UPGRADE YOUR BROWSER

We have detected your current browser version is not the latest one. Xilinx.com uses the latest web technologies to bring you the best online experience possible. Please upgrade to a Xilinx.com supported browser:Chrome, Firefox, Internet Explorer 11, Safari. Thank you!

cancel
Showing results for 
Search instead for 
Did you mean: 
Highlighted
Visitor baruchlerer
Visitor
313 Views
Registered: ‎05-10-2018

QSPI24 boot failures

Jump to solution

We have specific board using ZynqMP ultrascale XCZU4EV

When booting from QSPI24 mode, from time to time boot fails.

Following AR# 68656   FSBL doesn't print anything at this situation (I opened FSBL_DEBUG_DETAILED ) , 

So I focused on next step using jtag and boot_registers_log_revX_Y.tcl script.

Sometimes JTAG fails to connect to board in boot fail state ??? I attached two tcl outputs when jtag successfully connected.

I'm not sure how to parse tcl output response.

Attached two fails of tcl outputs, one boot OK tcl output,  and FSBL boot OK output.

What can be the reason for such inconsistent behavior?

Regards

Baruch

 

Tags (1)
0 Kudos
1 Solution

Accepted Solutions
Visitor baruchlerer
Visitor
233 Views
Registered: ‎05-10-2018

Re: QSPI24 boot failures

Jump to solution

We delayed the QSPI clk line, this seems to bypass the problem.

I can't use FSBL to debug, since it's not loaded,  If there was away to change/debug the bootrom.

 

Thanks for help

0 Kudos
5 Replies
Xilinx Employee
Xilinx Employee
263 Views
Registered: ‎09-01-2014

Re: QSPI24 boot failures

Jump to solution

CSU_BR_ERROR                    = FFD80528:   00009292

means Tamper event that is detected while in post boot

ERROR_STATUS_2                  = FFD80540:   00001600

Mean PLL lock error.

Are the input ps_clk or power stable on your board?

if there is no FSBL load error from BootRom error codes and  FSBL doesn't print anything with FSBL_DEBUG_DETAILED.  Basically FSBL might hang in psu_init 

0 Kudos
Visitor baruchlerer
Visitor
254 Views
Registered: ‎05-10-2018

Re: QSPI24 boot failures

Jump to solution

Thanks for quick reply.

trying to debug this issue I found that:

from psu_init.c 

if checking return value, than it takes much more time till IOPLL is locked than code does, only after three mask_poll() calling the IOPLL is locked, I don't think this is my problem but this is bad coding not to check mask_poll() return value. see below code from psu_init.c auto generated, wher mask_poll() return value is not verified.

it seems that my issues related to  PMU bootrom fails to load FSBL from QSPI24, FSBL is not runing at all.

when scope probe or even the touch of my finger is places on QSPI_clk pin it always resolve this issue, if none than somtimes boot fails as mentioned.

I assume it related to timing issue, any more suggestions how to fix it? or how to debug bootrom code?

Regards

Baruch

 

 

/*
* CHECK PLL STATUS
*/
/*
* Register : PLL_STATUS @ 0XFF5E0040

* IOPLL is locked
* PSU_CRL_APB_PLL_STATUS_IOPLL_LOCK 1
* (OFFSET, MASK, VALUE) (0XFF5E0040, 0x00000001U ,0x00000001U)
        */
    mask_poll(CRL_APB_PLL_STATUS_OFFSET, 0x00000001U);
0 Kudos
Moderator
Moderator
241 Views
Registered: ‎03-19-2014

Re: QSPI24 boot failures

Jump to solution

You can change to QSPI device clock as described in AR69831  You can create an FSBL that has the debug symbols as described on the WIKI page

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos
Visitor baruchlerer
Visitor
234 Views
Registered: ‎05-10-2018

Re: QSPI24 boot failures

Jump to solution

We delayed the QSPI clk line, this seems to bypass the problem.

I can't use FSBL to debug, since it's not loaded,  If there was away to change/debug the bootrom.

 

Thanks for help

0 Kudos
Moderator
Moderator
224 Views
Registered: ‎03-19-2014

Re: QSPI24 boot failures

Jump to solution

Sounds like you have some SI issues in your QSPI clock.    The BootROM is not editable.  

-------------------------------------------------------------------------
Don’t forget to reply, kudo, and accept as solution.
-------------------------------------------------------------------------
0 Kudos