2017年3月15日 08:23于 André Przywara andre.przywara@arm.com写道：

On 13/03/17 17:50, Icenowy Zheng wrote:

...

The DesignWare DRAM controller used by H3 and newer SoCs use a bit to identify whether the DRAM is half-width.

As H3 itself come with 32-bit DRAM, the two modes of the bit used to be named "MCTL_CR_32BIT" and "MCTL_CR_16BIT", but for SoCs with 16-bit DRAM they're really 8-bit and 16-bit.

Rename the bit's macro, and also rename the variable name in dram_sun8i_h3.c.

This commit do not add 16-bit DRAM controller support, but the support will be introduced in next commit.

Signed-off-by: Icenowy Zheng icenowy@aosc.xyz

arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h |  6 +++---   arch/arm/mach-sunxi/dram_sunxi_dw.c             | 11 ++++++-----   2 files changed, 9 insertions(+), 8 deletions(-)

diff --git a/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h b/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h index 25d07d9863..48bd6f7c0f 100644 --- a/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h +++ b/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h @@ -52,9 +52,9 @@ struct sunxi_mctl_com_reg {   #define MCTL_CR_SEQUENTIAL (0x1 << 15)   #define MCTL_CR_INTERLEAVED (0x0 << 15)   -#define MCTL_CR_32BIT (0x1 << 12) -#define MCTL_CR_16BIT (0x0 << 12) -#define MCTL_CR_BUS_WIDTH(x) ((x) == 32 ? MCTL_CR_32BIT : MCTL_CR_16BIT) +#define MCTL_CR_FULL_WIDTH (0x1 << 12) +#define MCTL_CR_HALF_WIDTH (0x0 << 12) +#define MCTL_CR_BUS_FULL_WIDTH(x) ((x) << 12)     #define MCTL_CR_PAGE_SIZE(x) ((fls(x) - 4) << 8)   #define MCTL_CR_ROW_BITS(x) (((x) - 1) << 4) diff --git a/arch/arm/mach-sunxi/dram_sunxi_dw.c b/arch/arm/mach-sunxi/dram_sunxi_dw.c index 9f7cc7fd4c..0c73a43075 100644 --- a/arch/arm/mach-sunxi/dram_sunxi_dw.c +++ b/arch/arm/mach-sunxi/dram_sunxi_dw.c @@ -28,7 +28,7 @@   #define LINES_PER_BYTE_LANE (BITS_PER_BYTE + 3)   struct dram_para {   u16 page_size;

• u8 bus_width;

• u8 bus_full_width;

So this could be a "bool" then?

Yes, thanks for this tips.

And wouldn't it be better to _add_ this variable and keep bus width, so it could be re-purposed as the actual bus width in the next patch?

The controller itself is configured with half/full width info, but not real bus widh.

The rest of the patch is fine.

Cheers, Andre.

...

u8 dual_rank;   u8 row_bits;   const u8 dx_read_delays[NR_OF_BYTE_LANES][LINES_PER_BYTE_LANE]; @@ -358,7 +358,8 @@ static void mctl_set_cr(struct dram_para *para)   (struct sunxi_mctl_com_reg *)SUNXI_DRAM_COM_BASE;     writel(MCTL_CR_BL8 | MCTL_CR_2T | MCTL_CR_DDR3 | MCTL_CR_INTERLEAVED |

• MCTL_CR_EIGHT_BANKS | MCTL_CR_BUS_WIDTH(para->bus_width) |

• MCTL_CR_EIGHT_BANKS |
• MCTL_CR_BUS_FULL_WIDTH(para->bus_full_width) |

(para->dual_rank ? MCTL_CR_DUAL_RANK : MCTL_CR_SINGLE_RANK) |          MCTL_CR_PAGE_SIZE(para->page_size) |          MCTL_CR_ROW_BITS(para->row_bits), &mctl_com->cr); @@ -471,7 +472,7 @@ static int mctl_channel_init(uint16_t socid, struct dram_para *para)   }     /* set half DQ */

• if (para->bus_width != 32) {

• if (!para->bus_full_width) {

writel(0x0, &mctl_ctl->dx[2].gcr);   writel(0x0, &mctl_ctl->dx[3].gcr);   } @@ -509,7 +510,7 @@ static int mctl_channel_init(uint16_t socid, struct dram_para *para)       ((readl(&mctl_ctl->dx[3].gsr[0]) >> 24) & 0x1)) {   writel(0x0, &mctl_ctl->dx[2].gcr);   writel(0x0, &mctl_ctl->dx[3].gcr);

• para->bus_width = 16;

• para->bus_full_width = 0;

}     mctl_set_cr(para); @@ -613,7 +614,7 @@ unsigned long sunxi_dram_init(void)     struct dram_para para = {   .dual_rank = 0,

• .bus_width = 32,

• .bus_full_width = 1,

.row_bits = 15,   .page_size = 4096,

-- You received this message because you are subscribed to the Google Groups "linux-sunxi" group. To unsubscribe from this group and stop receiving emails from it, send an email to linux-sunxi+unsubscribe@googlegroups.com. For more options, visit https://groups.google.com/d/optout.