altos/arm: Align data so that gcc 5.4 doesn't do byte-accesses. Add -Wcast-align

Gcc 5.4.1 tracks alignment of data through assignments, so that a
uint32_t pointer which comes from byte-aligned uint8_t data:

extern uint8_t foo[];

	uint32_t	*q = (void *) foo;

Fetches and stores through this pointer are done bytewise. This is
slow (meh), but if q references a device register, things to bad very
quickly.

This patch works around this bug in the compiler by adding
__attribute__((aligned(4))) tags to some variables, or changing them
from uint8_t to uint32_t. Places doing this will now be caught as I've
added -Wcast-align to the compiler flags. That required adding (void
*) casts, after the relevant code was checked to make sure the
compiler could tell that the addresses were aligned.

Signed-off-by: Keith Packard <keithp@keithp.com>

5 files changed, 5 insertions, 5 deletions

diff --git a/src/stmf0/Makefile-stmf0.defs b/src/stmf0/Makefile-stmf0.defs
index 4862f46e..f3296b69 100644
--- a/src/stmf0/Makefile-stmf0.defs
+++ b/src/stmf0/Makefile-stmf0.defs
@@ -25,7 +25,7 @@ endif
 ELFTOHEX=$(TOPDIR)/../ao-tools/ao-elftohex/ao-elftohex
 CC=$(ARM_CC)
 
-WARN_FLAGS=-Wall -Wextra -Werror
+WARN_FLAGS=-Wall -Wextra -Werror -Wcast-align
 
 AO_CFLAGS=-I. -I$(TOPDIR)/stmf0 -I$(TOPDIR)/kernel -I$(TOPDIR)/drivers -I$(TOPDIR)/product -I$(TOPDIR) -I$(TOPDIR)/math $(PDCLIB_INCLUDES) 
 STMF0_CFLAGS=-std=gnu99 -mlittle-endian -mcpu=cortex-m0 -mthumb\
diff --git a/src/stmf0/ao_adc_fast.h b/src/stmf0/ao_adc_fast.h
index b8b5e003..3f0b0547 100644
--- a/src/stmf0/ao_adc_fast.h
+++ b/src/stmf0/ao_adc_fast.h
@@ -28,7 +28,7 @@ ao_adc_init(void);
 /* Total ring size in samples */
 #define AO_ADC_RING_SIZE	256
 
-extern uint16_t	ao_adc_ring[AO_ADC_RING_SIZE];
+extern uint16_t	ao_adc_ring[AO_ADC_RING_SIZE] __attribute__((aligned(4)));
 
 #define ao_adc_ring_step(pos,inc)	(((pos) + (inc)) & (AO_ADC_RING_SIZE - 1))
 
diff --git a/src/stmf0/ao_arch_funcs.h b/src/stmf0/ao_arch_funcs.h
index d35bafbd..c38ce41a 100644
--- a/src/stmf0/ao_arch_funcs.h
+++ b/src/stmf0/ao_arch_funcs.h
@@ -366,7 +366,7 @@ ao_arch_memory_barrier() {
 static inline void
 ao_arch_init_stack(struct ao_task *task, void *start)
 {
-	uint32_t	*sp = (uint32_t *) (task->stack + AO_STACK_SIZE);
+	uint32_t	*sp = (uint32_t *) ((void *) task->stack + AO_STACK_SIZE);
 	uint32_t	a = (uint32_t) start;
 	int		i;
 
diff --git a/src/stmf0/ao_usb_stm.c b/src/stmf0/ao_usb_stm.c
index cbedb996..652b3b6c 100644
--- a/src/stmf0/ao_usb_stm.c
+++ b/src/stmf0/ao_usb_stm.c
@@ -185,7 +185,7 @@ static inline uint32_t set_toggle(uint32_t 	current_value,
 
 static inline uint16_t *ao_usb_packet_buffer_addr(uint16_t sram_addr)
 {
-	return (uint16_t *) (stm_usb_sram + sram_addr);
+	return (uint16_t *) (void *) (stm_usb_sram + sram_addr);
 }
 
 static inline uint16_t ao_usb_packet_buffer_offset(uint16_t *addr)
diff --git a/src/stmf0/stm32f0.h b/src/stmf0/stm32f0.h
index 182cd963..1c33f020 100644
--- a/src/stmf0/stm32f0.h
+++ b/src/stmf0/stm32f0.h
@@ -1996,7 +1996,7 @@ union stm_usb_bdt {
 
 #define STM_USB_BDT_SIZE	8
 
-extern uint8_t stm_usb_sram[];
+extern uint8_t stm_usb_sram[] __attribute__((aligned(4)));
 
 struct stm_exti {
 	vuint32_t	imr;