[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1492051237.2866.36.camel@mhfsdcap03>
Date: Thu, 13 Apr 2017 10:40:37 +0800
From: Guochun Mao <guochun.mao@...iatek.com>
To: Cyrille Pitchen <cyrille.pitchen@...ev4u.fr>
CC: Cyrille Pitchen <cyrille.pitchen@...el.com>,
Mark Rutland <mark.rutland@....com>,
<devicetree@...r.kernel.org>, Richard Weinberger <richard@....at>,
Russell King <linux@...linux.org.uk>,
<linux-kernel@...r.kernel.org>, Rob Herring <robh+dt@...nel.org>,
<linux-mtd@...ts.infradead.org>,
Matthias Brugger <matthias.bgg@...il.com>,
<linux-mediatek@...ts.infradead.org>,
David Woodhouse <dwmw2@...radead.org>,
<linux-arm-kernel@...ts.infradead.org>
Subject: Re: [PATCH v1 1/1] mtd: mtk-nor: set controller's address width
according to nor flash
Hi Cyrille,
On Wed, 2017-04-12 at 22:57 +0200, Cyrille Pitchen wrote:
> Hi Guochun,
>
> Le 05/04/2017 à 10:37, Guochun Mao a écrit :
> > When nor's size larger than 16MByte, nor's address width maybe
> > set to 3 or 4, and controller should change address width according
> > to nor's setting.
> >
> > Signed-off-by: Guochun Mao <guochun.mao@...iatek.com>st
> > ---
> > drivers/mtd/spi-nor/mtk-quadspi.c | 27 +++++++++++++++++++++++++++
> > 1 file changed, 27 insertions(+)
> >
> > diff --git a/drivers/mtd/spi-nor/mtk-quadspi.c b/drivers/mtd/spi-nor/mtk-quadspi.c
> > index e661877..b637770 100644
> > --- a/drivers/mtd/spi-nor/mtk-quadspi.c
> > +++ b/drivers/mtd/spi-nor/mtk-quadspi.c
> > @@ -104,6 +104,8 @@
> > #define MTK_NOR_MAX_RX_TX_SHIFT 6
> > /* can shift up to 56 bits (7 bytes) transfer by MTK_NOR_PRG_CMD */
> > #define MTK_NOR_MAX_SHIFT 7
> > +/* nor controller 4-byte address mode enable bit */
> > +#define MTK_NOR_4B_ADDR_EN BIT(4)
> >
> > /* Helpers for accessing the program data / shift data registers */
> > #define MTK_NOR_PRG_REG(n) (MTK_NOR_PRGDATA0_REG + 4 * (n))
> > @@ -230,10 +232,35 @@ static int mt8173_nor_write_buffer_disable(struct mt8173_nor *mt8173_nor)
> > 10000);
> > }
> >
> > +static void mt8173_nor_set_addr_width(struct mt8173_nor *mt8173_nor)
> > +{
> > + u8 val;
> > + struct spi_nor *nor = &mt8173_nor->nor;
> > +
> > + val = readb(mt8173_nor->base + MTK_NOR_DUAL_REG);
> > +
> > + switch (nor->addr_width) {
> > + case 3:
> > + val &= ~MTK_NOR_4B_ADDR_EN;
> > + break;
> > + case 4:
> > + val |= MTK_NOR_4B_ADDR_EN;
> > + break;
> > + default:
> > + dev_warn(mt8173_nor->dev, "Unexpected address width %u.\n",
> > + nor->addr_width);
> > + break;
> > + }
> > +
> > + writeb(val, mt8173_nor->base + MTK_NOR_DUAL_REG);
> > +}
> > +
> > static void mt8173_nor_set_addr(struct mt8173_nor *mt8173_nor, u32 addr)
> > {
> > int i;
> >
> > + mt8173_nor_set_addr_width(mt8173_nor);
> > +
> > for (i = 0; i < 3; i++) {
>
> Should it be 'i < nor->addr_width' instead of 'i < 3' ?
> Does it work when accessing data after 128Mbit ?
Yes, it can work.
Let's see the whole function,
static void mt8173_nor_set_addr(struct mt8173_nor *mt8173_nor, u32 addr)
{
int i;
mt8173_nor_set_addr_width(mt8173_nor);
for (i = 0; i < 3; i++) {
writeb(addr & 0xff, mt8173_nor->base + MTK_NOR_RADR0_REG
+ i * 4);
addr >>= 8;
}
/* Last register is non-contiguous */
writeb(addr & 0xff, mt8173_nor->base + MTK_NOR_RADR3_REG);
}
The nor controller has 4 registers for address.
This '3' indicates the number of contiguous address' registers
base + MTK_NOR_RADR0_REG(0x10)
base + MTK_NOR_RADR1_REG(0x14)
base + MTK_NOR_RADR2_REG(0x18),
but the last address register is non-contiguous,
it's base + MTK_NOR_RADR3_REG(0xc8)
mt8173_nor_set_addr will set addr into these 4 registers by Byte.
The bit MTK_NOR_4B_ADDR_EN will decide whether 3-byte(0x10,0x14,0x18)
or 4-byte(0x10,0x14,x018,0xc8) been sent to nor device.
and, it can access data after 128Mbit when sent 4-byte address.
Best regards,
Guochun
>
> Best regards,
>
> Cyrille
>
> > writeb(addr & 0xff, mt8173_nor->base + MTK_NOR_RADR0_REG + i * 4);
> > addr >>= 8;
> >
>
Powered by blists - more mailing lists