Dear Poonam,
in message 1BD5CFC378ED0946B688E0C9BA2EF095192E64@zin33exm24.fsl.freescale.net you wrote:
Please explain why - we've been using this on PowerPC (where it has been implemented right in the beginning) for more than 9 years now, without ever finding it difficult.
Could you please point me to some Freescale platform which is using this API so that I can use the implementation as a reference. I searched but could not find.
What do you mean by "Freescale platform'? Freescale board,m or board based on Freescale processors? I'm not responsible for the former, so I cannot comment. As for boards using Freescale processors, see for example:
board/tqc/tqm5200/tqm5200.c: test2 = get_ram_size((long *)(CONFIG_SYS_SDRAM_BASE + dramsize), 0x20000000); board/tqc/tqm5200/tqm5200.c: test1 = get_ram_size((long *)CONFIG_SYS_SDRAM_BASE, 0x20000000); board/tqc/tqm5200/tqm5200.c: test2 = get_ram_size((long *)CONFIG_SYS_SDRAM_BASE, 0x20000000); board/tqc/tqm5200/tqm5200.c: test2 = test1 = get_ram_size((long *)(CONFIG_SYS_SDRAM_BASE + dramsize), 0x20000000); board/tqc/tqm8260/tqm8260.c: size = get_ram_size((long *)base, maxsize); board/tqc/tqm8272/tqm8272.c: size = get_ram_size((long *)base, maxsize); board/tqc/tqm834x/tqm834x.c: size = get_ram_size(base, DDR_MAX_SIZE_PER_CS); board/tqc/tqm85xx/sdram.c: if (get_ram_size (0, ddr_cs_conf[i].size) == board/tqc/tqm8xx/tqm8xx.c: return (get_ram_size(base, maxsize));
...just to give you a small slection.
534 MHz? Not 533.333 = 4 x 133.333 ?
You did not comment this one.
The settings were suggested by board designer,might be just to take care of the infinite series boundary condition.
I recommend to ask him again. This looks wrong to me.
A very good suggestion!, This will save us from maintaining the redundant stuff at many places. May be this change would not be possible immediately, as we
need the
P2020RDB board support in mainline quickly at the moment.
Well, it will not got in before the next merge window opens, which means not before September. Which leaves you with at least 4 weeks of time for such a cleanup.
Okay, but I need to work on other patches also for P2020RDB :(
Hey, all the better - then you can integrate this for P2020RDB, too, while you are at it :-)
Best regards,
Wolfgang Denk