Date:   Tue, 29 May 2018 10:25:04 +0300
From:   Eugen Hristev <eugen.hristev@...rochip.com>
To:     Peter Rosin <peda@...ntia.se>
CC:     Alexandre Belloni <alexandre.belloni@...tlin.com>,
        Richard Weinberger <richard@....at>,
        Tudor Ambarus <tudor.ambarus@...rochip.com>,
        Josh Wu <rainyfeeling@...look.com>,
        Boris Brezillon <boris.brezillon@...tlin.com>,
        <linux-kernel@...r.kernel.org>,
        Marek Vasut <marek.vasut@...il.com>,
        Ludovic Desroches <ludovic.desroches@...rochip.com>,
        <linux-mtd@...ts.infradead.org>,
        Cyrille Pitchen <cyrille.pitchen@...ev4u.fr>,
        Brian Norris <computersforpeace@...il.com>,
        David Woodhouse <dwmw2@...radead.org>,
        <linux-arm-kernel@...ts.infradead.org>
Subject: Re: [PATCH] mtd: nand: raw: atmel: add module param to avoid using
 dma



On 29.05.2018 10:10, Peter Rosin wrote:
> On 2018-05-29 08:30, Eugen Hristev wrote:
>>
>>
>> On 28.05.2018 13:10, Peter Rosin wrote:
>>> On 2018-05-28 00:11, Peter Rosin wrote:
>>>> On 2018-05-27 11:18, Peter Rosin wrote:
>>>>> On 2018-05-25 16:51, Tudor Ambarus wrote:
>>>>>> We think the best way is to keep LCD on DDR Ports 2 and 3 (8th and 9th
>>>>>> slaves), to have maximum bandwidth and to use DMA on DDR port 1 for NAND
>>>>>> (7th slave).
>>>>>
>>>>> Exactly how do I accomplish that?
>>>>>
>>>>> I can see how I can move the LCD between slave DDR port 2 and 3 by
>>>>> selecting LCDC DMA master 8 or 9 (but according to the above it should
>>>>> not matter). The big question is how I control what slave the NAND flash
>>>>> is going to use? I find nothing in the datasheet, and the code is also
>>>>> non-transparent enough for me to figure it out by myself without
>>>>> throwing out this question first...
>>
>>   >> [...]
>>
>>>> and the output is
>>>>
>>>> atmel-nand-controller 10000000.ebi:nand-controller: using dma0chan5 for DMA transfers
>>>>
>>>> So, DMA controller 0 is in use. I still don't know if IF0, IF1 or IF2 is used
>>>> or how to find out. I guess IF2 is not in use since that does not allow any
>>>> DDR2 port as slave...
>>
>> Hello Peter,
>>
>> Thank you for all the information, I will chip in to help a little bit.
>> The Master/channel is described in the device tree. The channel has a
>> controller, a mem/periph interface and a periph ID, plus a FIFO
>> configuration.
>>
>> The dma chan number reported in the dmesg is just software.
> 
> Got that, that was why I added the various additional traces in that
> horrid patch in the mail you are responding to :-)
> 
>>                                                              Here is an
>> example from DT:
>> dmas = <&dma0 2 AT91_DMA_CFG_PER_ID(1)>,
>>          <&dma0 2 AT91_DMA_CFG_PER_ID(2)>;
>>
>> you can match this with the help from
>> Documentation/devicetree/bindings/dma/atmel-dma.txt:
>>
>> 1. A phandle pointing to the DMA controller.
>>
>> 2. The memory interface (16 most significant bits), the peripheral
>> interface
>> (16 less significant bits).
>>
>> 3. Parameters for the at91 DMA configuration register which are device
>>
>> dependent:
>>
>>     - bit 7-0: peripheral identifier for the hardware handshaking
>> interface. The
>>     identifier can be different for tx and rx.
>>
>>     - bit 11-8: FIFO configuration. 0 for half FIFO, 1 for ALAP, 2 for ASAP.
>>
>>
>> So , what was Tudor asking for, is your DT for the ebi node (if you are
>> using ebi), or, your NFC SRAM (Nand Flash Controller SRAM) DMA
>> devicetree chunk, so, we can figure out which type of DMA are you using.
>>
>> Normally, the ebi should be connected to both DMA0 and DMA1 on those
>> interfaces specified in DT. Which ones you want to use, depends on your
>> setup (and contention on the bus/accesses, like in your case, the HLCDC)
>>
>> Thats why we have multiple choices, to pick the right one for each case.
>> In our vanilla DT sama5d3.dtsi we do not have DMA described for ebi
>> interface.
> 
> Ahh, *that* NAND DMA configuration. Of course, how silly of me...
> 
> This is a setup based on the at91-linea.dtsi "CPU" board. That dtsi should
> have the relevant NAND/DMA info. Also, at91-nattis-2-natte-2.dts describes
> the older HW (with a 1024x768 panel) that is also affected, if you want a
> full device tree to look at.
> 
> Looks like I didn't make a selection, quoting from at91-linea.dtsi:

Ok, so try to force the nand to use DDR port 1 : use DMAC0 or DMAC1 on 
the interfaces corresponding just to DDR port 1, and see if helps (while 
leaving HLCDC on both masters - DDR port 2 and 3).

One more thing: what are the actual nand commands which you use when you 
get the glitches? read/write/erase ... ?
What happens if you try to minimize the nand access? you also said at 
some point that only *some* nand accesses cause glitches.

Another thing : even if the LCD displays a still image, the DMA still 
feeds data to the LCD right ?

> 
> &ebi {
> 	pinctrl-0 = <&pinctrl_ebi_nand_addr>;
> 	pinctrl-names = "default";
> 	status = "okay";
> };
> 
> &nand_controller {
> 	status = "okay";
> 
> 	nand: nand@3 {
> 		reg = <0x3 0x0 0x2>;
> 		atmel,rb = <0>;
> 		nand-bus-width = <8>;
> 		nand-ecc-mode = "hw";
> 		nand-ecc-strength = <4>;
> 		nand-ecc-step-size = <512>;
> 		nand-on-flash-bbt;
> 		label = "atmel_nand";
> 	};
> };
> 
> The reason is probably because the sama5d3xek device-trees didn't at the
> time of "fork". Does anybody have any suggestion for some extra properties
> to try in the above nodes?
> 
> Further, I forgot that I had actually upstreamed linea support for
> at91bootstrap, so relevant NAND timings etc can be found in lpddr1_init()
> in
> 
> 	at91bootstrap/contrib/board/axentia/sama5d3_linea/sama5d3_linea.c
> 
> Cheers,
> Peter
> 
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@...ts.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
> 

Powered by blists - more mailing lists