diff options
Diffstat (limited to 'arch/sh/lib64')
| -rw-r--r-- | arch/sh/lib64/Makefile | 4 | ||||
| -rw-r--r-- | arch/sh/lib64/c-checksum.c | 214 | ||||
| -rw-r--r-- | arch/sh/lib64/clear_page.S | 54 | ||||
| -rw-r--r-- | arch/sh/lib64/copy_user_memcpy.S | 2 | ||||
| -rw-r--r-- | arch/sh/lib64/dbg.c | 248 | ||||
| -rw-r--r-- | arch/sh/lib64/memcpy.S | 2 |
6 files changed, 4 insertions, 520 deletions
diff --git a/arch/sh/lib64/Makefile b/arch/sh/lib64/Makefile index 4bacb9e8347..69779ff741d 100644 --- a/arch/sh/lib64/Makefile +++ b/arch/sh/lib64/Makefile @@ -10,8 +10,8 @@ # # Panic should really be compiled as PIC -lib-y := udelay.o c-checksum.o dbg.o panic.o memcpy.o memset.o \ - copy_user_memcpy.o copy_page.o clear_page.o strcpy.o strlen.o +lib-y := udelay.o panic.o memcpy.o memset.o \ + copy_user_memcpy.o copy_page.o strcpy.o strlen.o # Extracted from libgcc lib-y += udivsi3.o udivdi3.o sdivsi3.o diff --git a/arch/sh/lib64/c-checksum.c b/arch/sh/lib64/c-checksum.c deleted file mode 100644 index 73c0877e3a2..00000000000 --- a/arch/sh/lib64/c-checksum.c +++ /dev/null @@ -1,214 +0,0 @@ -/* - * arch/sh/lib64/c-checksum.c - * - * This file contains network checksum routines that are better done - * in an architecture-specific manner due to speed.. - */ -#include <linux/string.h> -#include <linux/errno.h> -#include <linux/kernel.h> -#include <linux/module.h> -#include <asm/byteorder.h> -#include <asm/uaccess.h> - -static inline unsigned short from64to16(unsigned long long x) -{ - /* add up 32-bit words for 33 bits */ - x = (x & 0xffffffff) + (x >> 32); - /* add up 16-bit and 17-bit words for 17+c bits */ - x = (x & 0xffff) + (x >> 16); - /* add up 16-bit and 2-bit for 16+c bit */ - x = (x & 0xffff) + (x >> 16); - /* add up carry.. */ - x = (x & 0xffff) + (x >> 16); - return x; -} - -static inline unsigned short foldto16(unsigned long x) -{ - /* add up 16-bit for 17 bits */ - x = (x & 0xffff) + (x >> 16); - /* add up carry.. */ - x = (x & 0xffff) + (x >> 16); - return x; -} - -static inline unsigned short myfoldto16(unsigned long long x) -{ - /* Fold down to 32-bits so we don't lose in the typedef-less - network stack. */ - /* 64 to 33 */ - x = (x & 0xffffffff) + (x >> 32); - /* 33 to 32 */ - x = (x & 0xffffffff) + (x >> 32); - - /* add up 16-bit for 17 bits */ - x = (x & 0xffff) + (x >> 16); - /* add up carry.. */ - x = (x & 0xffff) + (x >> 16); - return x; -} - -#define odd(x) ((x)&1) -#define U16(x) ntohs(x) - -static unsigned long do_csum(const unsigned char *buff, int len) -{ - int odd, count; - unsigned long result = 0; - - pr_debug("do_csum buff %p, len %d (0x%x)\n", buff, len, len); -#ifdef DEBUG - for (i = 0; i < len; i++) { - if ((i % 26) == 0) - printk("\n"); - printk("%02X ", buff[i]); - } -#endif - - if (len <= 0) - goto out; - - odd = 1 & (unsigned long) buff; - if (odd) { - result = *buff << 8; - len--; - buff++; - } - count = len >> 1; /* nr of 16-bit words.. */ - if (count) { - if (2 & (unsigned long) buff) { - result += *(unsigned short *) buff; - count--; - len -= 2; - buff += 2; - } - count >>= 1; /* nr of 32-bit words.. */ - if (count) { - unsigned long carry = 0; - do { - unsigned long w = *(unsigned long *) buff; - buff += 4; - count--; - result += carry; - result += w; - carry = (w > result); - } while (count); - result += carry; - result = (result & 0xffff) + (result >> 16); - } - if (len & 2) { - result += *(unsigned short *) buff; - buff += 2; - } - } - if (len & 1) - result += *buff; - result = foldto16(result); - if (odd) - result = ((result >> 8) & 0xff) | ((result & 0xff) << 8); - - pr_debug("\nCHECKSUM is 0x%lx\n", result); - - out: - return result; -} - -/* computes the checksum of a memory block at buff, length len, - and adds in "sum" (32-bit) */ -__wsum csum_partial(const void *buff, int len, __wsum sum) -{ - unsigned long long result = do_csum(buff, len); - - /* add in old sum, and carry.. */ - result += (__force u32)sum; - /* 32+c bits -> 32 bits */ - result = (result & 0xffffffff) + (result >> 32); - - pr_debug("csum_partial, buff %p len %d sum 0x%x result=0x%016Lx\n", - buff, len, sum, result); - - return (__force __wsum)result; -} - -/* Copy while checksumming, otherwise like csum_partial. */ -__wsum -csum_partial_copy_nocheck(const void *src, void *dst, int len, __wsum sum) -{ - sum = csum_partial(src, len, sum); - memcpy(dst, src, len); - - return sum; -} - -/* Copy from userspace and compute checksum. If we catch an exception - then zero the rest of the buffer. */ -__wsum -csum_partial_copy_from_user(const void __user *src, void *dst, int len, - __wsum sum, int *err_ptr) -{ - int missing; - - pr_debug - ("csum_partial_copy_from_user src %p, dest %p, len %d, sum %08x, err_ptr %p\n", - src, dst, len, sum, err_ptr); - missing = copy_from_user(dst, src, len); - pr_debug(" access_ok %d\n", __access_ok((unsigned long) src, len)); - pr_debug(" missing %d\n", missing); - if (missing) { - memset(dst + len - missing, 0, missing); - *err_ptr = -EFAULT; - } - - return csum_partial(dst, len, sum); -} - -/* Copy to userspace and compute checksum. */ -__wsum -csum_partial_copy_to_user(const unsigned char *src, unsigned char *dst, int len, - __wsum sum, int *err_ptr) -{ - sum = csum_partial(src, len, sum); - - if (copy_to_user(dst, src, len)) - *err_ptr = -EFAULT; - - return sum; -} - -/* - * This is a version of ip_compute_csum() optimized for IP headers, - * which always checksum on 4 octet boundaries. - */ -__sum16 ip_fast_csum(const void *iph, unsigned int ihl) -{ - pr_debug("ip_fast_csum %p,%d\n", iph, ihl); - - return (__force __sum16)~do_csum(iph, ihl * 4); -} - -__wsum csum_tcpudp_nofold(__be32 saddr, __be32 daddr, - unsigned short len, - unsigned short proto, __wsum sum) -{ - unsigned long long result; - - pr_debug("ntohs(0x%x)=0x%x\n", 0xdead, ntohs(0xdead)); - pr_debug("htons(0x%x)=0x%x\n", 0xdead, htons(0xdead)); - - result = (__force u64) saddr + (__force u64) daddr + - (__force u64) sum + ((len + proto) << 8); - - /* Fold down to 32-bits so we don't lose in the typedef-less - network stack. */ - /* 64 to 33 */ - result = (result & 0xffffffff) + (result >> 32); - /* 33 to 32 */ - result = (result & 0xffffffff) + (result >> 32); - - pr_debug("%s saddr %x daddr %x len %x proto %x sum %x result %08Lx\n", - __func__, saddr, daddr, len, proto, sum, result); - - return (__wsum)result; -} -EXPORT_SYMBOL(csum_tcpudp_nofold); diff --git a/arch/sh/lib64/clear_page.S b/arch/sh/lib64/clear_page.S deleted file mode 100644 index 007ab48ecc1..00000000000 --- a/arch/sh/lib64/clear_page.S +++ /dev/null @@ -1,54 +0,0 @@ -/* - Copyright 2003 Richard Curnow, SuperH (UK) Ltd. - - This file is subject to the terms and conditions of the GNU General Public - License. See the file "COPYING" in the main directory of this archive - for more details. - - Tight version of memset for the case of just clearing a page. It turns out - that having the alloco's spaced out slightly due to the increment/branch - pair causes them to contend less for access to the cache. Similarly, - keeping the stores apart from the allocos causes less contention. => Do two - separate loops. Do multiple stores per loop to amortise the - increment/branch cost a little. - - Parameters: - r2 : source effective address (start of page) - - Always clears 4096 bytes. - - Note : alloco guarded by synco to avoid TAKum03020 erratum - -*/ - - .section .text..SHmedia32,"ax" - .little - - .balign 8 - .global clear_page -clear_page: - pta/l 1f, tr1 - pta/l 2f, tr2 - ptabs/l r18, tr0 - - movi 4096, r7 - add r2, r7, r7 - add r2, r63, r6 -1: - alloco r6, 0 - synco ! TAKum03020 - addi r6, 32, r6 - bgt/l r7, r6, tr1 - - add r2, r63, r6 -2: - st.q r6, 0, r63 - st.q r6, 8, r63 - st.q r6, 16, r63 - st.q r6, 24, r63 - addi r6, 32, r6 - bgt/l r7, r6, tr2 - - blink tr0, r63 - - diff --git a/arch/sh/lib64/copy_user_memcpy.S b/arch/sh/lib64/copy_user_memcpy.S index 2a62816d2dd..49aeabeba2c 100644 --- a/arch/sh/lib64/copy_user_memcpy.S +++ b/arch/sh/lib64/copy_user_memcpy.S @@ -27,7 +27,7 @@ ! 2.: When there are two or three bytes in the last word of an 11-or-more ! bytes memory chunk to b copied, the rest of the word can be read ! without side effects. -! This could be easily changed by increasing the minumum size of +! This could be easily changed by increasing the minimum size of ! a fast memcpy and the amount subtracted from r7 before L_2l_loop be 2, ! however, this would cost a few extra cyles on average. ! For SHmedia, the assumption is that any quadword can be read in its diff --git a/arch/sh/lib64/dbg.c b/arch/sh/lib64/dbg.c deleted file mode 100644 index 6152a6a6d9c..00000000000 --- a/arch/sh/lib64/dbg.c +++ /dev/null @@ -1,248 +0,0 @@ -/*-------------------------------------------------------------------------- --- --- Identity : Linux50 Debug Funcions --- --- File : arch/sh/lib64/dbg.c --- --- Copyright 2000, 2001 STMicroelectronics Limited. --- Copyright 2004 Richard Curnow (evt_debug etc) --- ---------------------------------------------------------------------------*/ -#include <linux/types.h> -#include <linux/kernel.h> -#include <linux/sched.h> -#include <linux/mm.h> -#include <linux/fs.h> -#include <asm/mmu_context.h> - -typedef u64 regType_t; - -static regType_t getConfigReg(u64 id) -{ - register u64 reg __asm__("r2"); - asm volatile ("getcfg %1, 0, %0":"=r" (reg):"r"(id)); - return (reg); -} - -/* ======================================================================= */ - -static char *szTab[] = { "4k", "64k", "1M", "512M" }; -static char *protTab[] = { "----", - "---R", - "--X-", - "--XR", - "-W--", - "-W-R", - "-WX-", - "-WXR", - "U---", - "U--R", - "U-X-", - "U-XR", - "UW--", - "UW-R", - "UWX-", - "UWXR" -}; -#define ITLB_BASE 0x00000000 -#define DTLB_BASE 0x00800000 -#define MAX_TLBs 64 -/* PTE High */ -#define GET_VALID(pte) ((pte) & 0x1) -#define GET_SHARED(pte) ((pte) & 0x2) -#define GET_ASID(pte) ((pte >> 2) & 0x0ff) -#define GET_EPN(pte) ((pte) & 0xfffff000) - -/* PTE Low */ -#define GET_CBEHAVIOR(pte) ((pte) & 0x3) -#define GET_PAGE_SIZE(pte) szTab[((pte >> 3) & 0x3)] -#define GET_PROTECTION(pte) protTab[((pte >> 6) & 0xf)] -#define GET_PPN(pte) ((pte) & 0xfffff000) - -#define PAGE_1K_MASK 0x00000000 -#define PAGE_4K_MASK 0x00000010 -#define PAGE_64K_MASK 0x00000080 -#define MMU_PAGESIZE_MASK (PAGE_64K_MASK | PAGE_4K_MASK) -#define PAGE_1MB_MASK MMU_PAGESIZE_MASK -#define PAGE_1K (1024) -#define PAGE_4K (1024 * 4) -#define PAGE_64K (1024 * 64) -#define PAGE_1MB (1024 * 1024) - -#define HOW_TO_READ_TLB_CONTENT \ - "[ ID] PPN EPN ASID Share CB P.Size PROT.\n" - -void print_single_tlb(unsigned long tlb, int single_print) -{ - regType_t pteH; - regType_t pteL; - unsigned int valid, shared, asid, epn, cb, ppn; - char *pSize; - char *pProt; - - /* - ** in case of single print <single_print> is true, this implies: - ** 1) print the TLB in any case also if NOT VALID - ** 2) print out the header - */ - - pteH = getConfigReg(tlb); - valid = GET_VALID(pteH); - if (single_print) - printk(HOW_TO_READ_TLB_CONTENT); - else if (!valid) - return; - - pteL = getConfigReg(tlb + 1); - - shared = GET_SHARED(pteH); - asid = GET_ASID(pteH); - epn = GET_EPN(pteH); - cb = GET_CBEHAVIOR(pteL); - pSize = GET_PAGE_SIZE(pteL); - pProt = GET_PROTECTION(pteL); - ppn = GET_PPN(pteL); - printk("[%c%2ld] 0x%08x 0x%08x %03d %02x %02x %4s %s\n", - ((valid) ? ' ' : 'u'), ((tlb & 0x0ffff) / TLB_STEP), - ppn, epn, asid, shared, cb, pSize, pProt); -} - -void print_dtlb(void) -{ - int count; - unsigned long tlb; - - printk(" ================= SH-5 D-TLBs Status ===================\n"); - printk(HOW_TO_READ_TLB_CONTENT); - tlb = DTLB_BASE; - for (count = 0; count < MAX_TLBs; count++, tlb += TLB_STEP) - print_single_tlb(tlb, 0); - printk - (" =============================================================\n"); -} - -void print_itlb(void) -{ - int count; - unsigned long tlb; - - printk(" ================= SH-5 I-TLBs Status ===================\n"); - printk(HOW_TO_READ_TLB_CONTENT); - tlb = ITLB_BASE; - for (count = 0; count < MAX_TLBs; count++, tlb += TLB_STEP) - print_single_tlb(tlb, 0); - printk - (" =============================================================\n"); -} - -void show_excp_regs(char *from, int trapnr, int signr, struct pt_regs *regs) -{ - - unsigned long long ah, al, bh, bl, ch, cl; - - printk("\n"); - printk("EXCEPTION - %s: task %d; Linux trap # %d; signal = %d\n", - ((from) ? from : "???"), current->pid, trapnr, signr); - - asm volatile ("getcon " __EXPEVT ", %0":"=r"(ah)); - asm volatile ("getcon " __EXPEVT ", %0":"=r"(al)); - ah = (ah) >> 32; - al = (al) & 0xffffffff; - asm volatile ("getcon " __KCR1 ", %0":"=r"(bh)); - asm volatile ("getcon " __KCR1 ", %0":"=r"(bl)); - bh = (bh) >> 32; - bl = (bl) & 0xffffffff; - asm volatile ("getcon " __INTEVT ", %0":"=r"(ch)); - asm volatile ("getcon " __INTEVT ", %0":"=r"(cl)); - ch = (ch) >> 32; - cl = (cl) & 0xffffffff; - printk("EXPE: %08Lx%08Lx KCR1: %08Lx%08Lx INTE: %08Lx%08Lx\n", - ah, al, bh, bl, ch, cl); - - asm volatile ("getcon " __PEXPEVT ", %0":"=r"(ah)); - asm volatile ("getcon " __PEXPEVT ", %0":"=r"(al)); - ah = (ah) >> 32; - al = (al) & 0xffffffff; - asm volatile ("getcon " __PSPC ", %0":"=r"(bh)); - asm volatile ("getcon " __PSPC ", %0":"=r"(bl)); - bh = (bh) >> 32; - bl = (bl) & 0xffffffff; - asm volatile ("getcon " __PSSR ", %0":"=r"(ch)); - asm volatile ("getcon " __PSSR ", %0":"=r"(cl)); - ch = (ch) >> 32; - cl = (cl) & 0xffffffff; - printk("PEXP: %08Lx%08Lx PSPC: %08Lx%08Lx PSSR: %08Lx%08Lx\n", - ah, al, bh, bl, ch, cl); - - ah = (regs->pc) >> 32; - al = (regs->pc) & 0xffffffff; - bh = (regs->regs[18]) >> 32; - bl = (regs->regs[18]) & 0xffffffff; - ch = (regs->regs[15]) >> 32; - cl = (regs->regs[15]) & 0xffffffff; - printk("PC : %08Lx%08Lx LINK: %08Lx%08Lx SP : %08Lx%08Lx\n", - ah, al, bh, bl, ch, cl); - - ah = (regs->sr) >> 32; - al = (regs->sr) & 0xffffffff; - asm volatile ("getcon " __TEA ", %0":"=r"(bh)); - asm volatile ("getcon " __TEA ", %0":"=r"(bl)); - bh = (bh) >> 32; - bl = (bl) & 0xffffffff; - asm volatile ("getcon " __KCR0 ", %0":"=r"(ch)); - asm volatile ("getcon " __KCR0 ", %0":"=r"(cl)); - ch = (ch) >> 32; - cl = (cl) & 0xffffffff; - printk("SR : %08Lx%08Lx TEA : %08Lx%08Lx KCR0: %08Lx%08Lx\n", - ah, al, bh, bl, ch, cl); - - ah = (regs->regs[0]) >> 32; - al = (regs->regs[0]) & 0xffffffff; - bh = (regs->regs[1]) >> 32; - bl = (regs->regs[1]) & 0xffffffff; - ch = (regs->regs[2]) >> 32; - cl = (regs->regs[2]) & 0xffffffff; - printk("R0 : %08Lx%08Lx R1 : %08Lx%08Lx R2 : %08Lx%08Lx\n", - ah, al, bh, bl, ch, cl); - - ah = (regs->regs[3]) >> 32; - al = (regs->regs[3]) & 0xffffffff; - bh = (regs->regs[4]) >> 32; - bl = (regs->regs[4]) & 0xffffffff; - ch = (regs->regs[5]) >> 32; - cl = (regs->regs[5]) & 0xffffffff; - printk("R3 : %08Lx%08Lx R4 : %08Lx%08Lx R5 : %08Lx%08Lx\n", - ah, al, bh, bl, ch, cl); - - ah = (regs->regs[6]) >> 32; - al = (regs->regs[6]) & 0xffffffff; - bh = (regs->regs[7]) >> 32; - bl = (regs->regs[7]) & 0xffffffff; - ch = (regs->regs[8]) >> 32; - cl = (regs->regs[8]) & 0xffffffff; - printk("R6 : %08Lx%08Lx R7 : %08Lx%08Lx R8 : %08Lx%08Lx\n", - ah, al, bh, bl, ch, cl); - - ah = (regs->regs[9]) >> 32; - al = (regs->regs[9]) & 0xffffffff; - bh = (regs->regs[10]) >> 32; - bl = (regs->regs[10]) & 0xffffffff; - ch = (regs->regs[11]) >> 32; - cl = (regs->regs[11]) & 0xffffffff; - printk("R9 : %08Lx%08Lx R10 : %08Lx%08Lx R11 : %08Lx%08Lx\n", - ah, al, bh, bl, ch, cl); - printk("....\n"); - - ah = (regs->tregs[0]) >> 32; - al = (regs->tregs[0]) & 0xffffffff; - bh = (regs->tregs[1]) >> 32; - bl = (regs->tregs[1]) & 0xffffffff; - ch = (regs->tregs[2]) >> 32; - cl = (regs->tregs[2]) & 0xffffffff; - printk("T0 : %08Lx%08Lx T1 : %08Lx%08Lx T2 : %08Lx%08Lx\n", - ah, al, bh, bl, ch, cl); - printk("....\n"); - - print_dtlb(); - print_itlb(); -} diff --git a/arch/sh/lib64/memcpy.S b/arch/sh/lib64/memcpy.S index dd300c372ce..5d682e0ee24 100644 --- a/arch/sh/lib64/memcpy.S +++ b/arch/sh/lib64/memcpy.S @@ -29,7 +29,7 @@ ! 2.: When there are two or three bytes in the last word of an 11-or-more ! bytes memory chunk to b copied, the rest of the word can be read ! without side effects. -! This could be easily changed by increasing the minumum size of +! This could be easily changed by increasing the minimum size of ! a fast memcpy and the amount subtracted from r7 before L_2l_loop be 2, ! however, this would cost a few extra cyles on average. ! For SHmedia, the assumption is that any quadword can be read in its |
