anotherlife View Post
Ok. So i have to try different mmu libs?
Changing the 68060.library would probably be the easiest possible solution. If ThoR's Mu 68060.library works then it could be helpful. MuScan output from the Mu package may give an indication of how the hardware is mapped with the MMU. A different 68060.library may setup the MMU tables differently and all of a sudden everything might work. The Phase 5 68060.library would be my 2nd choice to try.

Last edited by matthey; 08 October 2015 at 23:59.
matthey is offline  
Old 09 October 2015, 00:01   #19
anotherlife
Registered User

 
Join Date: Oct 2015
Location: Sweden
Posts: 19
Quote:
Originally Posted by matthey View Post
Wow! I would have never suspected these upper/negative addresses to be used. It should be alright to use these addresses if the pci.library and all drivers are carefully written but these addresses would probably not be safe to add as AmigaOS system memory. My Mediator maps the PCI memory to a different address for system use.

Voodoo 4
Pci Device 1:
VendorID: $121a
VendorName: 3dfx Interactive Inc
DeviceID: $9
DeviceName: Voodoo5 AGP X2, or PCI. Chip: 2x VSA100 with 32Mo
ClassCode: $30000
RevisionID: $1
InterruptPin: 2
SubsystemVendorID: $121a
SubsystemID: $5
MemSpace0: $60000000 - $67ffffff
MemSpace1: $68000000 - $6fffffff
IOSpace2: $40c06000 - $40c060ff
RomSpace: $70000000 - $7000ffff

Only 16 MB of the 32 MB of Voodoo 4 memory gets added to the system at $7000000 - $7fffffff (mh_Lower: $7000000, mh_Upper: $8000000). The mh_Upper for 16MB is right at the limit of what is safe for system use and may have something to do with my missing 16MB of memory. My system seems stable although the mapping is much different that for anotherlife.



Maybe. The 1200 support could be complicated with a memory window and MMU mapping.



Changing the 68060.library would probably be the easiest possible solution. If ThoR's Mu 68060.library works then it could be helpful. MuScan output from the Mu package may give an indication of how the hardware is mapped with the MMU. A different 68060.library may setup the MMU tables differently and all of a sudden everything might work. The Phase 5 68060.library would be my 2nd choice to try.

I'm using ThoR's libs. MuScan indicate OK. Don't know if Phase 5 works with apollo.. Oh I hate mmu right now..

Wonder why my system is mapping everything over 80000000..

Last edited by anotherlife; 09 October 2015 at 00:07.
anotherlife is offline  
Old 09 October 2015, 00:10   #20
matthey
Banned
 
Join Date: Jan 2010
Location: Kansas
Posts: 1,284
Quote:
Originally Posted by anotherlife View Post
I'm using ThoR's libs. MuScan indicate OK. Don't know if Phase 5 works with apollo.. Oh I hate mmu right now..
Can you post your MuScan output?
Mine looks like this:

Code:
MuScan 40.3 (28.11.99) © THOR

68060 MMU detected.
MMU page size is 0x1000 bytes.

Memory map:
0x00000000 - 0x00003FFF CopyBack Single Remapped to 0x087D4000 
0x00004000 - 0x001FFFFF CacheInhibit Imprecise NonSerial 
0x00200000 - 0x00BBFFFF Blank 
0x00BC0000 - 0x00BFFFFF CacheInhibit I/O space 
0x00C00000 - 0x00D7FFFF Blank 
0x00D80000 - 0x00DFFFFF CacheInhibit I/O space 
0x00E00000 - 0x00EFFFFF Blank 
0x00F00000 - 0x00F00FFF CacheInhibit 
0x00F01000 - 0x00F21FFF CacheInhibit I/O space 
0x00F22000 - 0x00F7FFFF CacheInhibit 
0x00F80000 - 0x00FFFFFF CopyBack 
0x01000000 - 0x06FFFFFF Blank 
0x07000000 - 0x0801CFFF CopyBack 
0x0801D000 - 0x0801FFFF CacheInhibit 
0x08020000 - 0x0853BFFF CopyBack 
0x0853C000 - 0x087CBFFF CacheInhibit Imprecise NonSerial 
0x087CC000 - 0x0DF7FFFF CopyBack 
0x0DF80000 - 0x0DFFFFFF CacheInhibit 
0x0E000000 - 0x3FFFFFFF Blank 
0x40000000 - 0x40FFFFFF CacheInhibit I/O space 
0x41000000 - 0x5FFFFFFF Blank 
0x60000000 - 0x7FFFFFFF CacheInhibit Imprecise I/O space 
0x80000000 - 0xFFFFFFFF Blank
It would be better if we had the output from someone with a 1200 Mediator, preferably with Radeon memory added to the system.

Also, where does Scout say the graphics board memory is added to the system? For me I run Scout, click "Memory" and look for the address of expansion memory with a lower priority than the accelerator memory (ln_Pri: 30 for me).

Last edited by matthey; 09 October 2015 at 00:15.
matthey is offline  
 


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools

Similar Threads
Thread Thread Starter Forum Replies Last Post
radeon.card 2.17 or later for Mediator JohnFante request.Apps 4 08 November 2013 11:37
Mediator/Radeon Question moxavius support.Hardware 5 31 January 2012 17:51
my 256mb +3.3V ATI Radeon in v1 1200 mediator johnim Hardware mods 5 01 May 2011 13:43
Radeon on Mediator 1200 PCI Megatron-UK support.Hardware 3 09 September 2010 05:52
Mediator TX & Radeon 9200 Sallinen support.Hardware 87 01 February 2010 20:50

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +2. The time now is 07:16.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2018, vBulletin Solutions Inc.
Page generated in 0.09006 seconds with 13 queries