freebsd-skq/sys/mips/rmi/rmi_mips_exts.h
jchandra 71fb5fea95 Fix asm for XKPHYS read for o32, constriants needed for input registers
that are clobbered.  Fixes a crash when compiled without DEBUG
2010-10-05 05:49:38 +00:00

580 lines
13 KiB
C

/*-
* Copyright (c) 2003-2009 RMI Corporation
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of RMI Corporation, nor the names of its contributors,
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* RMI_BSD
* $FreeBSD$
*/
#ifndef __MIPS_EXTS_H__
#define __MIPS_EXTS_H__
#define CPU_BLOCKID_IFU 0
#define CPU_BLOCKID_ICU 1
#define CPU_BLOCKID_IEU 2
#define CPU_BLOCKID_LSU 3
#define CPU_BLOCKID_MMU 4
#define CPU_BLOCKID_PRF 5
#define LSU_CERRLOG_REGID 9
#if defined(__mips_n64) || defined(__mips_n32)
static __inline uint64_t
read_xlr_ctrl_register(int block, int reg)
{
uint64_t res;
__asm__ __volatile__(
".set push\n\t"
".set noreorder\n\t"
"move $9, %1\n\t"
".word 0x71280018\n\t" /* mfcr $8, $9 */
"move %0, $8\n\t"
".set pop\n"
: "=r" (res) : "r"((block << 8) | reg)
: "$8", "$9"
);
return (res);
}
static __inline void
write_xlr_ctrl_register(int block, int reg, uint64_t value)
{
__asm__ __volatile__(
".set push\n\t"
".set noreorder\n\t"
"move $8, %0\n"
"move $9, %1\n"
".word 0x71280019\n" /* mtcr $8, $9 */
".set pop\n"
:
: "r" (value), "r" ((block << 8) | reg)
: "$8", "$9"
);
}
#else /* !(defined(__mips_n64) || defined(__mips_n32)) */
static __inline uint64_t
read_xlr_ctrl_register(int block, int reg)
{
uint32_t high, low;
__asm__ __volatile__(
".set push\n\t"
".set noreorder\n\t"
".set mips64\n\t"
"move $9, %2\n"
".word 0x71280018\n" /* "mfcr $8, $9\n" */
"dsra32 %0, $8, 0\n\t"
"sll %1, $8, 0\n\t"
".set pop"
: "=r" (high), "=r"(low)
: "r" ((block << 8) | reg)
: "$8", "$9");
return ( (((uint64_t)high) << 32) | low);
}
static __inline void
write_xlr_ctrl_register(int block, int reg, uint64_t value)
{
uint32_t low, high;
high = value >> 32;
low = value & 0xffffffff;
__asm__ __volatile__(
".set push\n\t"
".set noreorder\n\t"
".set mips64\n\t"
"dsll32 $9, %0, 0\n\t"
"dsll32 $8, %1, 0\n\t"
"dsrl32 $8, $8, 0\n\t"
"or $8, $9, $8\n\t"
"move $9, %2\n\t"
".word 0x71280019\n\t" /* mtcr $8, $9 */
".set pop\n"
: /* No outputs */
: "r" (high), "r" (low), "r"((block << 8) | reg)
: "$8", "$9");
}
#endif /* defined(__mips_n64) || defined(__mips_n32) */
/*
* 32 bit read write for c0
*/
#define read_c0_register32(reg, sel) \
({ \
uint32_t __rv; \
__asm__ __volatile__( \
".set push\n\t" \
".set mips32\n\t" \
"mfc0 %0, $%1, %2\n\t" \
".set pop\n" \
: "=r" (__rv) : "i" (reg), "i" (sel) ); \
__rv; \
})
#define write_c0_register32(reg, sel, value) \
__asm__ __volatile__( \
".set push\n\t" \
".set mips32\n\t" \
"mtc0 %0, $%1, %2\n\t" \
".set pop\n" \
: : "r" (value), "i" (reg), "i" (sel) );
#define read_c2_register32(reg, sel) \
({ \
uint32_t __rv; \
__asm__ __volatile__( \
".set push\n\t" \
".set mips32\n\t" \
"mfc2 %0, $%1, %2\n\t" \
".set pop\n" \
: "=r" (__rv) : "i" (reg), "i" (sel) ); \
__rv; \
})
#define write_c2_register32(reg, sel, value) \
__asm__ __volatile__( \
".set push\n\t" \
".set mips32\n\t" \
"mtc2 %0, $%1, %2\n\t" \
".set pop\n" \
: : "r" (value), "i" (reg), "i" (sel) );
#if defined(__mips_n64) || defined(__mips_n32)
/*
* On 64 bit compilation, the operations are simple
*/
#define read_c0_register64(reg, sel) \
({ \
uint64_t __rv; \
__asm__ __volatile__( \
".set push\n\t" \
".set mips64\n\t" \
"dmfc0 %0, $%1, %2\n\t" \
".set pop\n" \
: "=r" (__rv) : "i" (reg), "i" (sel) ); \
__rv; \
})
#define write_c0_register64(reg, sel, value) \
__asm__ __volatile__( \
".set push\n\t" \
".set mips64\n\t" \
"dmtc0 %0, $%1, %2\n\t" \
".set pop\n" \
: : "r" (value), "i" (reg), "i" (sel) );
#define read_c2_register64(reg, sel) \
({ \
uint64_t __rv; \
__asm__ __volatile__( \
".set push\n\t" \
".set mips64\n\t" \
"dmfc2 %0, $%1, %2\n\t" \
".set pop\n" \
: "=r" (__rv) : "i" (reg), "i" (sel) ); \
__rv; \
})
#define write_c2_register64(reg, sel, value) \
__asm__ __volatile__( \
".set push\n\t" \
".set mips64\n\t" \
"dmtc2 %0, $%1, %2\n\t" \
".set pop\n" \
: : "r" (value), "i" (reg), "i" (sel) );
#else /* ! (defined(__mips_n64) || defined(__mips_n32)) */
/*
* 32 bit compilation, 64 bit values has to split
*/
#define read_c0_register64(reg, sel) \
({ \
uint32_t __high, __low; \
__asm__ __volatile__( \
".set push\n\t" \
".set noreorder\n\t" \
".set mips64\n\t" \
"dmfc0 $8, $%2, %3\n\t" \
"dsra32 %0, $8, 0\n\t" \
"sll %1, $8, 0\n\t" \
".set pop\n" \
: "=r"(__high), "=r"(__low): "i"(reg), "i"(sel) \
: "$8"); \
((uint64_t)__high << 32) | __low; \
})
#define write_c0_register64(reg, sel, value) \
do { \
uint32_t __high = value >> 32; \
uint32_t __low = value & 0xffffffff; \
__asm__ __volatile__( \
".set push\n\t" \
".set noreorder\n\t" \
".set mips64\n\t" \
"dsll32 $8, %1, 0\n\t" \
"dsll32 $9, %0, 0\n\t" \
"dsrl32 $8, $8, 0\n\t" \
"or $8, $8, $9\n\t" \
"dmtc0 $8, $%2, %3\n\t" \
".set pop" \
:: "r"(__high), "r"(__low), "i"(reg), "i"(sel) \
:"$8", "$9"); \
} while(0)
#define read_c2_register64(reg, sel) \
({ \
uint32_t __high, __low; \
__asm__ __volatile__( \
".set push\n\t" \
".set noreorder\n\t" \
".set mips64\n\t" \
"dmfc2 $8, $%2, %3\n\t" \
"dsra32 %0, $8, 0\n\t" \
"sll %1, $8, 0\n\t" \
".set pop\n" \
: "=r"(__high), "=r"(__low): "i"(reg), "i"(sel) \
: "$8"); \
((uint64_t)__high << 32) | __low; \
})
#define write_c2_register64(reg, sel, value) \
do { \
uint32_t __high = value >> 32; \
uint32_t __low = value & 0xffffffff; \
__asm__ __volatile__( \
".set push\n\t" \
".set noreorder\n\t" \
".set mips64\n\t" \
"dsll32 $8, %1, 0\n\t" \
"dsll32 $9, %0, 0\n\t" \
"dsrl32 $8, $8, 0\n\t" \
"or $8, $8, $9\n\t" \
"dmtc2 $8, $%2, %3\n\t" \
".set pop" \
:: "r"(__high), "r"(__low), "i"(reg), "i"(sel) \
:"$8", "$9"); \
} while(0)
#endif /* defined(__mips_n64) || defined(__mips_n32) */
static __inline int
xlr_cpu_id(void)
{
return (read_c0_register32(15, 1) & 0x1f);
}
static __inline int
xlr_core_id(void)
{
return (xlr_cpu_id() / 4);
}
static __inline int
xlr_thr_id(void)
{
return (read_c0_register32(15, 1) & 0x3);
}
/* Additional registers on the XLR */
#define MIPS_COP_0_OSSCRATCH 22
#define XLR_CACHELINE_SIZE 32
/* functions to write to and read from the extended
* cp0 registers.
* EIRR : Extended Interrupt Request Register
* cp0 register 9 sel 6
* bits 0...7 are same as cause register 8...15
* EIMR : Extended Interrupt Mask Register
* cp0 register 9 sel 7
* bits 0...7 are same as status register 8...15
*/
static __inline uint64_t
read_c0_eirr64(void)
{
return (read_c0_register64(9, 6));
}
static __inline void
write_c0_eirr64(uint64_t val)
{
write_c0_register64(9, 6, val);
}
static __inline uint64_t
read_c0_eimr64(void)
{
return (read_c0_register64(9, 7));
}
static __inline void
write_c0_eimr64(uint64_t val)
{
write_c0_register64(9, 7, val);
}
static __inline int
xlr_test_and_set(int *lock)
{
int oldval = 0;
__asm__ __volatile__(
".set push\n"
".set noreorder\n"
"move $9, %2\n"
"li $8, 1\n"
// "swapw $8, $9\n"
".word 0x71280014\n"
"move %1, $8\n"
".set pop\n"
: "+m"(*lock), "=r"(oldval)
: "r"((unsigned long)lock)
: "$8", "$9"
);
return (oldval == 0 ? 1 /* success */ : 0 /* failure */);
}
static __inline uint32_t
xlr_mfcr(uint32_t reg)
{
uint32_t val;
__asm__ __volatile__(
"move $8, %1\n"
".word 0x71090018\n"
"move %0, $9\n"
: "=r"(val)
: "r"(reg):"$8", "$9");
return val;
}
static __inline void
xlr_mtcr(uint32_t reg, uint32_t val)
{
__asm__ __volatile__(
"move $8, %1\n"
"move $9, %0\n"
".word 0x71090019\n"
:: "r"(val), "r"(reg)
: "$8", "$9");
}
/*
* Atomic increment a unsigned int
*/
static __inline unsigned int
xlr_ldaddwu(unsigned int value, unsigned int *addr)
{
__asm__ __volatile__(
".set push\n"
".set noreorder\n"
"move $8, %2\n"
"move $9, %3\n"
".word 0x71280011\n" /* ldaddwu $8, $9 */
"move %0, $8\n"
".set pop\n"
: "=&r"(value), "+m"(*addr)
: "0"(value), "r" ((unsigned long)addr)
: "$8", "$9");
return (value);
}
#if defined(__mips_n64)
static __inline uint32_t
xlr_paddr_lw(uint64_t paddr)
{
paddr |= 0x9800000000000000ULL;
return (*(uint32_t *)(uintptr_t)paddr);
}
static __inline uint64_t
xlr_paddr_ld(uint64_t paddr)
{
paddr |= 0x9800000000000000ULL;
return (*(uint64_t *)(uintptr_t)paddr);
}
#elif defined(__mips_n32)
static __inline uint32_t
xlr_paddr_lw(uint64_t paddr)
{
uint32_t val;
paddr |= 0x9800000000000000ULL;
__asm__ __volatile__(
".set push \n\t"
".set mips64 \n\t"
"lw %0, 0(%1) \n\t"
".set pop \n"
: "=r"(val)
: "r"(paddr));
return (val);
}
static __inline uint64_t
xlr_paddr_ld(uint64_t paddr)
{
uint64_t val;
paddr |= 0x9800000000000000ULL;
__asm__ __volatile__(
".set push \n\t"
".set mips64 \n\t"
"ld %0, 0(%1) \n\t"
".set pop \n"
: "=r"(val)
: "r"(paddr));
return (val);
}
#else /* o32 compilation */
static __inline uint32_t
xlr_paddr_lw(uint64_t paddr)
{
uint32_t addrh, addrl;
uint32_t val;
addrh = 0x98000000 | (paddr >> 32);
addrl = paddr & 0xffffffff;
__asm__ __volatile__(
".set push \n\t"
".set mips64 \n\t"
"dsll32 $8, %1, 0 \n\t"
"dsll32 $9, %2, 0 \n\t" /* get rid of the */
"dsrl32 $9, $9, 0 \n\t" /* sign extend */
"or $9, $8, $8 \n\t"
"lw %0, 0($9) \n\t"
".set pop \n"
: "=r"(val)
: "r"(addrh), "r"(addrl)
: "$8", "$9");
return (val);
}
static __inline uint64_t
xlr_paddr_ld(uint64_t paddr)
{
uint32_t addrh, addrl;
uint32_t valh, vall;
addrh = 0x98000000 | (paddr >> 32);
addrl = paddr & 0xffffffff;
__asm__ __volatile__(
".set push \n\t"
".set mips64 \n\t"
"dsll32 %0, %2, 0 \n\t"
"dsll32 %1, %3, 0 \n\t" /* get rid of the */
"dsrl32 %1, %1, 0 \n\t" /* sign extend */
"or %0, %0, %1 \n\t"
"lw %1, 4(%0) \n\t"
"lw %0, 0(%0) \n\t"
".set pop \n"
: "=&r"(valh), "=&r"(vall)
: "r"(addrh), "r"(addrl));
return (((uint64_t)valh << 32) | vall);
}
#endif
/*
* XXX: Not really needed in n32 or n64, retain for now
*/
#if defined(__mips_n64) || defined(__mips_n32)
static __inline uint32_t
xlr_enable_kx(void)
{
return (0);
}
static __inline void
xlr_restore_kx(uint32_t sr)
{
}
#else /* !defined(__mips_n64) && !defined(__mips_n32) */
/*
* o32 compilation, we will disable interrupts and enable
* the KX bit so that we can use XKPHYS to access any 40bit
* physical address
*/
static __inline uint32_t
xlr_enable_kx(void)
{
uint32_t sr = mips_rd_status();
mips_wr_status((sr & ~MIPS_SR_INT_IE) | MIPS_SR_KX);
return (sr);
}
static __inline void
xlr_restore_kx(uint32_t sr)
{
mips_wr_status(sr);
}
#endif /* defined(__mips_n64) || defined(__mips_n32) */
/*
* XLR/XLS processors have maximum 8 cores, and maximum 4 threads
* per core
*/
#define XLR_MAX_CORES 8
#define XLR_NTHREADS 4
/*
* FreeBSD can be started with few threads and cores turned off,
* so have a hardware thread id to FreeBSD cpuid mapping.
*/
extern int xlr_ncores;
extern int xlr_threads_per_core;
extern uint32_t xlr_hw_thread_mask;
extern int xlr_cpuid_to_hwtid[];
extern int xlr_hwtid_to_cpuid[];
#endif