qemu/softmmu/memory_mapping.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

378 lines
11 KiB
C
Raw Normal View History

/*
* QEMU memory mapping
*
* Copyright Fujitsu, Corp. 2011, 2012
*
* Authors:
* Wen Congyang <wency@cn.fujitsu.com>
*
* This work is licensed under the terms of the GNU GPL, version 2 or later.
* See the COPYING file in the top-level directory.
*
*/
#include "qemu/osdep.h"
2016-03-14 11:01:28 +03:00
#include "qapi/error.h"
#include "sysemu/memory_mapping.h"
#include "exec/memory.h"
#include "exec/address-spaces.h"
#include "hw/core/cpu.h"
//#define DEBUG_GUEST_PHYS_REGION_ADD
static void memory_mapping_list_add_mapping_sorted(MemoryMappingList *list,
MemoryMapping *mapping)
{
MemoryMapping *p;
QTAILQ_FOREACH(p, &list->head, next) {
if (p->phys_addr >= mapping->phys_addr) {
QTAILQ_INSERT_BEFORE(p, mapping, next);
return;
}
}
QTAILQ_INSERT_TAIL(&list->head, mapping, next);
}
static void create_new_memory_mapping(MemoryMappingList *list,
hwaddr phys_addr,
hwaddr virt_addr,
ram_addr_t length)
{
MemoryMapping *memory_mapping;
memory_mapping = g_new(MemoryMapping, 1);
memory_mapping->phys_addr = phys_addr;
memory_mapping->virt_addr = virt_addr;
memory_mapping->length = length;
list->last_mapping = memory_mapping;
list->num++;
memory_mapping_list_add_mapping_sorted(list, memory_mapping);
}
static inline bool mapping_contiguous(MemoryMapping *map,
hwaddr phys_addr,
hwaddr virt_addr)
{
return phys_addr == map->phys_addr + map->length &&
virt_addr == map->virt_addr + map->length;
}
/*
* [map->phys_addr, map->phys_addr + map->length) and
* [phys_addr, phys_addr + length) have intersection?
*/
static inline bool mapping_have_same_region(MemoryMapping *map,
hwaddr phys_addr,
ram_addr_t length)
{
return !(phys_addr + length < map->phys_addr ||
phys_addr >= map->phys_addr + map->length);
}
/*
* [map->phys_addr, map->phys_addr + map->length) and
* [phys_addr, phys_addr + length) have intersection. The virtual address in the
* intersection are the same?
*/
static inline bool mapping_conflict(MemoryMapping *map,
hwaddr phys_addr,
hwaddr virt_addr)
{
return virt_addr - map->virt_addr != phys_addr - map->phys_addr;
}
/*
* [map->virt_addr, map->virt_addr + map->length) and
* [virt_addr, virt_addr + length) have intersection. And the physical address
* in the intersection are the same.
*/
static inline void mapping_merge(MemoryMapping *map,
hwaddr virt_addr,
ram_addr_t length)
{
if (virt_addr < map->virt_addr) {
map->length += map->virt_addr - virt_addr;
map->virt_addr = virt_addr;
}
if ((virt_addr + length) >
(map->virt_addr + map->length)) {
map->length = virt_addr + length - map->virt_addr;
}
}
void memory_mapping_list_add_merge_sorted(MemoryMappingList *list,
hwaddr phys_addr,
hwaddr virt_addr,
ram_addr_t length)
{
MemoryMapping *memory_mapping, *last_mapping;
if (QTAILQ_EMPTY(&list->head)) {
create_new_memory_mapping(list, phys_addr, virt_addr, length);
return;
}
last_mapping = list->last_mapping;
if (last_mapping) {
if (mapping_contiguous(last_mapping, phys_addr, virt_addr)) {
last_mapping->length += length;
return;
}
}
QTAILQ_FOREACH(memory_mapping, &list->head, next) {
if (mapping_contiguous(memory_mapping, phys_addr, virt_addr)) {
memory_mapping->length += length;
list->last_mapping = memory_mapping;
return;
}
if (phys_addr + length < memory_mapping->phys_addr) {
/* create a new region before memory_mapping */
break;
}
if (mapping_have_same_region(memory_mapping, phys_addr, length)) {
if (mapping_conflict(memory_mapping, phys_addr, virt_addr)) {
continue;
}
/* merge this region into memory_mapping */
mapping_merge(memory_mapping, virt_addr, length);
list->last_mapping = memory_mapping;
return;
}
}
/* this region can not be merged into any existed memory mapping. */
create_new_memory_mapping(list, phys_addr, virt_addr, length);
}
void memory_mapping_list_free(MemoryMappingList *list)
{
MemoryMapping *p, *q;
QTAILQ_FOREACH_SAFE(p, &list->head, next, q) {
QTAILQ_REMOVE(&list->head, p, next);
g_free(p);
}
list->num = 0;
list->last_mapping = NULL;
}
void memory_mapping_list_init(MemoryMappingList *list)
{
list->num = 0;
list->last_mapping = NULL;
QTAILQ_INIT(&list->head);
}
void guest_phys_blocks_free(GuestPhysBlockList *list)
{
GuestPhysBlock *p, *q;
QTAILQ_FOREACH_SAFE(p, &list->head, next, q) {
QTAILQ_REMOVE(&list->head, p, next);
memory_region_unref(p->mr);
g_free(p);
}
list->num = 0;
}
void guest_phys_blocks_init(GuestPhysBlockList *list)
{
list->num = 0;
QTAILQ_INIT(&list->head);
}
typedef struct GuestPhysListener {
GuestPhysBlockList *list;
MemoryListener listener;
} GuestPhysListener;
static void guest_phys_block_add_section(GuestPhysListener *g,
MemoryRegionSection *section)
{
const hwaddr target_start = section->offset_within_address_space;
const hwaddr target_end = target_start + int128_get64(section->size);
uint8_t *host_addr = memory_region_get_ram_ptr(section->mr) +
section->offset_within_region;
GuestPhysBlock *predecessor = NULL;
/* find continuity in guest physical address space */
if (!QTAILQ_EMPTY(&g->list->head)) {
hwaddr predecessor_size;
predecessor = QTAILQ_LAST(&g->list->head);
predecessor_size = predecessor->target_end - predecessor->target_start;
/* the memory API guarantees monotonically increasing traversal */
g_assert(predecessor->target_end <= target_start);
/* we want continuity in both guest-physical and host-virtual memory */
if (predecessor->target_end < target_start ||
predecessor->host_addr + predecessor_size != host_addr ||
predecessor->mr != section->mr) {
predecessor = NULL;
}
}
if (predecessor == NULL) {
/* isolated mapping, allocate it and add it to the list */
GuestPhysBlock *block = g_malloc0(sizeof *block);
block->target_start = target_start;
block->target_end = target_end;
block->host_addr = host_addr;
block->mr = section->mr;
memory_region_ref(section->mr);
QTAILQ_INSERT_TAIL(&g->list->head, block, next);
++g->list->num;
} else {
/* expand predecessor until @target_end; predecessor's start doesn't
* change
*/
predecessor->target_end = target_end;
}
#ifdef DEBUG_GUEST_PHYS_REGION_ADD
fprintf(stderr, "%s: target_start=" HWADDR_FMT_plx " target_end="
HWADDR_FMT_plx ": %s (count: %u)\n", __func__, target_start,
target_end, predecessor ? "joined" : "added", g->list->num);
#endif
}
softmmu/memory_mapping: optimize for RamDiscardManager sections virtio-mem logically plugs/unplugs memory within a sparse memory region and notifies via the RamDiscardManager interface when parts become plugged (populated) or unplugged (discarded). Currently, we end up (via the two users) 1) zeroing all logically unplugged/discarded memory during TPM resets. 2) reading all logically unplugged/discarded memory when dumping, to figure out the content is zero. 1) is always bad, because we assume unplugged memory stays discarded (and is already implicitly zero). 2) isn't that bad with anonymous memory, we end up reading the zero page (slow and unnecessary, though). However, once we use some file-backed memory (future use case), even reading will populate memory. Let's cut out all parts marked as not-populated (discarded) via the RamDiscardManager. As virtio-mem is the single user, this now means that logically unplugged memory ranges will no longer be included in the dump, which results in smaller dump files and faster dumping. virtio-mem has a minimum granularity of 1 MiB (and the default is usually 2 MiB). Theoretically, we can see quite some fragmentation, in practice we won't have it completely fragmented in 1 MiB pieces. Still, we might end up with many physical ranges. Both, the ELF format and kdump seem to be ready to support many individual ranges (e.g., for ELF it seems to be UINT32_MAX, kdump has a linear bitmap). Reviewed-by: Peter Xu <peterx@redhat.com> Cc: Marc-André Lureau <marcandre.lureau@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Eduardo Habkost <ehabkost@redhat.com> Cc: Alex Williamson <alex.williamson@redhat.com> Cc: Dr. David Alan Gilbert <dgilbert@redhat.com> Cc: Igor Mammedov <imammedo@redhat.com> Cc: Claudio Fontana <cfontana@suse.de> Cc: Thomas Huth <thuth@redhat.com> Cc: "Alex Bennée" <alex.bennee@linaro.org> Cc: Peter Xu <peterx@redhat.com> Cc: Laurent Vivier <lvivier@redhat.com> Cc: Stefan Berger <stefanb@linux.ibm.com> Signed-off-by: David Hildenbrand <david@redhat.com> Message-Id: <20210727082545.17934-5-david@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2021-07-27 11:25:45 +03:00
static int guest_phys_ram_populate_cb(MemoryRegionSection *section,
void *opaque)
{
GuestPhysListener *g = opaque;
guest_phys_block_add_section(g, section);
return 0;
}
static void guest_phys_blocks_region_add(MemoryListener *listener,
MemoryRegionSection *section)
{
GuestPhysListener *g = container_of(listener, GuestPhysListener, listener);
/* we only care about RAM */
if (!memory_region_is_ram(section->mr) ||
memory_region_is_ram_device(section->mr) ||
memory_region_is_nonvolatile(section->mr)) {
return;
}
softmmu/memory_mapping: optimize for RamDiscardManager sections virtio-mem logically plugs/unplugs memory within a sparse memory region and notifies via the RamDiscardManager interface when parts become plugged (populated) or unplugged (discarded). Currently, we end up (via the two users) 1) zeroing all logically unplugged/discarded memory during TPM resets. 2) reading all logically unplugged/discarded memory when dumping, to figure out the content is zero. 1) is always bad, because we assume unplugged memory stays discarded (and is already implicitly zero). 2) isn't that bad with anonymous memory, we end up reading the zero page (slow and unnecessary, though). However, once we use some file-backed memory (future use case), even reading will populate memory. Let's cut out all parts marked as not-populated (discarded) via the RamDiscardManager. As virtio-mem is the single user, this now means that logically unplugged memory ranges will no longer be included in the dump, which results in smaller dump files and faster dumping. virtio-mem has a minimum granularity of 1 MiB (and the default is usually 2 MiB). Theoretically, we can see quite some fragmentation, in practice we won't have it completely fragmented in 1 MiB pieces. Still, we might end up with many physical ranges. Both, the ELF format and kdump seem to be ready to support many individual ranges (e.g., for ELF it seems to be UINT32_MAX, kdump has a linear bitmap). Reviewed-by: Peter Xu <peterx@redhat.com> Cc: Marc-André Lureau <marcandre.lureau@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Eduardo Habkost <ehabkost@redhat.com> Cc: Alex Williamson <alex.williamson@redhat.com> Cc: Dr. David Alan Gilbert <dgilbert@redhat.com> Cc: Igor Mammedov <imammedo@redhat.com> Cc: Claudio Fontana <cfontana@suse.de> Cc: Thomas Huth <thuth@redhat.com> Cc: "Alex Bennée" <alex.bennee@linaro.org> Cc: Peter Xu <peterx@redhat.com> Cc: Laurent Vivier <lvivier@redhat.com> Cc: Stefan Berger <stefanb@linux.ibm.com> Signed-off-by: David Hildenbrand <david@redhat.com> Message-Id: <20210727082545.17934-5-david@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2021-07-27 11:25:45 +03:00
/* for special sparse regions, only add populated parts */
if (memory_region_has_ram_discard_manager(section->mr)) {
RamDiscardManager *rdm;
rdm = memory_region_get_ram_discard_manager(section->mr);
ram_discard_manager_replay_populated(rdm, section,
guest_phys_ram_populate_cb, g);
return;
}
guest_phys_block_add_section(g, section);
}
void guest_phys_blocks_append(GuestPhysBlockList *list)
{
GuestPhysListener g = { 0 };
g.list = list;
g.listener.region_add = &guest_phys_blocks_region_add;
memory_listener_register(&g.listener, &address_space_memory);
memory_listener_unregister(&g.listener);
}
static CPUState *find_paging_enabled_cpu(CPUState *start_cpu)
{
CPUState *cpu;
CPU_FOREACH(cpu) {
if (cpu_paging_enabled(cpu)) {
return cpu;
}
}
return NULL;
}
void qemu_get_guest_memory_mapping(MemoryMappingList *list,
const GuestPhysBlockList *guest_phys_blocks,
Error **errp)
{
CPUState *cpu, *first_paging_enabled_cpu;
GuestPhysBlock *block;
ram_addr_t offset, length;
first_paging_enabled_cpu = find_paging_enabled_cpu(first_cpu);
if (first_paging_enabled_cpu) {
for (cpu = first_paging_enabled_cpu; cpu != NULL;
cpu = CPU_NEXT(cpu)) {
Error *err = NULL;
cpu_get_memory_mapping(cpu, list, &err);
if (err) {
error_propagate(errp, err);
return;
}
}
return;
}
/*
* If the guest doesn't use paging, the virtual address is equal to physical
* address.
*/
QTAILQ_FOREACH(block, &guest_phys_blocks->head, next) {
offset = block->target_start;
length = block->target_end - block->target_start;
create_new_memory_mapping(list, offset, offset, length);
}
}
void qemu_get_guest_simple_memory_mapping(MemoryMappingList *list,
const GuestPhysBlockList *guest_phys_blocks)
{
GuestPhysBlock *block;
QTAILQ_FOREACH(block, &guest_phys_blocks->head, next) {
create_new_memory_mapping(list, block->target_start, 0,
block->target_end - block->target_start);
}
}
void memory_mapping_filter(MemoryMappingList *list, int64_t begin,
int64_t length)
{
MemoryMapping *cur, *next;
QTAILQ_FOREACH_SAFE(cur, &list->head, next, next) {
if (cur->phys_addr >= begin + length ||
cur->phys_addr + cur->length <= begin) {
QTAILQ_REMOVE(&list->head, cur, next);
g_free(cur);
list->num--;
continue;
}
if (cur->phys_addr < begin) {
cur->length -= begin - cur->phys_addr;
if (cur->virt_addr) {
cur->virt_addr += begin - cur->phys_addr;
}
cur->phys_addr = begin;
}
if (cur->phys_addr + cur->length > begin + length) {
cur->length -= cur->phys_addr + cur->length - begin - length;
}
}
}