mirror of
https://github.com/AuxXxilium/linux_dsm_epyc7002.git
synced 2024-11-30 14:36:46 +07:00
e7cc9a7340
The idea is that we want to get rid of the in/out/readb/writeb callbacks from the machvec and replace that with simple inline read and write operations to memory. Fast and simple for most hardware devices (think pci). Some devices require special treatment though - like 16-bit only CF devices - so we need to have some method to hook in callbacks. This patch makes it possible to add a per-device trap generating filter. This way we can get maximum performance of sane hardware - which doesn't need this filter - and crappy hardware works but gets punished by a performance hit. V2 changes things around a bit and replaces io access callbacks with a simple minimum_bus_width value. In the future we can add stride as well. Signed-off-by: Magnus Damm <damm@igel.co.jp> Signed-off-by: Paul Mundt <lethal@linux-sh.org>
103 lines
2.7 KiB
C
103 lines
2.7 KiB
C
#ifndef __ASM_SH_SYSTEM_32_H
|
|
#define __ASM_SH_SYSTEM_32_H
|
|
|
|
#include <linux/types.h>
|
|
|
|
struct task_struct *__switch_to(struct task_struct *prev,
|
|
struct task_struct *next);
|
|
|
|
/*
|
|
* switch_to() should switch tasks to task nr n, first
|
|
*/
|
|
#define switch_to(prev, next, last) \
|
|
do { \
|
|
register u32 *__ts1 __asm__ ("r1") = (u32 *)&prev->thread.sp; \
|
|
register u32 *__ts2 __asm__ ("r2") = (u32 *)&prev->thread.pc; \
|
|
register u32 *__ts4 __asm__ ("r4") = (u32 *)prev; \
|
|
register u32 *__ts5 __asm__ ("r5") = (u32 *)next; \
|
|
register u32 *__ts6 __asm__ ("r6") = (u32 *)&next->thread.sp; \
|
|
register u32 __ts7 __asm__ ("r7") = next->thread.pc; \
|
|
struct task_struct *__last; \
|
|
\
|
|
__asm__ __volatile__ ( \
|
|
".balign 4\n\t" \
|
|
"stc.l gbr, @-r15\n\t" \
|
|
"sts.l pr, @-r15\n\t" \
|
|
"mov.l r8, @-r15\n\t" \
|
|
"mov.l r9, @-r15\n\t" \
|
|
"mov.l r10, @-r15\n\t" \
|
|
"mov.l r11, @-r15\n\t" \
|
|
"mov.l r12, @-r15\n\t" \
|
|
"mov.l r13, @-r15\n\t" \
|
|
"mov.l r14, @-r15\n\t" \
|
|
"mov.l r15, @r1\t! save SP\n\t" \
|
|
"mov.l @r6, r15\t! change to new stack\n\t" \
|
|
"mova 1f, %0\n\t" \
|
|
"mov.l %0, @r2\t! save PC\n\t" \
|
|
"mov.l 2f, %0\n\t" \
|
|
"jmp @%0\t! call __switch_to\n\t" \
|
|
" lds r7, pr\t! with return to new PC\n\t" \
|
|
".balign 4\n" \
|
|
"2:\n\t" \
|
|
".long __switch_to\n" \
|
|
"1:\n\t" \
|
|
"mov.l @r15+, r14\n\t" \
|
|
"mov.l @r15+, r13\n\t" \
|
|
"mov.l @r15+, r12\n\t" \
|
|
"mov.l @r15+, r11\n\t" \
|
|
"mov.l @r15+, r10\n\t" \
|
|
"mov.l @r15+, r9\n\t" \
|
|
"mov.l @r15+, r8\n\t" \
|
|
"lds.l @r15+, pr\n\t" \
|
|
"ldc.l @r15+, gbr\n\t" \
|
|
: "=z" (__last) \
|
|
: "r" (__ts1), "r" (__ts2), "r" (__ts4), \
|
|
"r" (__ts5), "r" (__ts6), "r" (__ts7) \
|
|
: "r3", "t"); \
|
|
\
|
|
last = __last; \
|
|
} while (0)
|
|
|
|
#define __uses_jump_to_uncached __attribute__ ((__section__ (".uncached.text")))
|
|
|
|
/*
|
|
* Jump to uncached area.
|
|
* When handling TLB or caches, we need to do it from an uncached area.
|
|
*/
|
|
#define jump_to_uncached() \
|
|
do { \
|
|
unsigned long __dummy; \
|
|
\
|
|
__asm__ __volatile__( \
|
|
"mova 1f, %0\n\t" \
|
|
"add %1, %0\n\t" \
|
|
"jmp @%0\n\t" \
|
|
" nop\n\t" \
|
|
".balign 4\n" \
|
|
"1:" \
|
|
: "=&z" (__dummy) \
|
|
: "r" (cached_to_uncached)); \
|
|
} while (0)
|
|
|
|
/*
|
|
* Back to cached area.
|
|
*/
|
|
#define back_to_cached() \
|
|
do { \
|
|
unsigned long __dummy; \
|
|
ctrl_barrier(); \
|
|
__asm__ __volatile__( \
|
|
"mov.l 1f, %0\n\t" \
|
|
"jmp @%0\n\t" \
|
|
" nop\n\t" \
|
|
".balign 4\n" \
|
|
"1: .long 2f\n" \
|
|
"2:" \
|
|
: "=&r" (__dummy)); \
|
|
} while (0)
|
|
|
|
int handle_unaligned_access(opcode_t instruction, struct pt_regs *regs,
|
|
struct mem_access *ma);
|
|
|
|
#endif /* __ASM_SH_SYSTEM_32_H */
|