[U-Boot] Newbie SPL question for socfpga_sockit

[Dinh Nguyen <dinguyen@opensource.altera.com>]

On 03/02/2016 05:24 PM, Marek Vasut wrote:
> On 03/03/2016 12:08 AM, Dinh Nguyen wrote:
>> On 03/02/2016 04:54 PM, Dinh Nguyen wrote:
>>> CC: Marek Vasut
>>>
>>> On 03/01/2016 08:40 PM, George Broz wrote:
>>>> On 17 February 2016 at 18:45, Phil Reid <preid@electromag.com.au> wrote:
>>>>> G'day George
>>>>>
>>>>>
>>>>> On 18/02/2016 5:54 AM, George Broz wrote:
>>>>>>
>>>>>> Hello,
>>>>>>
>>>>>> Sorry for the newbie question...
>>>>>>
>>>>>> I have an Altera/Terasic board (socfpga_sockit) that has issues
>>>>>> recognizing
>>>>>> USB storage devices (roughly 60% good / 40% bad):
>>>>>>
>>>>
>>>>>> If I try a later release (e.g. v2016.01 which seems to support the Terasic
>>>>>> board explicitly) the boot process stops just after loading the SPL.
>>>>>>
>>>>>> At this point I have only been changing the u-boot.img component, not the
>>>>>> SPL (in the a2 partition of the MMC).
>>>>>>
>>>>>> The documentation from Altera about generating the SPL seems to require
>>>>>> using Qsys/Quartus tools which I'd really like to avoid.
>>>>>>
>>>>>> I can produce u-boot-spl.bin / u-boot.img from
>>>>>> make socfpga_sockit_defconfig; make all.
>>>>>>
>>>>>> My newbie question ... should be I able to use directly the u-boot-spl.bin
>>>>>> generated by the build to replace the SPL on the board?
>>>>>>
>>>>>
>>>>> I've just gone thru the process (this week) of getting uboot 2016.01 to boot
>>>>> on our custom socfpga board.
>>>>> There where a few issues in getting things going.
>>>>> If your using the sd card to boot then the current memory layout is not per
>>>>> the altera documentation.
>>>>> I had to make the changes below to config to match the altera docs.
>>>>> Otherwise it hangs just after the SPL.
>>>>> You also need to use the -dtb image versions with 2016.01 for things to
>>>>> work.
>>>>> Again you may need to make sure the uboot device trees match your hardware.
>>>>> Not all uboot drivers appear to be using the DT as yet.
>>>>>
>>>>> Note the SPL is tightly coupled to the board design and potentially the FPGA
>>>>> image.
>>>>> If you using bridges or routing HPS resources (eg i2c etc) to the fpga the
>>>>> SPL configures the muxes.
>>>>>
>>>>> There's a script in uboot src at arch\arm\mach-socfpga\qts-filter.sh
>>>>> That will generate the qts files for you arch.
>>>>>
>>>>> I haven't tried the USB stuff as yet.
>>>>> There are some issues witht eh altera USB port trigger an Over Current event
>>>>> on device insertation.
>>>>> They've modified the linux kenerl driver to use external OV current
>>>>> detection to get around the problem.
>>>>> My intial workaround was to insert USB device prior to power on.
>>>>> I haven't looked at the uboot USB driver yet to see what's in there.
>>>>>
>>>>>
>>>>> --
>>>>> Regards
>>>>> Phil Reid
>>>>>
>>>>>
>>>>> diff --git a/include/configs/socfpga_common.h
>>>>> b/include/configs/socfpga_common.h
>>>>> index a09e906..3a1b59b 100644 (file)
>>>>> --- a/include/configs/socfpga_common.h
>>>>> +++ b/include/configs/socfpga_common.h
>>>>> @@ -357,13 +357,13 @@ unsigned int cm_get_qspi_controller_clk_hz(void);
>>>>>
>>>>> /* SPL SDMMC boot support */
>>>>> #ifdef CONFIG_SPL_MMC_SUPPORT
>>>>> +#define CONFIG_SPL_LIBDISK_SUPPORT
>>>>> #if defined(CONFIG_SPL_FAT_SUPPORT) || defined(CONFIG_SPL_EXT_SUPPORT)
>>>>> #define CONFIG_SYS_MMCSD_FS_BOOT_PARTITION 2
>>>>> #define CONFIG_SPL_FS_LOAD_PAYLOAD_NAME "u-boot-dtb.img"
>>>>> -#define CONFIG_SPL_LIBDISK_SUPPORT
>>>>> #else
>>>>> -#define CONFIG_SYS_MMCSD_FS_BOOT_PARTITION 3
>>>>> -#define CONFIG_SYS_MMCSD_RAW_MODE_U_BOOT_SECTOR 0xa00 /* offset 2560 sect
>>>>> (1M+256k) */
>>>>> +#define CONFIG_SYS_MMCSD_RAW_MODE_U_BOOT_PARTITION 3
>>>>> +#define CONFIG_SYS_MMCSD_RAW_MODE_U_BOOT_SECTOR 0x200
>>>>> #define CONFIG_SYS_U_BOOT_MAX_SIZE_SECTORS 800 /* 400 KB */
>>>>> #endif
>>>>> #endif
>>>>>
>>>>>
>>>>
>>>> Thanks, Phil!
>>>>
>>>> Sorry for the delayed response - got called away, but am back to this
>>>> now. I patched
>>>> socfpga_common.h and re-built the project. I picked up
>>>> spl/u-boot-spl-dtb.sfp and
>>>> u-boot-dtb.img and transferred them to the SD card with:
>>>>
>>>> dd if=u-boot-spl-dtb.sfp of=/dev/sdf3 bs=64k seek=0
>>>> dd if=u-boot-dtb.img of=/dev/sdf3 bs=64k seek=4
>>>>
>>>
>>> You can just do a 'make u-boot-with-spl.sfp' and burn the
>>> u-boot-with-spl.sfp file straight to the a2 partition, or sdf3.
>>>
>>>> Tried this with both the original DT set (socfpga.dtsi, socfpga_cyclone.dtsi,
>>>> socfpga_cyclone5_sockit.dts) that came with the u-boot v2016.01 download and
>>>> also an Altera-patched DT set that I've used to boot into Linux numerous times.
>>>>
>>>> When I start up the board I get:
>>>>
>>>> U-Boot SPL 2016.01 (Mar 01 2016 - 17:28:14)
>>>> drivers/ddr/altera/sequencer.c: Preparing to start memory calibration
>>>> drivers/ddr/altera/sequencer.c: CALIBRATION FAILED
>>>> drivers/ddr/altera/sequencer.c: Calibration complete
>>>> SDRAM calibration failed.
>>>> ### ERROR ### Please RESET the board ###
>>>>
>>>> I'm not a Quartus user, so I haven't done anything with the
>>>> qts-filter.sh script you
>>>> mentioned. Do I need to? I don't have any custom FPGA logic - it's
>>>> just the Terasic
>>>> board out of the box.
>>>>
>>>
>>> I just tested U-Boot(v2016.03-rc3) and USB is not working on the devkit
>>> or sockit. It's failing to detect my mass storage device.
>>>
>>> Debugging...
>>>
>>
>> Sorry, I mis-spoke for the sockit. I have to turn dcache off for USB to
>> work reliably.
>>
>> => dcache off
>> => usb start
>> starting USB...
>> USB0:   Core Release: 2.93a
>> scanning bus 0 for devices... 2 USB Device(s) found
>> => usb tree
>> USB device tree:
>>   1  Hub (480 Mb/s, 0mA)
>>   |   U-Boot Root Hub
>>   |
>>   +-2  Mass Storage (480 Mb/s, 200mA)
>>        SanDisk Firebird USB Flash Drive 4C532000040115111005
>>
>> => version
>>
>> U-Boot 2016.03-rc3 (Mar 02 2016 - 17:00:25 -0600)
> 
> Well, that's our usual USB/QSPI cache issue that's tormenting your soul.
> CCing Chin ;-)
> 
> Does the issue by any chance magically disappear if you apply this patch:
> 
> diff --git a/arch/arm/include/asm/system.h b/arch/arm/include/asm/system.h
> index 026e7ef..06802c6 100644
> --- a/arch/arm/include/asm/system.h
> +++ b/arch/arm/include/asm/system.h
> @@ -274,7 +274,7 @@ static inline void set_dacr(unsigned int val)
> 
>  /* options available for data cache on each page */
>  enum dcache_option {
> -       DCACHE_OFF = TTB_SECT_DOMAIN(0) | TTB_SECT_XN_MASK | TTB_SECT,
> +       DCACHE_OFF = TTB_SECT_S_MASK | TTB_SECT_DOMAIN(0) |
> TTB_SECT_XN_MASK | TTB_SECT,
>         DCACHE_WRITETHROUGH = DCACHE_OFF | TTB_SECT_C_MASK,
>         DCACHE_WRITEBACK = DCACHE_WRITETHROUGH | TTB_SECT_B_MASK,
>         DCACHE_WRITEALLOC = DCACHE_WRITEBACK | TTB_SECT_TEX(1),
> 

Yes, this patch fixes USB with dcache left on.

Dinh