Skip to contentSkip to conversation

SystemACE driver problem

DISCUSSION aauer1 3 participants 7 messages Oct 19, 2007 – Oct 22, 2007

lore
Share & tools
lore

aauer1

author
#0 lore raw

Hello

I'm trying to boot a Linux Kernel 2.6.23-rc2 from Grant Likely (thanks for the great work) on a Xilinx ML403 board. I want to use the second partition of the Compact Flash Card as root filesystem. But the boot process hangs at the initialization of the SystemACE module. A strange effect is that the DONE LED (configuration of the FPGA is done) turns off. So, this means that the FPGA loses his configuration.
Has anybody recognized a similar problem with the SystemACE module??

The boot log:
========

[    0.000000] Linux version 2.6.23-rc2 (bbauer@soclin) (gcc version 4.0.0
(DEN7
[    0.000000] Xilinx ML403 Reference System (Virtex-4 FX)
[    0.000000] Zone PFN ranges:
[    0.000000]   DMA             0 ->    16384
[    0.000000]   Normal      16384 ->    16384
[    0.000000] Movable zone start PFN for each node
[    0.000000] early_node_map[1] active PFN ranges
[    0.000000]     0:        0 ->    16384
[    0.000000] Built 1 zonelists in Zone order.  Total pages: 16256
[    0.000000] Kernel command line: console=ttyUL0 root=/dev/xsa2 rw
[    0.000000] Xilinx INTC #0 at 0x41200000 mapped to 0xFDFFF000
[    0.000000] PID hash table entries: 256 (order: 8, 1024 bytes)
[    0.000328] Console: colour dummy device 80x25
[    0.001390] Dentry cache hash table entries: 8192 (order: 3, 32768 bytes)
[    0.002934] Inode-cache hash table entries: 4096 (order: 2, 16384 bytes)
[    0.029759] Memory: 63244k available (1176k kernel code, 388k data, 84k
init)
[    0.120774] Mount-cache hash table entries: 512
[    0.171986] io scheduler noop registered
[    0.172248] io scheduler anticipatory registered (default)
[    0.172300] io scheduler deadline registered
[    0.172508] io scheduler cfq registered
[    0.268807] uartlite.0: ttyUL0 at MMIO 0x40600000 (irq = 2) is a uartlite
[    0.270399] console [ttyUL0] enabled
[    1.762725] RAMDISK driver initialized: 8 RAM disks of 8192K size 1024
blocke
[    1.852062] Registering Xilinx SystemACE driver, major=254
[    1.919734] xsysace xsysace.0: ace_probe(c01638e0)

Thanks,
Andreas

Signature · 2 lines
View this message in context: http://www.nabble.com/SystemACE-driver-problem-tf4651846.html#a13290652
Sent from the linuxppc-embedded mailing list archive at Nabble.com.
#1 lore raw
Quoted 34 lines · On 10/19/07, aauer1 <[email protected]> wrote:

Hello

I'm trying to boot a Linux Kernel 2.6.23-rc2 from Grant Likely (thanks for the great work) on a Xilinx ML403 board. I want to use the second partition of the Compact Flash Card as root filesystem. But the boot process hangs at the initialization of the SystemACE module. A strange effect is that the DONE LED (configuration of the FPGA is done) turns off. So, this means that the FPGA loses his configuration.
Has anybody recognized a similar problem with the SystemACE module??

The boot log:
========

[    0.000000] Linux version 2.6.23-rc2 (bbauer@soclin) (gcc version 4.0.0
(DEN7
[    0.000000] Xilinx ML403 Reference System (Virtex-4 FX)
[    0.000000] Zone PFN ranges:
[    0.000000]   DMA             0 ->    16384
[    0.000000]   Normal      16384 ->    16384
[    0.000000] Movable zone start PFN for each node
[    0.000000] early_node_map[1] active PFN ranges
[    0.000000]     0:        0 ->    16384
[    0.000000] Built 1 zonelists in Zone order.  Total pages: 16256
[    0.000000] Kernel command line: console=ttyUL0 root=/dev/xsa2 rw
[    0.000000] Xilinx INTC #0 at 0x41200000 mapped to 0xFDFFF000
[    0.000000] PID hash table entries: 256 (order: 8, 1024 bytes)
[    0.000328] Console: colour dummy device 80x25
[    0.001390] Dentry cache hash table entries: 8192 (order: 3, 32768 bytes)
[    0.002934] Inode-cache hash table entries: 4096 (order: 2, 16384 bytes)
[    0.029759] Memory: 63244k available (1176k kernel code, 388k data, 84k
init)
[    0.120774] Mount-cache hash table entries: 512
[    0.171986] io scheduler noop registered
[    0.172248] io scheduler anticipatory registered (default)
Quoted 8 lines
[    0.172300] io scheduler deadline registered
[    0.172508] io scheduler cfq registered
[    0.268807] uartlite.0: ttyUL0 at MMIO 0x40600000 (irq = 2) is a uartlite
[    0.270399] console [ttyUL0] enabled
[    1.762725] RAMDISK driver initialized: 8 RAM disks of 8192K size 1024
blocke
[    1.852062] Registering Xilinx SystemACE driver, major=254
[    1.919734] xsysace xsysace.0: ace_probe(c01638e0)

Yes, I've seen that. It means something went wrong with the setup and caused the sysace to reload the FPGA. Is the systemace base address and bus width set correctly?

Cheers,
g.

Signature · 4 lines
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
[email protected]
(403) 399-0195
author
#2 lore raw
Quoted 28 lines · Grant Likely-2 wrote:
Quoted 17 lines · On 10/19/07, aauer1 <[email protected]> wrote:

Hello

I'm trying to boot a Linux Kernel 2.6.23-rc2 from Grant Likely (thanks for
the great work) on a Xilinx ML403 board. I want to use the second partition
of the Compact Flash Card as root filesystem. But the boot process hangs at
the initialization of the SystemACE module. A strange effect is that the DONE LED (configuration of the FPGA is done) turns off. So, this means that
the FPGA loses his configuration.
Has anybody recognized a similar problem with the SystemACE module??

The boot log:
========

[snipped]

Quoted 7 lines
[    0.268807] uartlite.0: ttyUL0 at MMIO 0x40600000 (irq = 2) is a
uartlite
[    0.270399] console [ttyUL0] enabled
[    1.762725] RAMDISK driver initialized: 8 RAM disks of 8192K size 1024
blocke
[    1.852062] Registering Xilinx SystemACE driver, major=254
[    1.919734] xsysace xsysace.0: ace_probe(c01638e0)
Quoted 5 lines

Yes, I've seen that. It means something went wrong with the setup and caused the sysace to reload the FPGA. Is the systemace base address and bus width set correctly?

The base address is the same as in the system.mhs file of my hardware design. So, I think it should be correct. The bus width is set to 16 bits (hardware design). So far as I can see, the sysace driver defines a 16 bit data bus (hard coded), too.

A paper of BlueCat Linux says that there is a problem with the System ACE hardware module:
"Due to a known problem of the opb_sysace IP Core programmed into the FPGA (incorrect operation in the interrupt mode) the SystemACE Compact Flash device is supported in the polling mode."

Does anyone know something about the bug in the SystemACE IP core??

Thanks,
Andreas

Signature · 2 lines
View this message in context: http://www.nabble.com/SystemACE-driver-problem-tf4651846.html#a13300089
Sent from the linuxppc-embedded mailing list archive at Nabble.com.
#3 lore raw
Quoted 38 lines · On 10/19/07, aauer1 <[email protected]> wrote:
Quoted 28 lines · Grant Likely-2 wrote:
Quoted 17 lines · On 10/19/07, aauer1 <[email protected]> wrote:

Hello

I'm trying to boot a Linux Kernel 2.6.23-rc2 from Grant Likely (thanks for
the great work) on a Xilinx ML403 board. I want to use the second partition
of the Compact Flash Card as root filesystem. But the boot process hangs at
the initialization of the SystemACE module. A strange effect is that the DONE LED (configuration of the FPGA is done) turns off. So, this means that
the FPGA loses his configuration.
Has anybody recognized a similar problem with the SystemACE module??

The boot log:
========

[snipped]

Quoted 7 lines
[    0.268807] uartlite.0: ttyUL0 at MMIO 0x40600000 (irq = 2) is a
uartlite
[    0.270399] console [ttyUL0] enabled
[    1.762725] RAMDISK driver initialized: 8 RAM disks of 8192K size 1024
blocke
[    1.852062] Registering Xilinx SystemACE driver, major=254
[    1.919734] xsysace xsysace.0: ace_probe(c01638e0)
Quoted 5 lines

Yes, I've seen that. It means something went wrong with the setup and caused the sysace to reload the FPGA. Is the systemace base address and bus width set correctly?

Quoted 4 lines

The base address is the same as in the system.mhs file of my hardware design. So, I think it should be correct. The bus width is set to 16 bits (hardware design). So far as I can see, the sysace driver defines a 16 bit data bus (hard coded), too.

Hmmm; that's all very odd. I don't know what could be causing that.

Quoted 7 lines

A paper of BlueCat Linux says that there is a problem with the System ACE hardware module:
"Due to a known problem of the opb_sysace IP Core programmed into the FPGA (incorrect operation in the interrupt mode) the SystemACE Compact Flash device is supported in the polling mode."

Does anyone know something about the bug in the SystemACE IP core??

I don't know about the old xilinx supplied sysace driver, but the new xsysace.c driver is robust enough to work around missed interrupts. I don't know of any other problems with the opb_sysace IP core.

Cheers,
g.

Signature · 4 lines
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
[email protected]
(403) 399-0195
#4 lore raw

Grant

Regarding your comment about the new xsysace.c driver vs. the old Xilinx one, is that chosen at the kernel config time ?

i.e.
cp arch/ppc/configs/ml403_defconfig
make ARCH=ppc CROSS_COMPILE=/opt/eldk4.1/usr/bin/ppc_4xx- oldconfig

...
Xilinx SystemACE support (XILINX_SYSACE) [N/m/y/?] (NEW) y Xilinx SystemACE support (old driver) (XILINX_SYSACE_OLD) [N/m/y/?] (NEW) N

I am also about build and test with the 2.6.23-rc2 same source. Just wanted to be sure I am choosing the right driver for the ml403 reference design bits from Xilinx.

Thanks

John Hahn
BCF Semiconductor
---

P.S. Sorry for taking the thread off on a slight tangent ...

Quoted 3 lines

From: Grant Likely
Sent: Friday, October 19, 2007 11:21 AM

... <snip>

Quoted 17 lines

I don't know about the old xilinx supplied sysace driver, but the new xsysace.c driver is robust enough to work around missed interrupts. I don't know of any other problems with the opb_sysace IP core.

Cheers,
g.

--
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
[email protected]
(403) 399-0195
_______________________________________________
Linuxppc-embedded mailing list
[email protected]
https://ozlabs.org/mailman/listinfo/linuxppc-embedded

#5 lore raw
Quoted 9 lines · John Hahn

Grant

Regarding your comment about the new xsysace.c driver vs. the old Xilinx one, is that chosen at the kernel config time ?

i.e.
cp arch/ppc/configs/ml403_defconfig
make ARCH=ppc CROSS_COMPILE=/opt/eldk4.1/usr/bin/ppc_4xx- oldconfig

Quoted 7 lines

...
Xilinx SystemACE support (XILINX_SYSACE) [N/m/y/?] (NEW) y Xilinx SystemACE support (old driver) (XILINX_SYSACE_OLD) [N/m/y/?] (NEW) N

I am also about build and test with the 2.6.23-rc2 same source. Just wanted to be sure I am choosing the right driver for the ml403 reference design bits from Xilinx.

Both drivers work. The new driver is faster, but the old driver is more featurefull. If one isn't working for you, try the other and report your experiences with both here to the list.

Cheers,
g.

Signature · 4 lines
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
[email protected]
(403) 399-0195
author
#6 lore raw
In reply to Grant Likely

Hello,

I have partly solved my problem with the System ACE controller. And I think it may be interesting for. I don't know the real cause of the problem, because it is very difficult to reproduce it. But I think the problem is anywhere in the initialization of the System ACE controller.

So, I started a new design with the Xilinx platform studio and wrote my own code to read the first sector of the CF card. This code is not very beautiful, but it works. So, I configured the FPGA and downloaded my program. And wow... I get the first sector of the card. Afterwards, I reseted the System ACE controller and downloaded the zImage.elf from my kernel. Now, the kernel can initialize the CF card and mount the root filesystem. Wonderful!

If the problem occurs again, I will have a closer look.

For the sake of completeness, my init code:
================================
XStatus status;
XSysAce SysAce;
int i=0;
volatile short *ptr = (short *)0x41800000; // BASE Address of SYSACE volatile char *cptr = (char *) 0x41800000; // BASE Address of SYSACE

  print("\r\nTesting SysAce for SysACE_CompactFlash...\r\n");
  status = XSysAce_Initialize(&SysAce, XPAR_SYSACE_COMPACTFLASH_DEVICE_ID);
  if(status != XST_SUCCESS)
  {
    print("Init... FAILED\r\n");
  }
  else 
  {
    print("Init... PASSED\r\n");
    printf("Base Address: %x\r\n",SysAce.BaseAddress);
    printf("Is Ready: %x\r\n",SysAce.IsReady);
    printf("sizeof(short): %d\r\n", sizeof(short));
    printf("Status0 value: %x\r\n",ptr[2]);
    printf("Status1 value: %x\r\n",ptr[3]);
    printf("Error0 value: %x\r\n",ptr[4]);
    printf("Error1 value: %x\r\n",ptr[5]);
    printf("Version value: %x\r\n",ptr[0xb]);
    printf("Ctrl0 value: %x\r\n",cptr[0x18]);
    printf("Ctrl0 value: %x\r\n",cptr[0x18]);
    ptr[0xc] = 0x0300;
    printf("Status0 value: %x\r\n",ptr[2]);
    while(!(ptr[2] & 0x0200));
    print("Lock granted\r\n");
    while(!(ptr[2] & 0x0001));
    print("Ready for command\r\n");
			
    ptr[8] = 0x0000; // LBA
    ptr[9] = 0x0000; // LBA
    ptr[10] = 0x0103;
			
    ptr[0xc] = 0x8000;
    while(!(ptr[3] & 0x0200));
    print("Busy cleared\r\n");
    printf("Status0 value: %x\r\n",ptr[2]);
    while(!(ptr[2] & 0x2000));
    print("Data Buffer is ready\r\n");
			
    for(i=0;i<256;i++)
    {
      printf("0x%04x\r\n", ptr[0x20]&0xffff);
    }
			
    ptr[0xc] = 0x0100;
    printf("Status0 value: %x\r\n",ptr[2]);
    printf("Ctrl0 value: %x\r\n",ptr[0xc]);
    printf("Ctrl1 value: %x\r\n",ptr[0xd]);

Cheers,
Andreas

Quoted 78 lines · Grant Likely-2 wrote:
Quoted 41 lines · On 10/19/07, aauer1 <[email protected]> wrote:
Quoted 10 lines · Grant Likely-2 wrote:
Quoted 8 lines · On 10/19/07, aauer1 <[email protected]> wrote:

Hello

I'm trying to boot a Linux Kernel 2.6.23-rc2 from Grant Likely (thanks for
the great work) on a Xilinx ML403 board. I want to use the second partition
of the Compact Flash Card as root filesystem. But the boot process

hangs

Quoted 2 lines
Quoted 2 lines

at
the initialization of the SystemACE module. A strange effect is that

the

Quoted 5 lines
Quoted 5 lines

DONE LED (configuration of the FPGA is done) turns off. So, this means that
the FPGA loses his configuration.
Has anybody recognized a similar problem with the SystemACE module??

Quoted 7 lines
Quoted 2 lines

The boot log:
========

[snipped]

Quoted 4 lines
[    0.268807] uartlite.0: ttyUL0 at MMIO 0x40600000 (irq = 2) is a
uartlite
[    0.270399] console [ttyUL0] enabled
[    1.762725] RAMDISK driver initialized: 8 RAM disks of 8192K size

1024

Quoted 9 lines
Quoted 3 lines

blocke

[    1.852062] Registering Xilinx SystemACE driver, major=254
[    1.919734] xsysace xsysace.0: ace_probe(c01638e0)

Yes, I've seen that. It means something went wrong with the setup and caused the sysace to reload the FPGA. Is the systemace base address and bus width set correctly?

Quoted 5 lines

The base address is the same as in the system.mhs file of my hardware design. So, I think it should be correct. The bus width is set to 16 bits (hardware design). So far as I can see, the sysace driver defines a 16 bit
data bus (hard coded), too.

Hmmm; that's all very odd. I don't know what could be causing that.

Quoted 8 lines

A paper of BlueCat Linux says that there is a problem with the System ACE hardware module:
"Due to a known problem of the opb_sysace IP Core programmed into the FPGA
(incorrect operation in the interrupt mode) the SystemACE Compact Flash device is supported in the polling mode."

Does anyone know something about the bug in the SystemACE IP core??

I don't know about the old xilinx supplied sysace driver, but the new xsysace.c driver is robust enough to work around missed interrupts. I don't know of any other problems with the opb_sysace IP core.

Cheers,
g.

Signature · 8 lines
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.
[email protected]
(403) 399-0195
_______________________________________________
Linuxppc-embedded mailing list
[email protected]
https://ozlabs.org/mailman/listinfo/linuxppc-embedded
Signature · 2 lines
View this message in context: http://www.nabble.com/SystemACE-driver-problem-tf4651846.html#a13342284
Sent from the linuxppc-embedded mailing list archive at Nabble.com.