linux-2.6 system ACE driver - need help

Tue Sep 26 04:47:12 EST 2006

Hi Ameet,

  Thanks so much for your reply.

  ISSUE Details :

  We are currently debugging System driver in 2.6.16-2 kernel in interrupt mode, for AMCC440SPe based custom board.

  We got the kernel and applied patch for Xilinx System ACE driver.

  We are using powerpc-eabi-  toolachain to compile the kernel and NOT EDK.

  We have done changes in the driver in regard to –

   System ACE Controller register configuration - BASE address is 64-bit and hence done ioremap64(SYSTEMACE_BASEADRESS).- adapter.c  
   Endianess (Xsysace_l.c)  
   IRQ - xsysace_intr.c; arch/ppc/platforms/4xx/xparameters/xparameters_ml300.h ;  arch/ppc/syslib/ppc4xx_pic.c;  
   Kernel configuration to support FAT file system  
    Format the Compact Flash with FAT16 filesystem using ‘mkdosfs’, since  System ACE supports only FAT16

  With all these changes, we are able to communicate with the system ACE controller, register the interrupt, create block device using the driver, format and partition the Compact Flash.

  But we see issues while mounting the compact flash – Meaning, mounting shows inconsistent results.

  When we are able to mount the CF successfully, we are able to read/write to the compact flash. And when it fails, it throws a kernel Oops sometimes or complains that it’s a ‘bogus’ filesystem. Please see the screen-dump for success and failure cases:

  We are looking for inputs to attack this issue. Please advice.

  Note : we have worked on System ACE driver on 2.4 kernel on a different platform and were successful.

  Thanks very much
  Junita

  SCREEN DUMP:

  CASE 1 : Success:

  BusyBox v(null) ((null)) Built-in shell (msh)
  Enter 'help' for a list of built-in commands.

  # ### Application running ...

  #
  # cat /proc/interrupts
             CPU0
   33:          0  UIC1 Level     MAL SERR
   34:          0  UIC1 Level     MAL TX DE
   35:          0  UIC1 Level     MAL RX DE
   38:          0  UIC1 Level     MAL TX EOB
   39:          0  UIC1 Level     MAL RX EOB
   89:          0  UIC2 Level     System ACE
  BAD:          0

  # mknod /dev/xsysace b 254 0

  # mkdir /mnt

  # mkdir /mnt/sysace

  # mount -t msdos /dev/xsysace /mnt/sysace/

  # ls
  bin      etc      home     linuxrc  proc     sbin     usr
  dev      ftp      lib      mnt      root     tmp      var

  # ls /mnt/sysace/
  rev0.ace
  # cp /sbin/ifconfig /mnt/sysace/

  # ls /mnt/sysace/
  ifconfig  rev0.ace
  #

  CASE : 2 – Failure – Kernel Oops.

   # mount -t msdos /dev/xsysace /root/cf # cd /root/
   > Oops: kernel access of bad area, sig: 11 [#1]
   > NIP: C00556B8 LR: C00557E4 CTR: 00000000
   > REGS: dfeb1e08 TRAP: 0300   Not tainted  (2.6.16.2)
   > MSR: 00021000 <ME>  CR: 227A8728  XER: 00000000
   > DAR: 00000000, DSISR: 00800000
   > TASK = dff6c030[4] 'events/0' THREAD: dfeb0000
   > GPR00: 00100100 DFEB1EB8 DFF6C030 C025AC60 DFF7BE10 00000018
  DFD96000
   > C0258D60
   > GPR08: 00000000 00200200 DFD96154 00000000 227A8788 800823E2
  1FFB9600
   > 00000000
   > GPR16: 00000001 FFFFFFFF 00000000 007FFF00 1FFB3540 1FF63920
  1FFCEE78
   > C01F0000
   > GPR24: C0240000 00100100 C0240000 00000000 DFF7BE10 00000018
  00000000
   > C025AC60
   > NIP [C00556B8] free_block+0xa8/0x148 LR [C00557E4] 
   > drain_array_locked+0x8c/0xd8 Call Trace:
   > [DFEB1EB8] [DFED53F0] 0xdfed53f0 (unreliable) [DFEB1ED8] [C00557E4] 
   > drain_array_locked+0x8c/0xd8 [DFEB1EF0] [C0056F80] 
   > cache_reap+0x74/0x18c [DFEB1F28] [C002B578] run_workqueue+0x9c/0x110 
   > [DFEB1F48] [C002B6E4] worker_thread+0xf8/0x13c [DFEB1FC0] [C002F6F0] 
   > kthread+0xf4/0x130 [DFEB1FF0] [C000413C] kernel_thread+0x44/0x60 
   > Instruction dump:
   > 7cfbfa14 3c000010 80e70014 3d2a4000 60000100 5529c9f4 7d295a14
  80c9001c
   > 3d200020 61290200 81060004 81660000 <91680000> 910b0004 3966001c
  90060000
   > BUG: events/0/4, lock held at task exit time!
   >  [c01f5d60] {cache_chain_mutex}
   > .. held by:          events/0:    4 [dff6c030, 110]
   > ... acquired at:               cache_reap+0x1c/0x18c
   >

Ameet Patil <ammubhai at gmail.com> wrote:

  agnel juni wrote:
> Hi Ameet
>
> I have posted a few messages regarding Ssytem ACE driver for Linux-2.6.
>
> We are working on a AMCC 440SPe based custom board.
>
> We applied the patch from
> //http://www.cs.york.ac.uk/rtslab/demos/amos/xupv2pro/patches/linuxppc-2.6.17.1-sysace-1.2.patch
> //and applied against 2.6.16-2 kernel.
>
> We are trying to make the driver work in interrupt mode.
>
> First, I would like to know if the driver tested in interrupt mode.
>
> We are able to mount the CF, but it is very inconsistent.
>
> Same is the case with fdisk command. When it fails,we get errors which you could see in the screen-dump below.//
> //
> Are we missing to apply the right patch? Please let us know your 
> inputs to go forward.
> Looking forwards for your reply.
>
> Thanks
> Junita
>
> /*************** Screen dump ********************/
> # fdisk /dev/xsysace
> 1. sector = 0 xsa_cur_req->sector=0
> System ACE: Error 0 when reading sector 2.
> 2. sector = 2 xsa_cur_req->sector=16
> end_request: I/O error, System ACE: Error 0 when reading sectoru dev 
> xsa, sector 16
> Buffer I/O error on device xsa, logical block 2
> 1. sector = 184 xsa_cur_req->sector=184
> 
> Command (m for help): p
> 
> Disk /dev/xsysace: 524 MB, 524869632 bytes
> 17 heads, 59 sectors/track, 1022 cylinders
> Units = cylinders of 1003 * 512 = 513536 bytes
> 
> Device Boot Start End Blocks Id System
> /dev/xsysace1 1 1022 512503+ 6 FAT16
> 
> Command (m for help): q
> And for 'mount'
> #
> #
> # mount -t msdos /dev/xsysace /root/cf
> 1. sector = 0 xsa_cur_req->sector=0
> 1. sector = 503 xsa_cur_req->sector=503
> 1. sector = 504 xsa_cur_req->sector=504
> 1. sector = 506 xsa_cur_req->sector=506
> 1. sector = 508 xsa_cur_req->sector=508
> 1. sector = 510 xsa_cur_req->sector=510
> # cd /root/cf
> < Here Prints Some Symbols like + - etc, which i am
> not able to capture/copy -------------ERROR
> b: No such file or directory--------------------------------ERROR
> pci.h
> #
> Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.n0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.0Minicom2.00.d#
> # cp pci.h /root/
> # cd ../
> # ls
> app cf hello1 pci.h
> #
> #/root
> /dev/xsysace /root/cf
> FAT: bogus number of reserved sectors
> VFS: Can't find a valid FAT filesystem on dev xsa.
> mount: Mounting /dev/xsysace on /root/cf failed: Invalid argument
> #
> 
>
>
> //
>
>
>
> //
>
> ------------------------------------------------------------------------
> Find out what India is talking about on - Yahoo! Answers India 
> 
> Send FREE SMS to your friend's mobile from Yahoo! Messenger Version 8. 
> Get it NOW 
> 

Hi Agnel,
The driver was last tested with 2.6.17 kernel and till date works 
fine in the interrupt mode. I have been using it since the last 4 months 
now. It should also work on 2.6.16-2 kernel I think, but I have not 
tested this. Have you made any changes to the driver? If so... please 
give the details so I can give you better feedback as to where things 
might have gone wrong. When time permits I shall try my patch on the 
2.6.16-2 kernel and let you know if it works for me.

Thanks,
-Ameet

---------------------------------
 Find out what India is talking about on  - Yahoo! Answers India 
 Send FREE SMS to your friend's mobile from Yahoo! Messenger Version 8. Get it NOW
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://ozlabs.org/pipermail/linuxppc-embedded/attachments/20060925/c2e0e987/attachment.htm 