ARM: Add support for FA526 v2

Adds support for Faraday FA526 core. This core is used at least by:
Cortina Systems Gemini and Centroid family
Cavium Networks ECONA family
Grain Media GM8120
Pixelplus ImageARM
Prolific PL-1029
Faraday IP evaluation boards

v2:
- move TLB_BTB to separate patch
- update copyrights

Signed-off-by: Paulius Zaleckas <paulius.zaleckas@teltonika.lt>

1 files changed, 220 insertions, 0 deletions

diff --git a/arch/arm/mm/cache-fa.S b/arch/arm/mm/cache-fa.S
new file mode 100644
index 000000000000..b63a8f7b95cf
--- /dev/null
+++ b/arch/arm/mm/cache-fa.S
@@ -0,0 +1,220 @@
+/*
+ *  linux/arch/arm/mm/cache-fa.S
+ *
+ *  Copyright (C) 2005 Faraday Corp.
+ *  Copyright (C) 2008-2009 Paulius Zaleckas <paulius.zaleckas@teltonika.lt>
+ *
+ * Based on cache-v4wb.S:
+ *  Copyright (C) 1997-2002 Russell king
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ *  Processors: FA520 FA526 FA626       
+ */
+#include <linux/linkage.h>
+#include <linux/init.h>
+#include <asm/memory.h>
+#include <asm/page.h>
+
+#include "proc-macros.S"
+
+/*
+ * The size of one data cache line.
+ */
+#define CACHE_DLINESIZE 16
+
+/*
+ * The total size of the data cache.
+ */
+#ifdef CONFIG_ARCH_GEMINI
+#define CACHE_DSIZE     8192
+#else
+#define CACHE_DSIZE     16384 
+#endif 
+
+/* FIXME: put optimal value here. Current one is just estimation */
+#define CACHE_DLIMIT    (CACHE_DSIZE * 2)
+
+/*
+ *      flush_user_cache_all()
+ *
+ *      Clean and invalidate all cache entries in a particular address
+ *      space.
+ */
+ENTRY(fa_flush_user_cache_all)
+        /* FALLTHROUGH */
+/*
+ *      flush_kern_cache_all()
+ *
+ *      Clean and invalidate the entire cache.
+ */
+ENTRY(fa_flush_kern_cache_all)
+        mov     ip, #0
+        mov     r2, #VM_EXEC
+__flush_whole_cache:
+        mcr     p15, 0, ip, c7, c14, 0          @ clean/invalidate D cache
+        tst     r2, #VM_EXEC
+        mcrne   p15, 0, ip, c7, c5, 0           @ invalidate I cache
+        mcrne   p15, 0, ip, c7, c5, 6           @ invalidate BTB
+        mcrne   p15, 0, ip, c7, c10, 4          @ drain write buffer
+        mcrne   p15, 0, ip, c7, c5, 4           @ prefetch flush
+        mov     pc, lr
+
+/*
+ *      flush_user_cache_range(start, end, flags)
+ *
+ *      Invalidate a range of cache entries in the specified
+ *      address space.
+ *
+ *      - start - start address (inclusive, page aligned)
+ *      - end   - end address (exclusive, page aligned)
+ *      - flags - vma_area_struct flags describing address space
+ */
+ENTRY(fa_flush_user_cache_range)
+        mov     ip, #0
+        sub     r3, r1, r0                      @ calculate total size
+        cmp     r3, #CACHE_DLIMIT               @ total size >= limit?
+        bhs     __flush_whole_cache             @ flush whole D cache
+
+1:      tst     r2, #VM_EXEC
+        mcrne   p15, 0, r0, c7, c5, 1           @ invalidate I line
+        mcr     p15, 0, r0, c7, c14, 1          @ clean and invalidate D entry
+        add     r0, r0, #CACHE_DLINESIZE
+        cmp     r0, r1
+        blo     1b
+        tst     r2, #VM_EXEC
+        mcrne   p15, 0, ip, c7, c5, 6           @ invalidate BTB
+        mcrne   p15, 0, ip, c7, c10, 4          @ data write barrier
+        mcrne   p15, 0, ip, c7, c5, 4           @ prefetch flush
+        mov     pc, lr
+
+/*
+ *      coherent_kern_range(start, end)
+ *
+ *      Ensure coherency between the Icache and the Dcache in the
+ *      region described by start.  If you have non-snooping
+ *      Harvard caches, you need to implement this function.
+ *
+ *      - start  - virtual start address
+ *      - end    - virtual end address
+ */
+ENTRY(fa_coherent_kern_range)
+        /* fall through */
+
+/*
+ *      coherent_user_range(start, end)
+ *
+ *      Ensure coherency between the Icache and the Dcache in the
+ *      region described by start.  If you have non-snooping
+ *      Harvard caches, you need to implement this function.
+ *
+ *      - start  - virtual start address
+ *      - end    - virtual end address
+ */
+ENTRY(fa_coherent_user_range)
+        bic     r0, r0, #CACHE_DLINESIZE - 1
+1:      mcr     p15, 0, r0, c7, c14, 1          @ clean and invalidate D entry
+        mcr     p15, 0, r0, c7, c5, 1           @ invalidate I entry
+        add     r0, r0, #CACHE_DLINESIZE
+        cmp     r0, r1
+        blo     1b
+        mov     r0, #0
+        mcr     p15, 0, r0, c7, c5, 6           @ invalidate BTB
+        mcr     p15, 0, r0, c7, c10, 4          @ drain write buffer
+        mcr     p15, 0, r0, c7, c5, 4           @ prefetch flush
+        mov     pc, lr
+
+/*
+ *      flush_kern_dcache_page(kaddr)
+ *
+ *      Ensure that the data held in the page kaddr is written back
+ *      to the page in question.
+ *
+ *      - kaddr   - kernel address (guaranteed to be page aligned)
+ */
+ENTRY(fa_flush_kern_dcache_page)
+        add     r1, r0, #PAGE_SZ
+1:      mcr     p15, 0, r0, c7, c14, 1          @ clean & invalidate D line
+        add     r0, r0, #CACHE_DLINESIZE
+        cmp     r0, r1
+        blo     1b
+        mov     r0, #0
+        mcr     p15, 0, r0, c7, c5, 0           @ invalidate I cache
+        mcr     p15, 0, r0, c7, c10, 4          @ drain write buffer
+        mov     pc, lr
+
+/*
+ *      dma_inv_range(start, end)
+ *
+ *      Invalidate (discard) the specified virtual address range.
+ *      May not write back any entries.  If 'start' or 'end'
+ *      are not cache line aligned, those lines must be written
+ *      back.
+ *
+ *      - start  - virtual start address
+ *      - end    - virtual end address
+ */
+ENTRY(fa_dma_inv_range)
+        tst     r0, #CACHE_DLINESIZE - 1
+        bic     r0, r0, #CACHE_DLINESIZE - 1
+        mcrne   p15, 0, r0, c7, c14, 1          @ clean & invalidate D entry
+        tst     r1, #CACHE_DLINESIZE - 1
+        bic     r1, r1, #CACHE_DLINESIZE - 1
+        mcrne   p15, 0, r1, c7, c14, 1          @ clean & invalidate D entry
+1:      mcr     p15, 0, r0, c7, c6, 1           @ invalidate D entry
+        add     r0, r0, #CACHE_DLINESIZE
+        cmp     r0, r1
+        blo     1b
+        mov     r0, #0
+        mcr     p15, 0, r0, c7, c10, 4          @ drain write buffer
+        mov     pc, lr
+
+/*
+ *      dma_clean_range(start, end)
+ *
+ *      Clean (write back) the specified virtual address range.
+ *
+ *      - start  - virtual start address
+ *      - end    - virtual end address
+ */
+ENTRY(fa_dma_clean_range)
+        bic     r0, r0, #CACHE_DLINESIZE - 1
+1:      mcr     p15, 0, r0, c7, c10, 1          @ clean D entry
+        add     r0, r0, #CACHE_DLINESIZE
+        cmp     r0, r1
+        blo     1b
+        mov     r0, #0  
+        mcr     p15, 0, r0, c7, c10, 4          @ drain write buffer
+        mov     pc, lr
+
+/*
+ *      dma_flush_range(start,end)
+ *      - start   - virtual start address of region
+ *      - end     - virtual end address of region
+ */
+ENTRY(fa_dma_flush_range)
+        bic     r0, r0, #CACHE_DLINESIZE - 1
+1:      mcr     p15, 0, r0, c7, c14, 1          @ clean & invalidate D entry
+        add     r0, r0, #CACHE_DLINESIZE
+        cmp     r0, r1
+        blo     1b
+        mov     r0, #0  
+        mcr     p15, 0, r0, c7, c10, 4          @ drain write buffer
+        mov     pc, lr
+
+        __INITDATA
+
+        .type   fa_cache_fns, #object
+ENTRY(fa_cache_fns)
+        .long   fa_flush_kern_cache_all
+        .long   fa_flush_user_cache_all
+        .long   fa_flush_user_cache_range
+        .long   fa_coherent_kern_range
+        .long   fa_coherent_user_range
+        .long   fa_flush_kern_dcache_page
+        .long   fa_dma_inv_range
+        .long   fa_dma_clean_range
+        .long   fa_dma_flush_range
+        .size   fa_cache_fns, . - fa_cache_fns