summaryrefslogtreecommitdiff
path: root/arch
diff options
context:
space:
mode:
authorHeiko Schocher <hs@denx.de>2010-09-17 13:10:31 +0200
committerWolfgang Denk <wd@denx.de>2010-09-19 19:29:51 +0200
commit95c6f6d34d4ff23f4d005488d84682eec5fa9ec8 (patch)
treedf364195d710c9af05e9fa3341dd6e13fc121173 /arch
parentc3330e9d6a11b6ead4a2346001338ce884b5832b (diff)
downloadu-boot-imx-95c6f6d34d4ff23f4d005488d84682eec5fa9ec8.zip
u-boot-imx-95c6f6d34d4ff23f4d005488d84682eec5fa9ec8.tar.gz
u-boot-imx-95c6f6d34d4ff23f4d005488d84682eec5fa9ec8.tar.bz2
ARM V7 (OMAP): add data cache support, test on Beagle board
Add data cache support for ARM V7 systems. Used cache flush functions from linux:arch/arm/mm/cache-v7.S developed from Catalin Marinas. Enable "cache" command on Beagle board and test performance. Test 1: Loading 127 MB of data from NAND flash into RAM: Instr. Cache off on on Data Cache off off on -------------------------------------------------- Beagle (Cortex A8) 116s 106s 30.3s = x 3.8 Test 2: uncompressing a gzipped image from RAM to RAM (size compressed: 6.5 MiB, uncompressed: 35 MiB): Instr. Cache off on on Data Cache off off on -------------------------------------------------- Beagle (Cortex A8) 1.84s 1.64s 0.12s = x 15.3 Portions of this work were supported by funding from the CE Linux Forum. Signed-off-by: Heiko Schocher <hs@denx.de> Reviewed-by: Ben Gardiner<bengardiner@nanometrics.ca>
Diffstat (limited to 'arch')
-rw-r--r--arch/arm/cpu/armv7/omap3/cache.S80
-rw-r--r--arch/arm/lib/cache.c5
2 files changed, 85 insertions, 0 deletions
diff --git a/arch/arm/cpu/armv7/omap3/cache.S b/arch/arm/cpu/armv7/omap3/cache.S
index fdc1666..cda87ba 100644
--- a/arch/arm/cpu/armv7/omap3/cache.S
+++ b/arch/arm/cpu/armv7/omap3/cache.S
@@ -181,3 +181,83 @@ setup_auxcr:
orrlt r0, r0, #1 << 27
.word 0xE1600070 @ SMC
bx lr
+
+.align 5
+.global v7_flush_dcache_all
+.global v7_flush_cache_all
+
+/*
+ * v7_flush_dcache_all()
+ *
+ * Flush the whole D-cache.
+ *
+ * Corrupted registers: r0-r7, r9-r11 (r6 only in Thumb mode)
+ *
+ * - mm - mm_struct describing address space
+ */
+v7_flush_dcache_all:
+# dmb @ ensure ordering with previous memory accesses
+ mrc p15, 1, r0, c0, c0, 1 @ read clidr
+ ands r3, r0, #0x7000000 @ extract loc from clidr
+ mov r3, r3, lsr #23 @ left align loc bit field
+ beq finished @ if loc is 0, then no need to clean
+ mov r10, #0 @ start clean at cache level 0
+loop1:
+ add r2, r10, r10, lsr #1 @ work out 3x current cache level
+ mov r1, r0, lsr r2 @ extract cache type bits from clidr
+ and r1, r1, #7 @ mask of the bits for current cache only
+ cmp r1, #2 @ see what cache we have at this level
+ blt skip @ skip if no cache, or just i-cache
+ mcr p15, 2, r10, c0, c0, 0 @ select current cache level in cssr
+ mcr p15, 0, r10, c7, c5, 4 @ flush prefetch buffer,
+ @ with armv7 this is 'isb',
+ @ but we compile with armv5
+ mrc p15, 1, r1, c0, c0, 0 @ read the new csidr
+ and r2, r1, #7 @ extract the length of the cache lines
+ add r2, r2, #4 @ add 4 (line length offset)
+ ldr r4, =0x3ff
+ ands r4, r4, r1, lsr #3 @ find maximum number on the way size
+ clz r5, r4 @ find bit position of way size increment
+ ldr r7, =0x7fff
+ ands r7, r7, r1, lsr #13 @ extract max number of the index size
+loop2:
+ mov r9, r4 @ create working copy of max way size
+loop3:
+ orr r11, r10, r9, lsl r5 @ factor way and cache number into r11
+ orr r11, r11, r7, lsl r2 @ factor index number into r11
+ mcr p15, 0, r11, c7, c14, 2 @ clean & invalidate by set/way
+ subs r9, r9, #1 @ decrement the way
+ bge loop3
+ subs r7, r7, #1 @ decrement the index
+ bge loop2
+skip:
+ add r10, r10, #2 @ increment cache number
+ cmp r3, r10
+ bgt loop1
+finished:
+ mov r10, #0 @ swith back to cache level 0
+ mcr p15, 2, r10, c0, c0, 0 @ select current cache level in cssr
+# dsb
+ mcr p15, 0, r10, c7, c5, 4 @ flush prefetch buffer,
+ @ with armv7 this is 'isb',
+ @ but we compile with armv5
+ mov pc, lr
+
+/*
+ * v7_flush_cache_all()
+ *
+ * Flush the entire cache system.
+ * The data cache flush is now achieved using atomic clean / invalidates
+ * working outwards from L1 cache. This is done using Set/Way based cache
+ * maintainance instructions.
+ * The instruction cache can still be invalidated back to the point of
+ * unification in a single instruction.
+ *
+ */
+v7_flush_cache_all:
+ stmfd sp!, {r0-r7, r9-r11, lr}
+ bl v7_flush_dcache_all
+ mov r0, #0
+ mcr p15, 0, r0, c7, c5, 0 @ I+BTB cache invalidate
+ ldmfd sp!, {r0-r7, r9-r11, lr}
+ mov pc, lr
diff --git a/arch/arm/lib/cache.c b/arch/arm/lib/cache.c
index b36fd24..3684cad 100644
--- a/arch/arm/lib/cache.c
+++ b/arch/arm/lib/cache.c
@@ -38,5 +38,10 @@ void flush_cache (unsigned long dummy1, unsigned long dummy2)
/* disable write buffer as well (page 2-22) */
asm("mcr p15, 0, %0, c7, c10, 4" : : "r" (0));
#endif
+#ifdef CONFIG_ARMCORTEXA8
+ void v7_flush_cache_all(void);
+
+ v7_flush_cache_all();
+#endif
return;
}