Recent changes to bugs

Major problem in sft

Leon — Sun, 24 Nov 2013 12:17:59 -0000

I am pretty sure I see a bug in the sfh code.
In brief (I am not sure that somebody reads this) - in ALL versions sfh programs/burns UBL/U-Boot WITHOUT(!!!) checking the current(!), not only the first, block to be bad (factory marked).
This may cause a lot of troubles - and already causes me.
Details to somebody who will answer...

the size of sft_DM6446_NAND.bin is larger than 0x3800

Anonymous — Thu, 19 May 2011 06:15:35 -0000

the version:
DM644x_FlashAndBootUtils_2_00

the tool:
MVL 5.0 Tools come with davinci DVSDK 2.00.00.22
[http://software-dl.ti.com/dsps/dsps_public_sw/sdo_sb/S1SDKLNX/DVSDK_2_00/index_FDS.html]

the steps to reproduce the bug
1. change to the directory DM644x/GNU/sft
2. make clean && make all
3. the size of re-generated sft_DM6446_NAND.bin file is 14356 bytes.

sfh_DM646x Output cannot be redirected

Anonymous — Tue, 10 Aug 2010 14:06:15 -0000

The output of the sfh_DM646x.exe file cannot be redirected due to an uncaught exception thrown by Console.Clear() when the program title is printed.

DM6467 - wrong setting in UBL speed definition

Anonymous — Wed, 15 Jul 2009 08:48:46 -0000

file DM646x\Common\src\device.c file should be changed from "#define DM6467_1GHz_MICRON" to #define DM6467_594MHz"

DM355 UBL: error in DDR setting

Massimo Martelli — Mon, 06 Apr 2009 15:38:41 -0000

There is a bug in the UBL that causes a low performance on the DDR2 bus.

In DM35x_FlashAndBootUtils_1_10\DM35x\Common\src\device.c line 432-435:

// Set the refresh rate
//DDR->SDRCR = 0x00000535;
DDR->SDRCR &= (0xFFFF0000 | DDR_RR);

With
static const Uint16 DDR_RR = 1336;

The mask is obviously wrong. If we force it to
DDR->SDRCR = 0x00000535;

We get the performance expected.