1/1 
|
problem: OR1200 on Altera board
by Unknown on Dec 28, 2003 |
Not available! | ||
|
Hi all
I'm running OR1200 on an Altera Stratix board, the data address from
OR1200 (dwb_adr_o) is incorrect.
The hardware include OR1200 (no IC, no DC, no IMMU, no DMMU, no DU,
no PM), UART, GPIO on-chip RAM and WISHBONE Conbus. following is
the program:
uart-nocache.or32: file format elf32-or32
00000100 :
100: 15 00 00 00 l.nop 0x0
104: 15 00 00 00 l.nop 0x0
108: 9c 40 00 00 l.addi r2,r0,0x0
10c: 9c 60 00 00 l.addi r3,r0,0x0
... ...
17c: 9f e0 00 00 l.addi r31,r0,0x0
180: 18 60 00 00 l.movhi r3,0x0
184: a8 63 23 a8 l.ori r3,r3,0x23a8
188: 44 00 18 00 l.jr r3
18c: 15 00 00 00 l.nop 0x0
... ...
00002324 :
2324: 9c 21 ff ec l.addi r1,r1,0xffffffec
2328: d4 01 48 00 l.sw 0x0(r1),r9
232c: d4 01 50 04 l.sw 0x4(r1),r10
2330: d4 01 60 08 l.sw 0x8(r1),r12
2334: d4 01 70 0c l.sw 0xc(r1),r14
2338: d4 01 80 10 l.sw 0x10(r1),r16
233c: 07 ff ff 36 l.jal 2014
... ...
000023a8 :
23a8: 18 20 00 00 l.movhi r1,0x0
23ac: a8 21 10 28 l.ori r1,r1,0x1028
23b0: 18 40 00 00 l.movhi r2,0x0
23b4: a8 42 23 24 l.ori r2,r2,0x2324
23b8: 44 00 10 00 l.jr r2
23bc: 15 00 00 00 l.nop 0x0
I use an oscillosope to observe the signal. When first store instruction
execute (address 2328), dwb_adr_o is 'ffffffd8', but I think it should
be '00001014'.
Is there anything wrong in the program or the hardware?
Thanks,
Roger Zhou
022021130@fudan.edu.cn>
|
|||
|
problem: OR1200 on Altera board
by Unknown on Dec 29, 2003 |
Not available! | ||
|
Hi Roger,
hard to say where the problem could be. I can tell you that the same
configuration of OR1200 (except on Xilinx Virtex) works. I'd suggest you
make sure when you look at the dwb_adr_o you take into account also the
dwb_cyc_o and dwb_stb_o. They have to be both active !! If they are not you
need to ignore dwb_adr_o. Also you can try a configuration with DC or/and IC
enabled, maybe it changes something...
I'd strongly advise that you run RTL simulation first and only later go to
real HW. You will save a lot of time this way.
regards,
Damjan
----- Original Message -----
From: "Roger Zhou" 022021130@fudan.edu.cn>
To: openrisc@opencores.org>
Sent: Sunday, December 28, 2003 12:04 PM
Subject: [openrisc] problem: OR1200 on Altera board
Hi all
I'm running OR1200 on an Altera Stratix board, the data address from
OR1200 (dwb_adr_o) is incorrect.
The hardware include OR1200 (no IC, no DC, no IMMU, no DMMU, no DU,
no PM), UART, GPIO on-chip RAM and WISHBONE Conbus. following is
the program:
uart-nocache.or32: file format elf32-or32
00000100 :
100: 15 00 00 00 l.nop 0x0
104: 15 00 00 00 l.nop 0x0
108: 9c 40 00 00 l.addi r2,r0,0x0
10c: 9c 60 00 00 l.addi r3,r0,0x0
... ...
17c: 9f e0 00 00 l.addi r31,r0,0x0
180: 18 60 00 00 l.movhi r3,0x0
184: a8 63 23 a8 l.ori r3,r3,0x23a8
188: 44 00 18 00 l.jr r3
18c: 15 00 00 00 l.nop 0x0
... ...
00002324 :
2324: 9c 21 ff ec l.addi r1,r1,0xffffffec
2328: d4 01 48 00 l.sw 0x0(r1),r9
232c: d4 01 50 04 l.sw 0x4(r1),r10
2330: d4 01 60 08 l.sw 0x8(r1),r12
2334: d4 01 70 0c l.sw 0xc(r1),r14
2338: d4 01 80 10 l.sw 0x10(r1),r16
233c: 07 ff ff 36 l.jal 2014
... ...
000023a8 :
23a8: 18 20 00 00 l.movhi r1,0x0
23ac: a8 21 10 28 l.ori r1,r1,0x1028
23b0: 18 40 00 00 l.movhi r2,0x0
23b4: a8 42 23 24 l.ori r2,r2,0x2324
23b8: 44 00 10 00 l.jr r2
23bc: 15 00 00 00 l.nop 0x0
I use an oscillosope to observe the signal. When first store instruction
execute (address 2328), dwb_adr_o is 'ffffffd8', but I think it should
be '00001014'.
Is there anything wrong in the program or the hardware?
Thanks,
Roger Zhou
022021130@fudan.edu.cn>
_______________________________________________
http://www.opencores.org/mailman/listinfo/openrisc
|
|||
1/1