This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

RTFS SATA example is "half" working

Other Parts Discussed in Thread: OMAPL138

Hello,

I have all required packages installed for CCS v5.1.1

  • rtfs_1_10_02_32
  • edma3_lld_01_11_03_01
  • pspdrivers_01_30_01
  • bios_5_41_11_38
  • biosusb_01_10_03

I'm able to compile and debug example SATA.

I try couples different HDD and ONLY Hitachi Model HTS727550A9E364 working, it hangs for others Seagate ones at

Get drive ID DONE. A:

While debugging, I found out a lot missing source for functions getdriveid(), fmtdrv(), bfs_shell() with error msg

 Can't find a source file at "/db/vtree/library/trees/rtfs/rtfs-b41x/src/ti/rtfs/posix/src/rtfsPosix.c"

1) Any idea where to find this source ?

2) How to make it work for other SATA beside Hitachi ?

Another concern is PRINTF on the console!

In the Release version, I still get unexpected printout on the console. Does this slow down the real-time execute in stand-alone boot mode ?

Thanks,

~Duy-Ky

What should I do ?

Thanks,

~Duy-Ky




  • I have moved this to the BIOS forums.

    Regards

    Mukul

  • Hi Duy-Ky,

    I'm forwarding this thread to the group who can help you out here. But I'll answer what I can.

    What device are you using?

    Duy-Ky Nguyen88210 said:
    1) Any idea where to find this source ?

    The source file you are looking for is within the rtfs product installation. Within the rtfs_1_10_02_32 directory you will find a "packages" directory. Under this directory you can find the source file as indicated by CCS.

    Duy-Ky Nguyen88210 said:

    Another concern is PRINTF on the console!

    In the Release version, I still get unexpected printout on the console. Does this slow down the real-time execute in stand-alone boot mode ?

    What is being printed to the console?

    In general, a printf to the CCS console consumes a lot of time due to the data transfer process and because it halts the CPU. When the target is not connected via the debugger to CCS, printf will still go through the formatting routine but since it can't send the data, it won't halt the CPU and the execution time will be faster. Instead of using printf, we recommend using LOG_printf(), which stores log events in a buffer to be read in by the host.

  • Hi Tom,

    I searched through the package so many time, even few minute ago, and NOT able to locate the file.


    Can't find a source file at "/db/vtree/library/trees/rtfs/rtfs-b41x/src/ti/rtfs/posix/src/rtfsPosix.c"

    I'm using the package rtfs_1_10_02_32.tar.gz of size 22,470 KB and unziped to 72.3 MB

    Could you have a look and let me know what your finding.

    Thanks,

    ~Duy-Ky

  • Hi Duy-Ky,

    Unfortunately, the RTFS sources are not available without a license.  RTFS is made by a 3rd party (who contracted with TI) called EBS.

    If you wish to obtain the RTFS sources, you would need to work with EBS to buy a source license.  I would be able to help you get into contact with EBS if you like.

    Regarding the Hitachi drive support, this is an issue for the PSP drivers team to help with.  I will see to it that your question is brought to their attention.

    Steve

  • Duy-Ky Nguyen88210 said:

    I try couples different HDD and ONLY Hitachi Model HTS727550A9E364 working, it hangs for others Seagate ones at

    Duy-Ky Nguyen88210 said:

    2) How to make it work for other SATA beside Hitachi ?

     

    All our release testing/validation has been done using the seagate 500GB HDD, 3.5'. So, the sample application should have worked on seagate HDD.

    Regards,

    Sandeep K

  • Hi Steven,

    Could I have attention from your PSP drivers team on my issues below

    1)

    I'm quite the SATA III problem in the errata document and I confirm the Hitachi I'm using is SATA II. However I still have quite some problems below

    For some reason, I'm NOT able to create partition larger than 128 GB while I'm ABLE to get full 500 GB using Linux DosTools or Window FAT32FORMAT tool. Needless to say native windows tool allows maximum 32 GB for FAT32

    You may  check for yourself by using the windows FAT32FORMAT from

    http://www.ridgecrop.demon.co.uk/index.htm?fat32format.htm

    and it could format FAT32 my 750GB Seagate to max capacity 750 GB with even smallest Clusster of 512 B, formatting is too slow though

    2) For some reason, I'm NOT able to use USB-SATA cable to verify SATA after partitioned-formatted-written,  It happened to work with my old cable I bought few year ago, but not for new purchases. I first bought ViewCon, it failed, then bought a well-known StarTech, it failed too!

    Any solution to thess TWO problems ?

    Thanks,

    ~Duy-Ky


  • Hi Sandeep,

    I recall that there is a table somewhere in the PSP drivers which lists the serial numbers of supported devices.  Do you recall this?  Would that table show whether or not the Hitachi drive is supported?

    Thanks,

    Steve

  • Duy-Ky,

    Duy-Ky Nguyen88210 said:
    I'm quite the SATA III problem in the errata document and I confirm the Hitachi I'm using is SATA II

    I didn't completely understand what you meant by this.  What is the SATA III problem?

    Duy-Ky Nguyen88210 said:
    For some reason, I'm NOT able to create partition larger than 128 GB while I'm ABLE to get full 500 GB using Linux DosTools or Window FAT32FORMAT tool. Needless to say native windows tool allows maximum 32 GB for FAT32

    can you provide your application code that attempts to do this using RTFS?

    One idea is to try the API:

    fmtxdrv() - Perform an extended volume format

    Duy-Ky Nguyen88210 said:
    2) For some reason, I'm NOT able to use USB-SATA cable to verify SATA after partitioned-formatted-written,  It happened to work with my old cable I bought few year ago, but not for new purchases. I first bought ViewCon, it failed, then bought a well-known StarTech, it failed too!

    This sounds to me like an issue with the cable you are using ...

    Steve

  • Hi Steve,

    1) SATA III

    OMAPL138 Errata document mention SATA GEN 3 and one of co-worker mentioned something about SATA GEN3 over the phone when I contacted for quick help.

    That's why I ignore the problem in using Seagate SATA I have, and focus on Hitachi one which is the most successful one I have

    2) Cable

    I'm afraid nothing wrong with the cable as I can use it successfully to partition / format / write / read for 500GB SATA with usage of full capacity of 500GB in FAT32

    3) Partition

    I'm using the example SATA from TI package RTFS rtfs_1_10_02_32. This example creates 2 partitions with only 200 cylinder, leaves the rest for 2nd partition. I have 50MB for 1st partition and 127GB for 2nd one

    I increase number cylinders for 1st partition size of 60 GB, the 2nd one is created with only 60 GB, even all remaining disk is used for 2nd partition!

    I create only ONE partition with full disk space, its size is only 128 GB

    So, the total max disk usage is 128 GB, for 1 or 2 partitions.

    In 3 problems above, I'm really after PROBLEM 3) OF PARTITION

    Any idea how to use TI RTFS in creating partition larger than 128GB, say 5000GB ?

    Thanks,

    ~Duy-Ky

    Attached is the main file from SATA example of TI RTFS package

  • Duy-Ky,

    Ok, thanks for the extra info.  Just so you know, I am trying to reproduce your problem here.  However, I am having some problems with our SATA drive.

    I'll get back to you on it as soon as I know more.

    Thanks,

    Steve

  • Duy-Ky,

    Since we last spoke I have been in contact with several people about the problem you are facing, including the creator of the RTFS file system and the PSP drivers team.

    This problem has been encountered before, and is due to a limitation with in the SATA driver.  The limitation is that the driver was written to only support 28 bit LBA; as such the maximum disk size is limited to 128GB.

    The driver team also informed me that the SATA controller on the evmOMAPL138 supports up to 32 bit LBA.  Would that be sufficient for you?

    If so, they will be willing to help you with the code changes necessary to do that.

    Steve

  • Hello Steve,

    Thanks so much for your excellent support, Steve

    We have to stream data at sampling rate 64 Mbit/sec for  a default test time of 30 minutes, so storage data is of 864 GByte(not bit), it's just for only 1 test.

    And we have 2 OMAP's on each board and a single system has 4 boards, so each system has 8 OMAPs with 8 HDDs

    So 28-bit is NOT enough for us, please to have code change to support 32 bit LBA

    Thanks so much for your very great suppport, Steve

    ~Duy-Ky

  • Hi Duy-Ky,

    I can provide you the couple of pointers to support the 32 bit LBA in SATA driver.

    Can you please send me the private message mentioning your email ID?

    Regards,

    Sandeep K

  • Hi Sandeep,

    Could you get my email from TI system or you may want to call me at

    707-524-8368 x323

    Thanks,

    ~Duy-Ky

  • Duy-Ky,

    Email ID is the part of the private information in the profile. I do not have the access to it. :(

    Regards,

    Sandeep K

  • Hi Sandeep,

    Could you send to

    duykyng@unitedthc.com

    Thanks so much for your great support, Sandeep,

    ~Duy-Ky

  • Duy-Ky,

    As promised, I had shared the example code to support the 32 bit LBA addressing mode in the SATA driver.

    Thanks and Regards,

    Sandeep K