[v4 12/12] spi: aspeed: Fix bug when SPI_NOR_4B_OPCODES flag is set

[Chin-Ting Kuo]

Hi Cédric,

> -----Original Message-----
> From: Cédric Le Goater <clg at kaod.org>
> Sent: Friday, July 1, 2022 5:45 PM
> Subject: Re: [v4 12/12] spi: aspeed: Fix bug when SPI_NOR_4B_OPCODES flag
> is set
> 
> On 5/24/22 07:56, Chin-Ting Kuo wrote:
> > When SPI_NOR_4B_OPCODES flag is set for a specific flash in
> > spi_nor_ids[] table, EN4B command will not be issued and 4-byte
> > dedicated commands will be used to access SPI flash.
> > Thus, instead of EN4B command, address length is more suitable for
> > deciding whether the controller should be set to 4-byte mode.
> >
> > Signed-off-by: Chin-Ting Kuo <chin-ting_kuo at aspeedtech.com>
> > ---
> >   drivers/spi/spi-aspeed.c | 7 +++----
> >   1 file changed, 3 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/spi/spi-aspeed.c b/drivers/spi/spi-aspeed.c index
> > e5e348eb7b..687ddc370b 100644
> > --- a/drivers/spi/spi-aspeed.c
> > +++ b/drivers/spi/spi-aspeed.c
> > @@ -504,10 +504,6 @@ static int aspeed_spi_exec_op_user_mode(struct
> spi_slave *slave,
> >   	/* Restore controller setting. */
> >   	writel(flash->ce_ctrl_read, ctrl_reg);
> >
> > -	/* Set controller to 4-byte mode when flash is in 4-byte mode. */
> > -	if (op->cmd.opcode == SPINOR_OP_EN4B)
> > -		priv->info->set_4byte(bus, cs);
> > -
> >   	return 0;
> >   }
> 
> 
> This should be part of the patch adding the driver.
> 

Okay.

Thanks.

Chin-Ting

> Thanks,
> 
> C.
> 
> 
> > @@ -561,6 +557,9 @@ static int aspeed_spi_dirmap_create(struct
> spi_mem_dirmap_desc *desc)
> >   		       plat->ctrl_base + REG_CE0_CTRL_REG + cs * 4);
> >   		priv->flashes[cs].ce_ctrl_read = reg_val;
> >
> > +		if (op_tmpl.addr.nbytes == 4)
> > +			priv->info->set_4byte(bus, cs);
> > +
> >   		dev_dbg(dev, "read bus width: %d [0x%08x]\n",
> >   			op_tmpl.data.buswidth, priv->flashes[cs].ce_ctrl_read);
> >   	} else {