Skip to contentSkip to conversation

ARM: imx: i.mx6d/q: disable the double linefill feature of PL310

PATCH New Jason Liu 3 participants 3 messages Sep 16, 2013 – Oct 17, 2013

Subsystems (3)
lore
Share & tools
lore

Jason Liu

author
#0 lore raw

PATCH v2 1/1 ARM: imx: i.mx6d/q: disable the double linefill feature of PL310

The L2 cache controller(PL310) version on the i.MX6D/Q is r3p1-50rel0 The L2 cache controller(PL310) version on the i.MX6DL/SOLO/SL is r3p2 But according to ARM PL310 errata: 752271
ID: 752271: Double linefill feature can cause data corruption Fault Status: Present in: r3p0, r3p1, r3p1-50rel0. Fixed in r3p2
Workaround: The only workaround to this erratum is to disable the double linefill feature. This is the default behavior.

without this patch, you will meet the following error when run the memtester application at: http://pyropus.ca/software/memtester/

FAILURE: 0x00100000 != 0x00200000 at offset 0x01365664.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365668.
FAILURE: 0x00100000 != 0x00200000 at offset 0x0136566c.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365670.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365674.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365678.

---
V2:
- disable the L2 doulbe linefill explicitly on i.MX6Q by using &= mask
- using the cpu_is_imx6q() instead of checking through DT

---
 arch/arm/mach-imx/system.c |   11 +++++++++++
 1 file changed, 11 insertions(+)
arch/arm/mach-imx/system.c +11 -0
index 64ff37e..80c177c 100644
@@ -117,6 +117,17 @@ void __init imx_init_l2cache(void)
 	/* Configure the L2 PREFETCH and POWER registers */
 	val = readl_relaxed(l2x0_base + L2X0_PREFETCH_CTRL);
 	val |= 0x70800000;
+	/*
+	 * The L2 cache controller(PL310) version on the i.MX6D/Q is r3p1-50rel0
+	 * The L2 cache controller(PL310) version on the i.MX6DL/SOLO/SL is r3p2
+	 * But according to ARM PL310 errata: 752271
+	 * ID: 752271: Double linefill feature can cause data corruption
+	 * Fault Status: Present in: r3p0, r3p1, r3p1-50rel0. Fixed in r3p2
+	 * Workaround: The only workaround to this erratum is to disable the
+	 * double linefill feature. This is the default behavior.
+	 */
+	if (cpu_is_imx6q())
+		val &= ~(1 << 30 | 1 << 23);
 	writel_relaxed(val, l2x0_base + L2X0_PREFETCH_CTRL);
 	val = L2X0_DYNAMIC_CLK_GATING_EN | L2X0_STNDBY_MODE_EN;
 	writel_relaxed(val, l2x0_base + L2X0_POWER_CTRL);
Signature · 1 line
1.7.10.4
Applied
#1 lore raw

PATCH v2 1/1 ARM: imx: i.mx6d/q: disable the double linefill feature of PL310

Quoted 11 lines · Jason Liu

The L2 cache controller(PL310) version on the i.MX6D/Q is r3p1-50rel0 The L2 cache controller(PL310) version on the i.MX6DL/SOLO/SL is r3p2 But according to ARM PL310 errata: 752271
ID: 752271: Double linefill feature can cause data corruption Fault Status: Present in: r3p0, r3p1, r3p1-50rel0. Fixed in r3p2
Workaround: The only workaround to this erratum is to disable the double linefill feature. This is the default behavior.

without this patch, you will meet the following error when run the memtester application at: http://pyropus.ca/software/memtester/

Quoted 8 lines

FAILURE: 0x00100000 != 0x00200000 at offset 0x01365664.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365668.
FAILURE: 0x00100000 != 0x00200000 at offset 0x0136566c.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365670.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365674.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365678.

Applied, thanks.

Shawn

#2 lore raw

PATCH v2 1/1 ARM: imx: i.mx6d/q: disable the double linefill feature of PL310

In reply to Jason Liu
Quoted 37 lines · Jason Liu

The L2 cache controller(PL310) version on the i.MX6D/Q is r3p1-50rel0 The L2 cache controller(PL310) version on the i.MX6DL/SOLO/SL is r3p2 But according to ARM PL310 errata: 752271
ID: 752271: Double linefill feature can cause data corruption Fault Status: Present in: r3p0, r3p1, r3p1-50rel0. Fixed in r3p2
Workaround: The only workaround to this erratum is to disable the double linefill feature. This is the default behavior.

without this patch, you will meet the following error when run the memtester application at: http://pyropus.ca/software/memtester/

FAILURE: 0x00100000 != 0x00200000 at offset 0x01365664.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365668.
FAILURE: 0x00100000 != 0x00200000 at offset 0x0136566c.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365670.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365674.
FAILURE: 0x00100000 != 0x00200000 at offset 0x01365678.

---
V2:
- disable the L2 doulbe linefill explicitly on i.MX6Q by using &= mask
- using the cpu_is_imx6q() instead of checking through DT

---
  arch/arm/mach-imx/system.c |   11 +++++++++++
  1 file changed, 11 insertions(+)
arch/arm/mach-imx/system.c +1 -0
index 64ff37e..80c177c 100644
@@ -117,6 +117,17 @@ void __init imx_init_l2cache(void)
  	/* Configure the L2 PREFETCH and POWER registers */
  	val = readl_relaxed(l2x0_base + L2X0_PREFETCH_CTRL);
  	val |= 0x70800000;
+	/*

arch/arm/mach-imx/system.c void __init imx_init_l2cache(void) · line 130

Quoted 10 lines
arch/arm/mach-imx/system.c +10 -0
+	 * The L2 cache controller(PL310) version on the i.MX6D/Q is r3p1-50rel0
+	 * The L2 cache controller(PL310) version on the i.MX6DL/SOLO/SL is r3p2
+	 * But according to ARM PL310 errata: 752271
+	 * ID: 752271: Double linefill feature can cause data corruption
+	 * Fault Status: Present in: r3p0, r3p1, r3p1-50rel0. Fixed in r3p2
+	 * Workaround: The only workaround to this erratum is to disable the
+	 * double linefill feature. This is the default behavior.
+	 */
+	if (cpu_is_imx6q())
+		val &= ~(1 << 30 | 1 << 23);

Since you sent this patch I was wondering why you touch Bit 23, too.

Having access to the manual, now (CoreLink? Level 2 Cache Controller L2C-310 Revision: r3p3 ARM DDI 0246H (ID080112)) it states on page 2-39 (section Double linefill issuing):

"You can control this feature (the double linefill) using Bits 30, 27, and 23 of the Prefetch Control Register. Bit 23, and Bit 27 are only used if you set Bit 30 HIGH."

I.e. to my understanding

val &= ~(1 << 30);

should be sufficient here?

Best regards

Dirk