ARM: OMAP3-DRA7: CP15 erratum workarounds and improvements

PATCH Superseded Nishanth Menon 3 participants 21 messages Mar 3, 2015 – Mar 6, 2015

Subsystems (2)

Arm (32-bit) Arm/actions semi architecture

lore

Share & tools

lore

Nishanth Menon

author

2015-03-03 22:26 UTC #0 lore raw

v4 cover ARM: OMAP3-DRA7: CP15 erratum workarounds and improvements

The fourth incarnation of this series to address review comments on V3

With all the usual disclaimers and request to see V1 of the series for a detailed blurb.. As usual additional testing preferred.. Sorry, I dont have access to all possible variants atm..

changes since v3:
- few corrections - i have tried to do a push-pop of register params. hopefully, they should do the job
- smc with a makefile handling of secure-ext enablement.. (stolen from kernel).

V3: http://thread.gmane.org/gmane.comp.boot-loaders.u-boot/213207/focus=213307
V2: http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/213060
V1: http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/212174

Testing: with http://paste.ubuntu.org.cn/2522971 (4.0-rc1 patch)

BeagleBoard-X15: http://pastebin.ubuntu.com/10518934/

	Before: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000000 L2PFR=0x000009b0 ACTLR=0x00000040
	After: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000198 L2PFR=0x000009b0 ACTLR=0x00000040

OMAP5uEVM: http://pastebin.ubuntu.com/10518958/

	Before: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000000 L2PFR=0x000009b0 ACTLR=0x00000040
	After: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000198 L2PFR=0x000009b0 ACTLR=0x00000040

Beagle-XM: http://pastebin.ubuntu.com/10519417/ (this is a r2p3 device)

	Before: CPUID=0x413fc082 ACR=0x000000e2 L2AUXCR=0x00000042
	After: CPUID=0x413fc082 ACR=0x00000042 L2AUXCR=0x00000042

I dont have access to other omap3 platforms to give a better coverage

Sanity check:
OMAP4Panda-ES: http://pastebin.ubuntu.com/10518971/

Nishanth Menon (10):
ARM: Introduce erratum workaround for 798870
ARM: Introduce erratum workaround for 454179
ARM: Introduce erratum workaround for 430973
ARM: Introduce erratum workaround for 621766
ARM: OMAP: Change set_pl310_ctrl_reg to be generic
ARM: OMAP3: Rename omap3.h to omap.h to be generic as all SoCs
ARM: OMAP3: Get rid of omap3_gp_romcode_call and replace with
omap_smc1
ARM: OMAP5 / DRA7: Setup L2 Aux Control Register with recommended
configuration
ARM: OMAP3: Enable workaround for ARM errata 454179, 430973, 621766
ARM: OMAP3: rx51: Enable workaround for ARM errata 454179, 430973,

Praveen Rao (1):
ARM: DRA7 / OMAP5: Add workaround for ARM errata 798870

 README                                             |    8 +++
 arch/arm/cpu/armv7/Makefile                        |    2 +-
 arch/arm/cpu/armv7/cp15.c                          |   29 ++++++++++
 arch/arm/cpu/armv7/omap-common/Makefile            |    5 +-
 arch/arm/cpu/armv7/omap-common/lowlevel_init.S     |   19 +++---
 arch/arm/cpu/armv7/omap3/board.c                   |   60 +++++++------------
 arch/arm/cpu/armv7/omap3/lowlevel_init.S           |   11 ----
 arch/arm/cpu/armv7/omap4/hwinit.c                  |    4 +-
 arch/arm/cpu/armv7/omap5/hwinit.c                  |   23 ++++++++
 arch/arm/cpu/armv7/start.S                         |   61 ++++++++++++++++++++
 .../arm/include/asm/arch-omap3/{omap3.h => omap.h} |    0
 arch/arm/include/asm/arch-omap3/sys_proto.h        |    3 +-
 arch/arm/include/asm/arch-omap4/sys_proto.h        |    5 +-
 arch/arm/include/asm/arch-omap5/sys_proto.h        |    4 ++
 arch/arm/include/asm/armv7.h                       |    5 ++
 board/nokia/rx51/rx51.c                            |   19 +++---
 include/configs/am3517_crane.h                     |    6 +-
 include/configs/am3517_evm.h                       |    6 +-
 include/configs/cm_t35.h                           |    6 +-
 include/configs/cm_t3517.h                         |    6 +-
 include/configs/dig297.h                           |    6 +-
 include/configs/mcx.h                              |    6 +-
 include/configs/nokia_rx51.h                       |    6 +-
 include/configs/omap3_evm.h                        |    2 +-
 include/configs/omap3_evm_common.h                 |    4 ++
 include/configs/omap3_evm_quick_mmc.h              |    2 +-
 include/configs/omap3_evm_quick_nand.h             |    2 +-
 include/configs/omap3_logic.h                      |    6 +-
 include/configs/omap3_mvblx.h                      |    6 +-
 include/configs/omap3_pandora.h                    |    6 +-
 include/configs/omap3_sdp3430.h                    |    6 +-
 include/configs/omap3_zoom1.h                      |    2 +-
 include/configs/tam3517-common.h                   |    6 +-
 include/configs/tao3530.h                          |    6 +-
 include/configs/ti_omap3_common.h                  |    7 ++-
 include/configs/ti_omap5_common.h                  |    3 +
 include/configs/tricorder.h                        |    6 +-
 37 files changed, 275 insertions(+), 89 deletions(-)
 create mode 100644 arch/arm/cpu/armv7/cp15.c
 rename arch/arm/include/asm/arch-omap3/{omap3.h => omap.h} (100%)

Signature · 1 line

1.7.9.5

Nishanth Menon

author

2015-03-03 22:26 UTC #1 lore raw

PATCH v4 1/11 ARM: Introduce erratum workaround for 798870

Add workaround for Cortex-A15 ARM erratum 798870 which says "If back-to-back speculative cache line fills (fill A and fill B) are issued from the L1 data cache of a CPU to the L2 cache, the second request (fill B) is then cancelled, and the second request would have detected a hazard against a recent write or eviction (write B) to the same cache line as fill B then the L2 logic might deadlock."

Implementations for SoC families such as Exynos, OMAP5/DRA7 etc will be widely different.

Every SoC has slightly different manner of setting up access to L2ACLR and similar registers since the Secure Monitor handling of Secure Monitor Call(smc) is diverse. Hence an weak function is introduced which may be overriden to implement SoC specific accessor implementation.

Based on ARM errata Document revision 18.0 (22 Nov 2013)

Signed-off-by Nishanth Menon <[email protected]>

---
 README                       |    5 +++++
 arch/arm/cpu/armv7/Makefile  |    2 +-
 arch/arm/cpu/armv7/cp15.c    |   23 +++++++++++++++++++++++
 arch/arm/cpu/armv7/start.S   |   22 ++++++++++++++++++++++
 arch/arm/include/asm/armv7.h |    3 +++
 5 files changed, 54 insertions(+), 1 deletion(-)
 create mode 100644 arch/arm/cpu/armv7/cp15.c

README +5 -0

index a28ff133ee05..11ec37ba4a00 100644
@@ -621,6 +621,11 @@ The following options need to be configured:
 		exists, unlike the similar options in the Linux kernel. Do not
 		set these options unless they apply!
 
+		NOTE: The following can be machine specific errata. These
+		do have ability to provide rudimentary version and machine
+		specific checks, but expect no product checks.
+		CONFIG_ARM_ERRATA_798870
+
 - Driver Model
 		Driver model is a new framework for devices in U-Boot
 		introduced in early 2014. U-Boot is being progressively

arch/arm/cpu/armv7/Makefile +1 -1

index 409e6f5651b6..43da3e586f71 100644
@@ -9,7 +9,7 @@ extra-y	:= start.o
 
 obj-y	+= cache_v7.o
 
-obj-y	+= cpu.o
+obj-y	+= cpu.o cp15.o
 obj-y	+= syslib.o
 
 ifneq ($(CONFIG_AM43XX)$(CONFIG_AM33XX)$(CONFIG_OMAP44XX)$(CONFIG_OMAP54XX)$(CONFIG_TEGRA)$(CONFIG_MX6)$(CONFIG_TI81XX)$(CONFIG_AT91FAMILY)$(CONFIG_SUNXI),)

arch/arm/cpu/armv7/cp15.c new file +23 -0

new file mode 100644
index 000000000000..8ac81c9ba147
@@ -0,0 +1,23 @@
+/*
+ * (C) Copyright 2015 Texas Insturments
+ *
+ * SPDX-License-Identifier:	GPL-2.0+
+ */
+
+/*
+ * CP15 specific code
+ */
+
+#include <common.h>
+#include <command.h>
+#include <asm/system.h>
+#include <asm/cache.h>
+#include <asm/armv7.h>
+#include <linux/compiler.h>
+
+void __weak v7_arch_cp15_set_l2aux_ctrl(u32 l2actlr, u32 cpu_midr,
+				     u32 cpu_rev_comb, u32 cpu_variant,
+				     u32 cpu_rev)
+{
+	asm volatile ("mcr p15, 1, %0, c15, c0, 0\n\t" : : "r"(l2actlr));
+}

arch/arm/cpu/armv7/start.S +22 -0

index 70048c10aee6..6fbc44e5bf74 100644
@@ -163,6 +163,28 @@ ENTRY(cpu_init_cp15)
 	mcr	p15, 0, r0, c15, c0, 1	@ write diagnostic register
 #endif
 
+	mrc p15, 0, r1, c0, c0, 0	@ r1 has Read Main ID Register (MIDR)
+	mov r3, r1, lsr #20		@ get variant field
+	and r3, r3, #0xf		@ r3 has CPU variant
+	and r4, r1, #0xf                @ r4 has CPU revision
+	mov r2, r3, lsl #4		@ shift variant field for combined value
+	orr r2, r4, r2                  @ r2 has combined CPU variant + revision
+
+#ifdef CONFIG_ARM_ERRATA_798870
+	cmp	r2, #0x30		@ Applies to lower than R3p0
+	bge	skip_errata_798870      @ skip if not affected rev
+	cmp	r2, #0x20		@ Applies to including and above R2p0
+	blt	skip_errata_798870      @ skip if not affected rev
+
+	mrc	p15, 1, r0, c15, c0, 0  @ read l2 aux ctrl reg
+	orr	r0, r0, #1 << 7         @ Enable hazard-detect timeout
+	push	{r1-r4}			@ Save the cpu info registers
+	b	v7_arch_cp15_set_l2aux_ctrl
+	isb				@ Recommended ISB after l2actlr update
+	pop	{r1-r4}			@ Restore the cpu info - fall through
+skip_errata_798870:
+#endif
+
 	mov	pc, lr			@ back to my caller
 ENDPROC(cpu_init_cp15)

arch/arm/include/asm/armv7.h +3 -0

index a13da23cf172..30e0cc3cf3d5 100644
@@ -93,6 +93,9 @@ extern char __secure_end[];
 
 #endif /* CONFIG_ARMV7_NONSEC || CONFIG_ARMV7_VIRT */
 
+void v7_arch_cp15_set_l2aux_ctrl(u32 l2auxctrl, u32 cpu_midr,
+				 u32 cpu_rev_comb, u32 cpu_variant,
+				 u32 cpu_rev);
 #endif /* ! __ASSEMBLY__ */
 
 #endif

Signature · 1 line

1.7.9.5

Nishanth Menon

author

2015-03-03 22:26 UTC #2 lore raw

PATCH v4 2/11 ARM: Introduce erratum workaround for 454179

In reply to Nishanth Menon

454179: Stale prediction may inhibit target address misprediction on
next predicted taken branch
Impacts: Every Cortex-A8 processors with revision lower than r2p1 Work around: Set IBE and disable branch size mispredict to 1

Also provide a hook for SoC specific handling to take place if needed.

Based on ARM errata Document revision 20.0 (13 Nov 2010)

Signed-off-by Nishanth Menon <[email protected]>

---
 README                       |    1 +
 arch/arm/cpu/armv7/cp15.c    |    6 ++++++
 arch/arm/cpu/armv7/start.S   |   13 +++++++++++++
 arch/arm/include/asm/armv7.h |    2 ++
 4 files changed, 22 insertions(+)

README +1 -0

index 11ec37ba4a00..e538cf061286 100644
@@ -624,6 +624,7 @@ The following options need to be configured:
 		NOTE: The following can be machine specific errata. These
 		do have ability to provide rudimentary version and machine
 		specific checks, but expect no product checks.
+		CONFIG_ARM_ERRATA_454179
 		CONFIG_ARM_ERRATA_798870
 
 - Driver Model

arch/arm/cpu/armv7/cp15.c +6 -0

index 8ac81c9ba147..b44c9f94a822 100644
@@ -21,3 +21,9 @@ void __weak v7_arch_cp15_set_l2aux_ctrl(u32 l2actlr, u32 cpu_midr,
 {
 	asm volatile ("mcr p15, 1, %0, c15, c0, 0\n\t" : : "r"(l2actlr));
 }
+
+void __weak v7_arch_cp15_set_acr(u32 acr, u32 cpu_midr, u32 cpu_rev_comb,
+				 u32 cpu_variant, u32 cpu_rev)
+{
+	asm volatile ("mcr p15, 0, %0, c1, c0, 1\n\t" : : "r"(acr));
+}

arch/arm/cpu/armv7/start.S +13 -0

index 6fbc44e5bf74..1516ed675dac 100644
@@ -185,6 +185,19 @@ ENTRY(cpu_init_cp15)
 skip_errata_798870:
 #endif
 
+#ifdef CONFIG_ARM_ERRATA_454179
+	cmp	r2, #0x21		@ Only on < r2p1
+	bge	skip_errata_454179
+
+	mrc	p15, 0, r0, c1, c0, 1	@ Read ACR
+	orr	r0, r0, #(0x3 << 6)	@ Set DBSM(BIT7) and IBE(BIT6) bits
+	push	{r1-r4}			@ Save the cpu info registers
+	b	v7_arch_cp15_set_acr
+	pop	{r1-r4}			@ Restore the cpu info - fall through
+
+skip_errata_454179:
+#endif
+
 	mov	pc, lr			@ back to my caller
 ENDPROC(cpu_init_cp15)

arch/arm/include/asm/armv7.h +2 -0

index 30e0cc3cf3d5..434b34de0ce9 100644
@@ -96,6 +96,8 @@ extern char __secure_end[];
 void v7_arch_cp15_set_l2aux_ctrl(u32 l2auxctrl, u32 cpu_midr,
 				 u32 cpu_rev_comb, u32 cpu_variant,
 				 u32 cpu_rev);
+void v7_arch_cp15_set_acr(u32 acr, u32 cpu_midr, u32 cpu_rev_comb,
+			  u32 cpu_variant, u32 cpu_rev);
 #endif /* ! __ASSEMBLY__ */
 
 #endif

Signature · 1 line

1.7.9.5

Nishanth Menon

author

2015-03-03 22:26 UTC #3 lore raw

PATCH v4 3/11 ARM: Introduce erratum workaround for 430973

In reply to Nishanth Menon

430973: Stale prediction on replaced inter working branch causes
Cortex-A8 to execute in the wrong ARM/Thumb state
Impacts: Every Cortex-A8 processors with revision lower than r2p1 Work around: Set IBE to 1

Based on ARM errata Document revision 20.0 (13 Nov 2010)

Signed-off-by Nishanth Menon <[email protected]>

---
 README                     |    1 +
 arch/arm/cpu/armv7/start.S |   13 +++++++++++++
 2 files changed, 14 insertions(+)

README +1 -0

index e538cf061286..484ae9ee39a8 100644
@@ -624,6 +624,7 @@ The following options need to be configured:
 		NOTE: The following can be machine specific errata. These
 		do have ability to provide rudimentary version and machine
 		specific checks, but expect no product checks.
+		CONFIG_ARM_ERRATA_430973
 		CONFIG_ARM_ERRATA_454179
 		CONFIG_ARM_ERRATA_798870

arch/arm/cpu/armv7/start.S +13 -0

index 1516ed675dac..99220ce76a25 100644
@@ -198,6 +198,19 @@ skip_errata_798870:
 skip_errata_454179:
 #endif
 
+#ifdef CONFIG_ARM_ERRATA_430973
+	cmp	r2, #0x21		@ Only on < r2p1
+	bge	skip_errata_430973
+
+	mrc	p15, 0, r0, c1, c0, 1	@ Read ACR
+	orr	r0, r0, #(0x1 << 6)	@ Set IBE bit
+	push	{r1-r4}			@ Save the cpu info registers
+	b	v7_arch_cp15_set_acr
+	pop	{r1-r4}			@ Restore the cpu info - fall through
+
+skip_errata_430973:
+#endif
+
 	mov	pc, lr			@ back to my caller
 ENDPROC(cpu_init_cp15)

Signature · 1 line

1.7.9.5

Nishanth Menon

author

2015-03-03 22:26 UTC #4 lore raw

PATCH v4 4/11 ARM: Introduce erratum workaround for 621766

In reply to Nishanth Menon

621766: Under a specific set of conditions, executing a sequence of
NEON or vfp load instructions can cause processor deadlock
Impacts: Every Cortex-A8 processors with revision lower than r2p1 Work around: Set L1NEON to 1

Based on ARM errata Document revision 20.0 (13 Nov 2010)

Signed-off-by Nishanth Menon <[email protected]>

---
 README                     |    1 +
 arch/arm/cpu/armv7/start.S |   13 +++++++++++++
 2 files changed, 14 insertions(+)

README +1 -0

index 484ae9ee39a8..5ee789aea627 100644
@@ -626,6 +626,7 @@ The following options need to be configured:
 		specific checks, but expect no product checks.
 		CONFIG_ARM_ERRATA_430973
 		CONFIG_ARM_ERRATA_454179
+		CONFIG_ARM_ERRATA_621766
 		CONFIG_ARM_ERRATA_798870
 
 - Driver Model

arch/arm/cpu/armv7/start.S +13 -0

index 99220ce76a25..2b1e312f9129 100644
@@ -211,6 +211,19 @@ skip_errata_454179:
 skip_errata_430973:
 #endif
 
+#ifdef CONFIG_ARM_ERRATA_621766
+	cmp	r2, #0x21		@ Only on < r2p1
+	bge	skip_errata_621766
+
+	mrc	p15, 0, r0, c1, c0, 1	@ Read ACR
+	orr	r0, r0, #(0x1 << 5)	@ Set L1NEON bit
+	push	{r1-r4}			@ Save the cpu info registers
+	b	v7_arch_cp15_set_acr
+	pop	{r1-r4}			@ Restore the cpu info - fall through
+
+skip_errata_621766:
+#endif
+
 	mov	pc, lr			@ back to my caller
 ENDPROC(cpu_init_cp15)

Signature · 1 line

1.7.9.5

Nishanth Menon

author

2015-03-03 22:26 UTC #5 lore raw

PATCH v4 5/11 ARM: OMAP: Change set_pl310_ctrl_reg to be generic

In reply to Nishanth Menon

set_pl310_ctrl_reg does use the Secure Monitor Call (SMC) to setup PL310 control register, however, that is something that is generic enough to be used for OMAP5 generation of processors as well. The only difference being the service being invoked for the function.

So, convert the service to a macro and use a generic name (same as that used in Linux for some consistency). While at that, also add a data barrier which is necessary as per recommendation.

While at this, switch over to smc #0 instead of handcoded assembly. To ensure gcc compatibility, steal the strategy used by Linux kernel for sec extension builds (NOTE: we no longer use '-march=armv5' as the legacy comment claims).

Signed-off-by Nishanth Menon <[email protected]>

---
 arch/arm/cpu/armv7/omap-common/Makefile        |    3 +++
 arch/arm/cpu/armv7/omap-common/lowlevel_init.S |   17 ++++++++++-------
 arch/arm/cpu/armv7/omap4/hwinit.c              |    4 ++--
 arch/arm/include/asm/arch-omap4/sys_proto.h    |    5 ++++-
 4 files changed, 19 insertions(+), 10 deletions(-)

arch/arm/cpu/armv7/omap-common/Makefile +3 -0

index 7695e16d36f5..b36e85d25f3f 100644
@@ -26,6 +26,9 @@ ifeq ($(CONFIG_SYS_DCACHE_OFF),)
 obj-y	+= omap-cache.o
 endif
 
+plus_sec := $(call as-instr,.arch_extension sec,+sec)
+AFLAGS_lowlevel_init.o	:=-Wa,-march=armv7-a$(plus_sec)
+
 ifeq ($(CONFIG_OMAP34XX),)
 obj-y	+= boot-common.o
 obj-y	+= lowlevel_init.o

arch/arm/cpu/armv7/omap-common/lowlevel_init.S +10 -7

index 86c0e4217478..83426291b22d 100644
@@ -22,11 +22,14 @@ ENTRY(save_boot_params)
 	bx	lr
 ENDPROC(save_boot_params)
 
-ENTRY(set_pl310_ctrl_reg)
-	PUSH	{r4-r11, lr}	@ save registers - ROM code may pollute
+ENTRY(omap_smc1)
+	PUSH	{r4-r12, lr}	@ save registers - ROM code may pollute
 				@ our registers
-	LDR	r12, =0x102	@ Set PL310 control register - value in R0
-	.word	0xe1600070	@ SMC #0 - hand assembled because -march=armv5
-				@ call ROM Code API to set control register
-	POP	{r4-r11, pc}
-ENDPROC(set_pl310_ctrl_reg)
+	MOV	r12, r0		@ Service
+	MOV	r0, r1		@ Argument
+	DSB
+	DMB
+	SMC	#0		@ Call the secure monitor for the service
+
+	POP	{r4-r12, pc}
+ENDPROC(omap_smc1)

arch/arm/cpu/armv7/omap4/hwinit.c +2 -2

index db16548fac49..9792761d40a0 100644
@@ -159,11 +159,11 @@ void init_omap_revision(void)
 #ifndef CONFIG_SYS_L2CACHE_OFF
 void v7_outer_cache_enable(void)
 {
-	set_pl310_ctrl_reg(1);
+	omap_smc1(OMAP4_SERVICE_PL310_CONTROL_REG_SET, 1);
 }
 
 void v7_outer_cache_disable(void)
 {
-	set_pl310_ctrl_reg(0);
+	omap_smc1(OMAP4_SERVICE_PL310_CONTROL_REG_SET, 0);
 }
 #endif /* !CONFIG_SYS_L2CACHE_OFF */

arch/arm/include/asm/arch-omap4/sys_proto.h +4 -1

index e19975efaf50..f425e3af54f5 100644
@@ -37,7 +37,7 @@ void do_set_mux(u32 base, struct pad_conf_entry const *array, int size);
 void set_muxconf_regs_essential(void);
 u32 wait_on_value(u32, u32, void *, u32);
 void sdelay(unsigned long);
-void set_pl310_ctrl_reg(u32 val);
+void omap_smc1(u32 service, u32 val);
 void setup_clocks_for_console(void);
 void prcm_init(void);
 void bypass_dpll(u32 const base);

@@ -57,4 +57,7 @@ int omap_vc_bypass_send_value(u8 sa, u8 reg_addr, u8 reg_data);
 u32 warm_reset(void);
 void force_emif_self_refresh(void);
 void setup_warmreset_time(void);
+
+#define OMAP4_SERVICE_PL310_CONTROL_REG_SET	0x102
+
 #endif

Signature · 1 line

1.7.9.5

Matt Porter

2015-03-05 14:00 UTC #6 lore raw

PATCH v4 5/11 ARM: OMAP: Change set_pl310_ctrl_reg to be generic

Quoted 5 lines · Nishanth Menon

set_pl310_ctrl_reg does use the Secure Monitor Call (SMC) to setup PL310 control register, however, that is something that is generic enough to be used for OMAP5 generation of processors as well. The only difference being the service being invoked for the function.

Quoted 8 lines

So, convert the service to a macro and use a generic name (same as that used in Linux for some consistency). While at that, also add a data barrier which is necessary as per recommendation.
While at this, switch over to smc #0 instead of handcoded assembly. To ensure gcc compatibility, steal the strategy used by Linux kernel for sec extension builds (NOTE: we no longer use '-march=armv5' as the legacy comment claims).

Hi Nishanth,

I applied this series with fuzz and fixed a minor conflict on master. I ran into a build issue for omap3 beagle with the sec extension scheme on the gcc version 4.7.4 (Ubuntu/Linaro 4.7.4-2ubuntu1) toolchain:

arm-linux-gnueabi-gcc
-Wp,-MD,arch/arm/cpu/armv7/omap3/.lowlevel_init.o.d -nostdinc -isystem /usr/lib/gcc-cross/arm-linux-gnueabi/4.7/include -Iinclude -I../include -I../arch/arm/include -include ../include/linux/kconfig.h -D__KERNEL__ -D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000 -D__ASSEMBLY__ -g -D__ARM__ -marm -mno-thumb-interwork -mabi=aapcs-linux -mword-relocations -march=armv7-a -mno-unaligned-access -ffunction-sections -fdata-sections -fno-common -ffixed-r9 -msoft-float
-pipe -c -o arch/arm/cpu/armv7/omap3/lowlevel_init.o
../arch/arm/cpu/armv7/omap3/lowlevel_init.S

../arch/arm/cpu/armv7/omap-common/lowlevel_init.S: Assembler messages:
../arch/arm/cpu/armv7/omap-common/lowlevel_init.S:34: Error: selected processor does not support ARM mode `smc #0'

I've worked around this for the moment by placing an explicit .arch_extension sec in lowlevel_init.S but hopefully you have some thoughts on why those flags don't seem to be picked up. I'll continue to take a look at it in the meantime.

-Matt

Quoted 94 lines

Signed-off-by Nishanth Menon <[email protected]>

---
 arch/arm/cpu/armv7/omap-common/Makefile        |    3 +++
 arch/arm/cpu/armv7/omap-common/lowlevel_init.S |   17 ++++++++++-------
 arch/arm/cpu/armv7/omap4/hwinit.c              |    4 ++--
 arch/arm/include/asm/arch-omap4/sys_proto.h    |    5 ++++-
 4 files changed, 19 insertions(+), 10 deletions(-)

arch/arm/cpu/armv7/omap-common/Makefile +3 -0

index 7695e16d36f5..b36e85d25f3f 100644
@@ -26,6 +26,9 @@ ifeq ($(CONFIG_SYS_DCACHE_OFF),)
 obj-y	+= omap-cache.o
 endif
 
+plus_sec := $(call as-instr,.arch_extension sec,+sec)
+AFLAGS_lowlevel_init.o	:=-Wa,-march=armv7-a$(plus_sec)
+
 ifeq ($(CONFIG_OMAP34XX),)
 obj-y	+= boot-common.o
 obj-y	+= lowlevel_init.o

arch/arm/cpu/armv7/omap-common/lowlevel_init.S +10 -7

index 86c0e4217478..83426291b22d 100644
@@ -22,11 +22,14 @@ ENTRY(save_boot_params)
 	bx	lr
 ENDPROC(save_boot_params)
 
-ENTRY(set_pl310_ctrl_reg)
-	PUSH	{r4-r11, lr}	@ save registers - ROM code may pollute
+ENTRY(omap_smc1)
+	PUSH	{r4-r12, lr}	@ save registers - ROM code may pollute
 				@ our registers
-	LDR	r12, =0x102	@ Set PL310 control register - value in R0
-	.word	0xe1600070	@ SMC #0 - hand assembled because -march=armv5
-				@ call ROM Code API to set control register
-	POP	{r4-r11, pc}
-ENDPROC(set_pl310_ctrl_reg)
+	MOV	r12, r0		@ Service
+	MOV	r0, r1		@ Argument
+	DSB
+	DMB
+	SMC	#0		@ Call the secure monitor for the service
+
+	POP	{r4-r12, pc}
+ENDPROC(omap_smc1)

arch/arm/cpu/armv7/omap4/hwinit.c +2 -2

index db16548fac49..9792761d40a0 100644
@@ -159,11 +159,11 @@ void init_omap_revision(void)
 #ifndef CONFIG_SYS_L2CACHE_OFF
 void v7_outer_cache_enable(void)
 {
-	set_pl310_ctrl_reg(1);
+	omap_smc1(OMAP4_SERVICE_PL310_CONTROL_REG_SET, 1);
 }
 
 void v7_outer_cache_disable(void)
 {
-	set_pl310_ctrl_reg(0);
+	omap_smc1(OMAP4_SERVICE_PL310_CONTROL_REG_SET, 0);
 }
 #endif /* !CONFIG_SYS_L2CACHE_OFF */

arch/arm/include/asm/arch-omap4/sys_proto.h +4 -1

index e19975efaf50..f425e3af54f5 100644
@@ -37,7 +37,7 @@ void do_set_mux(u32 base, struct pad_conf_entry const *array, int size);
 void set_muxconf_regs_essential(void);
 u32 wait_on_value(u32, u32, void *, u32);
 void sdelay(unsigned long);
-void set_pl310_ctrl_reg(u32 val);
+void omap_smc1(u32 service, u32 val);
 void setup_clocks_for_console(void);
 void prcm_init(void);
 void bypass_dpll(u32 const base);

@@ -57,4 +57,7 @@ int omap_vc_bypass_send_value(u8 sa, u8 reg_addr, u8 reg_data);
 u32 warm_reset(void);
 void force_emif_self_refresh(void);
 void setup_warmreset_time(void);
+
+#define OMAP4_SERVICE_PL310_CONTROL_REG_SET	0x102
+
 #endif

Signature · 6 lines

1.7.9.5

_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

Nishanth Menon

author

2015-03-05 17:49 UTC #7 lore raw

PATCH v4 5/11 ARM: OMAP: Change set_pl310_ctrl_reg to be generic

Quoted 32 lines · Matt Porter

Quoted 5 lines · Nishanth Menon
set_pl310_ctrl_reg does use the Secure Monitor Call (SMC) to setup PL310 control register, however, that is something that is generic enough to be used for OMAP5 generation of processors as well. The only difference being the service being invoked for the function.
Quoted 8 lines
So, convert the service to a macro and use a generic name (same as that used in Linux for some consistency). While at that, also add a data barrier which is necessary as per recommendation.
While at this, switch over to smc #0 instead of handcoded assembly. To ensure gcc compatibility, steal the strategy used by Linux kernel for sec extension builds (NOTE: we no longer use '-march=armv5' as the legacy comment claims).
Hi Nishanth,
I applied this series with fuzz and fixed a minor conflict on master. I ran into a build issue for omap3 beagle with the sec extension scheme on the gcc version 4.7.4 (Ubuntu/Linaro 4.7.4-2ubuntu1) toolchain:
arm-linux-gnueabi-gcc
-Wp,-MD,arch/arm/cpu/armv7/omap3/.lowlevel_init.o.d -nostdinc -isystem /usr/lib/gcc-cross/arm-linux-gnueabi/4.7/include -Iinclude -I../include
(quoted diff) +0 -6
-I../arch/arm/include -include ../include/linux/kconfig.h -D__KERNEL__
-D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000 -D__ASSEMBLY__ -g
-D__ARM__ -marm -mno-thumb-interwork -mabi=aapcs-linux
-mword-relocations -march=armv7-a -mno-unaligned-access
-ffunction-sections -fdata-sections -fno-common -ffixed-r9 -msoft-float
-pipe   -c -o arch/arm/cpu/armv7/omap3/lowlevel_init.o
../arch/arm/cpu/armv7/omap3/lowlevel_init.S

Quoted 9 lines

../arch/arm/cpu/armv7/omap-common/lowlevel_init.S: Assembler messages:
../arch/arm/cpu/armv7/omap-common/lowlevel_init.S:34: Error: selected processor does not support ARM mode `smc #0'
I've worked around this for the moment by placing an explicit .arch_extension sec in lowlevel_init.S but hopefully you have some thoughts on why those flags don't seem to be picked up. I'll continue to take a look at it in the meantime.

Uggh.. this is weird. I had considered ".arch_extension sec in lowlevel_init.S"

arch/arm/cpu/armv7/omap-common/Makefile ifeq ($(CONFIG_SYS_DCACHE_OFF),) · line 30

Quoted 2 lines

arch/arm/cpu/armv7/omap-common/Makefile +2 -0

+plus_sec := $(call as-instr,.arch_extension sec,+sec)
+AFLAGS_lowlevel_init.o       :=-Wa,-march=armv7-a$(plus_sec)

seems to be what we have in kernel and seems to do the job for me on

$ arm-linux-gnueabi-gcc --version
arm-linux-gnueabi-gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3

$ git clean -fdx; make omap3_beagle_defconfig;
make V=1 arch/arm/cpu/armv7/omap-common/lowlevel_init.o

with gcc 4.6:
arm-linux-gnueabi-gcc -Wp,-MD,arch/arm/cpu/armv7/omap-common/.lowlevel_init.o.d -nostdinc -isystem /usr/lib/gcc/arm-linux-gnueabi/4.6/include -Iinclude -I./arch/arm/include -include ./include/linux/kconfig.h -D__KERNEL__ -D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000 -D__ASSEMBLY__ -g -D__ARM__ -marm -mno-thumb-interwork -mabi=aapcs-linux -mword-relocations -march=armv7-a -mno-unaligned-access -ffunction-sections -fdata-sections -fno-common -ffixed-r9
-msoft-float -pipe -Wa,-march=armv7-a+sec -c -o
arch/arm/cpu/armv7/omap-common/lowlevel_init.o arch/arm/cpu/armv7/omap-common/lowlevel_init.S

with gcc 4.7:
arm-linux-gnueabi-gcc -Wp,-MD,arch/arm/cpu/armv7/omap-common/.lowlevel_init.o.d -nostdinc -isystem /usr/lib/gcc-cross/arm-linux-gnueabi/4.7/include -Iinclude -I./arch/arm/include -include ./include/linux/kconfig.h -D__KERNEL__ -D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000 -D__ASSEMBLY__ -g -D__ARM__ -marm -mno-thumb-interwork -mabi=aapcs-linux -mword-relocations -march=armv7-a -mno-unaligned-access -ffunction-sections -fdata-sections -fno-common -ffixed-r9
-msoft-float -pipe -Wa,-march=armv7-a -c -o
arch/arm/cpu/armv7/omap-common/lowlevel_init.o arch/arm/cpu/armv7/omap-common/lowlevel_init.S arch/arm/cpu/armv7/omap-common/lowlevel_init.S: Assembler messages:
arch/arm/cpu/armv7/omap-common/lowlevel_init.S:34: Error: selected processor does not support ARM mode `smc #0'
make[1]: *** [arch/arm/cpu/armv7/omap-common/lowlevel_init.o] Error 1
make: *** [arch/arm/cpu/armv7/omap-common/lowlevel_init.o] Error 2

I thought I stole the exact code from kernel, but as you can probably see -march=armv7-a+sec was generated for gcc 4.6 but -march=armv7-a without +sec for gcc 4.7!

https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/arch/arm/mach-omap2/Makefile#n44

It will be nice to have a solution that works on gcc 4.4 and above.. if we want to ignore gcc 4.4, then we can embed ".arch_extension sec in lowlevel_init.S"

https://gcc.gnu.org/ml/gcc-help/2012-07/msg00181.html

any suggestions?

[...]

Signature · 2 lines

Regards,
Nishanth Menon

Matt Porter

2015-03-05 20:22 UTC #8 lore raw

PATCH v4 5/11 ARM: OMAP: Change set_pl310_ctrl_reg to be generic

Quoted 37 lines · Nishanth Menon

Quoted 36 lines · Matt Porter
Quoted 5 lines · Nishanth Menon
set_pl310_ctrl_reg does use the Secure Monitor Call (SMC) to setup PL310 control register, however, that is something that is generic enough to be used for OMAP5 generation of processors as well. The only difference being the service being invoked for the function.
Quoted 8 lines
So, convert the service to a macro and use a generic name (same as that used in Linux for some consistency). While at that, also add a data barrier which is necessary as per recommendation.
While at this, switch over to smc #0 instead of handcoded assembly. To ensure gcc compatibility, steal the strategy used by Linux kernel for sec extension builds (NOTE: we no longer use '-march=armv5' as the legacy comment claims).
Hi Nishanth,
I applied this series with fuzz and fixed a minor conflict on master. I ran into a build issue for omap3 beagle with the sec extension scheme on the gcc version 4.7.4 (Ubuntu/Linaro 4.7.4-2ubuntu1) toolchain:
arm-linux-gnueabi-gcc
-Wp,-MD,arch/arm/cpu/armv7/omap3/.lowlevel_init.o.d -nostdinc -isystem /usr/lib/gcc-cross/arm-linux-gnueabi/4.7/include -Iinclude -I../include
(quoted diff) +0 -6
-I../arch/arm/include -include ../include/linux/kconfig.h -D__KERNEL__
-D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000 -D__ASSEMBLY__ -g
-D__ARM__ -marm -mno-thumb-interwork -mabi=aapcs-linux
-mword-relocations -march=armv7-a -mno-unaligned-access
-ffunction-sections -fdata-sections -fno-common -ffixed-r9 -msoft-float
-pipe   -c -o arch/arm/cpu/armv7/omap3/lowlevel_init.o
../arch/arm/cpu/armv7/omap3/lowlevel_init.S
../arch/arm/cpu/armv7/omap-common/lowlevel_init.S: Assembler messages:
../arch/arm/cpu/armv7/omap-common/lowlevel_init.S:34: Error: selected processor does not support ARM mode `smc #0'

Quoted 9 lines

Quoted 5 lines
I've worked around this for the moment by placing an explicit .arch_extension sec in lowlevel_init.S but hopefully you have some thoughts on why those flags don't seem to be picked up. I'll continue to take a look at it in the meantime.
Uggh.. this is weird. I had considered ".arch_extension sec in lowlevel_init.S"

Yeah, no worries, I'm not suggesting that. Just a temporary workaround.

Quoted 4 lines

arch/arm/cpu/armv7/omap-common/Makefile ifeq ($(CONFIG_SYS_DCACHE_OFF),) · line 30
Quoted 2 lines
arch/arm/cpu/armv7/omap-common/Makefile +2 -0
+plus_sec := $(call as-instr,.arch_extension sec,+sec)
+AFLAGS_lowlevel_init.o       :=-Wa,-march=armv7-a$(plus_sec)
seems to be what we have in kernel and seems to do the job for me on

right

Quoted 16 lines

$ arm-linux-gnueabi-gcc --version
arm-linux-gnueabi-gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3

$ git clean -fdx; make omap3_beagle_defconfig;
make V=1 arch/arm/cpu/armv7/omap-common/lowlevel_init.o

with gcc 4.6:
arm-linux-gnueabi-gcc

(quoted diff) +0 -8

-Wp,-MD,arch/arm/cpu/armv7/omap-common/.lowlevel_init.o.d  -nostdinc
-isystem /usr/lib/gcc/arm-linux-gnueabi/4.6/include -Iinclude
-I./arch/arm/include -include ./include/linux/kconfig.h -D__KERNEL__
-D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000  -D__ASSEMBLY__ -g
-D__ARM__ -marm -mno-thumb-interwork  -mabi=aapcs-linux
-mword-relocations  -march=armv7-a  -mno-unaligned-access
-ffunction-sections -fdata-sections -fno-common -ffixed-r9
-msoft-float  -pipe   -Wa,-march=armv7-a+sec   -c -o

Quoted 2 lines

arch/arm/cpu/armv7/omap-common/lowlevel_init.o
arch/arm/cpu/armv7/omap-common/lowlevel_init.S

I also succeed here..and on a gcc 4.8.2 toolchain

Quoted 19 lines

with gcc 4.7:
arm-linux-gnueabi-gcc
(quoted diff) +0 -8
-Wp,-MD,arch/arm/cpu/armv7/omap-common/.lowlevel_init.o.d  -nostdinc
-isystem /usr/lib/gcc-cross/arm-linux-gnueabi/4.7/include -Iinclude
-I./arch/arm/include -include ./include/linux/kconfig.h -D__KERNEL__
-D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000  -D__ASSEMBLY__ -g
-D__ARM__ -marm -mno-thumb-interwork  -mabi=aapcs-linux
-mword-relocations  -march=armv7-a  -mno-unaligned-access
-ffunction-sections -fdata-sections -fno-common -ffixed-r9
-msoft-float  -pipe  -Wa,-march=armv7-a   -c -o
arch/arm/cpu/armv7/omap-common/lowlevel_init.o arch/arm/cpu/armv7/omap-common/lowlevel_init.S arch/arm/cpu/armv7/omap-common/lowlevel_init.S: Assembler messages:
arch/arm/cpu/armv7/omap-common/lowlevel_init.S:34: Error: selected processor does not support ARM mode `smc #0'
make[1]: *** [arch/arm/cpu/armv7/omap-common/lowlevel_init.o] Error 1
make: *** [arch/arm/cpu/armv7/omap-common/lowlevel_init.o] Error 2

Quoted 3 lines

I thought I stole the exact code from kernel, but as you can probably see -march=armv7-a+sec was generated for gcc 4.6 but -march=armv7-a without +sec for gcc 4.7!

Yeah, so I played a bit with the low-level checks and noted the following results.

gcc version 4.7.4 (Ubuntu/Linaro 4.7.4-2ubuntu1):

$ printf "%b\n" ".arch_extension sec" | arm-linux-gnueabi-gcc -c -x
assembler -; echo $?
{standard input}: Assembler messages:
{standard input}:1: Error: architectural extension `sec' is not allowed
for the current base architecture
1

gcc version 4.8.2 (Ubuntu/Linaro 4.8.2-16ubuntu4)

$ printf "%b\n" ".arch_extension sec" | arm-linux-gnueabihf-gcc -c -x
assembler -; echo $?
0

My 4.7.4 toolchain's gas does support -march=armv7-a+sec however I need to pass that architecture in specifically to make it pass:

$ printf "%b\n" ".arch_extension sec" | arm-linux-gnueabi-gcc
-Wa,-march=armv7-a+sec -c -x assembler -; echo $?
0

Quoted 11 lines

https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/arch/arm/mach-omap2/Makefile#n44
It will be nice to have a solution that works on gcc 4.4 and above.. if we want to ignore gcc 4.4, then we can embed ".arch_extension sec in lowlevel_init.S"
https://gcc.gnu.org/ml/gcc-help/2012-07/msg00181.html
any suggestions?

Not sure of the best approach here. It seems if the actual check engages the -march option for gas (on an omap build) we should build for any of them. That's a bit hacky though. But it's just as easy to just add the .arch_extension sec as you suggest and we have to quit supporting gcc 4.4 in either case AFAICT.

-Matt

Tom Rini

2015-03-05 21:36 UTC #9 lore raw

PATCH v4 5/11 ARM: OMAP: Change set_pl310_ctrl_reg to be generic

In reply to Nishanth Menon

Quoted 93 lines · Nishanth Menon

Quoted 32 lines · Matt Porter
Quoted 5 lines · Nishanth Menon
set_pl310_ctrl_reg does use the Secure Monitor Call (SMC) to setup PL310 control register, however, that is something that is generic enough to be used for OMAP5 generation of processors as well. The only difference being the service being invoked for the function.
Quoted 8 lines
So, convert the service to a macro and use a generic name (same as that used in Linux for some consistency). While at that, also add a data barrier which is necessary as per recommendation.
While at this, switch over to smc #0 instead of handcoded assembly. To ensure gcc compatibility, steal the strategy used by Linux kernel for sec extension builds (NOTE: we no longer use '-march=armv5' as the legacy comment claims).
Hi Nishanth,
I applied this series with fuzz and fixed a minor conflict on master. I ran into a build issue for omap3 beagle with the sec extension scheme on the gcc version 4.7.4 (Ubuntu/Linaro 4.7.4-2ubuntu1) toolchain:
arm-linux-gnueabi-gcc
-Wp,-MD,arch/arm/cpu/armv7/omap3/.lowlevel_init.o.d -nostdinc -isystem /usr/lib/gcc-cross/arm-linux-gnueabi/4.7/include -Iinclude -I../include
(quoted diff) +0 -6
-I../arch/arm/include -include ../include/linux/kconfig.h -D__KERNEL__
-D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000 -D__ASSEMBLY__ -g
-D__ARM__ -marm -mno-thumb-interwork -mabi=aapcs-linux
-mword-relocations -march=armv7-a -mno-unaligned-access
-ffunction-sections -fdata-sections -fno-common -ffixed-r9 -msoft-float
-pipe   -c -o arch/arm/cpu/armv7/omap3/lowlevel_init.o
../arch/arm/cpu/armv7/omap3/lowlevel_init.S
Quoted 9 lines
../arch/arm/cpu/armv7/omap-common/lowlevel_init.S: Assembler messages:
../arch/arm/cpu/armv7/omap-common/lowlevel_init.S:34: Error: selected processor does not support ARM mode `smc #0'
I've worked around this for the moment by placing an explicit .arch_extension sec in lowlevel_init.S but hopefully you have some thoughts on why those flags don't seem to be picked up. I'll continue to take a look at it in the meantime.
Uggh.. this is weird. I had considered ".arch_extension sec in lowlevel_init.S"
arch/arm/cpu/armv7/omap-common/Makefile ifeq ($(CONFIG_SYS_DCACHE_OFF),) · line 30
Quoted 2 lines
arch/arm/cpu/armv7/omap-common/Makefile +2 -0
+plus_sec := $(call as-instr,.arch_extension sec,+sec)
+AFLAGS_lowlevel_init.o       :=-Wa,-march=armv7-a$(plus_sec)
seems to be what we have in kernel and seems to do the job for me on
$ arm-linux-gnueabi-gcc --version
arm-linux-gnueabi-gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3
$ git clean -fdx; make omap3_beagle_defconfig;
make V=1 arch/arm/cpu/armv7/omap-common/lowlevel_init.o
with gcc 4.6:
arm-linux-gnueabi-gcc
(quoted diff) +0 -8
-Wp,-MD,arch/arm/cpu/armv7/omap-common/.lowlevel_init.o.d  -nostdinc
-isystem /usr/lib/gcc/arm-linux-gnueabi/4.6/include -Iinclude
-I./arch/arm/include -include ./include/linux/kconfig.h -D__KERNEL__
-D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000  -D__ASSEMBLY__ -g
-D__ARM__ -marm -mno-thumb-interwork  -mabi=aapcs-linux
-mword-relocations  -march=armv7-a  -mno-unaligned-access
-ffunction-sections -fdata-sections -fno-common -ffixed-r9
-msoft-float  -pipe   -Wa,-march=armv7-a+sec   -c -o
arch/arm/cpu/armv7/omap-common/lowlevel_init.o
arch/arm/cpu/armv7/omap-common/lowlevel_init.S
with gcc 4.7:
arm-linux-gnueabi-gcc
(quoted diff) +0 -8
-Wp,-MD,arch/arm/cpu/armv7/omap-common/.lowlevel_init.o.d  -nostdinc
-isystem /usr/lib/gcc-cross/arm-linux-gnueabi/4.7/include -Iinclude
-I./arch/arm/include -include ./include/linux/kconfig.h -D__KERNEL__
-D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000  -D__ASSEMBLY__ -g
-D__ARM__ -marm -mno-thumb-interwork  -mabi=aapcs-linux
-mword-relocations  -march=armv7-a  -mno-unaligned-access
-ffunction-sections -fdata-sections -fno-common -ffixed-r9
-msoft-float  -pipe  -Wa,-march=armv7-a   -c -o
arch/arm/cpu/armv7/omap-common/lowlevel_init.o arch/arm/cpu/armv7/omap-common/lowlevel_init.S arch/arm/cpu/armv7/omap-common/lowlevel_init.S: Assembler messages:
arch/arm/cpu/armv7/omap-common/lowlevel_init.S:34: Error: selected processor does not support ARM mode `smc #0'
make[1]: *** [arch/arm/cpu/armv7/omap-common/lowlevel_init.o] Error 1
make: *** [arch/arm/cpu/armv7/omap-common/lowlevel_init.o] Error 2
I thought I stole the exact code from kernel, but as you can probably see -march=armv7-a+sec was generated for gcc 4.6 but -march=armv7-a without +sec for gcc 4.7!

Quoted 7 lines

https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/arch/arm/mach-omap2/Makefile#n44
It will be nice to have a solution that works on gcc 4.4 and above.. if we want to ignore gcc 4.4, then we can embed ".arch_extension sec in lowlevel_init.S"
https://gcc.gnu.org/ml/gcc-help/2012-07/msg00181.html

Sigh. I guess I'll give up on getting rid of the hand-crafted smc #0 call, it's being more troulbe than it's worth. Thanks for digging into it tho!

Signature · 8 lines

Tom
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20150305/e8402ad2/attachment.sig>

Nishanth Menon

author

2015-03-05 22:27 UTC #10 lore raw

PATCH v4 5/11 ARM: OMAP: Change set_pl310_ctrl_reg to be generic

Quoted 100 lines · Tom Rini

Quoted 99 lines · Nishanth Menon
Quoted 32 lines · Matt Porter
Quoted 5 lines · Nishanth Menon
set_pl310_ctrl_reg does use the Secure Monitor Call (SMC) to setup PL310 control register, however, that is something that is generic enough to be used for OMAP5 generation of processors as well. The only difference being the service being invoked for the function.
Quoted 8 lines
So, convert the service to a macro and use a generic name (same as that used in Linux for some consistency). While at that, also add a data barrier which is necessary as per recommendation.
While at this, switch over to smc #0 instead of handcoded assembly. To ensure gcc compatibility, steal the strategy used by Linux kernel for sec extension builds (NOTE: we no longer use '-march=armv5' as the legacy comment claims).
Hi Nishanth,
I applied this series with fuzz and fixed a minor conflict on master. I ran into a build issue for omap3 beagle with the sec extension scheme on the gcc version 4.7.4 (Ubuntu/Linaro 4.7.4-2ubuntu1) toolchain:
arm-linux-gnueabi-gcc
-Wp,-MD,arch/arm/cpu/armv7/omap3/.lowlevel_init.o.d -nostdinc -isystem /usr/lib/gcc-cross/arm-linux-gnueabi/4.7/include -Iinclude -I../include
(quoted diff) +0 -6
-I../arch/arm/include -include ../include/linux/kconfig.h -D__KERNEL__
-D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000 -D__ASSEMBLY__ -g
-D__ARM__ -marm -mno-thumb-interwork -mabi=aapcs-linux
-mword-relocations -march=armv7-a -mno-unaligned-access
-ffunction-sections -fdata-sections -fno-common -ffixed-r9 -msoft-float
-pipe   -c -o arch/arm/cpu/armv7/omap3/lowlevel_init.o
../arch/arm/cpu/armv7/omap3/lowlevel_init.S
Quoted 9 lines
../arch/arm/cpu/armv7/omap-common/lowlevel_init.S: Assembler messages:
../arch/arm/cpu/armv7/omap-common/lowlevel_init.S:34: Error: selected processor does not support ARM mode `smc #0'
I've worked around this for the moment by placing an explicit .arch_extension sec in lowlevel_init.S but hopefully you have some thoughts on why those flags don't seem to be picked up. I'll continue to take a look at it in the meantime.
Uggh.. this is weird. I had considered ".arch_extension sec in lowlevel_init.S"
arch/arm/cpu/armv7/omap-common/Makefile ifeq ($(CONFIG_SYS_DCACHE_OFF),) · line 30
Quoted 2 lines
arch/arm/cpu/armv7/omap-common/Makefile +2 -0
+plus_sec := $(call as-instr,.arch_extension sec,+sec)
+AFLAGS_lowlevel_init.o       :=-Wa,-march=armv7-a$(plus_sec)
seems to be what we have in kernel and seems to do the job for me on
$ arm-linux-gnueabi-gcc --version
arm-linux-gnueabi-gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3
$ git clean -fdx; make omap3_beagle_defconfig;
make V=1 arch/arm/cpu/armv7/omap-common/lowlevel_init.o
with gcc 4.6:
arm-linux-gnueabi-gcc
(quoted diff) +0 -8
-Wp,-MD,arch/arm/cpu/armv7/omap-common/.lowlevel_init.o.d  -nostdinc
-isystem /usr/lib/gcc/arm-linux-gnueabi/4.6/include -Iinclude
-I./arch/arm/include -include ./include/linux/kconfig.h -D__KERNEL__
-D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000  -D__ASSEMBLY__ -g
-D__ARM__ -marm -mno-thumb-interwork  -mabi=aapcs-linux
-mword-relocations  -march=armv7-a  -mno-unaligned-access
-ffunction-sections -fdata-sections -fno-common -ffixed-r9
-msoft-float  -pipe   -Wa,-march=armv7-a+sec   -c -o
arch/arm/cpu/armv7/omap-common/lowlevel_init.o
arch/arm/cpu/armv7/omap-common/lowlevel_init.S
with gcc 4.7:
arm-linux-gnueabi-gcc
(quoted diff) +0 -8
-Wp,-MD,arch/arm/cpu/armv7/omap-common/.lowlevel_init.o.d  -nostdinc
-isystem /usr/lib/gcc-cross/arm-linux-gnueabi/4.7/include -Iinclude
-I./arch/arm/include -include ./include/linux/kconfig.h -D__KERNEL__
-D__UBOOT__ -DCONFIG_SYS_TEXT_BASE=0x80100000  -D__ASSEMBLY__ -g
-D__ARM__ -marm -mno-thumb-interwork  -mabi=aapcs-linux
-mword-relocations  -march=armv7-a  -mno-unaligned-access
-ffunction-sections -fdata-sections -fno-common -ffixed-r9
-msoft-float  -pipe  -Wa,-march=armv7-a   -c -o
arch/arm/cpu/armv7/omap-common/lowlevel_init.o arch/arm/cpu/armv7/omap-common/lowlevel_init.S arch/arm/cpu/armv7/omap-common/lowlevel_init.S: Assembler messages:
arch/arm/cpu/armv7/omap-common/lowlevel_init.S:34: Error: selected processor does not support ARM mode `smc #0'
make[1]: *** [arch/arm/cpu/armv7/omap-common/lowlevel_init.o] Error 1
make: *** [arch/arm/cpu/armv7/omap-common/lowlevel_init.o] Error 2
I thought I stole the exact code from kernel, but as you can probably see -march=armv7-a+sec was generated for gcc 4.6 but -march=armv7-a without +sec for gcc 4.7!
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/tree/arch/arm/mach-omap2/Makefile#n44
It will be nice to have a solution that works on gcc 4.4 and above.. if we want to ignore gcc 4.4, then we can embed ".arch_extension sec in lowlevel_init.S"

Quoted 5 lines

Quoted 1 line
https://gcc.gnu.org/ml/gcc-help/2012-07/msg00181.html
Sigh. I guess I'll give up on getting rid of the hand-crafted smc #0 call, it's being more troulbe than it's worth. Thanks for digging into it tho!

Alright, I will handcraft the smc code back in and keep us gcc safe :D... will wait till we figure out what the heck i screwed up[1] in the cpuid detect logic before i post a new rev. Sadly, i have'nt found a board that kind of matches what Matt has :(.

[1] http://lists.denx.de/pipermail/u-boot/2015-March/207313.html

Signature · 3 lines

---
Regards,
Nishanth Menon

Nishanth Menon

author

2015-03-03 22:26 UTC #11 lore raw

PATCH v4 6/11 ARM: OMAP3: Rename omap3.h to omap.h to be generic as all SoCs

In reply to Nishanth Menon

This is in preperation of using generic cross OMAP code.

Signed-off-by Nishanth Menon <[email protected]>

---
 .../arm/include/asm/arch-omap3/{omap3.h => omap.h} |    0
 include/configs/am3517_crane.h                     |    2 +-
 include/configs/am3517_evm.h                       |    2 +-
 include/configs/cm_t35.h                           |    2 +-
 include/configs/cm_t3517.h                         |    2 +-
 include/configs/dig297.h                           |    2 +-
 include/configs/mcx.h                              |    2 +-
 include/configs/nokia_rx51.h                       |    2 +-
 include/configs/omap3_evm.h                        |    2 +-
 include/configs/omap3_evm_quick_mmc.h              |    2 +-
 include/configs/omap3_evm_quick_nand.h             |    2 +-
 include/configs/omap3_logic.h                      |    2 +-
 include/configs/omap3_mvblx.h                      |    2 +-
 include/configs/omap3_pandora.h                    |    2 +-
 include/configs/omap3_sdp3430.h                    |    2 +-
 include/configs/omap3_zoom1.h                      |    2 +-
 include/configs/tam3517-common.h                   |    2 +-
 include/configs/tao3530.h                          |    2 +-
 include/configs/ti_omap3_common.h                  |    2 +-
 include/configs/tricorder.h                        |    2 +-
 20 files changed, 19 insertions(+), 19 deletions(-)
 rename arch/arm/include/asm/arch-omap3/{omap3.h => omap.h} (100%)

arch/arm/include/asm/arch-omap3/omap.h renamed from arch/arm/include/asm/arch-omap3/omap3.h +0 -0

similarity index 100%
rename from arch/arm/include/asm/arch-omap3/omap3.h
rename to arch/arm/include/asm/arch-omap3/omap.h

include/configs/am3517_crane.h +1 -1

index 09ee10c0590b..a4128773ddb2 100644
@@ -23,7 +23,7 @@
 #define CONFIG_EMIF4	/* The chip has EMIF4 controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /*
  * Display CPU and Board information

include/configs/am3517_evm.h +1 -1

index 190ef0e71bfb..94398c625eb0 100644
@@ -23,7 +23,7 @@
 #define CONFIG_EMIF4	/* The chip has EMIF4 controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /*
  * Display CPU and Board information

include/configs/cm_t35.h +1 -1

index 9767512a5cf0..56a5506f5c78 100644
@@ -30,7 +30,7 @@
 #define CONFIG_SDRC	/* The chip has SDRC controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /*
  * Display CPU and Board information

include/configs/cm_t3517.h +1 -1

index 918032bd753a..ee1b7a0cb5a7 100644
@@ -30,7 +30,7 @@
 #define CONFIG_EMIF4	/* The chip has EMIF4 controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /*
  * Display CPU and Board information

include/configs/dig297.h +1 -1

index c8739ed29490..8791199fc61d 100644
@@ -36,7 +36,7 @@
 #define CONFIG_SDRC	/* The chip has SDRC controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /*
  * Display CPU and Board information

include/configs/mcx.h +1 -1

index 26eb2203540f..2cf66c40d6fe 100644
@@ -26,7 +26,7 @@
 #define CONFIG_EMIF4	/* The chip has EMIF4 controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 #define CONFIG_OF_LIBFDT
 #define CONFIG_FIT

include/configs/nokia_rx51.h +1 -1

index 46fc91e5e197..442e16ae686b 100644
@@ -42,7 +42,7 @@
 #define CONFIG_SDRC			/* The chip has SDRC controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 #include <asm/arch/mem.h>
 #include <linux/stringify.h>

include/configs/omap3_evm.h +1 -1

index 8bdc08f5864c..4e587e10ffd3 100644
@@ -18,7 +18,7 @@
 #define __OMAP3EVM_CONFIG_H
 
 #include <asm/arch/cpu.h>
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /* ----------------------------------------------------------------------------
  * Supported U-boot commands

include/configs/omap3_evm_quick_mmc.h +1 -1

index 1185f425501f..a7acc1becbeb 100644
@@ -13,7 +13,7 @@
 #define __OMAP3_EVM_QUICK_MMC_H
 
 #include <asm/arch/cpu.h>
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /* ----------------------------------------------------------------------------
  * Supported U-boot commands

include/configs/omap3_evm_quick_nand.h +1 -1

index 4427e88b7e4b..c6cad9388961 100644
@@ -13,7 +13,7 @@
 #define __OMAP3_EVM_QUICK_NAND_H
 
 #include <asm/arch/cpu.h>
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /* ----------------------------------------------------------------------------
  * Supported U-boot commands

include/configs/omap3_logic.h +1 -1

index aeb385f5ff6b..fcef46735835 100644
@@ -24,7 +24,7 @@
 #define CONFIG_SDRC	/* The chip has SDRC controller */
 
 #include <asm/arch/cpu.h>	/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /*
  * Display CPU and Board information

include/configs/omap3_mvblx.h +1 -1

index bf1d34dedb4d..20ec3ad15439 100644
@@ -27,7 +27,7 @@
 #define CONFIG_SDRC	/* The chip has SDRC controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /*
  * Display CPU and Board information

include/configs/omap3_pandora.h +1 -1

index 45feeb577320..b92d67abccf9 100644
@@ -21,7 +21,7 @@
 #define CONFIG_SDRC	/* The chip has SDRC controller */
 
 #include <asm/arch/cpu.h>	/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /*
  * Display CPU and Board information

include/configs/omap3_sdp3430.h +1 -1

index ac307eb45629..bf5c7a8faf77 100644
@@ -27,7 +27,7 @@
 #define CONFIG_SDRC	/* The chip has SDRC controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /*
  * NOTE:  these #defines presume standard SDP jumper settings.

include/configs/omap3_zoom1.h +1 -1

index c5d742c2bd80..611cd5e1a350 100644
@@ -22,7 +22,7 @@
 #define CONFIG_NAND
 #define CONFIG_NR_DRAM_BANKS	2	/* CS1 may or may not be populated */
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 #include <configs/ti_omap3_common.h>
 
 /* Remove SPL boot option - we do not support that on LDP yet */

include/configs/tam3517-common.h +1 -1

index 9fbe68a8a2f2..bff9687b3c9c 100644
@@ -25,7 +25,7 @@
 #define CONFIG_EMIF4	/* The chip has EMIF4 controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /*
  * Display CPU and Board information

include/configs/tao3530.h +1 -1

index 7d2c0d2fa75c..4dfd5cb9fbb7 100644
@@ -27,7 +27,7 @@
 #define CONFIG_SDRC			/* Has an SDRC controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 /*
  * Display CPU and Board information

include/configs/ti_omap3_common.h +1 -1

index 3c634ee680d0..f909f6b94a18 100644
@@ -16,7 +16,7 @@
 
 
 #include <asm/arch/cpu.h>
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 #ifndef CONFIG_SPL_BUILD
 # define CONFIG_DM

include/configs/tricorder.h +1 -1

index 36621a553cdf..f27cd1b9b061 100644
@@ -32,7 +32,7 @@
 #define CONFIG_SDRC			/* The chip has SDRC controller */
 
 #include <asm/arch/cpu.h>		/* get chip and board defs */
-#include <asm/arch/omap3.h>
+#include <asm/arch/omap.h>
 
 #define CONFIG_SYS_GENERIC_BOARD

Signature · 1 line

1.7.9.5

Nishanth Menon

author

2015-03-03 22:26 UTC #12 lore raw

PATCH v4 7/11 ARM: OMAP3: Get rid of omap3_gp_romcode_call and replace with omap_smc1

In reply to Nishanth Menon

omap_smc1 is now generic enough to remove duplicate omap3_gp_romcode_call logic that omap3 introduced.

As part of this change, move to using the generic lowlevel_init.S for omap3 as well.

Signed-off-by Nishanth Menon <[email protected]>

---
 arch/arm/cpu/armv7/omap-common/Makefile        |    2 +-
 arch/arm/cpu/armv7/omap-common/lowlevel_init.S |    2 ++
 arch/arm/cpu/armv7/omap3/board.c               |    6 ++----
 arch/arm/cpu/armv7/omap3/lowlevel_init.S       |   11 -----------
 arch/arm/include/asm/arch-omap3/sys_proto.h    |    2 +-
 5 files changed, 6 insertions(+), 17 deletions(-)

arch/arm/cpu/armv7/omap-common/Makefile +1 -1

index b36e85d25f3f..1d5970f340d0 100644
@@ -31,7 +31,7 @@ AFLAGS_lowlevel_init.o	:=-Wa,-march=armv7-a$(plus_sec)
 
 ifeq ($(CONFIG_OMAP34XX),)
 obj-y	+= boot-common.o
-obj-y	+= lowlevel_init.o
 endif
+obj-y	+= lowlevel_init.o
 
 obj-y	+= mem-common.o

arch/arm/cpu/armv7/omap-common/lowlevel_init.S +2 -0

index 83426291b22d..23c9e197f0e8 100644
@@ -16,11 +16,13 @@
 #include <asm/arch/spl.h>
 #include <linux/linkage.h>
 
+#ifndef CONFIG_OMAP34XX
 ENTRY(save_boot_params)
 	ldr	r1, =OMAP_SRAM_SCRATCH_BOOT_PARAMS
 	str	r0, [r1]
 	bx	lr
 ENDPROC(save_boot_params)
+#endif
 
 ENTRY(omap_smc1)
 	PUSH	{r4-r12, lr}	@ save registers - ROM code may pollute

arch/arm/cpu/armv7/omap3/board.c +2 -4

index 90d6ae7bb5f5..51abc4b09e36 100644
@@ -418,8 +418,7 @@ static void omap3_update_aux_cr_secure(u32 set_bits, u32 clear_bits)
 	acr |= set_bits;
 
 	if (get_device_type() == GP_DEVICE) {
-		omap3_gp_romcode_call(OMAP3_GP_ROMCODE_API_WRITE_ACR,
-				       acr);
+		omap_smc1(OMAP3_GP_ROMCODE_API_WRITE_ACR, acr);
 	} else {
 		struct emu_hal_params emu_romcode_params;
 		emu_romcode_params.num_params = 1;

@@ -459,8 +458,7 @@ static void omap3_update_aux_cr(u32 set_bits, u32 clear_bits)
 static void omap3_invalidate_l2_cache_secure(void)
 {
 	if (get_device_type() == GP_DEVICE) {
-		omap3_gp_romcode_call(OMAP3_GP_ROMCODE_API_L2_INVAL,
-				      0);
+		omap_smc1(OMAP3_GP_ROMCODE_API_L2_INVAL, 0);
 	} else {
 		struct emu_hal_params emu_romcode_params;
 		emu_romcode_params.num_params = 1;

arch/arm/cpu/armv7/omap3/lowlevel_init.S +0 ~~-11~~

index 78577b1d1c75..84591b8b5bab 100644
@@ -27,17 +27,6 @@ ENTRY(save_boot_params)
 ENDPROC(save_boot_params)
 #endif
 
-ENTRY(omap3_gp_romcode_call)
-	PUSH {r4-r12, lr} @ Save all registers from ROM code!
-	MOV r12, r0	@ Copy the Service ID in R12
-	MOV r0, r1	@ Copy parameter to R0
-	mcr     p15, 0, r0, c7, c10, 4	@ DSB
-	mcr     p15, 0, r0, c7, c10, 5	@ DMB
-	.word	0xe1600070	@ SMC #0 to enter monitor - hand assembled
-				@ because we use -march=armv5
-	POP {r4-r12, pc}
-ENDPROC(omap3_gp_romcode_call)
-
 /*
  * Funtion for making PPA HAL API calls in secure devices
  * Input:

arch/arm/include/asm/arch-omap3/sys_proto.h +1 -1

index bcf92fbe658b..c06605d533d3 100644
@@ -73,6 +73,6 @@ void power_init_r(void);
 void dieid_num_r(void);
 void get_dieid(u32 *id);
 void do_omap3_emu_romcode_call(u32 service_id, u32 parameters);
-void omap3_gp_romcode_call(u32 service_id, u32 parameter);
+void omap_smc1(u32 service, u32 val);
 u32 warm_reset(void);
 #endif

Signature · 1 line

1.7.9.5

Nishanth Menon

author

2015-03-03 22:26 UTC #13 lore raw

PATCH v4 8/11 ARM: DRA7 / OMAP5: Add workaround for ARM errata 798870

In reply to Nishanth Menon

From: Praveen Rao <[email protected]>

This patch enables the workaround for ARM errata 798870 for OMAP5 / DRA7 which says "If back-to-back speculative cache line fills (fill A and fill B) are issued from the L1 data cache of a CPU to the L2 cache, the second request (fill B) is then cancelled, and the second request would have detected a hazard against a recent write or eviction (write B) to the same cache line as fill B then the L2 logic might deadlock."

An l2auxctlr accessor implementation for OMAP5 and DRA7 is introduced here as well.

Signed-off-by Praveen Rao <[email protected]>

Signed-off-by Angela Stegmaier <[email protected]>

Signed-off-by Nishanth Menon <[email protected]>

---
 arch/arm/cpu/armv7/omap5/hwinit.c           |    7 +++++++
 arch/arm/include/asm/arch-omap5/sys_proto.h |    4 ++++
 include/configs/ti_omap5_common.h           |    3 +++
 3 files changed, 14 insertions(+)

arch/arm/cpu/armv7/omap5/hwinit.c +7 -0

index a8a474a88be9..f8060555b680 100644
@@ -381,3 +381,10 @@ void setup_warmreset_time(void)
 	rst_val |= rst_time;
 	writel(rst_val, (*prcm)->prm_rsttime);
 }
+
+void v7_arch_cp15_set_l2aux_ctrl(u32 l2auxctrl, u32 cpu_midr,
+				 u32 cpu_rev_comb, u32 cpu_variant,
+				 u32 cpu_rev)
+{
+	omap_smc1(OMAP5_SERVICE_L2ACTLR_SET, l2auxctrl);
+}

arch/arm/include/asm/arch-omap5/sys_proto.h +4 -0

index 103830319a41..37dc9925eae2 100644
@@ -66,4 +66,8 @@ static inline u32 usec_to_32k(u32 usec)
 {
 	return div_round_up(32768 * usec, 1000000);
 }
+
+void omap_smc1(u32 service, u32 val);
+#define OMAP5_SERVICE_L2ACTLR_SET    0x104
+
 #endif

include/configs/ti_omap5_common.h +3 -0

index 925cb42dd38d..09f05f18a75d 100644
@@ -21,6 +21,9 @@
 #define CONFIG_DISPLAY_BOARDINFO
 #define CONFIG_ARCH_CPU_INIT
 
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_798870
+
 #define CONFIG_SYS_CACHELINE_SIZE	64
 
 /* Use General purpose timer 1 */

Signature · 1 line

1.7.9.5

Nishanth Menon

author

2015-03-03 22:26 UTC #14 lore raw

PATCH v4 9/11 ARM: OMAP5 / DRA7: Setup L2 Aux Control Register with recommended configuration

In reply to Nishanth Menon

Update to existing recommendation for L2ACTLR configuration to prevent system instability and optimize performance.

These apply to both OMAP5 and DRA7.

Reported-by Vivek Chengalvala <[email protected]>

Signed-off-by Nishanth Menon <[email protected]>

---
 arch/arm/cpu/armv7/omap5/hwinit.c |   16 ++++++++++++++++
 1 file changed, 16 insertions(+)

arch/arm/cpu/armv7/omap5/hwinit.c +16 -0

index f8060555b680..8d6b59eeb044 100644
@@ -304,6 +304,21 @@ void config_data_eye_leveling_samples(u32 emif_base)
 		       (*ctrl)->control_emif2_sdram_config_ext);
 }
 
+void init_cpu_configuration(void)
+{
+	u32 l2actlr;
+
+	asm volatile("mrc p15, 1, %0, c15, c0, 0" : "=r"(l2actlr));
+	/*
+	 * L2ACTLR: Ensure to enable the following:
+	 * 3: Disable clean/evict push to external
+	 * 4: Disable WriteUnique and WriteLineUnique transactions from master
+	 * 8: Disable DVM/CMO message broadcast
+	 */
+	l2actlr |= 0x118;
+	omap_smc1(OMAP5_SERVICE_L2ACTLR_SET, l2actlr);
+}
+
 void init_omap_revision(void)
 {
 	/*

@@ -342,6 +357,7 @@ void init_omap_revision(void)
 	default:
 		*omap_si_rev = OMAP5430_SILICON_ID_INVALID;
 	}
+	init_cpu_configuration();
 }
 
 void reset_cpu(ulong ignored)

Signature · 1 line

1.7.9.5

Nishanth Menon

author

2015-03-03 22:26 UTC #15 lore raw

PATCH v4 10/11 ARM: OMAP3: Enable workaround for ARM errata 454179, 430973, 621766

In reply to Nishanth Menon

Enable the OMAP3 specific errata code for 454179, 430973, 621766 and while at it, remove legacy non-revision checked errata logic.

Signed-off-by Nishanth Menon <[email protected]>

---
 arch/arm/cpu/armv7/omap3/board.c   |   31 ++++++++++---------------------
 include/configs/am3517_crane.h     |    4 ++++
 include/configs/am3517_evm.h       |    4 ++++
 include/configs/cm_t35.h           |    4 ++++
 include/configs/cm_t3517.h         |    4 ++++
 include/configs/dig297.h           |    4 ++++
 include/configs/mcx.h              |    4 ++++
 include/configs/omap3_evm_common.h |    4 ++++
 include/configs/omap3_logic.h      |    4 ++++
 include/configs/omap3_mvblx.h      |    4 ++++
 include/configs/omap3_pandora.h    |    4 ++++
 include/configs/omap3_sdp3430.h    |    4 ++++
 include/configs/tam3517-common.h   |    4 ++++
 include/configs/tao3530.h          |    4 ++++
 include/configs/ti_omap3_common.h  |    5 +++++
 include/configs/tricorder.h        |    4 ++++
 16 files changed, 71 insertions(+), 21 deletions(-)

arch/arm/cpu/armv7/omap3/board.c +10 ~~-21~~

index 51abc4b09e36..5587a10cdafd 100644
@@ -35,7 +35,6 @@ DECLARE_GLOBAL_DATA_PTR;
 
 /* Declarations */
 extern omap3_sysinfo sysinfo;
-static void omap3_setup_aux_cr(void);
 #ifndef CONFIG_SYS_L2CACHE_OFF
 static void omap3_invalidate_l2_cache_secure(void);
 #endif

@@ -246,9 +245,6 @@ void s_init(void)
 
 	try_unlock_memory();
 
-	/* Errata workarounds */
-	omap3_setup_aux_cr();
-
 #ifndef CONFIG_SYS_L2CACHE_OFF
 	/* Invalidate L2-cache from secure mode */
 	omap3_invalidate_l2_cache_secure();

@@ -408,15 +404,9 @@ static void omap3_emu_romcode_call(u32 service_id, u32 *parameters)
 	do_omap3_emu_romcode_call(service_id, OMAP3_PUBLIC_SRAM_SCRATCH_AREA);
 }
 
-static void omap3_update_aux_cr_secure(u32 set_bits, u32 clear_bits)
+void v7_arch_cp15_set_acr(u32 acr, u32 cpu_midr, u32 cpu_rev_comb,
+			  u32 cpu_variant, u32 cpu_rev)
 {
-	u32 acr;
-
-	/* Read ACR */
-	asm volatile ("mrc p15, 0, %0, c1, c0, 1" : "=r" (acr));
-	acr &= ~clear_bits;
-	acr |= set_bits;
-
 	if (get_device_type() == GP_DEVICE) {
 		omap_smc1(OMAP3_GP_ROMCODE_API_WRITE_ACR, acr);
 	} else {

@@ -428,16 +418,15 @@ static void omap3_update_aux_cr_secure(u32 set_bits, u32 clear_bits)
 	}
 }
 
-static void omap3_setup_aux_cr(void)
+static void omap3_update_aux_cr_secure(u32 set_bits, u32 clear_bits)
 {
-	/* Workaround for Cortex-A8 errata: #454179 #430973
-	 *	Set "IBE" bit
-	 *	Set "Disable Branch Size Mispredicts" bit
-	 * Workaround for erratum #621766
-	 *	Enable L1NEON bit
-	 * ACR |= (IBE | DBSM | L1NEON) => ACR |= 0xE0
-	 */
-	omap3_update_aux_cr_secure(0xE0, 0);
+	u32 acr;
+
+	/* Read ACR */
+	asm volatile ("mrc p15, 0, %0, c1, c0, 1" : "=r" (acr));
+	acr &= ~clear_bits;
+	acr |= set_bits;
+	v7_arch_cp15_set_acr(acr, 0, 0, 0, 0);
 }
 
 #ifndef CONFIG_SYS_L2CACHE_OFF

include/configs/am3517_crane.h +4 -0

index a4128773ddb2..39e121d9bc99 100644
@@ -19,6 +19,10 @@
 #define CONFIG_OMAP		1	/* in a TI OMAP core */
 #define CONFIG_OMAP3_AM3517CRANE	1	/* working with CRANEBOARD */
 #define CONFIG_OMAP_COMMON
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_EMIF4	/* The chip has EMIF4 controller */

include/configs/am3517_evm.h +4 -0

index 94398c625eb0..930d7af31f18 100644
@@ -19,6 +19,10 @@
 #define CONFIG_OMAP		1	/* in a TI OMAP core */
 #define CONFIG_OMAP3_AM3517EVM	1	/* working with AM3517EVM */
 #define CONFIG_OMAP_COMMON
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_EMIF4	/* The chip has EMIF4 controller */

include/configs/cm_t35.h +4 -0

index 56a5506f5c78..451cb5e2fb4f 100644
@@ -26,6 +26,10 @@
 #define CONFIG_CM_T3X	/* working with CM-T35 and CM-T3730 */
 #define CONFIG_OMAP_COMMON
 #define CONFIG_SYS_GENERIC_BOARD
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_SDRC	/* The chip has SDRC controller */

include/configs/cm_t3517.h +4 -0

index ee1b7a0cb5a7..8c6313832250 100644
@@ -17,6 +17,10 @@
 #define CONFIG_CM_T3517	/* working with CM-T3517 */
 #define CONFIG_OMAP_COMMON
 #define CONFIG_SYS_GENERIC_BOARD
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_SYS_TEXT_BASE	0x80008000

include/configs/dig297.h +4 -0

index 8791199fc61d..932640112b3a 100644
@@ -23,6 +23,10 @@
 #define MACH_TYPE_OMAP3_CPS 2751
 #endif
 #define CONFIG_MACH_TYPE MACH_TYPE_OMAP3_CPS
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 /*
  * High Level Configuration Options

include/configs/mcx.h +4 -0

index 2cf66c40d6fe..3fd31842451d 100644
@@ -16,6 +16,10 @@
 #define CONFIG_OMAP3_MCX		/* working with mcx */
 #define CONFIG_OMAP_GPIO
 #define CONFIG_OMAP_COMMON
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define MACH_TYPE_MCX			3656
 #define CONFIG_MACH_TYPE	MACH_TYPE_MCX

include/configs/omap3_evm_common.h +4 -0

index 8885e1758106..aeb769548f92 100644
@@ -15,6 +15,10 @@
 #define CONFIG_OMAP			/* This is TI OMAP core */
 #define CONFIG_OMAP_GPIO
 #define CONFIG_OMAP_COMMON
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_SDRC			/* The chip has SDRC controller */

include/configs/omap3_logic.h +4 -0

index fcef46735835..13cd35b16a03 100644
@@ -18,6 +18,10 @@
 #define CONFIG_OMAP3_LOGIC		/* working with Logic OMAP boards */
 #define CONFIG_OMAP_GPIO
 #define CONFIG_OMAP_COMMON
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_SYS_TEXT_BASE	0x80400000

include/configs/omap3_mvblx.h +4 -0

index 20ec3ad15439..b61297f9ee94 100644
@@ -23,6 +23,10 @@
 #define CONFIG_MACH_TYPE	MACH_TYPE_MVBLX
 #define CONFIG_OMAP_GPIO
 #define CONFIG_OMAP_COMMON
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_SDRC	/* The chip has SDRC controller */

include/configs/omap3_pandora.h +4 -0

index b92d67abccf9..11d7b86eb10d 100644
@@ -17,6 +17,10 @@
 #define CONFIG_OMAP3_PANDORA	1	/* working with pandora */
 #define CONFIG_OMAP_GPIO
 #define CONFIG_OMAP_COMMON
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_SDRC	/* The chip has SDRC controller */

include/configs/omap3_sdp3430.h +4 -0

index bf5c7a8faf77..1ca79d4ee4dd 100644
@@ -23,6 +23,10 @@
 #define CONFIG_OMAP		1	/* in a TI OMAP core */
 #define CONFIG_OMAP3_3430SDP	1	/* working with SDP Rev2 */
 #define CONFIG_OMAP_COMMON
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_SDRC	/* The chip has SDRC controller */

include/configs/tam3517-common.h +4 -0

index bff9687b3c9c..af259e20bbd2 100644
@@ -17,6 +17,10 @@
 #define CONFIG_OMAP_GPIO
 #define CONFIG_OMAP_COMMON
 #define CONFIG_SYS_GENERIC_BOARD
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_SYS_TEXT_BASE 0x80008000

include/configs/tao3530.h +4 -0

index 4dfd5cb9fbb7..634b441b5768 100644
@@ -21,6 +21,10 @@
 #define CONFIG_OMAP_GPIO
 #define CONFIG_OMAP_COMMON
 #define CONFIG_SYS_GENERIC_BOARD
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define MACH_TYPE_OMAP3_TAO3530		2836

include/configs/ti_omap3_common.h +5 -0

index f909f6b94a18..22f103c8dd8c 100644
@@ -27,6 +27,11 @@
 # define CONFIG_SYS_MALLOC_F_LEN	(1 << 10)
 #endif
 
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
+
 /* The chip has SDRC controller */
 #define CONFIG_SDRC

include/configs/tricorder.h +4 -0

index f27cd1b9b061..1954e62b4097 100644
@@ -19,6 +19,10 @@
 /* High Level Configuration Options */
 #define CONFIG_OMAP			/* in a TI OMAP core */
 #define CONFIG_OMAP_COMMON
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_MACH_TYPE		MACH_TYPE_TRICORDER
 /*

Signature · 1 line

1.7.9.5

Nishanth Menon

author

2015-03-03 22:26 UTC #16 lore raw

PATCH v4 11/11 ARM: OMAP3: rx51: Enable workaround for ARM errata 454179, 430973, 621766

In reply to Nishanth Menon

RX51 has a secure logic which uses different parameters compared to traditional implementation. So, make the generic secure acr write over-ride-able by board file and refactor rx51 code to use this.

While at it, enable the OMAP3 specific errata code for 454179, 430973, 621766.

Signed-off-by Nishanth Menon <[email protected]>

---
 arch/arm/cpu/armv7/omap3/board.c            |   45 ++++++++++++---------------
 arch/arm/include/asm/arch-omap3/sys_proto.h |    1 +
 board/nokia/rx51/rx51.c                     |   19 ++++++-----
 include/configs/nokia_rx51.h                |    4 +++
 4 files changed, 37 insertions(+), 32 deletions(-)

arch/arm/cpu/armv7/omap3/board.c +20 ~~-25~~

index 5587a10cdafd..bfbc197bf99c 100644
@@ -404,31 +404,30 @@ static void omap3_emu_romcode_call(u32 service_id, u32 *parameters)
 	do_omap3_emu_romcode_call(service_id, OMAP3_PUBLIC_SRAM_SCRATCH_AREA);
 }
 
-void v7_arch_cp15_set_acr(u32 acr, u32 cpu_midr, u32 cpu_rev_comb,
-			  u32 cpu_variant, u32 cpu_rev)
+void __weak omap3_set_aux_cr_secure(u32 acr)
 {
-	if (get_device_type() == GP_DEVICE) {
-		omap_smc1(OMAP3_GP_ROMCODE_API_WRITE_ACR, acr);
-	} else {
-		struct emu_hal_params emu_romcode_params;
-		emu_romcode_params.num_params = 1;
-		emu_romcode_params.param1 = acr;
-		omap3_emu_romcode_call(OMAP3_EMU_HAL_API_WRITE_ACR,
-				       (u32 *)&emu_romcode_params);
-	}
+	struct emu_hal_params emu_romcode_params;
+
+	emu_romcode_params.num_params = 1;
+	emu_romcode_params.param1 = acr;
+	omap3_emu_romcode_call(OMAP3_EMU_HAL_API_WRITE_ACR,
+			       (u32 *)&emu_romcode_params);
 }
 
-static void omap3_update_aux_cr_secure(u32 set_bits, u32 clear_bits)
+void v7_arch_cp15_set_acr(u32 acr, u32 cpu_midr, u32 cpu_rev_comb,
+			  u32 cpu_variant, u32 cpu_rev)
 {
-	u32 acr;
+	/* Write ACR - affects secure banked bits */
+	if (get_device_type() == GP_DEVICE)
+		omap_smc1(OMAP3_GP_ROMCODE_API_WRITE_ACR, acr);
+	else
+		omap3_set_aux_cr_secure(acr);
 
-	/* Read ACR */
-	asm volatile ("mrc p15, 0, %0, c1, c0, 1" : "=r" (acr));
-	acr &= ~clear_bits;
-	acr |= set_bits;
-	v7_arch_cp15_set_acr(acr, 0, 0, 0, 0);
+	/* Write ACR - affects non-secure banked bits - some erratas need it */
+	asm volatile ("mcr p15, 0, %0, c1, c0, 1" : : "r" (acr));
 }
 
+
 #ifndef CONFIG_SYS_L2CACHE_OFF
 static void omap3_update_aux_cr(u32 set_bits, u32 clear_bits)
 {

@@ -438,9 +437,8 @@ static void omap3_update_aux_cr(u32 set_bits, u32 clear_bits)
 	asm volatile ("mrc p15, 0, %0, c1, c0, 1" : "=r" (acr));
 	acr &= ~clear_bits;
 	acr |= set_bits;
+	v7_arch_cp15_set_acr(acr, 0, 0, 0, 0);
 
-	/* Write ACR - affects non-secure banked bits */
-	asm volatile ("mcr p15, 0, %0, c1, c0, 1" : : "r" (acr));
 }
 
 /* Invalidate the entire L2 cache from secure mode */

@@ -459,10 +457,9 @@ static void omap3_invalidate_l2_cache_secure(void)
 
 void v7_outer_cache_enable(void)
 {
-	/* Set L2EN */
-	omap3_update_aux_cr_secure(0x2, 0);
 
 	/*
+	 * Set L2EN
 	 * On some revisions L2EN bit is banked on some revisions it's not
 	 * No harm in setting both banked bits(in fact this is required
 	 * by an erratum)

@@ -472,10 +469,8 @@ void v7_outer_cache_enable(void)
 
 void omap3_outer_cache_disable(void)
 {
-	/* Clear L2EN */
-	omap3_update_aux_cr_secure(0, 0x2);
-
 	/*
+	 * Clear L2EN
 	 * On some revisions L2EN bit is banked on some revisions it's not
 	 * No harm in clearing both banked bits(in fact this is required
 	 * by an erratum)

arch/arm/include/asm/arch-omap3/sys_proto.h +1 -0

index c06605d533d3..ff1bf736cf06 100644
@@ -74,5 +74,6 @@ void dieid_num_r(void);
 void get_dieid(u32 *id);
 void do_omap3_emu_romcode_call(u32 service_id, u32 parameters);
 void omap_smc1(u32 service, u32 val);
+void omap3_set_aux_cr_secure(u32 acr);
 u32 warm_reset(void);
 #endif

board/nokia/rx51/rx51.c +12 -7

index 08fcaf21b3c6..3d019b01428b 100644
@@ -341,6 +341,17 @@ static void omap3_emu_romcode_call(u32 service_id, u32 *parameters)
 	do_omap3_emu_romcode_call(service_id, OMAP3_PUBLIC_SRAM_SCRATCH_AREA);
 }
 
+void omap3_set_aux_cr_secure(u32 acr)
+{
+	struct emu_hal_params_rx51 emu_romcode_params = { 0, };
+
+	emu_romcode_params.num_params = 2;
+	emu_romcode_params.param1 = acr;
+
+	omap3_emu_romcode_call(OMAP3_EMU_HAL_API_WRITE_ACR,
+			       (u32 *)&emu_romcode_params);
+}
+
 /*
  * Routine: omap3_update_aux_cr_secure_rx51
  * Description: Modify the contents Auxiliary Control Register.

@@ -350,19 +361,13 @@ static void omap3_emu_romcode_call(u32 service_id, u32 *parameters)
  */
 static void omap3_update_aux_cr_secure_rx51(u32 set_bits, u32 clear_bits)
 {
-	struct emu_hal_params_rx51 emu_romcode_params = { 0, };
 	u32 acr;
 
 	/* Read ACR */
 	asm volatile ("mrc p15, 0, %0, c1, c0, 1" : "=r" (acr));
 	acr &= ~clear_bits;
 	acr |= set_bits;
-
-	emu_romcode_params.num_params = 2;
-	emu_romcode_params.param1 = acr;
-
-	omap3_emu_romcode_call(OMAP3_EMU_HAL_API_WRITE_ACR,
-				(u32 *)&emu_romcode_params);
+	omap3_set_aux_cr_secure(acr);
 }
 
 /*

include/configs/nokia_rx51.h +4 -0

index 442e16ae686b..152a6e564b00 100644
@@ -29,6 +29,10 @@
 #define CONFIG_SYS_L2CACHE_OFF		/* pretend there is no L2 CACHE */
 #define CONFIG_OMAP_COMMON
 #define CONFIG_SYS_GENERIC_BOARD
+/* Common ARM Erratas */
+#define CONFIG_ARM_ERRATA_454179
+#define CONFIG_ARM_ERRATA_430973
+#define CONFIG_ARM_ERRATA_621766
 
 #define CONFIG_MACH_TYPE		MACH_TYPE_NOKIA_RX51

Signature · 1 line

1.7.9.5

Matt Porter

2015-03-05 16:21 UTC #17 lore raw

v4 cover ARM: OMAP3-DRA7: CP15 erratum workarounds and improvements

In reply to Nishanth Menon

Quoted 21 lines · Nishanth Menon

The fourth incarnation of this series to address review comments on V3
With all the usual disclaimers and request to see V1 of the series for a detailed blurb.. As usual additional testing preferred.. Sorry, I dont have access to all possible variants atm..
changes since v3:
- few corrections - i have tried to do a push-pop of register params. hopefully, they should do the job
- smc with a makefile handling of secure-ext enablement.. (stolen from kernel).
V3: http://thread.gmane.org/gmane.comp.boot-loaders.u-boot/213207/focus=213307
V2: http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/213060
V1: http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/212174
Testing: with http://paste.ubuntu.org.cn/2522971 (4.0-rc1 patch)
BeagleBoard-X15: http://pastebin.ubuntu.com/10518934/
	Before: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000000 L2PFR=0x000009b0 ACTLR=0x00000040
	After: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000198 L2PFR=0x000009b0 ACTLR=0x00000040

Quoted 7 lines

OMAP5uEVM: http://pastebin.ubuntu.com/10518958/

	Before: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000000 L2PFR=0x000009b0 ACTLR=0x00000040
	After: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000198 L2PFR=0x000009b0 ACTLR=0x00000040

Beagle-XM: http://pastebin.ubuntu.com/10519417/ (this is a r2p3 device)

	Before: CPUID=0x413fc082 ACR=0x000000e2 L2AUXCR=0x00000042
	After: CPUID=0x413fc082 ACR=0x00000042 L2AUXCR=0x00000042

I also got the same results on a Beagle-XM Rev. C1

Quoted 1 line

I dont have access to other omap3 platforms to give a better coverage

Beagle Rev. C2 (OMAP3530): http://pastebin.com/f5JcvRf4

	Before: CPUID=0x411fc083 ACR=0x000000e2 L2AUXCR=0x00000042
	After: CPUID=0x411fc083 ACR=0x00000042 L2AUXCR=0x00000042

Tested-by Matt Porter <[email protected]>

[With build workaround I noted elsewhere in the thread]

-Matt

Quoted 69 lines

Sanity check:
OMAP4Panda-ES: http://pastebin.ubuntu.com/10518971/

Praveen Rao (1):
ARM: DRA7 / OMAP5: Add workaround for ARM errata 798870

 README                                             |    8 +++
 arch/arm/cpu/armv7/Makefile                        |    2 +-
 arch/arm/cpu/armv7/cp15.c                          |   29 ++++++++++
 arch/arm/cpu/armv7/omap-common/Makefile            |    5 +-
 arch/arm/cpu/armv7/omap-common/lowlevel_init.S     |   19 +++---
 arch/arm/cpu/armv7/omap3/board.c                   |   60 +++++++------------
 arch/arm/cpu/armv7/omap3/lowlevel_init.S           |   11 ----
 arch/arm/cpu/armv7/omap4/hwinit.c                  |    4 +-
 arch/arm/cpu/armv7/omap5/hwinit.c                  |   23 ++++++++
 arch/arm/cpu/armv7/start.S                         |   61 ++++++++++++++++++++
 .../arm/include/asm/arch-omap3/{omap3.h => omap.h} |    0
 arch/arm/include/asm/arch-omap3/sys_proto.h        |    3 +-
 arch/arm/include/asm/arch-omap4/sys_proto.h        |    5 +-
 arch/arm/include/asm/arch-omap5/sys_proto.h        |    4 ++
 arch/arm/include/asm/armv7.h                       |    5 ++
 board/nokia/rx51/rx51.c                            |   19 +++---
 include/configs/am3517_crane.h                     |    6 +-
 include/configs/am3517_evm.h                       |    6 +-
 include/configs/cm_t35.h                           |    6 +-
 include/configs/cm_t3517.h                         |    6 +-
 include/configs/dig297.h                           |    6 +-
 include/configs/mcx.h                              |    6 +-
 include/configs/nokia_rx51.h                       |    6 +-
 include/configs/omap3_evm.h                        |    2 +-
 include/configs/omap3_evm_common.h                 |    4 ++
 include/configs/omap3_evm_quick_mmc.h              |    2 +-
 include/configs/omap3_evm_quick_nand.h             |    2 +-
 include/configs/omap3_logic.h                      |    6 +-
 include/configs/omap3_mvblx.h                      |    6 +-
 include/configs/omap3_pandora.h                    |    6 +-
 include/configs/omap3_sdp3430.h                    |    6 +-
 include/configs/omap3_zoom1.h                      |    2 +-
 include/configs/tam3517-common.h                   |    6 +-
 include/configs/tao3530.h                          |    6 +-
 include/configs/ti_omap3_common.h                  |    7 ++-
 include/configs/ti_omap5_common.h                  |    3 +
 include/configs/tricorder.h                        |    6 +-
 37 files changed, 275 insertions(+), 89 deletions(-)
 create mode 100644 arch/arm/cpu/armv7/cp15.c
 rename arch/arm/include/asm/arch-omap3/{omap3.h => omap.h} (100%)

Signature · 6 lines

1.7.9.5

_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

Nishanth Menon

author

2015-03-05 17:56 UTC #18 lore raw

v4 cover ARM: OMAP3-DRA7: CP15 erratum workarounds and improvements

Quoted 32 lines · Matt Porter

Quoted 21 lines · Nishanth Menon
The fourth incarnation of this series to address review comments on V3
With all the usual disclaimers and request to see V1 of the series for a detailed blurb.. As usual additional testing preferred.. Sorry, I dont have access to all possible variants atm..
changes since v3:
- few corrections - i have tried to do a push-pop of register params. hopefully, they should do the job
- smc with a makefile handling of secure-ext enablement.. (stolen from kernel).
V3: http://thread.gmane.org/gmane.comp.boot-loaders.u-boot/213207/focus=213307
V2: http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/213060
V1: http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/212174
Testing: with http://paste.ubuntu.org.cn/2522971 (4.0-rc1 patch)
BeagleBoard-X15: http://pastebin.ubuntu.com/10518934/
	Before: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000000 L2PFR=0x000009b0 ACTLR=0x00000040
	After: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000198 L2PFR=0x000009b0 ACTLR=0x00000040
Quoted 7 lines
OMAP5uEVM: http://pastebin.ubuntu.com/10518958/
	Before: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000000 L2PFR=0x000009b0 ACTLR=0x00000040
	After: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000198 L2PFR=0x000009b0 ACTLR=0x00000040
Beagle-XM: http://pastebin.ubuntu.com/10519417/ (this is a r2p3 device)
	Before: CPUID=0x413fc082 ACR=0x000000e2 L2AUXCR=0x00000042
	After: CPUID=0x413fc082 ACR=0x00000042 L2AUXCR=0x00000042
I also got the same results on a Beagle-XM Rev. C1

Quoted 7 lines

Quoted 1 line
I dont have access to other omap3 platforms to give a better coverage
Beagle Rev. C2 (OMAP3530): http://pastebin.com/f5JcvRf4
	Before: CPUID=0x411fc083 ACR=0x000000e2 L2AUXCR=0x00000042
	After: CPUID=0x411fc083 ACR=0x00000042 L2AUXCR=0x00000042
Tested-by Matt Porter <[email protected]>

Thanks for testing.

Quoted 2 lines

[With build workaround I noted elsewhere in the thread]

that should have been a r1p3 device(needs errata), right and mine should really be a r3p2?

Did i get the code wrong here? Need some additional eyes here :(

http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.ddi0432c/Bhccjgga.html

mrc p15, 0, r1, c0, c0, 0       @ r1 has Read Main ID Register (MIDR)
mov r3, r1, lsr #20             @ get variant field
and r3, r3, #0xf                @ r3 has CPU variant
and r4, r1, #0xf                @ r4 has CPU revision

mov r2, r3, lsl #4 @ shift variant field for combined value
orr r2, r4, r2 @ r2 has combined CPU variant + revision

         cmp     r2, #0x21               @ Only on < r2p1
         bge     skip_errata_621766

Signature · 2 lines

Regards,
Nishanth Menon

Nishanth Menon

author

2015-03-06 02:52 UTC #19 lore raw

v4 cover ARM: OMAP3-DRA7: CP15 erratum workarounds and improvements

Quoted 57 lines · Nishanth Menon

Quoted 32 lines · Matt Porter
Quoted 21 lines · Nishanth Menon
The fourth incarnation of this series to address review comments on V3
With all the usual disclaimers and request to see V1 of the series for a detailed blurb.. As usual additional testing preferred.. Sorry, I dont have access to all possible variants atm..
changes since v3:
- few corrections - i have tried to do a push-pop of register params. hopefully, they should do the job
- smc with a makefile handling of secure-ext enablement.. (stolen from kernel).
V3: http://thread.gmane.org/gmane.comp.boot-loaders.u-boot/213207/focus=213307
V2: http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/213060
V1: http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/212174
Testing: with http://paste.ubuntu.org.cn/2522971 (4.0-rc1 patch)
BeagleBoard-X15: http://pastebin.ubuntu.com/10518934/
	Before: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000000 L2PFR=0x000009b0 ACTLR=0x00000040
	After: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000198 L2PFR=0x000009b0 ACTLR=0x00000040
Quoted 7 lines
OMAP5uEVM: http://pastebin.ubuntu.com/10518958/
	Before: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000000 L2PFR=0x000009b0 ACTLR=0x00000040
	After: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000198 L2PFR=0x000009b0 ACTLR=0x00000040
Beagle-XM: http://pastebin.ubuntu.com/10519417/ (this is a r2p3 device)
	Before: CPUID=0x413fc082 ACR=0x000000e2 L2AUXCR=0x00000042
	After: CPUID=0x413fc082 ACR=0x00000042 L2AUXCR=0x00000042
I also got the same results on a Beagle-XM Rev. C1
Quoted 7 lines
Quoted 1 line
I dont have access to other omap3 platforms to give a better coverage
Beagle Rev. C2 (OMAP3530): http://pastebin.com/f5JcvRf4
	Before: CPUID=0x411fc083 ACR=0x000000e2 L2AUXCR=0x00000042
	After: CPUID=0x411fc083 ACR=0x00000042 L2AUXCR=0x00000042
Tested-by Matt Porter <[email protected]>
Thanks for testing.
Quoted 2 lines
[With build workaround I noted elsewhere in the thread]
that should have been a r1p3 device(needs errata), right and mine should really be a r3p2?
Did i get the code wrong here? Need some additional eyes here :(
http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.ddi0432c/Bhccjgga.html
mrc p15, 0, r1, c0, c0, 0       @ r1 has Read Main ID Register (MIDR)
mov r3, r1, lsr #20             @ get variant field
and r3, r3, #0xf                @ r3 has CPU variant
and r4, r1, #0xf                @ r4 has CPU revision

Quoted 7 lines

mov r2, r3, lsl #4 @ shift variant field for combined value
orr r2, r4, r2 @ r2 has combined CPU variant + revision
         cmp     r2, #0x21               @ Only on < r2p1
         bge     skip_errata_621766

Alright, thanks to Jtag debugger and replaced 'bgt skip_errata' with 'blt skip_errata (forcing errata flow on my "non-errata beagleboard-xm board") found the issue. (A bit embarrassed to be staring at it and not recollecting)..

I should have s/b v7_arch_cp15_set_acr/bl v7_arch_cp15_set_acr/

Since I was not branching and linking (just branching), I was returning back to the caller of cpu_init_cp15 when the call from v7_arch_cp15_set_acr was returning, instead of returning back to v7_arch_cp15_set_acr! elementary error! Sorry, did not see it previously with A15 erratum since it was a single one to walk down and follow on errata never took place.

Please standby for a v5 with proper code flow. Apologies on requesting testing again..

Signature · 2 lines

Uggh... very very embarrassedely yours,
Nishanth Menon

Matt Porter

2015-03-06 21:02 UTC #20 lore raw

v4 cover ARM: OMAP3-DRA7: CP15 erratum workarounds and improvements

In reply to Nishanth Menon

Quoted 39 lines · Nishanth Menon

Quoted 38 lines · Matt Porter
Quoted 21 lines · Nishanth Menon
The fourth incarnation of this series to address review comments on V3
With all the usual disclaimers and request to see V1 of the series for a detailed blurb.. As usual additional testing preferred.. Sorry, I dont have access to all possible variants atm..
changes since v3:
- few corrections - i have tried to do a push-pop of register params. hopefully, they should do the job
- smc with a makefile handling of secure-ext enablement.. (stolen from kernel).
V3: http://thread.gmane.org/gmane.comp.boot-loaders.u-boot/213207/focus=213307
V2: http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/213060
V1: http://comments.gmane.org/gmane.comp.boot-loaders.u-boot/212174
Testing: with http://paste.ubuntu.org.cn/2522971 (4.0-rc1 patch)
BeagleBoard-X15: http://pastebin.ubuntu.com/10518934/
	Before: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000000 L2PFR=0x000009b0 ACTLR=0x00000040
	After: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000198 L2PFR=0x000009b0 ACTLR=0x00000040
Quoted 7 lines
OMAP5uEVM: http://pastebin.ubuntu.com/10518958/
	Before: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000000 L2PFR=0x000009b0 ACTLR=0x00000040
	After: CPUID=0x412fc0f2 L2CTLR=0x01800083 L2ACLR=0x00000198 L2PFR=0x000009b0 ACTLR=0x00000040
Beagle-XM: http://pastebin.ubuntu.com/10519417/ (this is a r2p3 device)
	Before: CPUID=0x413fc082 ACR=0x000000e2 L2AUXCR=0x00000042
	After: CPUID=0x413fc082 ACR=0x00000042 L2AUXCR=0x00000042
I also got the same results on a Beagle-XM Rev. C1
Quoted 1 line
I dont have access to other omap3 platforms to give a better coverage
Beagle Rev. C2 (OMAP3530): http://pastebin.com/f5JcvRf4
	Before: CPUID=0x411fc083 ACR=0x000000e2 L2AUXCR=0x00000042
	After: CPUID=0x411fc083 ACR=0x00000042 L2AUXCR=0x00000042

Quoted 8 lines

Quoted 1 line
Tested-by Matt Porter <[email protected]>
Thanks for testing.
Quoted 2 lines
[With build workaround I noted elsewhere in the thread]
that should have been a r1p3 device(needs errata), right and mine should really be a r3p2?

Correct. My 3530 beagle is r1p3 and your xM is r3p2.

Quoted 1 line

Did i get the code wrong here? Need some additional eyes here :(

I verified by inspection that your code below is fine but I also shoved it into a command line app with some doctored r1 content and confirmed that it really does work for the r3p2 case. So I'm not yet clear as to why it's engaging the erratum fix for r3p2.

-Matt

Quoted 21 lines

http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.ddi0432c/Bhccjgga.html
mrc p15, 0, r1, c0, c0, 0       @ r1 has Read Main ID Register (MIDR)
mov r3, r1, lsr #20             @ get variant field
and r3, r3, #0xf                @ r3 has CPU variant
and r4, r1, #0xf                @ r4 has CPU revision
mov r2, r3, lsl #4 @ shift variant field for combined value
orr r2, r4, r2 @ r2 has combined CPU variant + revision
         cmp     r2, #0x21               @ Only on < r2p1
         bge     skip_errata_621766
Signature · 6 lines
Regards,
Nishanth Menon
_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot