2009-07-02 10:32:06 +00:00
|
|
|
/*
|
|
|
|
* defines ioport related functions
|
|
|
|
*
|
|
|
|
* Copyright (c) 2003 Fabrice Bellard
|
|
|
|
*
|
|
|
|
* This library is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU Lesser General Public
|
|
|
|
* License as published by the Free Software Foundation; either
|
|
|
|
* version 2 of the License, or (at your option) any later version.
|
|
|
|
*
|
|
|
|
* This library is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
|
|
* Lesser General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU Lesser General Public
|
2009-07-16 20:47:01 +00:00
|
|
|
* License along with this library; if not, see <http://www.gnu.org/licenses/>.
|
2009-07-02 10:32:06 +00:00
|
|
|
*/
|
|
|
|
|
|
|
|
/**************************************************************************
|
|
|
|
* IO ports API
|
|
|
|
*/
|
|
|
|
|
|
|
|
#ifndef IOPORT_H
|
|
|
|
#define IOPORT_H
|
|
|
|
|
|
|
|
#include "qemu-common.h"
|
|
|
|
|
ioport: use uint{32, 16, 8}_t for ioport value and pio_addr_t for ioport address.
Using int for cpu_{in, out}[bwl] is inconsistent with other part
because for address or value, uintN_t is used by other qemu part.
At least, softmmu, CPU{Read, Write}MemoryFunc, pci, target_phys_addr_t
and the callers of cpu_{in, out}[bwl]().
This patch removes the inconsistency.
IO port has its own address space so define pio_addr_t as uint32_t
because PCI io space width is 32bit.
And use uint{32, 16, 8}_t for ioport value.
Changing signedness of value might cause subtle issue. However
only a suspicious caller is kvm_handle_io() which is ok. And other callers
pass unsigned value in the first place.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Cc: Stuart Brady <sdbrady@ntlworld.com>
Cc: Anthony Liguori <anthony@codemonkey.ws>
Cc: Samuel Thibault <samuel.thibault@gnu.org>
Cc: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2009-07-14 10:10:43 +00:00
|
|
|
typedef uint32_t pio_addr_t;
|
|
|
|
#define FMT_pioaddr PRIx32
|
|
|
|
|
2009-07-02 10:32:06 +00:00
|
|
|
#define MAX_IOPORTS (64 * 1024)
|
2009-07-02 10:32:07 +00:00
|
|
|
#define IOPORTS_MASK (MAX_IOPORTS - 1)
|
2009-07-02 10:32:06 +00:00
|
|
|
|
|
|
|
/* These should really be in isa.h, but are here to make pc.h happy. */
|
|
|
|
typedef void (IOPortWriteFunc)(void *opaque, uint32_t address, uint32_t data);
|
|
|
|
typedef uint32_t (IOPortReadFunc)(void *opaque, uint32_t address);
|
|
|
|
|
ioport: use uint{32, 16, 8}_t for ioport value and pio_addr_t for ioport address.
Using int for cpu_{in, out}[bwl] is inconsistent with other part
because for address or value, uintN_t is used by other qemu part.
At least, softmmu, CPU{Read, Write}MemoryFunc, pci, target_phys_addr_t
and the callers of cpu_{in, out}[bwl]().
This patch removes the inconsistency.
IO port has its own address space so define pio_addr_t as uint32_t
because PCI io space width is 32bit.
And use uint{32, 16, 8}_t for ioport value.
Changing signedness of value might cause subtle issue. However
only a suspicious caller is kvm_handle_io() which is ok. And other callers
pass unsigned value in the first place.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Cc: Stuart Brady <sdbrady@ntlworld.com>
Cc: Anthony Liguori <anthony@codemonkey.ws>
Cc: Samuel Thibault <samuel.thibault@gnu.org>
Cc: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2009-07-14 10:10:43 +00:00
|
|
|
int register_ioport_read(pio_addr_t start, int length, int size,
|
2009-07-02 10:32:06 +00:00
|
|
|
IOPortReadFunc *func, void *opaque);
|
ioport: use uint{32, 16, 8}_t for ioport value and pio_addr_t for ioport address.
Using int for cpu_{in, out}[bwl] is inconsistent with other part
because for address or value, uintN_t is used by other qemu part.
At least, softmmu, CPU{Read, Write}MemoryFunc, pci, target_phys_addr_t
and the callers of cpu_{in, out}[bwl]().
This patch removes the inconsistency.
IO port has its own address space so define pio_addr_t as uint32_t
because PCI io space width is 32bit.
And use uint{32, 16, 8}_t for ioport value.
Changing signedness of value might cause subtle issue. However
only a suspicious caller is kvm_handle_io() which is ok. And other callers
pass unsigned value in the first place.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Cc: Stuart Brady <sdbrady@ntlworld.com>
Cc: Anthony Liguori <anthony@codemonkey.ws>
Cc: Samuel Thibault <samuel.thibault@gnu.org>
Cc: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2009-07-14 10:10:43 +00:00
|
|
|
int register_ioport_write(pio_addr_t start, int length, int size,
|
2009-07-02 10:32:06 +00:00
|
|
|
IOPortWriteFunc *func, void *opaque);
|
ioport: use uint{32, 16, 8}_t for ioport value and pio_addr_t for ioport address.
Using int for cpu_{in, out}[bwl] is inconsistent with other part
because for address or value, uintN_t is used by other qemu part.
At least, softmmu, CPU{Read, Write}MemoryFunc, pci, target_phys_addr_t
and the callers of cpu_{in, out}[bwl]().
This patch removes the inconsistency.
IO port has its own address space so define pio_addr_t as uint32_t
because PCI io space width is 32bit.
And use uint{32, 16, 8}_t for ioport value.
Changing signedness of value might cause subtle issue. However
only a suspicious caller is kvm_handle_io() which is ok. And other callers
pass unsigned value in the first place.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Cc: Stuart Brady <sdbrady@ntlworld.com>
Cc: Anthony Liguori <anthony@codemonkey.ws>
Cc: Samuel Thibault <samuel.thibault@gnu.org>
Cc: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2009-07-14 10:10:43 +00:00
|
|
|
void isa_unassign_ioport(pio_addr_t start, int length);
|
2009-07-02 10:32:06 +00:00
|
|
|
|
|
|
|
|
|
|
|
/* NOTE: as these functions may be even used when there is an isa
|
|
|
|
brige on non x86 targets, we always defined them */
|
|
|
|
#if !defined(NO_CPU_IO_DEFS) && defined(NEED_CPU_H)
|
ioport: use uint{32, 16, 8}_t for ioport value and pio_addr_t for ioport address.
Using int for cpu_{in, out}[bwl] is inconsistent with other part
because for address or value, uintN_t is used by other qemu part.
At least, softmmu, CPU{Read, Write}MemoryFunc, pci, target_phys_addr_t
and the callers of cpu_{in, out}[bwl]().
This patch removes the inconsistency.
IO port has its own address space so define pio_addr_t as uint32_t
because PCI io space width is 32bit.
And use uint{32, 16, 8}_t for ioport value.
Changing signedness of value might cause subtle issue. However
only a suspicious caller is kvm_handle_io() which is ok. And other callers
pass unsigned value in the first place.
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Cc: Stuart Brady <sdbrady@ntlworld.com>
Cc: Anthony Liguori <anthony@codemonkey.ws>
Cc: Samuel Thibault <samuel.thibault@gnu.org>
Cc: Tristan Gingold <gingold@adacore.com>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2009-07-14 10:10:43 +00:00
|
|
|
void cpu_outb(CPUState *env, pio_addr_t addr, uint8_t val);
|
|
|
|
void cpu_outw(CPUState *env, pio_addr_t addr, uint16_t val);
|
|
|
|
void cpu_outl(CPUState *env, pio_addr_t addr, uint32_t val);
|
|
|
|
uint8_t cpu_inb(CPUState *env, pio_addr_t addr);
|
|
|
|
uint16_t cpu_inw(CPUState *env, pio_addr_t addr);
|
|
|
|
uint32_t cpu_inl(CPUState *env, pio_addr_t addr);
|
2009-07-02 10:32:06 +00:00
|
|
|
#endif
|
|
|
|
|
|
|
|
#endif /* IOPORT_H */
|