* [PATCH v2 0/5] omap assembly fixes for CONFIG_DEBUG_RODATA
@ 2016-01-21 19:12 ` Tony Lindgren
0 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-21 19:12 UTC (permalink / raw)
To: linux-arm-kernel
Hi all,
Here are read-only data fixes for omaps discovered few weeks ago
with "[PATCH v2] ARM: mm: flip priority of CONFIG_DEBUG_RODATA"
that caused issues during deeper idle states on omaps [0].
Please review and test if possible. I'd like to merge these fixes
during v4.5-rc cycle to remove these blockers for making it possible
to have CONFIG_DEBUG_RODATA enabled by default.
Regards,
Tony
Changes since v1:
- Fix missing ldr in patch 2/5 as Noticed by Nicolas Pitre. This
had gone unnoticed as the code just happened to work on omap36xx
without actually toggling the L2 cache. And patches 4 and 5 are setting
the address, not the value unlike patch 2/5. I've added some comments
to patch 2/5 too.
- Fix a typo noticed by Sergei Shtylyov in patch 3/5.
[0] https://lkml.org/lkml/2015/12/2/908
Tony Lindgren (5):
ARM: OMAP2+: Fix wait_dll_lock_timed for rodata
ARM: OMAP2+: Fix l2dis_3630 for rodata
ARM: OMAP2+: Fix save_secure_ram_context for rodata
ARM: OMAP2+: Fix l2_inv_api_params for rodata
ARM: OMAP2+: Fix ppa_zero_params and ppa_por_params for rodata
arch/arm/mach-omap2/sleep34xx.S | 61 ++++++++++++++++++++---------------------
arch/arm/mach-omap2/sleep44xx.S | 25 +++++++++++------
2 files changed, 47 insertions(+), 39 deletions(-)
--
2.7.0.rc3
^ permalink raw reply [flat|nested] 20+ messages in thread
* [PATCH 1/5] ARM: OMAP2+: Fix wait_dll_lock_timed for rodata
2016-01-21 19:12 ` Tony Lindgren
@ 2016-01-21 19:12 ` Tony Lindgren
-1 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-21 19:12 UTC (permalink / raw)
To: linux-omap
Cc: Nicolas Pitre, Nishanth Menon, Richard Woodruff, Russell King,
Kees Cook, Sergei Shtylyov, Tero Kristo, Laura Abbott,
linux-arm-kernel
We don't want to be writing to .text so it can be set rodata.
Fix error "Unable to handle kernel paging request at virtual address
c012396c" in wait_dll_lock_timed if CONFIG_DEBUG_RODATA is selected.
As these counters are for debugging only and unused, we can just
remove them.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <nm@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep34xx.S | 22 ----------------------
1 file changed, 22 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
index eafd120..787cfda 100644
--- a/arch/arm/mach-omap2/sleep34xx.S
+++ b/arch/arm/mach-omap2/sleep34xx.S
@@ -289,12 +289,6 @@ wait_sdrc_ready:
bic r5, r5, #0x40
str r5, [r4]
-/*
- * PC-relative stores lead to undefined behaviour in Thumb-2: use a r7 as a
- * base instead.
- * Be careful not to clobber r7 when maintaing this code.
- */
-
is_dll_in_lock_mode:
/* Is dll in lock mode? */
ldr r4, sdrc_dlla_ctrl
@@ -302,11 +296,7 @@ is_dll_in_lock_mode:
tst r5, #0x4
bne exit_nonoff_modes @ Return if locked
/* wait till dll locks */
- adr r7, kick_counter
wait_dll_lock_timed:
- ldr r4, wait_dll_lock_counter
- add r4, r4, #1
- str r4, [r7, #wait_dll_lock_counter - kick_counter]
ldr r4, sdrc_dlla_status
/* Wait 20uS for lock */
mov r6, #8
@@ -330,9 +320,6 @@ kick_dll:
orr r6, r6, #(1<<3) @ enable dll
str r6, [r4]
dsb
- ldr r4, kick_counter
- add r4, r4, #1
- str r4, [r7] @ kick_counter
b wait_dll_lock_timed
exit_nonoff_modes:
@@ -360,15 +347,6 @@ sdrc_dlla_status:
.word SDRC_DLLA_STATUS_V
sdrc_dlla_ctrl:
.word SDRC_DLLA_CTRL_V
- /*
- * When exporting to userspace while the counters are in SRAM,
- * these 2 words need to be at the end to facilitate retrival!
- */
-kick_counter:
- .word 0
-wait_dll_lock_counter:
- .word 0
-
ENTRY(omap3_do_wfi_sz)
.word . - omap3_do_wfi
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread
* [PATCH 1/5] ARM: OMAP2+: Fix wait_dll_lock_timed for rodata
@ 2016-01-21 19:12 ` Tony Lindgren
0 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-21 19:12 UTC (permalink / raw)
To: linux-arm-kernel
We don't want to be writing to .text so it can be set rodata.
Fix error "Unable to handle kernel paging request at virtual address
c012396c" in wait_dll_lock_timed if CONFIG_DEBUG_RODATA is selected.
As these counters are for debugging only and unused, we can just
remove them.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <nm@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep34xx.S | 22 ----------------------
1 file changed, 22 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
index eafd120..787cfda 100644
--- a/arch/arm/mach-omap2/sleep34xx.S
+++ b/arch/arm/mach-omap2/sleep34xx.S
@@ -289,12 +289,6 @@ wait_sdrc_ready:
bic r5, r5, #0x40
str r5, [r4]
-/*
- * PC-relative stores lead to undefined behaviour in Thumb-2: use a r7 as a
- * base instead.
- * Be careful not to clobber r7 when maintaing this code.
- */
-
is_dll_in_lock_mode:
/* Is dll in lock mode? */
ldr r4, sdrc_dlla_ctrl
@@ -302,11 +296,7 @@ is_dll_in_lock_mode:
tst r5, #0x4
bne exit_nonoff_modes @ Return if locked
/* wait till dll locks */
- adr r7, kick_counter
wait_dll_lock_timed:
- ldr r4, wait_dll_lock_counter
- add r4, r4, #1
- str r4, [r7, #wait_dll_lock_counter - kick_counter]
ldr r4, sdrc_dlla_status
/* Wait 20uS for lock */
mov r6, #8
@@ -330,9 +320,6 @@ kick_dll:
orr r6, r6, #(1<<3) @ enable dll
str r6, [r4]
dsb
- ldr r4, kick_counter
- add r4, r4, #1
- str r4, [r7] @ kick_counter
b wait_dll_lock_timed
exit_nonoff_modes:
@@ -360,15 +347,6 @@ sdrc_dlla_status:
.word SDRC_DLLA_STATUS_V
sdrc_dlla_ctrl:
.word SDRC_DLLA_CTRL_V
- /*
- * When exporting to userspace while the counters are in SRAM,
- * these 2 words need to be@the end to facilitate retrival!
- */
-kick_counter:
- .word 0
-wait_dll_lock_counter:
- .word 0
-
ENTRY(omap3_do_wfi_sz)
.word . - omap3_do_wfi
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread
* [PATCH 2/5] ARM: OMAP2+: Fix l2dis_3630 for rodata
2016-01-21 19:12 ` Tony Lindgren
@ 2016-01-21 19:12 ` Tony Lindgren
-1 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-21 19:12 UTC (permalink / raw)
To: linux-omap
Cc: Nicolas Pitre, Nishanth Menon, Richard Woodruff, Russell King,
Kees Cook, Sergei Shtylyov, Tero Kristo, Laura Abbott,
linux-arm-kernel
We don't want to write to .text section. Let's move l2dis_3630
to .data and access it via a pointer.
For calculating the offset, let's optimize out the add and do it
in ldr/str as suggested by Nicolas Pitre <nicolas.pitre@linaro.org>.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <nm@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep34xx.S | 17 +++++++++++++----
1 file changed, 13 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
index 787cfda..5659d96 100644
--- a/arch/arm/mach-omap2/sleep34xx.S
+++ b/arch/arm/mach-omap2/sleep34xx.S
@@ -86,8 +86,9 @@ ENTRY(enable_omap3630_toggle_l2_on_restore)
stmfd sp!, {lr} @ save registers on stack
/* Setup so that we will disable and enable l2 */
mov r1, #0x1
- adrl r2, l2dis_3630 @ may be too distant for plain adr
- str r1, [r2]
+ adrl r3, l2dis_3630_offset @ may be too distant for plain adr
+ ldr r2, [r3] @ value for offset
+ str r1, [r2, r3] @ write to l2dis_3630
ldmfd sp!, {pc} @ restore regs and return
ENDPROC(enable_omap3630_toggle_l2_on_restore)
@@ -415,7 +416,9 @@ ENTRY(omap3_restore)
cmp r2, #0x0 @ Check if target power state was OFF or RET
bne logic_l1_restore
- ldr r0, l2dis_3630
+ adr r1, l2dis_3630_offset @ address for offset
+ ldr r0, [r1] @ value for offset
+ ldr r0, [r1, r0] @ value at l2dis_3630
cmp r0, #0x1 @ should we disable L2 on 3630?
bne skipl2dis
mrc p15, 0, r0, c1, c0, 1
@@ -484,7 +487,9 @@ l2_inv_gp:
mov r12, #0x2
smc #0 @ Call SMI monitor (smieq)
logic_l1_restore:
- ldr r1, l2dis_3630
+ adr r0, l2dis_3630_offset @ adress for offset
+ ldr r1, [r0] @ value for offset
+ ldr r1, [r0, r1] @ value at l2dis_3630
cmp r1, #0x1 @ Test if L2 re-enable needed on 3630
bne skipl2reen
mrc p15, 0, r1, c1, c0, 1
@@ -513,6 +518,10 @@ control_stat:
.word CONTROL_STAT
control_mem_rta:
.word CONTROL_MEM_RTA_CTRL
+l2dis_3630_offset:
+ .long l2dis_3630 - .
+
+ .data
l2dis_3630:
.word 0
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread* [PATCH 2/5] ARM: OMAP2+: Fix l2dis_3630 for rodata
@ 2016-01-21 19:12 ` Tony Lindgren
0 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-21 19:12 UTC (permalink / raw)
To: linux-arm-kernel
We don't want to write to .text section. Let's move l2dis_3630
to .data and access it via a pointer.
For calculating the offset, let's optimize out the add and do it
in ldr/str as suggested by Nicolas Pitre <nicolas.pitre@linaro.org>.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <nm@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep34xx.S | 17 +++++++++++++----
1 file changed, 13 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
index 787cfda..5659d96 100644
--- a/arch/arm/mach-omap2/sleep34xx.S
+++ b/arch/arm/mach-omap2/sleep34xx.S
@@ -86,8 +86,9 @@ ENTRY(enable_omap3630_toggle_l2_on_restore)
stmfd sp!, {lr} @ save registers on stack
/* Setup so that we will disable and enable l2 */
mov r1, #0x1
- adrl r2, l2dis_3630 @ may be too distant for plain adr
- str r1, [r2]
+ adrl r3, l2dis_3630_offset @ may be too distant for plain adr
+ ldr r2, [r3] @ value for offset
+ str r1, [r2, r3] @ write to l2dis_3630
ldmfd sp!, {pc} @ restore regs and return
ENDPROC(enable_omap3630_toggle_l2_on_restore)
@@ -415,7 +416,9 @@ ENTRY(omap3_restore)
cmp r2, #0x0 @ Check if target power state was OFF or RET
bne logic_l1_restore
- ldr r0, l2dis_3630
+ adr r1, l2dis_3630_offset @ address for offset
+ ldr r0, [r1] @ value for offset
+ ldr r0, [r1, r0] @ value at l2dis_3630
cmp r0, #0x1 @ should we disable L2 on 3630?
bne skipl2dis
mrc p15, 0, r0, c1, c0, 1
@@ -484,7 +487,9 @@ l2_inv_gp:
mov r12, #0x2
smc #0 @ Call SMI monitor (smieq)
logic_l1_restore:
- ldr r1, l2dis_3630
+ adr r0, l2dis_3630_offset @ adress for offset
+ ldr r1, [r0] @ value for offset
+ ldr r1, [r0, r1] @ value at l2dis_3630
cmp r1, #0x1 @ Test if L2 re-enable needed on 3630
bne skipl2reen
mrc p15, 0, r1, c1, c0, 1
@@ -513,6 +518,10 @@ control_stat:
.word CONTROL_STAT
control_mem_rta:
.word CONTROL_MEM_RTA_CTRL
+l2dis_3630_offset:
+ .long l2dis_3630 - .
+
+ .data
l2dis_3630:
.word 0
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread
* [PATCH 3/5] ARM: OMAP2+: Fix save_secure_ram_context for rodata
2016-01-21 19:12 ` Tony Lindgren
@ 2016-01-21 19:12 ` Tony Lindgren
-1 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-21 19:12 UTC (permalink / raw)
To: linux-omap
Cc: Nicolas Pitre, Nishanth Menon, Richard Woodruff, Russell King,
Kees Cook, Sergei Shtylyov, Tero Kristo, Laura Abbott,
linux-arm-kernel
We don't want to write to .text and we can move save_secure_ram_context
into .data as it all gets copied into SRAM anyways.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <nm@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep34xx.S | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
index 5659d96..8e2a7ac 100644
--- a/arch/arm/mach-omap2/sleep34xx.S
+++ b/arch/arm/mach-omap2/sleep34xx.S
@@ -92,8 +92,12 @@ ENTRY(enable_omap3630_toggle_l2_on_restore)
ldmfd sp!, {pc} @ restore regs and return
ENDPROC(enable_omap3630_toggle_l2_on_restore)
- .text
-/* Function to call rom code to save secure ram context */
+/*
+ * Function to call rom code to save secure ram context. This gets
+ * relocated to SRAM, so it can be all in .data section. Otherwise
+ * we need to initialize api_params separately.
+ */
+ .data
.align 3
ENTRY(save_secure_ram_context)
stmfd sp!, {r4 - r11, lr} @ save registers on stack
@@ -127,6 +131,8 @@ ENDPROC(save_secure_ram_context)
ENTRY(save_secure_ram_context_sz)
.word . - save_secure_ram_context
+ .text
+
/*
* ======================
* == Idle entry point ==
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread* [PATCH 3/5] ARM: OMAP2+: Fix save_secure_ram_context for rodata
@ 2016-01-21 19:12 ` Tony Lindgren
0 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-21 19:12 UTC (permalink / raw)
To: linux-arm-kernel
We don't want to write to .text and we can move save_secure_ram_context
into .data as it all gets copied into SRAM anyways.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <nm@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Sergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep34xx.S | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
index 5659d96..8e2a7ac 100644
--- a/arch/arm/mach-omap2/sleep34xx.S
+++ b/arch/arm/mach-omap2/sleep34xx.S
@@ -92,8 +92,12 @@ ENTRY(enable_omap3630_toggle_l2_on_restore)
ldmfd sp!, {pc} @ restore regs and return
ENDPROC(enable_omap3630_toggle_l2_on_restore)
- .text
-/* Function to call rom code to save secure ram context */
+/*
+ * Function to call rom code to save secure ram context. This gets
+ * relocated to SRAM, so it can be all in .data section. Otherwise
+ * we need to initialize api_params separately.
+ */
+ .data
.align 3
ENTRY(save_secure_ram_context)
stmfd sp!, {r4 - r11, lr} @ save registers on stack
@@ -127,6 +131,8 @@ ENDPROC(save_secure_ram_context)
ENTRY(save_secure_ram_context_sz)
.word . - save_secure_ram_context
+ .text
+
/*
* ======================
* == Idle entry point ==
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread
* [PATCH 4/5] ARM: OMAP2+: Fix l2_inv_api_params for rodata
2016-01-21 19:12 ` Tony Lindgren
@ 2016-01-21 19:12 ` Tony Lindgren
-1 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-21 19:12 UTC (permalink / raw)
To: linux-omap
Cc: Nicolas Pitre, Nishanth Menon, Richard Woodruff, Russell King,
Kees Cook, Sergei Shtylyov, Tero Kristo, Laura Abbott,
linux-arm-kernel
We don't want to write to .text, so let's move l2_inv_api_params
to .data and access it via a pointer.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <nm@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep34xx.S | 12 +++++++++---
1 file changed, 9 insertions(+), 3 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
index 8e2a7ac..1b9f052 100644
--- a/arch/arm/mach-omap2/sleep34xx.S
+++ b/arch/arm/mach-omap2/sleep34xx.S
@@ -436,12 +436,14 @@ skipl2dis:
and r1, #0x700
cmp r1, #0x300
beq l2_inv_gp
+ adr r0, l2_inv_api_params_offset
+ ldr r3, [r0]
+ add r3, r3, r0 @ r3 points to dummy parameters
mov r0, #40 @ set service ID for PPA
mov r12, r0 @ copy secure Service ID in r12
mov r1, #0 @ set task id for ROM code in r1
mov r2, #4 @ set some flags in r2, r6
mov r6, #0xff
- adr r3, l2_inv_api_params @ r3 points to dummy parameters
dsb @ data write barrier
dmb @ data memory barrier
smc #1 @ call SMI monitor (smi #1)
@@ -475,8 +477,8 @@ skipl2dis:
b logic_l1_restore
.align
-l2_inv_api_params:
- .word 0x1, 0x00
+l2_inv_api_params_offset:
+ .long l2_inv_api_params - .
l2_inv_gp:
/* Execute smi to invalidate L2 cache */
mov r12, #0x1 @ set up to invalidate L2
@@ -531,6 +533,10 @@ l2dis_3630_offset:
l2dis_3630:
.word 0
+ .data
+l2_inv_api_params:
+ .word 0x1, 0x00
+
/*
* Internal functions
*/
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread
* [PATCH 4/5] ARM: OMAP2+: Fix l2_inv_api_params for rodata
@ 2016-01-21 19:12 ` Tony Lindgren
0 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-21 19:12 UTC (permalink / raw)
To: linux-arm-kernel
We don't want to write to .text, so let's move l2_inv_api_params
to .data and access it via a pointer.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <nm@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep34xx.S | 12 +++++++++---
1 file changed, 9 insertions(+), 3 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
index 8e2a7ac..1b9f052 100644
--- a/arch/arm/mach-omap2/sleep34xx.S
+++ b/arch/arm/mach-omap2/sleep34xx.S
@@ -436,12 +436,14 @@ skipl2dis:
and r1, #0x700
cmp r1, #0x300
beq l2_inv_gp
+ adr r0, l2_inv_api_params_offset
+ ldr r3, [r0]
+ add r3, r3, r0 @ r3 points to dummy parameters
mov r0, #40 @ set service ID for PPA
mov r12, r0 @ copy secure Service ID in r12
mov r1, #0 @ set task id for ROM code in r1
mov r2, #4 @ set some flags in r2, r6
mov r6, #0xff
- adr r3, l2_inv_api_params @ r3 points to dummy parameters
dsb @ data write barrier
dmb @ data memory barrier
smc #1 @ call SMI monitor (smi #1)
@@ -475,8 +477,8 @@ skipl2dis:
b logic_l1_restore
.align
-l2_inv_api_params:
- .word 0x1, 0x00
+l2_inv_api_params_offset:
+ .long l2_inv_api_params - .
l2_inv_gp:
/* Execute smi to invalidate L2 cache */
mov r12, #0x1 @ set up to invalidate L2
@@ -531,6 +533,10 @@ l2dis_3630_offset:
l2dis_3630:
.word 0
+ .data
+l2_inv_api_params:
+ .word 0x1, 0x00
+
/*
* Internal functions
*/
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread
* [PATCH 5/5] ARM: OMAP2+: Fix ppa_zero_params and ppa_por_params for rodata
2016-01-21 19:12 ` Tony Lindgren
@ 2016-01-21 19:12 ` Tony Lindgren
-1 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-21 19:12 UTC (permalink / raw)
To: linux-omap
Cc: Nicolas Pitre, Nishanth Menon, Richard Woodruff, Russell King,
Kees Cook, Sergei Shtylyov, Tero Kristo, Laura Abbott,
linux-arm-kernel
We don't want to write to .text so let's move ppa_zero_params and
ppa_por_params to .data and access them via pointers.
Note that I have not been able to test as we I don't have a HS
omap4 to test with. The code has been changed in similar way as
for omap3 though.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <nm@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep44xx.S | 25 +++++++++++++++++--------
1 file changed, 17 insertions(+), 8 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep44xx.S b/arch/arm/mach-omap2/sleep44xx.S
index 9b09d85..c7a3b4a 100644
--- a/arch/arm/mach-omap2/sleep44xx.S
+++ b/arch/arm/mach-omap2/sleep44xx.S
@@ -29,12 +29,6 @@
dsb
.endm
-ppa_zero_params:
- .word 0x0
-
-ppa_por_params:
- .word 1, 0
-
#ifdef CONFIG_ARCH_OMAP4
/*
@@ -266,7 +260,9 @@ ENTRY(omap4_cpu_resume)
beq skip_ns_smp_enable
ppa_actrl_retry:
mov r0, #OMAP4_PPA_CPU_ACTRL_SMP_INDEX
- adr r3, ppa_zero_params @ Pointer to parameters
+ adr r1, ppa_zero_params_offset
+ ldr r3, [r1]
+ add r3, r3, r1 @ Pointer to ppa_zero_params
mov r1, #0x0 @ Process ID
mov r2, #0x4 @ Flag
mov r6, #0xff
@@ -303,7 +299,9 @@ skip_ns_smp_enable:
ldr r0, =OMAP4_PPA_L2_POR_INDEX
ldr r1, =OMAP44XX_SAR_RAM_BASE
ldr r4, [r1, #L2X0_PREFETCH_CTRL_OFFSET]
- adr r3, ppa_por_params
+ adr r1, ppa_por_params_offset
+ ldr r3, [r1]
+ add r3, r3, r1 @ Pointer to ppa_por_params
str r4, [r3, #0x04]
mov r1, #0x0 @ Process ID
mov r2, #0x4 @ Flag
@@ -328,6 +326,8 @@ skip_l2en:
#endif
b cpu_resume @ Jump to generic resume
+ppa_por_params_offset:
+ .long ppa_por_params - .
ENDPROC(omap4_cpu_resume)
#endif /* CONFIG_ARCH_OMAP4 */
@@ -380,4 +380,13 @@ ENTRY(omap_do_wfi)
nop
ldmfd sp!, {pc}
+ppa_zero_params_offset:
+ .long ppa_zero_params - .
ENDPROC(omap_do_wfi)
+
+ .data
+ppa_zero_params:
+ .word 0
+
+ppa_por_params:
+ .word 1, 0
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread* [PATCH 5/5] ARM: OMAP2+: Fix ppa_zero_params and ppa_por_params for rodata
@ 2016-01-21 19:12 ` Tony Lindgren
0 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-21 19:12 UTC (permalink / raw)
To: linux-arm-kernel
We don't want to write to .text so let's move ppa_zero_params and
ppa_por_params to .data and access them via pointers.
Note that I have not been able to test as we I don't have a HS
omap4 to test with. The code has been changed in similar way as
for omap3 though.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <nm@ti.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep44xx.S | 25 +++++++++++++++++--------
1 file changed, 17 insertions(+), 8 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep44xx.S b/arch/arm/mach-omap2/sleep44xx.S
index 9b09d85..c7a3b4a 100644
--- a/arch/arm/mach-omap2/sleep44xx.S
+++ b/arch/arm/mach-omap2/sleep44xx.S
@@ -29,12 +29,6 @@
dsb
.endm
-ppa_zero_params:
- .word 0x0
-
-ppa_por_params:
- .word 1, 0
-
#ifdef CONFIG_ARCH_OMAP4
/*
@@ -266,7 +260,9 @@ ENTRY(omap4_cpu_resume)
beq skip_ns_smp_enable
ppa_actrl_retry:
mov r0, #OMAP4_PPA_CPU_ACTRL_SMP_INDEX
- adr r3, ppa_zero_params @ Pointer to parameters
+ adr r1, ppa_zero_params_offset
+ ldr r3, [r1]
+ add r3, r3, r1 @ Pointer to ppa_zero_params
mov r1, #0x0 @ Process ID
mov r2, #0x4 @ Flag
mov r6, #0xff
@@ -303,7 +299,9 @@ skip_ns_smp_enable:
ldr r0, =OMAP4_PPA_L2_POR_INDEX
ldr r1, =OMAP44XX_SAR_RAM_BASE
ldr r4, [r1, #L2X0_PREFETCH_CTRL_OFFSET]
- adr r3, ppa_por_params
+ adr r1, ppa_por_params_offset
+ ldr r3, [r1]
+ add r3, r3, r1 @ Pointer to ppa_por_params
str r4, [r3, #0x04]
mov r1, #0x0 @ Process ID
mov r2, #0x4 @ Flag
@@ -328,6 +326,8 @@ skip_l2en:
#endif
b cpu_resume @ Jump to generic resume
+ppa_por_params_offset:
+ .long ppa_por_params - .
ENDPROC(omap4_cpu_resume)
#endif /* CONFIG_ARCH_OMAP4 */
@@ -380,4 +380,13 @@ ENTRY(omap_do_wfi)
nop
ldmfd sp!, {pc}
+ppa_zero_params_offset:
+ .long ppa_zero_params - .
ENDPROC(omap_do_wfi)
+
+ .data
+ppa_zero_params:
+ .word 0
+
+ppa_por_params:
+ .word 1, 0
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread
* Re: [PATCH v2 0/5] omap assembly fixes for CONFIG_DEBUG_RODATA
2016-01-21 19:12 ` Tony Lindgren
@ 2016-01-21 19:26 ` Nicolas Pitre
-1 siblings, 0 replies; 20+ messages in thread
From: Nicolas Pitre @ 2016-01-21 19:26 UTC (permalink / raw)
To: Tony Lindgren
Cc: Nishanth Menon, Richard Woodruff, Russell King, Kees Cook,
Sergei Shtylyov, Tero Kristo, linux-omap, Laura Abbott,
linux-arm-kernel
On Thu, 21 Jan 2016, Tony Lindgren wrote:
> Hi all,
>
> Here are read-only data fixes for omaps discovered few weeks ago
> with "[PATCH v2] ARM: mm: flip priority of CONFIG_DEBUG_RODATA"
> that caused issues during deeper idle states on omaps [0].
>
> Please review and test if possible. I'd like to merge these fixes
> during v4.5-rc cycle to remove these blockers for making it possible
> to have CONFIG_DEBUG_RODATA enabled by default.
>
> Regards,
For the series:
Acked-by: Nicolas Pitre <nico@linaro.org>
> Tony
>
> Changes since v1:
>
> - Fix missing ldr in patch 2/5 as Noticed by Nicolas Pitre. This
> had gone unnoticed as the code just happened to work on omap36xx
> without actually toggling the L2 cache. And patches 4 and 5 are setting
> the address, not the value unlike patch 2/5. I've added some comments
> to patch 2/5 too.
>
> - Fix a typo noticed by Sergei Shtylyov in patch 3/5.
>
> [0] https://lkml.org/lkml/2015/12/2/908
>
> Tony Lindgren (5):
> ARM: OMAP2+: Fix wait_dll_lock_timed for rodata
> ARM: OMAP2+: Fix l2dis_3630 for rodata
> ARM: OMAP2+: Fix save_secure_ram_context for rodata
> ARM: OMAP2+: Fix l2_inv_api_params for rodata
> ARM: OMAP2+: Fix ppa_zero_params and ppa_por_params for rodata
>
> arch/arm/mach-omap2/sleep34xx.S | 61 ++++++++++++++++++++---------------------
> arch/arm/mach-omap2/sleep44xx.S | 25 +++++++++++------
> 2 files changed, 47 insertions(+), 39 deletions(-)
>
> --
> 2.7.0.rc3
>
>
^ permalink raw reply [flat|nested] 20+ messages in thread
* [PATCH v2 0/5] omap assembly fixes for CONFIG_DEBUG_RODATA
@ 2016-01-21 19:26 ` Nicolas Pitre
0 siblings, 0 replies; 20+ messages in thread
From: Nicolas Pitre @ 2016-01-21 19:26 UTC (permalink / raw)
To: linux-arm-kernel
On Thu, 21 Jan 2016, Tony Lindgren wrote:
> Hi all,
>
> Here are read-only data fixes for omaps discovered few weeks ago
> with "[PATCH v2] ARM: mm: flip priority of CONFIG_DEBUG_RODATA"
> that caused issues during deeper idle states on omaps [0].
>
> Please review and test if possible. I'd like to merge these fixes
> during v4.5-rc cycle to remove these blockers for making it possible
> to have CONFIG_DEBUG_RODATA enabled by default.
>
> Regards,
For the series:
Acked-by: Nicolas Pitre <nico@linaro.org>
> Tony
>
> Changes since v1:
>
> - Fix missing ldr in patch 2/5 as Noticed by Nicolas Pitre. This
> had gone unnoticed as the code just happened to work on omap36xx
> without actually toggling the L2 cache. And patches 4 and 5 are setting
> the address, not the value unlike patch 2/5. I've added some comments
> to patch 2/5 too.
>
> - Fix a typo noticed by Sergei Shtylyov in patch 3/5.
>
> [0] https://lkml.org/lkml/2015/12/2/908
>
> Tony Lindgren (5):
> ARM: OMAP2+: Fix wait_dll_lock_timed for rodata
> ARM: OMAP2+: Fix l2dis_3630 for rodata
> ARM: OMAP2+: Fix save_secure_ram_context for rodata
> ARM: OMAP2+: Fix l2_inv_api_params for rodata
> ARM: OMAP2+: Fix ppa_zero_params and ppa_por_params for rodata
>
> arch/arm/mach-omap2/sleep34xx.S | 61 ++++++++++++++++++++---------------------
> arch/arm/mach-omap2/sleep44xx.S | 25 +++++++++++------
> 2 files changed, 47 insertions(+), 39 deletions(-)
>
> --
> 2.7.0.rc3
>
>
^ permalink raw reply [flat|nested] 20+ messages in thread
* [PATCH 2/5] ARM: OMAP2+: Fix l2dis_3630 for rodata
2016-01-19 17:48 [PATCH " Tony Lindgren
@ 2016-01-19 17:48 ` Tony Lindgren
0 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-19 17:48 UTC (permalink / raw)
To: linux-omap
Cc: Nicolas Pitre, Richard Woodruff, Russell King, Kees Cook,
Tero Kristo, Nishanth Menon, Laura Abbott, linux-arm-kernel
We don't want to write to .text section. Let's move l2dis_3630
to .data and access it via a pointer.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <menon.nishanth@gmail.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep34xx.S | 16 +++++++++++++---
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
index 787cfda..f7c7bf8 100644
--- a/arch/arm/mach-omap2/sleep34xx.S
+++ b/arch/arm/mach-omap2/sleep34xx.S
@@ -86,7 +86,9 @@ ENTRY(enable_omap3630_toggle_l2_on_restore)
stmfd sp!, {lr} @ save registers on stack
/* Setup so that we will disable and enable l2 */
mov r1, #0x1
- adrl r2, l2dis_3630 @ may be too distant for plain adr
+ adrl r3, l2dis_3630_offset @ may be too distant for plain adr
+ ldr r2, [r3]
+ add r2, r2, r3
str r1, [r2]
ldmfd sp!, {pc} @ restore regs and return
ENDPROC(enable_omap3630_toggle_l2_on_restore)
@@ -415,7 +417,9 @@ ENTRY(omap3_restore)
cmp r2, #0x0 @ Check if target power state was OFF or RET
bne logic_l1_restore
- ldr r0, l2dis_3630
+ adr r1, l2dis_3630_offset
+ ldr r0, [r1]
+ add r0, r0, r1
cmp r0, #0x1 @ should we disable L2 on 3630?
bne skipl2dis
mrc p15, 0, r0, c1, c0, 1
@@ -484,7 +488,9 @@ l2_inv_gp:
mov r12, #0x2
smc #0 @ Call SMI monitor (smieq)
logic_l1_restore:
- ldr r1, l2dis_3630
+ adr r0, l2dis_3630_offset
+ ldr r1, [r0]
+ add r1, r1, r0
cmp r1, #0x1 @ Test if L2 re-enable needed on 3630
bne skipl2reen
mrc p15, 0, r1, c1, c0, 1
@@ -513,6 +519,10 @@ control_stat:
.word CONTROL_STAT
control_mem_rta:
.word CONTROL_MEM_RTA_CTRL
+l2dis_3630_offset:
+ .long l2dis_3630 - .
+
+ .data
l2dis_3630:
.word 0
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread* [PATCH 2/5] ARM: OMAP2+: Fix l2dis_3630 for rodata
@ 2016-01-19 17:48 ` Tony Lindgren
0 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-19 17:48 UTC (permalink / raw)
To: linux-arm-kernel
We don't want to write to .text section. Let's move l2dis_3630
to .data and access it via a pointer.
Cc: Kees Cook <keescook@chromium.org>
Cc: Laura Abbott <labbott@redhat.com>
Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
Cc: Nishanth Menon <menon.nishanth@gmail.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Tero Kristo <t-kristo@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
arch/arm/mach-omap2/sleep34xx.S | 16 +++++++++++++---
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
index 787cfda..f7c7bf8 100644
--- a/arch/arm/mach-omap2/sleep34xx.S
+++ b/arch/arm/mach-omap2/sleep34xx.S
@@ -86,7 +86,9 @@ ENTRY(enable_omap3630_toggle_l2_on_restore)
stmfd sp!, {lr} @ save registers on stack
/* Setup so that we will disable and enable l2 */
mov r1, #0x1
- adrl r2, l2dis_3630 @ may be too distant for plain adr
+ adrl r3, l2dis_3630_offset @ may be too distant for plain adr
+ ldr r2, [r3]
+ add r2, r2, r3
str r1, [r2]
ldmfd sp!, {pc} @ restore regs and return
ENDPROC(enable_omap3630_toggle_l2_on_restore)
@@ -415,7 +417,9 @@ ENTRY(omap3_restore)
cmp r2, #0x0 @ Check if target power state was OFF or RET
bne logic_l1_restore
- ldr r0, l2dis_3630
+ adr r1, l2dis_3630_offset
+ ldr r0, [r1]
+ add r0, r0, r1
cmp r0, #0x1 @ should we disable L2 on 3630?
bne skipl2dis
mrc p15, 0, r0, c1, c0, 1
@@ -484,7 +488,9 @@ l2_inv_gp:
mov r12, #0x2
smc #0 @ Call SMI monitor (smieq)
logic_l1_restore:
- ldr r1, l2dis_3630
+ adr r0, l2dis_3630_offset
+ ldr r1, [r0]
+ add r1, r1, r0
cmp r1, #0x1 @ Test if L2 re-enable needed on 3630
bne skipl2reen
mrc p15, 0, r1, c1, c0, 1
@@ -513,6 +519,10 @@ control_stat:
.word CONTROL_STAT
control_mem_rta:
.word CONTROL_MEM_RTA_CTRL
+l2dis_3630_offset:
+ .long l2dis_3630 - .
+
+ .data
l2dis_3630:
.word 0
--
2.7.0.rc3
^ permalink raw reply related [flat|nested] 20+ messages in thread* Re: [PATCH 2/5] ARM: OMAP2+: Fix l2dis_3630 for rodata
2016-01-19 17:48 ` Tony Lindgren
@ 2016-01-19 18:10 ` Nicolas Pitre
-1 siblings, 0 replies; 20+ messages in thread
From: Nicolas Pitre @ 2016-01-19 18:10 UTC (permalink / raw)
To: Tony Lindgren
Cc: Richard Woodruff, Russell King, Kees Cook, Tero Kristo,
linux-omap, Nishanth Menon, Laura Abbott, linux-arm-kernel
On Tue, 19 Jan 2016, Tony Lindgren wrote:
> We don't want to write to .text section. Let's move l2dis_3630
> to .data and access it via a pointer.
>
> Cc: Kees Cook <keescook@chromium.org>
> Cc: Laura Abbott <labbott@redhat.com>
> Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
> Cc: Nishanth Menon <menon.nishanth@gmail.com>
> Cc: Richard Woodruff <r-woodruff2@ti.com>
> Cc: Russell King <linux@arm.linux.org.uk>
> Cc: Tero Kristo <t-kristo@ti.com>
> Signed-off-by: Tony Lindgren <tony@atomide.com>
> ---
> arch/arm/mach-omap2/sleep34xx.S | 16 +++++++++++++---
> 1 file changed, 13 insertions(+), 3 deletions(-)
>
> diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
> index 787cfda..f7c7bf8 100644
> --- a/arch/arm/mach-omap2/sleep34xx.S
> +++ b/arch/arm/mach-omap2/sleep34xx.S
> @@ -86,7 +86,9 @@ ENTRY(enable_omap3630_toggle_l2_on_restore)
> stmfd sp!, {lr} @ save registers on stack
> /* Setup so that we will disable and enable l2 */
> mov r1, #0x1
> - adrl r2, l2dis_3630 @ may be too distant for plain adr
> + adrl r3, l2dis_3630_offset @ may be too distant for plain adr
> + ldr r2, [r3]
> + add r2, r2, r3
> str r1, [r2]
You could combine the add with the following str:
str r1, [r2, r3]
> ldmfd sp!, {pc} @ restore regs and return
> ENDPROC(enable_omap3630_toggle_l2_on_restore)
> @@ -415,7 +417,9 @@ ENTRY(omap3_restore)
> cmp r2, #0x0 @ Check if target power state was OFF or RET
> bne logic_l1_restore
>
> - ldr r0, l2dis_3630
> + adr r1, l2dis_3630_offset
> + ldr r0, [r1]
> + add r0, r0, r1
> cmp r0, #0x1 @ should we disable L2 on 3630?
This looks wrong. You're testing the first bit of the address for
l2dis_3630 rather than its content.
> bne skipl2dis
> mrc p15, 0, r0, c1, c0, 1
> @@ -484,7 +488,9 @@ l2_inv_gp:
> mov r12, #0x2
> smc #0 @ Call SMI monitor (smieq)
> logic_l1_restore:
> - ldr r1, l2dis_3630
> + adr r0, l2dis_3630_offset
> + ldr r1, [r0]
> + add r1, r1, r0
> cmp r1, #0x1 @ Test if L2 re-enable needed on 3630
Ditto here.
> bne skipl2reen
> mrc p15, 0, r1, c1, c0, 1
> @@ -513,6 +519,10 @@ control_stat:
> .word CONTROL_STAT
> control_mem_rta:
> .word CONTROL_MEM_RTA_CTRL
> +l2dis_3630_offset:
> + .long l2dis_3630 - .
> +
> + .data
> l2dis_3630:
> .word 0
>
> --
> 2.7.0.rc3
>
>
^ permalink raw reply [flat|nested] 20+ messages in thread* [PATCH 2/5] ARM: OMAP2+: Fix l2dis_3630 for rodata
@ 2016-01-19 18:10 ` Nicolas Pitre
0 siblings, 0 replies; 20+ messages in thread
From: Nicolas Pitre @ 2016-01-19 18:10 UTC (permalink / raw)
To: linux-arm-kernel
On Tue, 19 Jan 2016, Tony Lindgren wrote:
> We don't want to write to .text section. Let's move l2dis_3630
> to .data and access it via a pointer.
>
> Cc: Kees Cook <keescook@chromium.org>
> Cc: Laura Abbott <labbott@redhat.com>
> Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
> Cc: Nishanth Menon <menon.nishanth@gmail.com>
> Cc: Richard Woodruff <r-woodruff2@ti.com>
> Cc: Russell King <linux@arm.linux.org.uk>
> Cc: Tero Kristo <t-kristo@ti.com>
> Signed-off-by: Tony Lindgren <tony@atomide.com>
> ---
> arch/arm/mach-omap2/sleep34xx.S | 16 +++++++++++++---
> 1 file changed, 13 insertions(+), 3 deletions(-)
>
> diff --git a/arch/arm/mach-omap2/sleep34xx.S b/arch/arm/mach-omap2/sleep34xx.S
> index 787cfda..f7c7bf8 100644
> --- a/arch/arm/mach-omap2/sleep34xx.S
> +++ b/arch/arm/mach-omap2/sleep34xx.S
> @@ -86,7 +86,9 @@ ENTRY(enable_omap3630_toggle_l2_on_restore)
> stmfd sp!, {lr} @ save registers on stack
> /* Setup so that we will disable and enable l2 */
> mov r1, #0x1
> - adrl r2, l2dis_3630 @ may be too distant for plain adr
> + adrl r3, l2dis_3630_offset @ may be too distant for plain adr
> + ldr r2, [r3]
> + add r2, r2, r3
> str r1, [r2]
You could combine the add with the following str:
str r1, [r2, r3]
> ldmfd sp!, {pc} @ restore regs and return
> ENDPROC(enable_omap3630_toggle_l2_on_restore)
> @@ -415,7 +417,9 @@ ENTRY(omap3_restore)
> cmp r2, #0x0 @ Check if target power state was OFF or RET
> bne logic_l1_restore
>
> - ldr r0, l2dis_3630
> + adr r1, l2dis_3630_offset
> + ldr r0, [r1]
> + add r0, r0, r1
> cmp r0, #0x1 @ should we disable L2 on 3630?
This looks wrong. You're testing the first bit of the address for
l2dis_3630 rather than its content.
> bne skipl2dis
> mrc p15, 0, r0, c1, c0, 1
> @@ -484,7 +488,9 @@ l2_inv_gp:
> mov r12, #0x2
> smc #0 @ Call SMI monitor (smieq)
> logic_l1_restore:
> - ldr r1, l2dis_3630
> + adr r0, l2dis_3630_offset
> + ldr r1, [r0]
> + add r1, r1, r0
> cmp r1, #0x1 @ Test if L2 re-enable needed on 3630
Ditto here.
> bne skipl2reen
> mrc p15, 0, r1, c1, c0, 1
> @@ -513,6 +519,10 @@ control_stat:
> .word CONTROL_STAT
> control_mem_rta:
> .word CONTROL_MEM_RTA_CTRL
> +l2dis_3630_offset:
> + .long l2dis_3630 - .
> +
> + .data
> l2dis_3630:
> .word 0
>
> --
> 2.7.0.rc3
>
>
^ permalink raw reply [flat|nested] 20+ messages in thread* Re: [PATCH 2/5] ARM: OMAP2+: Fix l2dis_3630 for rodata
2016-01-19 18:10 ` Nicolas Pitre
@ 2016-01-19 18:40 ` Tony Lindgren
-1 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-19 18:40 UTC (permalink / raw)
To: Nicolas Pitre
Cc: Richard Woodruff, Russell King, Kees Cook, Tero Kristo,
linux-omap, Nishanth Menon, Laura Abbott, linux-arm-kernel
* Nicolas Pitre <nicolas.pitre@linaro.org> [160119 10:10]:
> On Tue, 19 Jan 2016, Tony Lindgren wrote:
> > --- a/arch/arm/mach-omap2/sleep34xx.S
> > +++ b/arch/arm/mach-omap2/sleep34xx.S
> > @@ -86,7 +86,9 @@ ENTRY(enable_omap3630_toggle_l2_on_restore)
> > stmfd sp!, {lr} @ save registers on stack
> > /* Setup so that we will disable and enable l2 */
> > mov r1, #0x1
> > - adrl r2, l2dis_3630 @ may be too distant for plain adr
> > + adrl r3, l2dis_3630_offset @ may be too distant for plain adr
> > + ldr r2, [r3]
> > + add r2, r2, r3
> > str r1, [r2]
>
> You could combine the add with the following str:
OK.
> str r1, [r2, r3]
>
> > ldmfd sp!, {pc} @ restore regs and return
> > ENDPROC(enable_omap3630_toggle_l2_on_restore)
> > @@ -415,7 +417,9 @@ ENTRY(omap3_restore)
> > cmp r2, #0x0 @ Check if target power state was OFF or RET
> > bne logic_l1_restore
> >
> > - ldr r0, l2dis_3630
> > + adr r1, l2dis_3630_offset
> > + ldr r0, [r1]
> > + add r0, r0, r1
> > cmp r0, #0x1 @ should we disable L2 on 3630?
>
> This looks wrong. You're testing the first bit of the address for
> l2dis_3630 rather than its content.
Ugh yup a ldr missing missing there. Will check the whole series
and repost. Thanks for checking!
Regards,
Tony
^ permalink raw reply [flat|nested] 20+ messages in thread* [PATCH 2/5] ARM: OMAP2+: Fix l2dis_3630 for rodata
@ 2016-01-19 18:40 ` Tony Lindgren
0 siblings, 0 replies; 20+ messages in thread
From: Tony Lindgren @ 2016-01-19 18:40 UTC (permalink / raw)
To: linux-arm-kernel
* Nicolas Pitre <nicolas.pitre@linaro.org> [160119 10:10]:
> On Tue, 19 Jan 2016, Tony Lindgren wrote:
> > --- a/arch/arm/mach-omap2/sleep34xx.S
> > +++ b/arch/arm/mach-omap2/sleep34xx.S
> > @@ -86,7 +86,9 @@ ENTRY(enable_omap3630_toggle_l2_on_restore)
> > stmfd sp!, {lr} @ save registers on stack
> > /* Setup so that we will disable and enable l2 */
> > mov r1, #0x1
> > - adrl r2, l2dis_3630 @ may be too distant for plain adr
> > + adrl r3, l2dis_3630_offset @ may be too distant for plain adr
> > + ldr r2, [r3]
> > + add r2, r2, r3
> > str r1, [r2]
>
> You could combine the add with the following str:
OK.
> str r1, [r2, r3]
>
> > ldmfd sp!, {pc} @ restore regs and return
> > ENDPROC(enable_omap3630_toggle_l2_on_restore)
> > @@ -415,7 +417,9 @@ ENTRY(omap3_restore)
> > cmp r2, #0x0 @ Check if target power state was OFF or RET
> > bne logic_l1_restore
> >
> > - ldr r0, l2dis_3630
> > + adr r1, l2dis_3630_offset
> > + ldr r0, [r1]
> > + add r0, r0, r1
> > cmp r0, #0x1 @ should we disable L2 on 3630?
>
> This looks wrong. You're testing the first bit of the address for
> l2dis_3630 rather than its content.
Ugh yup a ldr missing missing there. Will check the whole series
and repost. Thanks for checking!
Regards,
Tony
^ permalink raw reply [flat|nested] 20+ messages in thread