Skip to contentSkip to conversation

sunxi: fix eMMC stability issues on A64

PATCH Superseded Vasily Khoruzhick 3 participants 13 messages Jun 6, 2018 – Jun 20, 2018

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

v5 cover sunxi: fix eMMC stability issues on A64

eMMC seems to require new clocking mode and calibration on A64, otherwise it is pretty unstable on some boards (e.g. Pinebook) with some eMMCs.

v2: - improve comment about calibration for eMMC on A64
- simplify ifdef-s around configuring delays
v3: - fix fallout due to ifdef simplification in v2
v4: - really fix ifdefs this time
v5: - add new option for SoCs without new mode switch in CCM

Vasily Khoruzhick (3):
sunxi-mmc: introduce new MMC_SUNXI_HAS_NEW_MODE_SWITCH option
sunxi-mmc: use new mode on A64
mmc: sunxi: run calibration on A64

 arch/arm/include/asm/arch-sunxi/mmc.h |  6 +++++-
 arch/arm/mach-sunxi/Kconfig           |  2 ++
 drivers/mmc/Kconfig                   |  4 ++++
 drivers/mmc/sunxi_mmc.c               | 13 +++++++++++++
 4 files changed, 24 insertions(+), 1 deletion(-)
Signature · 1 line
2.17.1
#1 lore raw

PATCH v5 1/3 sunxi-mmc: introduce new MMC_SUNXI_HAS_NEW_MODE_SWITCH option

A64 doesn't have a mode switch in CCM, so introduce new MMC_SUNXI_HAS_NEW_MODE_SWITCH option to support new mode on A64

---
 arch/arm/mach-sunxi/Kconfig | 1 +
 drivers/mmc/Kconfig         | 4 ++++
 drivers/mmc/sunxi_mmc.c     | 2 ++
 3 files changed, 7 insertions(+)
arch/arm/mach-sunxi/Kconfig +1 -0
index b9f778dd24..c4dc0f7672 100644
@@ -213,6 +213,7 @@ config MACH_SUN8I_A83T
 	select PHY_SUN4I_USB
 	select SUNXI_GEN_SUN6I
 	select MMC_SUNXI_HAS_NEW_MODE
+	select MMC_SUNXI_HAS_NEW_MODE_SWITCH
 	select SUPPORT_SPL
 
 config MACH_SUN8I_H3
drivers/mmc/Kconfig +4 -0
index 693b3ceaf0..f8e435a2a2 100644
@@ -536,6 +536,10 @@ config MMC_SUNXI_HAS_NEW_MODE
 	bool
 	depends on MMC_SUNXI
 
+config MMC_SUNXI_HAS_NEW_MODE_SWITCH
+	bool
+	depends on MMC_SUNXI
+
 config GENERIC_ATMEL_MCI
 	bool "Atmel Multimedia Card Interface support"
 	depends on DM_MMC && BLK && ARCH_AT91
drivers/mmc/sunxi_mmc.c +2 -0
index fe6d82c7b4..43c6d7c346 100644
@@ -166,7 +166,9 @@ static int mmc_set_mod_clk(struct sunxi_mmc_priv *priv, unsigned int hz)
 
 	if (new_mode) {
 #ifdef CONFIG_MMC_SUNXI_HAS_NEW_MODE
+#ifdef CONFIG_MMC_SUNXI_HAS_NEW_MODE_SWITCH
 		val = CCM_MMC_CTRL_MODE_SEL_NEW;
+#endif
 		setbits_le32(&priv->reg->ntsr, SUNXI_MMC_NTSR_MODE_SEL_NEW);
 #endif
 	} else {
Signature · 1 line
2.17.1
#2 lore raw

PATCH v5 2/3 sunxi-mmc: use new mode on A64

In reply to Vasily Khoruzhick

That is necessary for using automatic calibration on A64 eMMC.

---
 arch/arm/mach-sunxi/Kconfig | 1 +
 1 file changed, 1 insertion(+)
arch/arm/mach-sunxi/Kconfig +1 -0
index c4dc0f7672..fbba20ce10 100644
@@ -267,6 +267,7 @@ config MACH_SUN50I
 	select SUNXI_GEN_SUN6I
 	select SUN6I_PRCM
 	select SUNXI_HIGH_SRAM
+	select MMC_SUNXI_HAS_NEW_MODE
 	select SUPPORT_SPL
 	select SUNXI_DRAM_DW
 	select SUNXI_DRAM_DW_32BIT
Signature · 1 line
2.17.1
#3 lore raw

PATCH v5 3/3 mmc: sunxi: run calibration on A64

In reply to Vasily Khoruzhick

Along with using new mode it fixes eMMC instability on Pinebook

---
 arch/arm/include/asm/arch-sunxi/mmc.h |  6 +++++-
 drivers/mmc/sunxi_mmc.c               | 11 +++++++++++
 2 files changed, 16 insertions(+), 1 deletion(-)
arch/arm/include/asm/arch-sunxi/mmc.h +5 -1
index 1574b8e8fe..d6664a01f2 100644
@@ -46,7 +46,9 @@ struct sunxi_mmc {
 	u32 cbda;		/* 0x94 */
 	u32 res2[26];
 #ifdef CONFIG_SUNXI_GEN_SUN6I
-	u32 res3[64];
+	u32 res3[17];
+	u32 samp_dl;
+	u32 res4[46];
 #endif
 	u32 fifo;		/* 0x100 / 0x200 FIFO access address */
 };
@@ -130,5 +132,7 @@ struct sunxi_mmc {
 #define SUNXI_MMC_COMMON_CLK_GATE		(1 << 16)
 #define SUNXI_MMC_COMMON_RESET			(1 << 18)
 
+#define SUNXI_MMC_CAL_DL_SW_EN		(0x1 << 7)
+
 struct mmc *sunxi_mmc_init(int sdc_no);
 #endif /* _SUNXI_MMC_H */
drivers/mmc/sunxi_mmc.c +11 -0
index 43c6d7c346..624c907154 100644
@@ -225,6 +225,17 @@ static int mmc_config_clock(struct sunxi_mmc_priv *priv, struct mmc *mmc)
 	rval &= ~SUNXI_MMC_CLK_DIVIDER_MASK;
 	writel(rval, &priv->reg->clkcr);
 
+#ifdef CONFIG_MACH_SUN50I
+	/* A64 needs to run calibration on eMMC controller and we
+	 * have to set delay of zero before starting calibration.
+	 * Allwinner BSP driver sets a delay only in the case of
+	 * using HS400 which is not supported by mainline U-Boot or
+	 * Linux at the moment
+	 */
+	if (priv->mmc_no == 2)
+		writel(SUNXI_MMC_CAL_DL_SW_EN, &priv->reg->samp_dl);
+#endif
+
 	/* Re-enable Clock */
 	rval |= SUNXI_MMC_CLK_ENABLE;
 	writel(rval, &priv->reg->clkcr);
Signature · 1 line
2.17.1
#4 lore raw

v5 cover sunxi: fix eMMC stability issues on A64

In reply to Vasily Khoruzhick
Quoted 9 lines · Vasily Khoruzhick

eMMC seems to require new clocking mode and calibration on A64, otherwise it is pretty unstable on some boards (e.g. Pinebook) with some eMMCs.

v2: - improve comment about calibration for eMMC on A64
- simplify ifdef-s around configuring delays
v3: - fix fallout due to ifdef simplification in v2
v4: - really fix ifdefs this time
v5: - add new option for SoCs without new mode switch in CCM

Thanks!
Maxime

Signature · 10 lines
Maxime Ripard, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
https://bootlin.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20180606/c63a7d4c/attachment.sig>
#5 lore raw

v5 cover sunxi: fix eMMC stability issues on A64

In reply to Vasily Khoruzhick
Quoted 10 lines · Vasily Khoruzhick

eMMC seems to require new clocking mode and calibration on A64, otherwise it is pretty unstable on some boards (e.g. Pinebook) with some eMMCs.

v2: - improve comment about calibration for eMMC on A64
- simplify ifdef-s around configuring delays
v3: - fix fallout due to ifdef simplification in v2
v4: - really fix ifdefs this time
v5: - add new option for SoCs without new mode switch in CCM

Quoted 4 lines

Vasily Khoruzhick (3):
sunxi-mmc: introduce new MMC_SUNXI_HAS_NEW_MODE_SWITCH option
sunxi-mmc: use new mode on A64
mmc: sunxi: run calibration on A64

I have still observing eMMC boot issue, this need to debug further

U-Boot SPL 2018.07-rc1-00164-g117d5917d2 (Jun 16 2018 - 15:12:28 +0530)
DRAM: 1024 MiB
Trying to boot from MMC2
unable to select a mode
mmc_load_image_raw_sector: mmc block read error
SPL: failed to boot from all boot devices
### ERROR ### Please RESET the board ###

Jagan.

#6 lore raw

v5 cover sunxi: fix eMMC stability issues on A64

Quoted 11 lines · Jagan Teki
Quoted 10 lines · Vasily Khoruzhick

eMMC seems to require new clocking mode and calibration on A64, otherwise it is pretty unstable on some boards (e.g. Pinebook) with some eMMCs.

v2: - improve comment about calibration for eMMC on A64
- simplify ifdef-s around configuring delays
v3: - fix fallout due to ifdef simplification in v2
v4: - really fix ifdefs this time
v5: - add new option for SoCs without new mode switch in CCM

Quoted 6 lines
Quoted 4 lines

Vasily Khoruzhick (3):
sunxi-mmc: introduce new MMC_SUNXI_HAS_NEW_MODE_SWITCH option
sunxi-mmc: use new mode on A64
mmc: sunxi: run calibration on A64

I have still observing eMMC boot issue, this need to debug further

It boots for me from eMMC on Pinebook and Pine64-LTS, thus I have no hardware to debug it further.

Moreover, Linux also does calibration on A64 for eMMC, do you experience any issues with eMMC in Linux?

Quoted 9 lines

U-Boot SPL 2018.07-rc1-00164-g117d5917d2 (Jun 16 2018 - 15:12:28 +0530)
DRAM: 1024 MiB
Trying to boot from MMC2
unable to select a mode
mmc_load_image_raw_sector: mmc block read error
SPL: failed to boot from all boot devices
### ERROR ### Please RESET the board ###

Jagan.

#7 lore raw

v5 cover sunxi: fix eMMC stability issues on A64

Quoted 17 lines · Vasily Khoruzhick
Quoted 16 lines · Jagan Teki
Quoted 14 lines · Vasily Khoruzhick

eMMC seems to require new clocking mode and calibration on A64, otherwise it is pretty unstable on some boards (e.g. Pinebook) with some eMMCs.

v2: - improve comment about calibration for eMMC on A64
- simplify ifdef-s around configuring delays
v3: - fix fallout due to ifdef simplification in v2
v4: - really fix ifdefs this time
v5: - add new option for SoCs without new mode switch in CCM

Vasily Khoruzhick (3):
sunxi-mmc: introduce new MMC_SUNXI_HAS_NEW_MODE_SWITCH option
sunxi-mmc: use new mode on A64
mmc: sunxi: run calibration on A64

Quoted 4 lines
Quoted 1 line

I have still observing eMMC boot issue, this need to debug further

It boots for me from eMMC on Pinebook and Pine64-LTS, thus I have no hardware to debug it further.

Don't have these hardware with me.

Quoted 3 lines

Moreover, Linux also does calibration on A64 for eMMC, do you experience any issues with eMMC in Linux?

Yes I'm able to detect eMMC in Linux.

So we need to investigate what makes failing mode.

#8 lore raw

v5 cover sunxi: fix eMMC stability issues on A64

Quoted 2 lines · Jagan Teki

Don't have these hardware with me.

Could you test this patch on your hardware when you get some time?

https://github.com/anarsoul/u-boot-pine64/commit/a3f5d175a5ed0ece7a7720f1e665809c9f140850.patch

It uses calibration without using new clock mode and it fixes eMMC on Pinebook for me as well.
Also works fine on Pine64-LTS.

If it works for you I'll clean it up and send it out.

#10 lore raw

v5 cover sunxi: fix eMMC stability issues on A64

Quoted 10 lines · Jagan Teki
Quoted 7 lines · Vasily Khoruzhick
Quoted 2 lines · Jagan Teki

Don't have these hardware with me.

Could you test this patch on your hardware when you get some time?

https://github.com/anarsoul/u-boot-pine64/commit/a3f5d175a5ed0ece7a7720f1e665809c9f140850.patch

On top of your changes or separately?

This what[1] I changed on top of your changes, same issue.

[1] https://paste.ubuntu.com/p/NjYyMk68P6/

#12 lore raw

v5 cover sunxi: fix eMMC stability issues on A64

Quoted 7 lines · Vasily Khoruzhick
Quoted 4 lines · Jagan Teki

This what[1] I changed on top of your changes, same issue.

[1] https://paste.ubuntu.com/p/NjYyMk68P6/

It should be tested separately.

Works for me, with md 0x01c11144 value as 00000080