[PATCH 2/4] pinephone_defconfig: reduce boot delay

[PATCH 2/4] pinephone_defconfig: reduce boot delay

[Nicolas Boulenguez]

From: Arnaud Ferraris <arnaud.ferraris@collabora.com>

On a cellular phone, the vast majority of users can be expected to
have no serial console connected and prefer a short boot.

---
 configs/pinephone_defconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/configs/pinephone_defconfig b/configs/pinephone_defconfig
index 9967cddab9..9b51a7a100 100644
--- a/configs/pinephone_defconfig
+++ b/configs/pinephone_defconfig
@@ -10,4 +10,5 @@
 CONFIG_PINEPHONE_DT_SELECTION=y
 CONFIG_DEFAULT_DEVICE_TREE="sun50i-a64-pinephone-1.2"
 # CONFIG_SYS_MALLOC_CLEAR_ON_INIT is not set
+CONFIG_BOOTDELAY=0
 CONFIG_OF_LIST="sun50i-a64-pinephone-1.1 sun50i-a64-pinephone-1.2"
-- 
2.20.1

[PATCH 20/20] Enable led on boot to notify user of boot status

[Nicolas Boulenguez]

From: Marius Gripsgard <marius@ubports.com>

---
 arch/arm/mach-sunxi/Kconfig | 5 +++++
 board/sunxi/board.c         | 4 ++--
 configs/pinephone_defconfig | 1 +
 3 files changed, 8 insertions(+), 2 deletions(-)

diff --git a/arch/arm/mach-sunxi/Kconfig b/arch/arm/mach-sunxi/Kconfig
index 8421f3b685..2bfdf7738b 100644
--- a/arch/arm/mach-sunxi/Kconfig
+++ b/arch/arm/mach-sunxi/Kconfig
@@ -1,5 +1,10 @@
 if ARCH_SUNXI
 
+config PINEPHONE_LEDS
+	bool "Notify boot status via LEDs on PinePhone"
+	---help---
+	LED boot notification.
+
 config SPL_LDSCRIPT
 	default "arch/arm/cpu/armv7/sunxi/u-boot-spl.lds" if !ARM64
 
diff --git a/board/sunxi/board.c b/board/sunxi/board.c
index abd7e390b2..a117b89ba2 100644
--- a/board/sunxi/board.c
+++ b/board/sunxi/board.c
@@ -637,6 +638,12 @@ void sunxi_board_init(void)
 {
 	int power_failed = 0;
 
+#ifdef CONFIG_PINEPHONE_LEDS
+	/* PD18:G PD19:R PD20:B */
+	gpio_request(SUNXI_GPD(18), "led:green");
+	gpio_direction_output(SUNXI_GPD(18), 1);
+#endif
+
 #ifdef CONFIG_SY8106A_POWER
 	power_failed = sy8106a_set_vout1(CONFIG_SY8106A_VOUT1_VOLT);
 #endif
diff --git a/configs/pinephone_defconfig b/configs/pinephone_defconfig
index ff5da42ce0..9de6ab2316 100644
--- a/configs/pinephone_defconfig
+++ b/configs/pinephone_defconfig
@@ -1,6 +21,7 @@
 CONFIG_ARM=y
 CONFIG_ARCH_SUNXI=y
 CONFIG_SPL=y
+CONFIG_PINEPHONE_LEDS=y
 CONFIG_MACH_SUN50I=y
 CONFIG_SUNXI_DRAM_LPDDR3_STOCK=y
 CONFIG_DRAM_CLK=552
-- 
2.26.2

[PATCH 20/20] Enable led on boot to notify user of boot status

[André Przywara]

On 20/02/2021 12:14, Nicolas Boulenguez wrote:
> From: Marius Gripsgard <marius@ubports.com>

Hi,

This is not really Pinephone specific, actually not even sunxi specific.
I see two better ways of solving this:

- We introduce some generic code to find "gpio-leds" subnodes in the DT,
which have a default-state = "on" property. This could be either added
to drivers/led/led_gpio.c, or in some higher level.

- We introduce a generic boot LED GPIO config symbol. Maybe there is
already something, that sounds like a generally useful feature? That
would have the advantage of being already enabled in the SPL (admittedly
a sunxi specific problem).

Any opinions?

Cheers,
Andre

> ---
>  arch/arm/mach-sunxi/Kconfig | 5 +++++
>  board/sunxi/board.c         | 4 ++--
>  configs/pinephone_defconfig | 1 +
>  3 files changed, 8 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/arm/mach-sunxi/Kconfig b/arch/arm/mach-sunxi/Kconfig
> index 8421f3b685..2bfdf7738b 100644
> --- a/arch/arm/mach-sunxi/Kconfig
> +++ b/arch/arm/mach-sunxi/Kconfig
> @@ -1,5 +1,10 @@
>  if ARCH_SUNXI
>  
> +config PINEPHONE_LEDS
> +	bool "Notify boot status via LEDs on PinePhone"
> +	---help---
> +	LED boot notification.
> +
>  config SPL_LDSCRIPT
>  	default "arch/arm/cpu/armv7/sunxi/u-boot-spl.lds" if !ARM64
>  
> diff --git a/board/sunxi/board.c b/board/sunxi/board.c
> index abd7e390b2..a117b89ba2 100644
> --- a/board/sunxi/board.c
> +++ b/board/sunxi/board.c
> @@ -637,6 +638,12 @@ void sunxi_board_init(void)
>  {
>  	int power_failed = 0;
>  
> +#ifdef CONFIG_PINEPHONE_LEDS
> +	/* PD18:G PD19:R PD20:B */
> +	gpio_request(SUNXI_GPD(18), "led:green");
> +	gpio_direction_output(SUNXI_GPD(18), 1);
> +#endif
> +
>  #ifdef CONFIG_SY8106A_POWER
>  	power_failed = sy8106a_set_vout1(CONFIG_SY8106A_VOUT1_VOLT);
>  #endif
> diff --git a/configs/pinephone_defconfig b/configs/pinephone_defconfig
> index ff5da42ce0..9de6ab2316 100644
> --- a/configs/pinephone_defconfig
> +++ b/configs/pinephone_defconfig
> @@ -1,6 +21,7 @@
>  CONFIG_ARM=y
>  CONFIG_ARCH_SUNXI=y
>  CONFIG_SPL=y
> +CONFIG_PINEPHONE_LEDS=y
>  CONFIG_MACH_SUN50I=y
>  CONFIG_SUNXI_DRAM_LPDDR3_STOCK=y
>  CONFIG_DRAM_CLK=552
> 

[PATCH 20/20] Enable led on boot to notify user of boot status

[Peter Robinson]

On Thu, Feb 25, 2021 at 3:17 PM Andr? Przywara <andre.przywara@arm.com> wrote:
>
> On 20/02/2021 12:14, Nicolas Boulenguez wrote:
> > From: Marius Gripsgard <marius@ubports.com>
>
> Hi,
>
> This is not really Pinephone specific, actually not even sunxi specific.
> I see two better ways of solving this:
>
> - We introduce some generic code to find "gpio-leds" subnodes in the DT,
> which have a default-state = "on" property. This could be either added
> to drivers/led/led_gpio.c, or in some higher level.
>
> - We introduce a generic boot LED GPIO config symbol. Maybe there is
> already something, that sounds like a generally useful feature? That
> would have the advantage of being already enabled in the SPL (admittedly
> a sunxi specific problem).
>
> Any opinions?

I believe there's already a way to use LEDs for status during the boot
process using the CONFIG_LED_STATUS selection of config options. A
grep of the defconfig already shows a number of boards using this. I
wonder why the already existing infrastructure isn't useful.

Peter

> Cheers,
> Andre
>
> > ---
> >  arch/arm/mach-sunxi/Kconfig | 5 +++++
> >  board/sunxi/board.c         | 4 ++--
> >  configs/pinephone_defconfig | 1 +
> >  3 files changed, 8 insertions(+), 2 deletions(-)
> >
> > diff --git a/arch/arm/mach-sunxi/Kconfig b/arch/arm/mach-sunxi/Kconfig
> > index 8421f3b685..2bfdf7738b 100644
> > --- a/arch/arm/mach-sunxi/Kconfig
> > +++ b/arch/arm/mach-sunxi/Kconfig
> > @@ -1,5 +1,10 @@
> >  if ARCH_SUNXI
> >
> > +config PINEPHONE_LEDS
> > +     bool "Notify boot status via LEDs on PinePhone"
> > +     ---help---
> > +     LED boot notification.
> > +
> >  config SPL_LDSCRIPT
> >       default "arch/arm/cpu/armv7/sunxi/u-boot-spl.lds" if !ARM64
> >
> > diff --git a/board/sunxi/board.c b/board/sunxi/board.c
> > index abd7e390b2..a117b89ba2 100644
> > --- a/board/sunxi/board.c
> > +++ b/board/sunxi/board.c
> > @@ -637,6 +638,12 @@ void sunxi_board_init(void)
> >  {
> >       int power_failed = 0;
> >
> > +#ifdef CONFIG_PINEPHONE_LEDS
> > +     /* PD18:G PD19:R PD20:B */
> > +     gpio_request(SUNXI_GPD(18), "led:green");
> > +     gpio_direction_output(SUNXI_GPD(18), 1);
> > +#endif
> > +
> >  #ifdef CONFIG_SY8106A_POWER
> >       power_failed = sy8106a_set_vout1(CONFIG_SY8106A_VOUT1_VOLT);
> >  #endif
> > diff --git a/configs/pinephone_defconfig b/configs/pinephone_defconfig
> > index ff5da42ce0..9de6ab2316 100644
> > --- a/configs/pinephone_defconfig
> > +++ b/configs/pinephone_defconfig
> > @@ -1,6 +21,7 @@
> >  CONFIG_ARM=y
> >  CONFIG_ARCH_SUNXI=y
> >  CONFIG_SPL=y
> > +CONFIG_PINEPHONE_LEDS=y
> >  CONFIG_MACH_SUN50I=y
> >  CONFIG_SUNXI_DRAM_LPDDR3_STOCK=y
> >  CONFIG_DRAM_CLK=552
> >
>

[PATCH] sunxi: support asymmetric dual rank DRAM on A64/R40

[Nicolas Boulenguez]

From: Icenowy Zheng <icenowy@aosc.io>

Previously we have known that R40 has a configuration register for its
rank 1, which allows different configuration than rank 0. Reverse
engineering of newest libdram of A64 from Allwinner shows that A64 has
this register too. It's bit 0 (which enables dual rank in rank 0
configuration register) means a dedicated rank size setup is used for
rank 1.

Now, Pine64 scheduled to use a 3GiB LPDDR3 DRAM chip (which has 2GiB
rank 0 and 1GiB rank 1) on PinePhone, that makes asymmetric dual rank
DRAM support necessary.

Add this support. As we have gained knowledge of asymmetric dual rank,
we can now allow R40 dual rank memory setup to work too.

Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
---
 .../include/asm/arch-sunxi/dram_sunxi_dw.h    |  11 +-
 arch/arm/mach-sunxi/dram_sunxi_dw.c           | 100 +++++++++++++-----
 2 files changed, 84 insertions(+), 27 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 a5a7ebde44..e843c14202 100644
--- a/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h
+++ b/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h
@@ -215,12 +215,17 @@ struct sunxi_mctl_ctl_reg {
 #define NR_OF_BYTE_LANES	(32 / BITS_PER_BYTE)
 /* The eight data lines (DQn) plus DM, DQS and DQSN */
 #define LINES_PER_BYTE_LANE	(BITS_PER_BYTE + 3)
-struct dram_para {
+
+struct rank_para {
 	u16 page_size;
-	u8 bus_full_width;
-	u8 dual_rank;
 	u8 row_bits;
 	u8 bank_bits;
+};
+
+struct dram_para {
+	u8 dual_rank;
+	u8 bus_full_width;
+	struct rank_para ranks[2];
 	const u8 dx_read_delays[NR_OF_BYTE_LANES][LINES_PER_BYTE_LANE];
 	const u8 dx_write_delays[NR_OF_BYTE_LANES][LINES_PER_BYTE_LANE];
 	const u8 ac_delays[31];
diff --git a/arch/arm/mach-sunxi/dram_sunxi_dw.c b/arch/arm/mach-sunxi/dram_sunxi_dw.c
index 85e7a1874e..b679f92e70 100644
--- a/arch/arm/mach-sunxi/dram_sunxi_dw.c
+++ b/arch/arm/mach-sunxi/dram_sunxi_dw.c
@@ -346,18 +346,24 @@ static void mctl_set_cr(uint16_t socid, struct dram_para *para)
 #else
 #error Unsupported DRAM type!
 #endif
-	       (para->bank_bits == 3 ? MCTL_CR_EIGHT_BANKS : MCTL_CR_FOUR_BANKS) |
+	       (para->ranks[0].bank_bits == 3 ? MCTL_CR_EIGHT_BANKS : MCTL_CR_FOUR_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);
+	       MCTL_CR_PAGE_SIZE(para->ranks[0].page_size) |
+	       MCTL_CR_ROW_BITS(para->ranks[0].row_bits), &mctl_com->cr);
 
-	if (socid == SOCID_R40) {
-		if (para->dual_rank)
-			panic("Dual rank memory not supported\n");
+	if (socid == SOCID_A64 || socid == SOCID_R40) {
+		writel((para->ranks[1].bank_bits == 3 ? MCTL_CR_EIGHT_BANKS : MCTL_CR_FOUR_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->ranks[1].page_size) |
+		       MCTL_CR_ROW_BITS(para->ranks[1].row_bits), &mctl_com->cr_r1);
+	}
 
+	if (socid == SOCID_R40) {
 		/* Mux pin to A15 address line for single rank memory. */
-		setbits_le32(&mctl_com->cr_r1, MCTL_CR_R1_MUX_A15);
+		if (!para->dual_rank)
+			setbits_le32(&mctl_com->cr_r1, MCTL_CR_R1_MUX_A15);
 	}
 }
 
@@ -581,35 +587,63 @@ static int mctl_channel_init(uint16_t socid, struct dram_para *para)
 	return 0;
 }
 
-static void mctl_auto_detect_dram_size(uint16_t socid, struct dram_para *para)
+/*
+ * Test if memory at offset offset matches memory at a certain base
+ */
+static bool mctl_mem_matches_base(u32 offset, ulong base)
+{
+	/* Try to write different values to RAM at two addresses */
+	writel(0, base);
+	writel(0xaa55aa55, base + offset);
+	dsb();
+	/* Check if the same value is actually observed when reading back */
+	return readl(base) ==
+	       readl(base + offset);
+}
+
+static void mctl_auto_detect_dram_size_rank(uint16_t socid, struct dram_para *para, ulong base, struct rank_para *rank)
 {
 	/* detect row address bits */
-	para->page_size = 512;
-	para->row_bits = 16;
-	para->bank_bits = 2;
+	rank->page_size = 512;
+	rank->row_bits = 16;
+	rank->bank_bits = 2;
 	mctl_set_cr(socid, para);
 
-	for (para->row_bits = 11; para->row_bits < 16; para->row_bits++)
-		if (mctl_mem_matches((1 << (para->row_bits + para->bank_bits)) * para->page_size))
+	for (rank->row_bits = 11; rank->row_bits < 16; rank->row_bits++)
+		if (mctl_mem_matches_base((1 << (rank->row_bits + rank->bank_bits)) * rank->page_size, base))
 			break;
 
 	/* detect bank address bits */
-	para->bank_bits = 3;
+	rank->bank_bits = 3;
 	mctl_set_cr(socid, para);
 
-	for (para->bank_bits = 2; para->bank_bits < 3; para->bank_bits++)
-		if (mctl_mem_matches((1 << para->bank_bits) * para->page_size))
+	for (rank->bank_bits = 2; rank->bank_bits < 3; rank->bank_bits++)
+		if (mctl_mem_matches_base((1 << rank->bank_bits) * rank->page_size, base))
 			break;
 
 	/* detect page size */
-	para->page_size = 8192;
+	rank->page_size = 8192;
 	mctl_set_cr(socid, para);
 
-	for (para->page_size = 512; para->page_size < 8192; para->page_size *= 2)
-		if (mctl_mem_matches(para->page_size))
+	for (rank->page_size = 512; rank->page_size < 8192; rank->page_size *= 2)
+		if (mctl_mem_matches_base(rank->page_size, base))
 			break;
 }
 
+static unsigned long mctl_calc_rank_size(struct rank_para *rank)
+{
+	return (1UL << (rank->row_bits + rank->bank_bits)) * rank->page_size;
+}
+
+static void mctl_auto_detect_dram_size(uint16_t socid, struct dram_para *para)
+{
+	mctl_auto_detect_dram_size_rank(socid, para, (ulong)CONFIG_SYS_SDRAM_BASE, &para->ranks[0]);
+
+	if ((socid == SOCID_A64 || socid == SOCID_R40) && para->dual_rank) {
+		mctl_auto_detect_dram_size_rank(socid, para, (ulong)CONFIG_SYS_SDRAM_BASE + mctl_calc_rank_size(&para->ranks[0]), &para->ranks[1]);
+	}
+}
+
 /*
  * The actual values used here are taken from Allwinner provided boot0
  * binaries, though they are probably board specific, so would likely benefit
@@ -688,12 +722,23 @@ unsigned long sunxi_dram_init(void)
 	struct sunxi_mctl_ctl_reg * const mctl_ctl =
 			(struct sunxi_mctl_ctl_reg *)SUNXI_DRAM_CTL0_BASE;
 
+	unsigned long size;
+
 	struct dram_para para = {
 		.dual_rank = 1,
 		.bus_full_width = 1,
-		.row_bits = 15,
-		.bank_bits = 3,
-		.page_size = 4096,
+		.ranks = {
+			{
+				.row_bits = 15,
+				.bank_bits = 3,
+				.page_size = 4096,
+			},
+			{
+				.row_bits = 15,
+				.bank_bits = 3,
+				.page_size = 4096,
+			}
+		},
 
 #if defined(CONFIG_MACH_SUN8I_H3)
 		.dx_read_delays  = SUN8I_H3_DX_READ_DELAYS,
@@ -762,6 +807,13 @@ unsigned long sunxi_dram_init(void)
 	mctl_auto_detect_dram_size(socid, &para);
 	mctl_set_cr(socid, &para);
 
-	return (1UL << (para.row_bits + para.bank_bits)) * para.page_size *
-	       (para.dual_rank ? 2 : 1);
+	size = mctl_calc_rank_size(&para.ranks[0]);
+	if (socid == SOCID_A64 || socid == SOCID_R40) {
+		if (para.dual_rank)
+			size += mctl_calc_rank_size(&para.ranks[1]);
+	} else if (para.dual_rank) {
+		size *= 2;
+	}
+
+	return size;
 }
-- 
2.27.0

[PATCH] sunxi: support asymmetric dual rank DRAM on A64/R40

[André Przywara]

On 20/02/2021 12:14, Nicolas Boulenguez wrote:

Hi,

> From: Icenowy Zheng <icenowy@aosc.io>
> 
> Previously we have known that R40 has a configuration register for its
> rank 1, which allows different configuration than rank 0. Reverse
> engineering of newest libdram of A64 from Allwinner shows that A64 has
> this register too. It's bit 0 (which enables dual rank in rank 0
> configuration register) means a dedicated rank size setup is used for
> rank 1.
> 
> Now, Pine64 scheduled to use a 3GiB LPDDR3 DRAM chip (which has 2GiB
> rank 0 and 1GiB rank 1) on PinePhone, that makes asymmetric dual rank
> DRAM support necessary.
> 
> Add this support. As we have gained knowledge of asymmetric dual rank,
> we can now allow R40 dual rank memory setup to work too.

Icenowy just sent a new version, together with another patch. I will
review her version, so you can drop this one here.

Cheers,
Andre

> 
> Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
> ---
>  .../include/asm/arch-sunxi/dram_sunxi_dw.h    |  11 +-
>  arch/arm/mach-sunxi/dram_sunxi_dw.c           | 100 +++++++++++++-----
>  2 files changed, 84 insertions(+), 27 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 a5a7ebde44..e843c14202 100644
> --- a/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h
> +++ b/arch/arm/include/asm/arch-sunxi/dram_sunxi_dw.h
> @@ -215,12 +215,17 @@ struct sunxi_mctl_ctl_reg {
>  #define NR_OF_BYTE_LANES	(32 / BITS_PER_BYTE)
>  /* The eight data lines (DQn) plus DM, DQS and DQSN */
>  #define LINES_PER_BYTE_LANE	(BITS_PER_BYTE + 3)
> -struct dram_para {
> +
> +struct rank_para {
>  	u16 page_size;
> -	u8 bus_full_width;
> -	u8 dual_rank;
>  	u8 row_bits;
>  	u8 bank_bits;
> +};
> +
> +struct dram_para {
> +	u8 dual_rank;
> +	u8 bus_full_width;
> +	struct rank_para ranks[2];
>  	const u8 dx_read_delays[NR_OF_BYTE_LANES][LINES_PER_BYTE_LANE];
>  	const u8 dx_write_delays[NR_OF_BYTE_LANES][LINES_PER_BYTE_LANE];
>  	const u8 ac_delays[31];
> diff --git a/arch/arm/mach-sunxi/dram_sunxi_dw.c b/arch/arm/mach-sunxi/dram_sunxi_dw.c
> index 85e7a1874e..b679f92e70 100644
> --- a/arch/arm/mach-sunxi/dram_sunxi_dw.c
> +++ b/arch/arm/mach-sunxi/dram_sunxi_dw.c
> @@ -346,18 +346,24 @@ static void mctl_set_cr(uint16_t socid, struct dram_para *para)
>  #else
>  #error Unsupported DRAM type!
>  #endif
> -	       (para->bank_bits == 3 ? MCTL_CR_EIGHT_BANKS : MCTL_CR_FOUR_BANKS) |
> +	       (para->ranks[0].bank_bits == 3 ? MCTL_CR_EIGHT_BANKS : MCTL_CR_FOUR_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);
> +	       MCTL_CR_PAGE_SIZE(para->ranks[0].page_size) |
> +	       MCTL_CR_ROW_BITS(para->ranks[0].row_bits), &mctl_com->cr);
>  
> -	if (socid == SOCID_R40) {
> -		if (para->dual_rank)
> -			panic("Dual rank memory not supported\n");
> +	if (socid == SOCID_A64 || socid == SOCID_R40) {
> +		writel((para->ranks[1].bank_bits == 3 ? MCTL_CR_EIGHT_BANKS : MCTL_CR_FOUR_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->ranks[1].page_size) |
> +		       MCTL_CR_ROW_BITS(para->ranks[1].row_bits), &mctl_com->cr_r1);
> +	}
>  
> +	if (socid == SOCID_R40) {
>  		/* Mux pin to A15 address line for single rank memory. */
> -		setbits_le32(&mctl_com->cr_r1, MCTL_CR_R1_MUX_A15);
> +		if (!para->dual_rank)
> +			setbits_le32(&mctl_com->cr_r1, MCTL_CR_R1_MUX_A15);
>  	}
>  }
>  
> @@ -581,35 +587,63 @@ static int mctl_channel_init(uint16_t socid, struct dram_para *para)
>  	return 0;
>  }
>  
> -static void mctl_auto_detect_dram_size(uint16_t socid, struct dram_para *para)
> +/*
> + * Test if memory at offset offset matches memory at a certain base
> + */
> +static bool mctl_mem_matches_base(u32 offset, ulong base)
> +{
> +	/* Try to write different values to RAM at two addresses */
> +	writel(0, base);
> +	writel(0xaa55aa55, base + offset);
> +	dsb();
> +	/* Check if the same value is actually observed when reading back */
> +	return readl(base) ==
> +	       readl(base + offset);
> +}
> +
> +static void mctl_auto_detect_dram_size_rank(uint16_t socid, struct dram_para *para, ulong base, struct rank_para *rank)
>  {
>  	/* detect row address bits */
> -	para->page_size = 512;
> -	para->row_bits = 16;
> -	para->bank_bits = 2;
> +	rank->page_size = 512;
> +	rank->row_bits = 16;
> +	rank->bank_bits = 2;
>  	mctl_set_cr(socid, para);
>  
> -	for (para->row_bits = 11; para->row_bits < 16; para->row_bits++)
> -		if (mctl_mem_matches((1 << (para->row_bits + para->bank_bits)) * para->page_size))
> +	for (rank->row_bits = 11; rank->row_bits < 16; rank->row_bits++)
> +		if (mctl_mem_matches_base((1 << (rank->row_bits + rank->bank_bits)) * rank->page_size, base))
>  			break;
>  
>  	/* detect bank address bits */
> -	para->bank_bits = 3;
> +	rank->bank_bits = 3;
>  	mctl_set_cr(socid, para);
>  
> -	for (para->bank_bits = 2; para->bank_bits < 3; para->bank_bits++)
> -		if (mctl_mem_matches((1 << para->bank_bits) * para->page_size))
> +	for (rank->bank_bits = 2; rank->bank_bits < 3; rank->bank_bits++)
> +		if (mctl_mem_matches_base((1 << rank->bank_bits) * rank->page_size, base))
>  			break;
>  
>  	/* detect page size */
> -	para->page_size = 8192;
> +	rank->page_size = 8192;
>  	mctl_set_cr(socid, para);
>  
> -	for (para->page_size = 512; para->page_size < 8192; para->page_size *= 2)
> -		if (mctl_mem_matches(para->page_size))
> +	for (rank->page_size = 512; rank->page_size < 8192; rank->page_size *= 2)
> +		if (mctl_mem_matches_base(rank->page_size, base))
>  			break;
>  }
>  
> +static unsigned long mctl_calc_rank_size(struct rank_para *rank)
> +{
> +	return (1UL << (rank->row_bits + rank->bank_bits)) * rank->page_size;
> +}
> +
> +static void mctl_auto_detect_dram_size(uint16_t socid, struct dram_para *para)
> +{
> +	mctl_auto_detect_dram_size_rank(socid, para, (ulong)CONFIG_SYS_SDRAM_BASE, &para->ranks[0]);
> +
> +	if ((socid == SOCID_A64 || socid == SOCID_R40) && para->dual_rank) {
> +		mctl_auto_detect_dram_size_rank(socid, para, (ulong)CONFIG_SYS_SDRAM_BASE + mctl_calc_rank_size(&para->ranks[0]), &para->ranks[1]);
> +	}
> +}
> +
>  /*
>   * The actual values used here are taken from Allwinner provided boot0
>   * binaries, though they are probably board specific, so would likely benefit
> @@ -688,12 +722,23 @@ unsigned long sunxi_dram_init(void)
>  	struct sunxi_mctl_ctl_reg * const mctl_ctl =
>  			(struct sunxi_mctl_ctl_reg *)SUNXI_DRAM_CTL0_BASE;
>  
> +	unsigned long size;
> +
>  	struct dram_para para = {
>  		.dual_rank = 1,
>  		.bus_full_width = 1,
> -		.row_bits = 15,
> -		.bank_bits = 3,
> -		.page_size = 4096,
> +		.ranks = {
> +			{
> +				.row_bits = 15,
> +				.bank_bits = 3,
> +				.page_size = 4096,
> +			},
> +			{
> +				.row_bits = 15,
> +				.bank_bits = 3,
> +				.page_size = 4096,
> +			}
> +		},
>  
>  #if defined(CONFIG_MACH_SUN8I_H3)
>  		.dx_read_delays  = SUN8I_H3_DX_READ_DELAYS,
> @@ -762,6 +807,13 @@ unsigned long sunxi_dram_init(void)
>  	mctl_auto_detect_dram_size(socid, &para);
>  	mctl_set_cr(socid, &para);
>  
> -	return (1UL << (para.row_bits + para.bank_bits)) * para.page_size *
> -	       (para.dual_rank ? 2 : 1);
> +	size = mctl_calc_rank_size(&para.ranks[0]);
> +	if (socid == SOCID_A64 || socid == SOCID_R40) {
> +		if (para.dual_rank)
> +			size += mctl_calc_rank_size(&para.ranks[1]);
> +	} else if (para.dual_rank) {
> +		size *= 2;
> +	}
> +
> +	return size;
>  }
> 

[PATCH] sunxi-common: support loading compressed kernel images on ARMv8

[Nicolas Boulenguez]

From: Arnaud Ferraris <arnaud.ferraris@gmail.com>

---
 include/configs/sunxi-common.h | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/include/configs/sunxi-common.h b/include/configs/sunxi-common.h
index 5b0bec0561..227284cd29 100644
--- a/include/configs/sunxi-common.h
+++ b/include/configs/sunxi-common.h
@@ -291,6 +291,8 @@ extern int soft_i2c_gpio_scl;
  */
 #define BOOTM_SIZE	__stringify(0xa000000)
 #define KERNEL_ADDR_R	__stringify(SDRAM_OFFSET(0080000))
+#define KERNEL_COMP_ADDR_R	__stringify(SDRAM_OFFSET(4000000))
+#define KERNEL_COMP_SIZE	__stringify(0x4000000)
 #define FDT_ADDR_R	__stringify(SDRAM_OFFSET(FA00000))
 #define SCRIPT_ADDR_R	__stringify(SDRAM_OFFSET(FC00000))
 #define PXEFILE_ADDR_R	__stringify(SDRAM_OFFSET(FD00000))
@@ -332,6 +334,18 @@ extern int soft_i2c_gpio_scl;
 	"pxefile_addr_r=" PXEFILE_ADDR_R "\0" \
 	"ramdisk_addr_r=" RAMDISK_ADDR_R "\0"
 
+#ifdef CONFIG_ARM64
+
+#define MEM_LAYOUT_ENV_EXTRA_SETTINGS \
+	"kernel_comp_addr_r=" KERNEL_COMP_ADDR_R "\0" \
+	"kernel_comp_size=" KERNEL_COMP_SIZE "\0"
+
+#else
+
+#define MEM_LAYOUT_ENV_EXTRA_SETTINGS ""
+
+#endif
+
 #define DFU_ALT_INFO_RAM \
 	"dfu_alt_info_ram=" \
 	"kernel ram " KERNEL_ADDR_R " 0x1000000;" \
@@ -487,6 +501,7 @@ extern int soft_i2c_gpio_scl;
 #define CONFIG_EXTRA_ENV_SETTINGS \
 	CONSOLE_ENV_SETTINGS \
 	MEM_LAYOUT_ENV_SETTINGS \
+	MEM_LAYOUT_ENV_EXTRA_SETTINGS \
 	DFU_ALT_INFO_RAM \
 	"fdtfile=" FDTFILE "\0" \
 	"console=ttyS0,115200\0" \
-- 
2.28.0

[PATCH] sunxi-common: support loading compressed kernel images on ARMv8

[André Przywara]

On 20/02/2021 12:14, Nicolas Boulenguez wrote:

Hi,

> From: Arnaud Ferraris <arnaud.ferraris@gmail.com>

This looks OK on the first glance, but is missing a commit message and a
Signed-off-by:

Cheers,
Andre

> 
> ---
>  include/configs/sunxi-common.h | 15 +++++++++++++++
>  1 file changed, 15 insertions(+)
> 
> diff --git a/include/configs/sunxi-common.h b/include/configs/sunxi-common.h
> index 5b0bec0561..227284cd29 100644
> --- a/include/configs/sunxi-common.h
> +++ b/include/configs/sunxi-common.h
> @@ -291,6 +291,8 @@ extern int soft_i2c_gpio_scl;
>   */
>  #define BOOTM_SIZE	__stringify(0xa000000)
>  #define KERNEL_ADDR_R	__stringify(SDRAM_OFFSET(0080000))
> +#define KERNEL_COMP_ADDR_R	__stringify(SDRAM_OFFSET(4000000))
> +#define KERNEL_COMP_SIZE	__stringify(0x4000000)
>  #define FDT_ADDR_R	__stringify(SDRAM_OFFSET(FA00000))
>  #define SCRIPT_ADDR_R	__stringify(SDRAM_OFFSET(FC00000))
>  #define PXEFILE_ADDR_R	__stringify(SDRAM_OFFSET(FD00000))
> @@ -332,6 +334,18 @@ extern int soft_i2c_gpio_scl;
>  	"pxefile_addr_r=" PXEFILE_ADDR_R "\0" \
>  	"ramdisk_addr_r=" RAMDISK_ADDR_R "\0"
>  
> +#ifdef CONFIG_ARM64
> +
> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS \
> +	"kernel_comp_addr_r=" KERNEL_COMP_ADDR_R "\0" \
> +	"kernel_comp_size=" KERNEL_COMP_SIZE "\0"
> +
> +#else
> +
> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS ""
> +
> +#endif
> +
>  #define DFU_ALT_INFO_RAM \
>  	"dfu_alt_info_ram=" \
>  	"kernel ram " KERNEL_ADDR_R " 0x1000000;" \
> @@ -487,6 +501,7 @@ extern int soft_i2c_gpio_scl;
>  #define CONFIG_EXTRA_ENV_SETTINGS \
>  	CONSOLE_ENV_SETTINGS \
>  	MEM_LAYOUT_ENV_SETTINGS \
> +	MEM_LAYOUT_ENV_EXTRA_SETTINGS \
>  	DFU_ALT_INFO_RAM \
>  	"fdtfile=" FDTFILE "\0" \
>  	"console=ttyS0,115200\0" \
> 

[PATCH] sunxi-common: support loading compressed kernel images on ARMv8

[Andre Przywara]

On 20/02/2021 12:14, Nicolas Boulenguez wrote:

Hi Arnaud,

> From: Arnaud Ferraris <arnaud.ferraris@gmail.com>

as mentioned before, this looks useful, so can you re-send this with a
commit message and your Signed-off-by:?

Also I was wondering if you could increase the COMP_SIZE? The 64MB look
rather arbitrary and maybe unnecessarily limiting. I think we could go
up till 0xfa00000, but maybe should stop at 0xf800000, to leave some
space for further extensions?

Cheers,
Andre

> 
> ---
>  include/configs/sunxi-common.h | 15 +++++++++++++++
>  1 file changed, 15 insertions(+)
> 
> diff --git a/include/configs/sunxi-common.h b/include/configs/sunxi-common.h
> index 5b0bec0561..227284cd29 100644
> --- a/include/configs/sunxi-common.h
> +++ b/include/configs/sunxi-common.h
> @@ -291,6 +291,8 @@ extern int soft_i2c_gpio_scl;
>   */
>  #define BOOTM_SIZE	__stringify(0xa000000)
>  #define KERNEL_ADDR_R	__stringify(SDRAM_OFFSET(0080000))
> +#define KERNEL_COMP_ADDR_R	__stringify(SDRAM_OFFSET(4000000))
> +#define KERNEL_COMP_SIZE	__stringify(0x4000000)
>  #define FDT_ADDR_R	__stringify(SDRAM_OFFSET(FA00000))
>  #define SCRIPT_ADDR_R	__stringify(SDRAM_OFFSET(FC00000))
>  #define PXEFILE_ADDR_R	__stringify(SDRAM_OFFSET(FD00000))
> @@ -332,6 +334,18 @@ extern int soft_i2c_gpio_scl;
>  	"pxefile_addr_r=" PXEFILE_ADDR_R "\0" \
>  	"ramdisk_addr_r=" RAMDISK_ADDR_R "\0"
>  
> +#ifdef CONFIG_ARM64
> +
> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS \
> +	"kernel_comp_addr_r=" KERNEL_COMP_ADDR_R "\0" \
> +	"kernel_comp_size=" KERNEL_COMP_SIZE "\0"
> +
> +#else
> +
> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS ""
> +
> +#endif
> +
>  #define DFU_ALT_INFO_RAM \
>  	"dfu_alt_info_ram=" \
>  	"kernel ram " KERNEL_ADDR_R " 0x1000000;" \
> @@ -487,6 +501,7 @@ extern int soft_i2c_gpio_scl;
>  #define CONFIG_EXTRA_ENV_SETTINGS \
>  	CONSOLE_ENV_SETTINGS \
>  	MEM_LAYOUT_ENV_SETTINGS \
> +	MEM_LAYOUT_ENV_EXTRA_SETTINGS \
>  	DFU_ALT_INFO_RAM \
>  	"fdtfile=" FDTFILE "\0" \
>  	"console=ttyS0,115200\0" \
> 

[PATCH] sunxi-common: support loading compressed kernel images on ARMv8

[Samuel Holland]

Andre,

On 3/31/21 10:39 AM, Andre Przywara wrote:
> On 20/02/2021 12:14, Nicolas Boulenguez wrote:
> 
> Hi Arnaud,
> 
>> From: Arnaud Ferraris <arnaud.ferraris@gmail.com>
> 
> as mentioned before, this looks useful, so can you re-send this with a
> commit message and your Signed-off-by:?

It looks like you picked up this patch regardless?

> Also I was wondering if you could increase the COMP_SIZE? The 64MB look
> rather arbitrary and maybe unnecessarily limiting. I think we could go
> up till 0xfa00000, but maybe should stop at 0xf800000, to leave some
> space for further extensions?

And you changed KERNEL_COMP_SIZE to 0xb000000. How does that work when
U-Boot itself is at 0x4a000000 (KERNEL_COMP_ADDR_R + 0x6000000)?

Samuel

> Cheers,
> Andre
> 
>>
>> ---
>>  include/configs/sunxi-common.h | 15 +++++++++++++++
>>  1 file changed, 15 insertions(+)
>>
>> diff --git a/include/configs/sunxi-common.h b/include/configs/sunxi-common.h
>> index 5b0bec0561..227284cd29 100644
>> --- a/include/configs/sunxi-common.h
>> +++ b/include/configs/sunxi-common.h
>> @@ -291,6 +291,8 @@ extern int soft_i2c_gpio_scl;
>>   */
>>  #define BOOTM_SIZE	__stringify(0xa000000)
>>  #define KERNEL_ADDR_R	__stringify(SDRAM_OFFSET(0080000))
>> +#define KERNEL_COMP_ADDR_R	__stringify(SDRAM_OFFSET(4000000))
>> +#define KERNEL_COMP_SIZE	__stringify(0x4000000)
>>  #define FDT_ADDR_R	__stringify(SDRAM_OFFSET(FA00000))
>>  #define SCRIPT_ADDR_R	__stringify(SDRAM_OFFSET(FC00000))
>>  #define PXEFILE_ADDR_R	__stringify(SDRAM_OFFSET(FD00000))
>> @@ -332,6 +334,18 @@ extern int soft_i2c_gpio_scl;
>>  	"pxefile_addr_r=" PXEFILE_ADDR_R "\0" \
>>  	"ramdisk_addr_r=" RAMDISK_ADDR_R "\0"
>>  
>> +#ifdef CONFIG_ARM64
>> +
>> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS \
>> +	"kernel_comp_addr_r=" KERNEL_COMP_ADDR_R "\0" \
>> +	"kernel_comp_size=" KERNEL_COMP_SIZE "\0"
>> +
>> +#else
>> +
>> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS ""

This can be defined to nothing. No empty string is needed.

>> +
>> +#endif
>> +
>>  #define DFU_ALT_INFO_RAM \
>>  	"dfu_alt_info_ram=" \
>>  	"kernel ram " KERNEL_ADDR_R " 0x1000000;" \
>> @@ -487,6 +501,7 @@ extern int soft_i2c_gpio_scl;
>>  #define CONFIG_EXTRA_ENV_SETTINGS \
>>  	CONSOLE_ENV_SETTINGS \
>>  	MEM_LAYOUT_ENV_SETTINGS \
>> +	MEM_LAYOUT_ENV_EXTRA_SETTINGS \
>>  	DFU_ALT_INFO_RAM \
>>  	"fdtfile=" FDTFILE "\0" \
>>  	"console=ttyS0,115200\0" \
>>
> 

[PATCH] sunxi-common: support loading compressed kernel images on ARMv8

[Andre Przywara]

On Sat, 17 Apr 2021 12:47:23 -0500
Samuel Holland <samuel@sholland.org> wrote:

Hi Samuel,

> On 3/31/21 10:39 AM, Andre Przywara wrote:
> > On 20/02/2021 12:14, Nicolas Boulenguez wrote:
> > 
> > Hi Arnaud,
> >   
> >> From: Arnaud Ferraris <arnaud.ferraris@gmail.com>  
> > 
> > as mentioned before, this looks useful, so can you re-send this with a
> > commit message and your Signed-off-by:?  
> 
> It looks like you picked up this patch regardless?

Yes, I asked him twice, without any reply. At the end I figured that
the danger of a copyright infringement is low given the nature of the
change, and I didn't feel this should block this useful feature. The
alternative would be to have sent it as a patch of mine, but I felt
that I should preserve his authorship, so took the patch anyway.
I got a reply from him on another patch, and it's not his first U-Boot
patch, so I think it's just an oversight (from the sender of the patch?)

Chen-Yu asked about the feature lately, so if you think that approach
is dodgy, we can revert it and take another version of the patch.

> > Also I was wondering if you could increase the COMP_SIZE? The 64MB look
> > rather arbitrary and maybe unnecessarily limiting. I think we could go
> > up till 0xfa00000, but maybe should stop at 0xf800000, to leave some
> > space for further extensions?  
> 
> And you changed KERNEL_COMP_SIZE to 0xb000000. How does that work when
> U-Boot itself is at 0x4a000000 (KERNEL_COMP_ADDR_R + 0x6000000)?

It's not anymore at this time, that's just the load address, and it
relocates itself to the end of DRAM very early. So basically the whole
memory except the very end is available.
Btw, the reason we confine everything to below 0x50000000 is just that
we need to concatenate the base (0x4....) with that offset, and cannot
use arithmetic, because that doesn't work in the Makefiles.

Cheers,
Andre

> >> ---
> >>  include/configs/sunxi-common.h | 15 +++++++++++++++
> >>  1 file changed, 15 insertions(+)
> >>
> >> diff --git a/include/configs/sunxi-common.h b/include/configs/sunxi-common.h
> >> index 5b0bec0561..227284cd29 100644
> >> --- a/include/configs/sunxi-common.h
> >> +++ b/include/configs/sunxi-common.h
> >> @@ -291,6 +291,8 @@ extern int soft_i2c_gpio_scl;
> >>   */
> >>  #define BOOTM_SIZE	__stringify(0xa000000)
> >>  #define KERNEL_ADDR_R	__stringify(SDRAM_OFFSET(0080000))
> >> +#define KERNEL_COMP_ADDR_R	__stringify(SDRAM_OFFSET(4000000))
> >> +#define KERNEL_COMP_SIZE	__stringify(0x4000000)
> >>  #define FDT_ADDR_R	__stringify(SDRAM_OFFSET(FA00000))
> >>  #define SCRIPT_ADDR_R	__stringify(SDRAM_OFFSET(FC00000))
> >>  #define PXEFILE_ADDR_R	__stringify(SDRAM_OFFSET(FD00000))
> >> @@ -332,6 +334,18 @@ extern int soft_i2c_gpio_scl;
> >>  	"pxefile_addr_r=" PXEFILE_ADDR_R "\0" \
> >>  	"ramdisk_addr_r=" RAMDISK_ADDR_R "\0"
> >>  
> >> +#ifdef CONFIG_ARM64
> >> +
> >> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS \
> >> +	"kernel_comp_addr_r=" KERNEL_COMP_ADDR_R "\0" \
> >> +	"kernel_comp_size=" KERNEL_COMP_SIZE "\0"
> >> +
> >> +#else
> >> +
> >> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS ""  
> 
> This can be defined to nothing. No empty string is needed.
> 
> >> +
> >> +#endif
> >> +
> >>  #define DFU_ALT_INFO_RAM \
> >>  	"dfu_alt_info_ram=" \
> >>  	"kernel ram " KERNEL_ADDR_R " 0x1000000;" \
> >> @@ -487,6 +501,7 @@ extern int soft_i2c_gpio_scl;
> >>  #define CONFIG_EXTRA_ENV_SETTINGS \
> >>  	CONSOLE_ENV_SETTINGS \
> >>  	MEM_LAYOUT_ENV_SETTINGS \
> >> +	MEM_LAYOUT_ENV_EXTRA_SETTINGS \
> >>  	DFU_ALT_INFO_RAM \
> >>  	"fdtfile=" FDTFILE "\0" \
> >>  	"console=ttyS0,115200\0" \
> >>  
> >   
> 

[PATCH] sunxi-common: support loading compressed kernel images on ARMv8

[Samuel Holland]

Andre,

On 4/17/21 1:23 PM, Andre Przywara wrote:
> On Sat, 17 Apr 2021 12:47:23 -0500 Samuel Holland <samuel@sholland.org> wrote:
>> On 3/31/21 10:39 AM, Andre Przywara wrote:
>>> On 20/02/2021 12:14, Nicolas Boulenguez wrote:
>>>
>>> Hi Arnaud,
>>>   
>>>> From: Arnaud Ferraris <arnaud.ferraris@gmail.com>  
>>>
>>> as mentioned before, this looks useful, so can you re-send this with a
>>> commit message and your Signed-off-by:?  
>>
>> It looks like you picked up this patch regardless?
> 
> Yes, I asked him twice, without any reply. At the end I figured that
> the danger of a copyright infringement is low given the nature of the
> change, and I didn't feel this should block this useful feature. The
> alternative would be to have sent it as a patch of mine, but I felt
> that I should preserve his authorship, so took the patch anyway.
> I got a reply from him on another patch, and it's not his first U-Boot
> patch, so I think it's just an oversight (from the sender of the patch?)
> 
> Chen-Yu asked about the feature lately, so if you think that approach
> is dodgy, we can revert it and take another version of the patch.

Okay, I'll trust your judgement here.

>>> Also I was wondering if you could increase the COMP_SIZE? The 64MB look
>>> rather arbitrary and maybe unnecessarily limiting. I think we could go
>>> up till 0xfa00000, but maybe should stop at 0xf800000, to leave some
>>> space for further extensions?  
>>
>> And you changed KERNEL_COMP_SIZE to 0xb000000. How does that work when
>> U-Boot itself is at 0x4a000000 (KERNEL_COMP_ADDR_R + 0x6000000)?
> 
> It's not anymore at this time, that's just the load address, and it
> relocates itself to the end of DRAM very early. So basically the whole
> memory except the very end is available.
>
> Btw, the reason we confine everything to below 0x50000000 is just that
> we need to concatenate the base (0x4....) with that offset, and cannot
> use arithmetic, because that doesn't work in the Makefiles.

Ah, that makes sense. So we're assuming that U-Boot relocates itself
above 0x50000000, and so nothing important remains inside the address
range used by sunxi-common.h.

Thanks for the explanation,
Samuel

> Cheers,
> Andre
> 
>>>> ---
>>>>  include/configs/sunxi-common.h | 15 +++++++++++++++
>>>>  1 file changed, 15 insertions(+)
>>>>
>>>> diff --git a/include/configs/sunxi-common.h b/include/configs/sunxi-common.h
>>>> index 5b0bec0561..227284cd29 100644
>>>> --- a/include/configs/sunxi-common.h
>>>> +++ b/include/configs/sunxi-common.h
>>>> @@ -291,6 +291,8 @@ extern int soft_i2c_gpio_scl;
>>>>   */
>>>>  #define BOOTM_SIZE	__stringify(0xa000000)
>>>>  #define KERNEL_ADDR_R	__stringify(SDRAM_OFFSET(0080000))
>>>> +#define KERNEL_COMP_ADDR_R	__stringify(SDRAM_OFFSET(4000000))
>>>> +#define KERNEL_COMP_SIZE	__stringify(0x4000000)
>>>>  #define FDT_ADDR_R	__stringify(SDRAM_OFFSET(FA00000))
>>>>  #define SCRIPT_ADDR_R	__stringify(SDRAM_OFFSET(FC00000))
>>>>  #define PXEFILE_ADDR_R	__stringify(SDRAM_OFFSET(FD00000))
>>>> @@ -332,6 +334,18 @@ extern int soft_i2c_gpio_scl;
>>>>  	"pxefile_addr_r=" PXEFILE_ADDR_R "\0" \
>>>>  	"ramdisk_addr_r=" RAMDISK_ADDR_R "\0"
>>>>  
>>>> +#ifdef CONFIG_ARM64
>>>> +
>>>> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS \
>>>> +	"kernel_comp_addr_r=" KERNEL_COMP_ADDR_R "\0" \
>>>> +	"kernel_comp_size=" KERNEL_COMP_SIZE "\0"
>>>> +
>>>> +#else
>>>> +
>>>> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS ""  
>>
>> This can be defined to nothing. No empty string is needed.
>>
>>>> +
>>>> +#endif
>>>> +
>>>>  #define DFU_ALT_INFO_RAM \
>>>>  	"dfu_alt_info_ram=" \
>>>>  	"kernel ram " KERNEL_ADDR_R " 0x1000000;" \
>>>> @@ -487,6 +501,7 @@ extern int soft_i2c_gpio_scl;
>>>>  #define CONFIG_EXTRA_ENV_SETTINGS \
>>>>  	CONSOLE_ENV_SETTINGS \
>>>>  	MEM_LAYOUT_ENV_SETTINGS \
>>>> +	MEM_LAYOUT_ENV_EXTRA_SETTINGS \
>>>>  	DFU_ALT_INFO_RAM \
>>>>  	"fdtfile=" FDTFILE "\0" \
>>>>  	"console=ttyS0,115200\0" \
>>>>  
>>>   
>>
> 

Re: [PATCH] sunxi-common: support loading compressed kernel images on ARMv8

[Arnaud Ferraris]

Hi Andre,

Le 17/04/2021 à 20:23, Andre Przywara a écrit :
> On Sat, 17 Apr 2021 12:47:23 -0500
> Samuel Holland <samuel@sholland.org> wrote:
> 
> Hi Samuel,
> 
>> On 3/31/21 10:39 AM, Andre Przywara wrote:
>>> On 20/02/2021 12:14, Nicolas Boulenguez wrote:
>>>
>>> Hi Arnaud,
>>>   
>>>> From: Arnaud Ferraris <arnaud.ferraris@gmail.com>  
>>>
>>> as mentioned before, this looks useful, so can you re-send this with a
>>> commit message and your Signed-off-by:?  
>>
>> It looks like you picked up this patch regardless?
> 
> Yes, I asked him twice, without any reply. At the end I figured that
> the danger of a copyright infringement is low given the nature of the
> change, and I didn't feel this should block this useful feature. The
> alternative would be to have sent it as a patch of mine, but I felt
> that I should preserve his authorship, so took the patch anyway.
> I got a reply from him on another patch, and it's not his first U-Boot
> patch, so I think it's just an oversight (from the sender of the patch?)

Thanks for picking up and reworking this patch, and my apologies for not
replying earlier: I had personal stuff to deal with for a few months,
and am only now able to fully process my emails backlog.

I should be more reactive from now on, and might have a few more patches
for you in the near future.

All the best,
Arnaud

> 
> Chen-Yu asked about the feature lately, so if you think that approach
> is dodgy, we can revert it and take another version of the patch.
> 
>>> Also I was wondering if you could increase the COMP_SIZE? The 64MB look
>>> rather arbitrary and maybe unnecessarily limiting. I think we could go
>>> up till 0xfa00000, but maybe should stop at 0xf800000, to leave some
>>> space for further extensions?  
>>
>> And you changed KERNEL_COMP_SIZE to 0xb000000. How does that work when
>> U-Boot itself is at 0x4a000000 (KERNEL_COMP_ADDR_R + 0x6000000)?
> 
> It's not anymore at this time, that's just the load address, and it
> relocates itself to the end of DRAM very early. So basically the whole
> memory except the very end is available.
> Btw, the reason we confine everything to below 0x50000000 is just that
> we need to concatenate the base (0x4....) with that offset, and cannot
> use arithmetic, because that doesn't work in the Makefiles.
> 
> Cheers,
> Andre
> 
>>>> ---
>>>>  include/configs/sunxi-common.h | 15 +++++++++++++++
>>>>  1 file changed, 15 insertions(+)
>>>>
>>>> diff --git a/include/configs/sunxi-common.h b/include/configs/sunxi-common.h
>>>> index 5b0bec0561..227284cd29 100644
>>>> --- a/include/configs/sunxi-common.h
>>>> +++ b/include/configs/sunxi-common.h
>>>> @@ -291,6 +291,8 @@ extern int soft_i2c_gpio_scl;
>>>>   */
>>>>  #define BOOTM_SIZE	__stringify(0xa000000)
>>>>  #define KERNEL_ADDR_R	__stringify(SDRAM_OFFSET(0080000))
>>>> +#define KERNEL_COMP_ADDR_R	__stringify(SDRAM_OFFSET(4000000))
>>>> +#define KERNEL_COMP_SIZE	__stringify(0x4000000)
>>>>  #define FDT_ADDR_R	__stringify(SDRAM_OFFSET(FA00000))
>>>>  #define SCRIPT_ADDR_R	__stringify(SDRAM_OFFSET(FC00000))
>>>>  #define PXEFILE_ADDR_R	__stringify(SDRAM_OFFSET(FD00000))
>>>> @@ -332,6 +334,18 @@ extern int soft_i2c_gpio_scl;
>>>>  	"pxefile_addr_r=" PXEFILE_ADDR_R "\0" \
>>>>  	"ramdisk_addr_r=" RAMDISK_ADDR_R "\0"
>>>>  
>>>> +#ifdef CONFIG_ARM64
>>>> +
>>>> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS \
>>>> +	"kernel_comp_addr_r=" KERNEL_COMP_ADDR_R "\0" \
>>>> +	"kernel_comp_size=" KERNEL_COMP_SIZE "\0"
>>>> +
>>>> +#else
>>>> +
>>>> +#define MEM_LAYOUT_ENV_EXTRA_SETTINGS ""  
>>
>> This can be defined to nothing. No empty string is needed.
>>
>>>> +
>>>> +#endif
>>>> +
>>>>  #define DFU_ALT_INFO_RAM \
>>>>  	"dfu_alt_info_ram=" \
>>>>  	"kernel ram " KERNEL_ADDR_R " 0x1000000;" \
>>>> @@ -487,6 +501,7 @@ extern int soft_i2c_gpio_scl;
>>>>  #define CONFIG_EXTRA_ENV_SETTINGS \
>>>>  	CONSOLE_ENV_SETTINGS \
>>>>  	MEM_LAYOUT_ENV_SETTINGS \
>>>> +	MEM_LAYOUT_ENV_EXTRA_SETTINGS \
>>>>  	DFU_ALT_INFO_RAM \
>>>>  	"fdtfile=" FDTFILE "\0" \
>>>>  	"console=ttyS0,115200\0" \
>>>>  
>>>   
>>
> 

[PATCH] arm: dts: sun50i-a64-pinephone: fix random wifi mac address

[Nicolas Boulenguez]

From: Arnaud Ferraris <arnaud.ferraris@collabora.com>

The wi-fi mac-address was previously changing at every reboot.

---
 arch/arm/dts/sun50i-a64-pinephone.dtsi | 22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)

diff --git a/arch/arm/dts/sun50i-a64-pinephone.dtsi b/arch/arm/dts/sun50i-a64-pinephone.dtsi
index 25150aba74..86d55e38d8 100644
--- a/arch/arm/dts/sun50i-a64-pinephone.dtsi
+++ b/arch/arm/dts/sun50i-a64-pinephone.dtsi
@@ -14,6 +14,7 @@
 / {
 	aliases {
 		serial0 = &uart0;
+		ethernet0 = &rtl8723cs;
 	};
 
 	backlight: backlight {
@@ -49,6 +50,13 @@
 		};
 	};
 
+	reg_vbat_wifi: vbat-wifi {
+		compatible = "regulator-fixed";
+		regulator-min-microvolt = <3300000>;
+		regulator-max-microvolt = <3300000>;
+		regulator-name = "vbat-wifi";
+	};
+
 	speaker_amp: audio-amplifier {
 		compatible = "simple-audio-amplifier";
 		enable-gpios = <&pio 2 7 GPIO_ACTIVE_HIGH>; /* PC7 */
@@ -195,6 +203,20 @@
 	status = "okay";
 };
 
+&mmc1 {
+	pinctrl-names = "default";
+	pinctrl-0 = <&mmc1_pins>;
+	vmmc-supply = <&reg_vbat_wifi>;
+	vqmmc-supply = <&reg_dldo4>;
+	bus-width = <4>;
+	non-removable;
+	status = "okay";
+
+	rtl8723cs: wifi at 1 {
+		reg = <1>;
+	};
+};
+
 &mmc2 {
 	pinctrl-names = "default";
 	pinctrl-0 = <&mmc2_pins>;
-- 
2.29.2

[PATCH] arm: dts: sun50i-a64-pinephone: fix random wifi mac address

[André Przywara]

On 20/02/2021 12:14, Nicolas Boulenguez wrote:
> From: Arnaud Ferraris <arnaud.ferraris@collabora.com>
> 
> The wi-fi mac-address was previously changing at every reboot.

This is already part of the kernel .dtsi, which we will sync anyway
(with the new merge window). So we can drop this specific patch.

Cheers,
Andre

> 
> ---
>  arch/arm/dts/sun50i-a64-pinephone.dtsi | 22 ++++++++++++++++++++++
>  1 file changed, 22 insertions(+)
> 
> diff --git a/arch/arm/dts/sun50i-a64-pinephone.dtsi b/arch/arm/dts/sun50i-a64-pinephone.dtsi
> index 25150aba74..86d55e38d8 100644
> --- a/arch/arm/dts/sun50i-a64-pinephone.dtsi
> +++ b/arch/arm/dts/sun50i-a64-pinephone.dtsi
> @@ -14,6 +14,7 @@
>  / {
>  	aliases {
>  		serial0 = &uart0;
> +		ethernet0 = &rtl8723cs;
>  	};
>  
>  	backlight: backlight {
> @@ -49,6 +50,13 @@
>  		};
>  	};
>  
> +	reg_vbat_wifi: vbat-wifi {
> +		compatible = "regulator-fixed";
> +		regulator-min-microvolt = <3300000>;
> +		regulator-max-microvolt = <3300000>;
> +		regulator-name = "vbat-wifi";
> +	};
> +
>  	speaker_amp: audio-amplifier {
>  		compatible = "simple-audio-amplifier";
>  		enable-gpios = <&pio 2 7 GPIO_ACTIVE_HIGH>; /* PC7 */
> @@ -195,6 +203,20 @@
>  	status = "okay";
>  };
>  
> +&mmc1 {
> +	pinctrl-names = "default";
> +	pinctrl-0 = <&mmc1_pins>;
> +	vmmc-supply = <&reg_vbat_wifi>;
> +	vqmmc-supply = <&reg_dldo4>;
> +	bus-width = <4>;
> +	non-removable;
> +	status = "okay";
> +
> +	rtl8723cs: wifi at 1 {
> +		reg = <1>;
> +	};
> +};
> +
>  &mmc2 {
>  	pinctrl-names = "default";
>  	pinctrl-0 = <&mmc2_pins>;
> 

[PATCH 2/4] pinephone_defconfig: reduce boot delay

[André Przywara]

On 20/02/2021 12:14, Nicolas Boulenguez wrote:
> From: Arnaud Ferraris <arnaud.ferraris@collabora.com>
> 
> On a cellular phone, the vast majority of users can be expected to
> have no serial console connected and prefer a short boot.

It's a bit tricky to break in with a delay of 0, but indeed most users
won't care, so looks fine to me.

But it's missing a Signed-off-by: line.

Cheers,
Andre

> ---
>  configs/pinephone_defconfig | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/configs/pinephone_defconfig b/configs/pinephone_defconfig
> index 9967cddab9..9b51a7a100 100644
> --- a/configs/pinephone_defconfig
> +++ b/configs/pinephone_defconfig
> @@ -10,4 +10,5 @@
>  CONFIG_PINEPHONE_DT_SELECTION=y
>  CONFIG_DEFAULT_DEVICE_TREE="sun50i-a64-pinephone-1.2"
>  # CONFIG_SYS_MALLOC_CLEAR_ON_INIT is not set
> +CONFIG_BOOTDELAY=0
>  CONFIG_OF_LIST="sun50i-a64-pinephone-1.1 sun50i-a64-pinephone-1.2"
> 

[PATCH 2/4] pinephone_defconfig: reduce boot delay

[Maxime Ripard]

On Thu, Feb 25, 2021 at 05:02:40PM +0000, Andr? Przywara wrote:
> On 20/02/2021 12:14, Nicolas Boulenguez wrote:
> > From: Arnaud Ferraris <arnaud.ferraris@collabora.com>
> > 
> > On a cellular phone, the vast majority of users can be expected to
> > have no serial console connected and prefer a short boot.
> 
> It's a bit tricky to break in with a delay of 0, but indeed most users
> won't care, so looks fine to me.
> 
> But it's missing a Signed-off-by: line.

I'm not sure we should start accepting custom boards changes like it's
done here, this will only lead to inconsistencies between boards in the
long run.

And a defconfig is really easy to change anyway, even when it's
integrated in build systems

Maxime
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 228 bytes
Desc: not available
URL: <https://lists.denx.de/pipermail/u-boot/attachments/20210225/c584200a/attachment.sig>

[PATCH 2/4] pinephone_defconfig: reduce boot delay

[Andre Przywara]

On Thu, 25 Feb 2021 21:58:38 +0100
Maxime Ripard <maxime@cerno.tech> wrote:

Hi,

> On Thu, Feb 25, 2021 at 05:02:40PM +0000, Andr? Przywara wrote:
> > On 20/02/2021 12:14, Nicolas Boulenguez wrote:  
> > > From: Arnaud Ferraris <arnaud.ferraris@collabora.com>
> > > 
> > > On a cellular phone, the vast majority of users can be expected to
> > > have no serial console connected and prefer a short boot.  
> > 
> > It's a bit tricky to break in with a delay of 0, but indeed most users
> > won't care, so looks fine to me.
> > 
> > But it's missing a Signed-off-by: line.  
> 
> I'm not sure we should start accepting custom boards changes like it's
> done here, this will only lead to inconsistencies between boards in the
> long run.

I am a bit undecided, I see merits in both arguments.
 
> And a defconfig is really easy to change anyway, even when it's
> integrated in build systems

Nicolas, could you live with:

$ sed -i s/^CONFIG_BOOTDELAY=2/CONFIG_BOOTDELAY=0/ .config

in your build system? Or is the U-Boot build generic?

Cheers,
Andre

[PATCH 2/4] pinephone_defconfig: reduce boot delay

[Nicolas Boulenguez]

> Nicolas, could you live with:
> $ sed -i s/^CONFIG_BOOTDELAY=2/CONFIG_BOOTDELAY=0/ .config
> in your build system? Or is the U-Boot build generic?

No practical issue, the patch is already applied by Mobian.
My concern was only to reduce divergence between redistributors.

Anyway, thanks for asking before taking the decision.

[PATCH 2/4] pinephone_defconfig: reduce boot delay

[Arnaud Ferraris]

Hi Maxime,

Le 25/02/2021 ? 21:58, Maxime Ripard a ?crit?:
> On Thu, Feb 25, 2021 at 05:02:40PM +0000, Andr? Przywara wrote:
>> On 20/02/2021 12:14, Nicolas Boulenguez wrote:
>>> From: Arnaud Ferraris <arnaud.ferraris@collabora.com>
>>>
>>> On a cellular phone, the vast majority of users can be expected to
>>> have no serial console connected and prefer a short boot.
>>
>> It's a bit tricky to break in with a delay of 0, but indeed most users
>> won't care, so looks fine to me.
>>
>> But it's missing a Signed-off-by: line.
> 
> I'm not sure we should start accepting custom boards changes like it's
> done here, this will only lead to inconsistencies between boards in the
> long run.

The original reasoning behind this was, the PinePhone (and PineTab) is
both a dev board and a consumer device, which is why I assumed it could
use some kind of "special treatment".

However, I don't have a strong opinion on that matter and can certainly
live with the default boot delay.

Regards,
Arnaud

> 
> And a defconfig is really easy to change anyway, even when it's
> integrated in build systems
> 
> Maxime
>