freebsd-dev/sys/i386/include/vmparam.h

245 lines
7.4 KiB
C
Raw Normal View History

1993-06-12 14:58:17 +00:00
/*-
* SPDX-License-Identifier: BSD-3-Clause
*
1993-06-12 14:58:17 +00:00
* Copyright (c) 1990 The Regents of the University of California.
* All rights reserved.
* Copyright (c) 1994 John S. Dyson
* All rights reserved.
1993-06-12 14:58:17 +00:00
*
* This code is derived from software contributed to Berkeley by
* William Jolitz.
*
* 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 the University nor the names of its contributors
1993-06-12 14:58:17 +00:00
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS 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 REGENTS 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.
*
* from: @(#)vmparam.h 5.9 (Berkeley) 5/12/91
1999-08-28 01:08:13 +00:00
* $FreeBSD$
1993-06-12 14:58:17 +00:00
*/
#ifndef _MACHINE_VMPARAM_H_
#define _MACHINE_VMPARAM_H_ 1
1993-06-12 14:58:17 +00:00
/*
* Machine dependent constants for 386.
*/
/*
* Virtual memory related constants, all in bytes
*/
#define MAXTSIZ (128UL*1024*1024) /* max text size */
1993-06-12 14:58:17 +00:00
#ifndef DFLDSIZ
#define DFLDSIZ (128UL*1024*1024) /* initial data size limit */
1993-06-12 14:58:17 +00:00
#endif
#ifndef MAXDSIZ
#define MAXDSIZ (512UL*1024*1024) /* max data size */
1993-06-12 14:58:17 +00:00
#endif
#ifndef DFLSSIZ
#define DFLSSIZ (8UL*1024*1024) /* initial stack size limit */
1993-06-12 14:58:17 +00:00
#endif
#ifndef MAXSSIZ
#define MAXSSIZ (64UL*1024*1024) /* max stack size */
1993-06-12 14:58:17 +00:00
#endif
#ifndef SGROWSIZ
#define SGROWSIZ (128UL*1024) /* amount to grow stack */
#endif
1993-06-12 14:58:17 +00:00
/*
* Choose between DENSE and SPARSE based on whether lower execution time or
* lower kernel address space consumption is desired. Under PAE, kernel
* address space is often in short supply.
*/
#ifdef PAE
#define VM_PHYSSEG_SPARSE
#else
#define VM_PHYSSEG_DENSE
#endif
/*
* The number of PHYSSEG entries must be one greater than the number
* of phys_avail entries because the phys_avail entry that spans the
* largest physical address that is accessible by ISA DMA is split
* into two PHYSSEG entries.
*/
#define VM_PHYSSEG_MAX 17
/*
* Create one free page pool. Since the i386 kernel virtual address
* space does not include a mapping onto the machine's entire physical
* memory, VM_FREEPOOL_DIRECT is defined as an alias for the default
* pool, VM_FREEPOOL_DEFAULT.
*/
#define VM_NFREEPOOL 1
#define VM_FREEPOOL_DEFAULT 0
#define VM_FREEPOOL_DIRECT 0
/*
* Create up to three free page lists: VM_FREELIST_DMA32 is for physical pages
* that have physical addresses below 4G but are not accessible by ISA DMA,
* and VM_FREELIST_ISADMA is for physical pages that are accessible by ISA
* DMA.
*/
#define VM_NFREELIST 3
#define VM_FREELIST_DEFAULT 0
#define VM_FREELIST_DMA32 1
#define VM_FREELIST_LOWMEM 2
#define VM_LOWMEM_BOUNDARY (16 << 20) /* 16MB ISA DMA limit */
/*
* Always create DMA32 freelist if there is any memory above 4G.
* Bounce dma is extremely fragile and simultaneously intensively
* used.
*/
#define VM_DMA32_NPAGES_THRESHOLD 1
/*
* The largest allocation size is 2MB under PAE and 4MB otherwise.
*/
#define VM_NFREEORDER_PAE 10
#define VM_NFREEORDER_NOPAE 11
#define VM_NFREEORDER_MAX VM_NFREEORDER_NOPAE
#define VM_NFREEORDER i386_pmap_VM_NFREEORDER
/*
* Enable superpage reservations: 1 level.
*/
#ifndef VM_NRESERVLEVEL
#define VM_NRESERVLEVEL 1
#endif
/*
* Level 0 reservations consist of 512 pages when PAE pagetables are
* used, and 1024 pages otherwise.
*/
#ifndef VM_LEVEL_0_ORDER
#define VM_LEVEL_0_ORDER_PAE 9
#define VM_LEVEL_0_ORDER_NOPAE 10
#define VM_LEVEL_0_ORDER_MAX VM_LEVEL_0_ORDER_NOPAE
#define VM_LEVEL_0_ORDER i386_pmap_VM_LEVEL_0_ORDER
#else
#define VM_LEVEL_0_ORDER_MAX VM_LEVEL_0_ORDER
#endif
/*
* Kernel physical load address.
*/
#ifndef KERNLOAD
#define KERNLOAD (8 * 1024 * 1024)
#endif /* !defined(KERNLOAD) */
1993-06-12 14:58:17 +00:00
/*
* Virtual addresses of things. Derived from the page directory and
* page table indexes from pmap.h for precision.
* Because of the page that is both a PD and PT, it looks a little
* messy at times, but hey, we'll do anything to save a page :-)
1993-06-12 14:58:17 +00:00
*/
#define VM_MAX_KERNEL_ADDRESS (0xffffffffU - 16 * 1024 * 1024 + 1)
i386 4/4G split. The change makes the user and kernel address spaces on i386 independent, giving each almost the full 4G of usable virtual addresses except for one PDE at top used for trampoline and per-CPU trampoline stacks, and system structures that must be always mapped, namely IDT, GDT, common TSS and LDT, and process-private TSS and LDT if allocated. By using 1:1 mapping for the kernel text and data, it appeared possible to eliminate assembler part of the locore.S which bootstraps initial page table and KPTmap. The code is rewritten in C and moved into the pmap_cold(). The comment in vmparam.h explains the KVA layout. There is no PCID mechanism available in protected mode, so each kernel/user switch forth and back completely flushes the TLB, except for the trampoline PTD region. The TLB invalidations for userspace becomes trivial, because IPI handlers switch page tables. On the other hand, context switches no longer need to reload %cr3. copyout(9) was rewritten to use vm_fault_quick_hold(). An issue for new copyout(9) is compatibility with wiring user buffers around sysctl handlers. This explains two kind of locks for copyout ptes and accounting of the vslock() calls. The vm_fault_quick_hold() AKA slow path, is only tried after the 'fast path' failed, which temporary changes mapping to the userspace and copies the data to/from small per-cpu buffer in the trampoline. If a page fault occurs during the copy, it is short-circuit by exception.s to not even reach C code. The change was motivated by the need to implement the Meltdown mitigation, but instead of KPTI the full split is done. The i386 architecture already shows the sizing problems, in particular, it is impossible to link clang and lld with debugging. I expect that the issues due to the virtual address space limits would only exaggerate and the split gives more liveness to the platform. Tested by: pho Discussed with: bde Sponsored by: The FreeBSD Foundation MFC after: 1 month Differential revision: https://reviews.freebsd.org/D14633
2018-04-13 20:30:49 +00:00
#define VM_MIN_KERNEL_ADDRESS 0
i386 4/4G split. The change makes the user and kernel address spaces on i386 independent, giving each almost the full 4G of usable virtual addresses except for one PDE at top used for trampoline and per-CPU trampoline stacks, and system structures that must be always mapped, namely IDT, GDT, common TSS and LDT, and process-private TSS and LDT if allocated. By using 1:1 mapping for the kernel text and data, it appeared possible to eliminate assembler part of the locore.S which bootstraps initial page table and KPTmap. The code is rewritten in C and moved into the pmap_cold(). The comment in vmparam.h explains the KVA layout. There is no PCID mechanism available in protected mode, so each kernel/user switch forth and back completely flushes the TLB, except for the trampoline PTD region. The TLB invalidations for userspace becomes trivial, because IPI handlers switch page tables. On the other hand, context switches no longer need to reload %cr3. copyout(9) was rewritten to use vm_fault_quick_hold(). An issue for new copyout(9) is compatibility with wiring user buffers around sysctl handlers. This explains two kind of locks for copyout ptes and accounting of the vslock() calls. The vm_fault_quick_hold() AKA slow path, is only tried after the 'fast path' failed, which temporary changes mapping to the userspace and copies the data to/from small per-cpu buffer in the trampoline. If a page fault occurs during the copy, it is short-circuit by exception.s to not even reach C code. The change was motivated by the need to implement the Meltdown mitigation, but instead of KPTI the full split is done. The i386 architecture already shows the sizing problems, in particular, it is impossible to link clang and lld with debugging. I expect that the issues due to the virtual address space limits would only exaggerate and the split gives more liveness to the platform. Tested by: pho Discussed with: bde Sponsored by: The FreeBSD Foundation MFC after: 1 month Differential revision: https://reviews.freebsd.org/D14633
2018-04-13 20:30:49 +00:00
#define KERNBASE KERNLOAD
#define UPT_MAX_ADDRESS VADDR(PTDPTDI, PTDPTDI)
#define UPT_MIN_ADDRESS VADDR(PTDPTDI, 0)
#define VM_MAXUSER_ADDRESS (0xffffffff - 4 * 1024 * 1024 + 1)
#define SHAREDPAGE (VM_MAXUSER_ADDRESS - PAGE_SIZE)
#define USRSTACK SHAREDPAGE
i386 4/4G split. The change makes the user and kernel address spaces on i386 independent, giving each almost the full 4G of usable virtual addresses except for one PDE at top used for trampoline and per-CPU trampoline stacks, and system structures that must be always mapped, namely IDT, GDT, common TSS and LDT, and process-private TSS and LDT if allocated. By using 1:1 mapping for the kernel text and data, it appeared possible to eliminate assembler part of the locore.S which bootstraps initial page table and KPTmap. The code is rewritten in C and moved into the pmap_cold(). The comment in vmparam.h explains the KVA layout. There is no PCID mechanism available in protected mode, so each kernel/user switch forth and back completely flushes the TLB, except for the trampoline PTD region. The TLB invalidations for userspace becomes trivial, because IPI handlers switch page tables. On the other hand, context switches no longer need to reload %cr3. copyout(9) was rewritten to use vm_fault_quick_hold(). An issue for new copyout(9) is compatibility with wiring user buffers around sysctl handlers. This explains two kind of locks for copyout ptes and accounting of the vslock() calls. The vm_fault_quick_hold() AKA slow path, is only tried after the 'fast path' failed, which temporary changes mapping to the userspace and copies the data to/from small per-cpu buffer in the trampoline. If a page fault occurs during the copy, it is short-circuit by exception.s to not even reach C code. The change was motivated by the need to implement the Meltdown mitigation, but instead of KPTI the full split is done. The i386 architecture already shows the sizing problems, in particular, it is impossible to link clang and lld with debugging. I expect that the issues due to the virtual address space limits would only exaggerate and the split gives more liveness to the platform. Tested by: pho Discussed with: bde Sponsored by: The FreeBSD Foundation MFC after: 1 month Differential revision: https://reviews.freebsd.org/D14633
2018-04-13 20:30:49 +00:00
#define VM_MAX_ADDRESS VADDR(PTDPTDI, 0)
1993-06-12 14:58:17 +00:00
#define VM_MIN_ADDRESS ((vm_offset_t)0)
i386 4/4G split. The change makes the user and kernel address spaces on i386 independent, giving each almost the full 4G of usable virtual addresses except for one PDE at top used for trampoline and per-CPU trampoline stacks, and system structures that must be always mapped, namely IDT, GDT, common TSS and LDT, and process-private TSS and LDT if allocated. By using 1:1 mapping for the kernel text and data, it appeared possible to eliminate assembler part of the locore.S which bootstraps initial page table and KPTmap. The code is rewritten in C and moved into the pmap_cold(). The comment in vmparam.h explains the KVA layout. There is no PCID mechanism available in protected mode, so each kernel/user switch forth and back completely flushes the TLB, except for the trampoline PTD region. The TLB invalidations for userspace becomes trivial, because IPI handlers switch page tables. On the other hand, context switches no longer need to reload %cr3. copyout(9) was rewritten to use vm_fault_quick_hold(). An issue for new copyout(9) is compatibility with wiring user buffers around sysctl handlers. This explains two kind of locks for copyout ptes and accounting of the vslock() calls. The vm_fault_quick_hold() AKA slow path, is only tried after the 'fast path' failed, which temporary changes mapping to the userspace and copies the data to/from small per-cpu buffer in the trampoline. If a page fault occurs during the copy, it is short-circuit by exception.s to not even reach C code. The change was motivated by the need to implement the Meltdown mitigation, but instead of KPTI the full split is done. The i386 architecture already shows the sizing problems, in particular, it is impossible to link clang and lld with debugging. I expect that the issues due to the virtual address space limits would only exaggerate and the split gives more liveness to the platform. Tested by: pho Discussed with: bde Sponsored by: The FreeBSD Foundation MFC after: 1 month Differential revision: https://reviews.freebsd.org/D14633
2018-04-13 20:30:49 +00:00
#define PMAP_TRM_MIN_ADDRESS VM_MAXUSER_ADDRESS
#define PMAP_TRM_MAX_ADDRESS 0xffffffff
#define PMAP_MAP_LOW (4 * 1024 * 1024)
i386 4/4G split. The change makes the user and kernel address spaces on i386 independent, giving each almost the full 4G of usable virtual addresses except for one PDE at top used for trampoline and per-CPU trampoline stacks, and system structures that must be always mapped, namely IDT, GDT, common TSS and LDT, and process-private TSS and LDT if allocated. By using 1:1 mapping for the kernel text and data, it appeared possible to eliminate assembler part of the locore.S which bootstraps initial page table and KPTmap. The code is rewritten in C and moved into the pmap_cold(). The comment in vmparam.h explains the KVA layout. There is no PCID mechanism available in protected mode, so each kernel/user switch forth and back completely flushes the TLB, except for the trampoline PTD region. The TLB invalidations for userspace becomes trivial, because IPI handlers switch page tables. On the other hand, context switches no longer need to reload %cr3. copyout(9) was rewritten to use vm_fault_quick_hold(). An issue for new copyout(9) is compatibility with wiring user buffers around sysctl handlers. This explains two kind of locks for copyout ptes and accounting of the vslock() calls. The vm_fault_quick_hold() AKA slow path, is only tried after the 'fast path' failed, which temporary changes mapping to the userspace and copies the data to/from small per-cpu buffer in the trampoline. If a page fault occurs during the copy, it is short-circuit by exception.s to not even reach C code. The change was motivated by the need to implement the Meltdown mitigation, but instead of KPTI the full split is done. The i386 architecture already shows the sizing problems, in particular, it is impossible to link clang and lld with debugging. I expect that the issues due to the virtual address space limits would only exaggerate and the split gives more liveness to the platform. Tested by: pho Discussed with: bde Sponsored by: The FreeBSD Foundation MFC after: 1 month Differential revision: https://reviews.freebsd.org/D14633
2018-04-13 20:30:49 +00:00
/*
* KVA layout. The unit of the system allocation is single PDE, which
* represents NBPDR bytes, aligned to NBPDR. NBPDR is 4M for non-PAE
* page tables, and 2M for PAE, so PAE mode requires twice as many PTDs
* to create the same memory map as non-PAE.
i386 4/4G split. The change makes the user and kernel address spaces on i386 independent, giving each almost the full 4G of usable virtual addresses except for one PDE at top used for trampoline and per-CPU trampoline stacks, and system structures that must be always mapped, namely IDT, GDT, common TSS and LDT, and process-private TSS and LDT if allocated. By using 1:1 mapping for the kernel text and data, it appeared possible to eliminate assembler part of the locore.S which bootstraps initial page table and KPTmap. The code is rewritten in C and moved into the pmap_cold(). The comment in vmparam.h explains the KVA layout. There is no PCID mechanism available in protected mode, so each kernel/user switch forth and back completely flushes the TLB, except for the trampoline PTD region. The TLB invalidations for userspace becomes trivial, because IPI handlers switch page tables. On the other hand, context switches no longer need to reload %cr3. copyout(9) was rewritten to use vm_fault_quick_hold(). An issue for new copyout(9) is compatibility with wiring user buffers around sysctl handlers. This explains two kind of locks for copyout ptes and accounting of the vslock() calls. The vm_fault_quick_hold() AKA slow path, is only tried after the 'fast path' failed, which temporary changes mapping to the userspace and copies the data to/from small per-cpu buffer in the trampoline. If a page fault occurs during the copy, it is short-circuit by exception.s to not even reach C code. The change was motivated by the need to implement the Meltdown mitigation, but instead of KPTI the full split is done. The i386 architecture already shows the sizing problems, in particular, it is impossible to link clang and lld with debugging. I expect that the issues due to the virtual address space limits would only exaggerate and the split gives more liveness to the platform. Tested by: pho Discussed with: bde Sponsored by: The FreeBSD Foundation MFC after: 1 month Differential revision: https://reviews.freebsd.org/D14633
2018-04-13 20:30:49 +00:00
*
* 0x00000000 - 0x003fffff Transient identity map of low memory (0-4M),
* normally disabled to catch NULL derefs.
2018-07-14 19:35:41 +00:00
* 0x00400000 - 0x007fffff Fixed mapping of the low memory (4-8M).
i386 4/4G split. The change makes the user and kernel address spaces on i386 independent, giving each almost the full 4G of usable virtual addresses except for one PDE at top used for trampoline and per-CPU trampoline stacks, and system structures that must be always mapped, namely IDT, GDT, common TSS and LDT, and process-private TSS and LDT if allocated. By using 1:1 mapping for the kernel text and data, it appeared possible to eliminate assembler part of the locore.S which bootstraps initial page table and KPTmap. The code is rewritten in C and moved into the pmap_cold(). The comment in vmparam.h explains the KVA layout. There is no PCID mechanism available in protected mode, so each kernel/user switch forth and back completely flushes the TLB, except for the trampoline PTD region. The TLB invalidations for userspace becomes trivial, because IPI handlers switch page tables. On the other hand, context switches no longer need to reload %cr3. copyout(9) was rewritten to use vm_fault_quick_hold(). An issue for new copyout(9) is compatibility with wiring user buffers around sysctl handlers. This explains two kind of locks for copyout ptes and accounting of the vslock() calls. The vm_fault_quick_hold() AKA slow path, is only tried after the 'fast path' failed, which temporary changes mapping to the userspace and copies the data to/from small per-cpu buffer in the trampoline. If a page fault occurs during the copy, it is short-circuit by exception.s to not even reach C code. The change was motivated by the need to implement the Meltdown mitigation, but instead of KPTI the full split is done. The i386 architecture already shows the sizing problems, in particular, it is impossible to link clang and lld with debugging. I expect that the issues due to the virtual address space limits would only exaggerate and the split gives more liveness to the platform. Tested by: pho Discussed with: bde Sponsored by: The FreeBSD Foundation MFC after: 1 month Differential revision: https://reviews.freebsd.org/D14633
2018-04-13 20:30:49 +00:00
* 0x00800000 - 0xffbfffff KERNBASE (VA) == KERNLOAD (PA), kernel
* text + data and all kernel maps. Managed
* by MI VM.
* 0xffc00000 - 0xffdfffff Recursive kernel page table mapping, pointed
2018-07-14 19:35:41 +00:00
* to by PTmap. PTD[] recursively points
i386 4/4G split. The change makes the user and kernel address spaces on i386 independent, giving each almost the full 4G of usable virtual addresses except for one PDE at top used for trampoline and per-CPU trampoline stacks, and system structures that must be always mapped, namely IDT, GDT, common TSS and LDT, and process-private TSS and LDT if allocated. By using 1:1 mapping for the kernel text and data, it appeared possible to eliminate assembler part of the locore.S which bootstraps initial page table and KPTmap. The code is rewritten in C and moved into the pmap_cold(). The comment in vmparam.h explains the KVA layout. There is no PCID mechanism available in protected mode, so each kernel/user switch forth and back completely flushes the TLB, except for the trampoline PTD region. The TLB invalidations for userspace becomes trivial, because IPI handlers switch page tables. On the other hand, context switches no longer need to reload %cr3. copyout(9) was rewritten to use vm_fault_quick_hold(). An issue for new copyout(9) is compatibility with wiring user buffers around sysctl handlers. This explains two kind of locks for copyout ptes and accounting of the vslock() calls. The vm_fault_quick_hold() AKA slow path, is only tried after the 'fast path' failed, which temporary changes mapping to the userspace and copies the data to/from small per-cpu buffer in the trampoline. If a page fault occurs during the copy, it is short-circuit by exception.s to not even reach C code. The change was motivated by the need to implement the Meltdown mitigation, but instead of KPTI the full split is done. The i386 architecture already shows the sizing problems, in particular, it is impossible to link clang and lld with debugging. I expect that the issues due to the virtual address space limits would only exaggerate and the split gives more liveness to the platform. Tested by: pho Discussed with: bde Sponsored by: The FreeBSD Foundation MFC after: 1 month Differential revision: https://reviews.freebsd.org/D14633
2018-04-13 20:30:49 +00:00
* into PTmap.
* 0xffe00000 - 0xffffffff Kernel/User mode shared PDE, contains GDT,
* IDT, TSS, LDT, trampoline code and stacks.
* Managed by pmap_trm_alloc().
*/
/*
* How many physical pages per kmem arena virtual page.
*/
#ifndef VM_KMEM_SIZE_SCALE
#define VM_KMEM_SIZE_SCALE (1)
#endif
/*
* Optional floor (in bytes) on the size of the kmem arena.
*/
#ifndef VM_KMEM_SIZE_MIN
#define VM_KMEM_SIZE_MIN (12 * 1024 * 1024)
#endif
/*
* Optional ceiling (in bytes) on the size of the kmem arena: 40% of the
* kernel map rounded to the nearest multiple of the superpage size.
*/
#ifndef VM_KMEM_SIZE_MAX
#define VM_KMEM_SIZE_MAX (((((VM_MAX_KERNEL_ADDRESS - \
VM_MIN_KERNEL_ADDRESS) >> (PDRSHIFT - 2)) + 5) / 10) << PDRSHIFT)
#endif
1993-06-12 14:58:17 +00:00
/* initial pagein size of beginning of executable file */
#ifndef VM_INITIAL_PAGEIN
#define VM_INITIAL_PAGEIN 16
#endif
#define ZERO_REGION_SIZE (64 * 1024) /* 64KB */
#ifndef VM_MAX_AUTOTUNE_MAXUSERS
#define VM_MAX_AUTOTUNE_MAXUSERS 384
#endif
#define SFBUF
#define SFBUF_MAP
#define SFBUF_CPUSET
#define SFBUF_PROCESS_PAGE
#define PMAP_HAS_DMAP 0
#define PHYS_TO_DMAP(x) ({ panic("No direct map exists"); 0; })
#define DMAP_TO_PHYS(x) ({ panic("No direct map exists"); 0; })
#endif /* _MACHINE_VMPARAM_H_ */