Samuel Thibault
2008-Jul-31 13:18 UTC
[Xen-devel] [PATCH] hvm: Use main memory for video memory. [Was: xend, 8MB video memory and ballooning]
hvm: Use main memory for video memory.
When creating an HVM domain, if e.g. another domain is created before
qemu allocates video memory, the extra 8MB memory ballooning is not
available any more, because it got consumed by the other domain.
This fixes it by taking video memory from the main memory:
- make hvmloader use e820_malloc to reserve some of the main memory and
notify ioemu of its address through the Xen platform PCI card.
- add XENMAPSPACE_mfn to the xen_add_to_physmap memory op, to allow
ioemu to move the MFNs between the original position and the PCI
mapping, when LFB acceleration is disabled/enabled
- add a remove_from_physmap memory op, to allow ioemu to unmap it
completely for the case of old guests with acceleration disabled.
- add xc_domain_memory_translate_gpfn_list to libxc to allow ioemu to
get the MFNs of the video memory.
- have xend save the PCI memory space instead of ioemu: if a memory page
is there, the guest can access it like usual memory, so xend can
safely be responsible to save it. The extra benefit is that live
migration will apply the logdirty optimization there too.
- handle old saved images, populating the video memory from ioemu if
really needed.
diff -r 83b3a8a2a4f3 tools/firmware/hvmloader/hvmloader.c
--- a/tools/firmware/hvmloader/hvmloader.c Thu Jul 31 12:58:07 2008 +0100
+++ b/tools/firmware/hvmloader/hvmloader.c Thu Jul 31 13:20:02 2008 +0100
@@ -430,11 +430,12 @@ static void cmos_write_memory_size(void)
cmos_outb(0x35, (uint8_t)( alt_mem >> 8));
}
-static void init_xen_platform_io_base(void)
+static uint16_t init_xen_platform_io_base(void)
{
struct bios_info *bios_info = (struct bios_info *)ACPI_PHYSICAL_ADDRESS;
uint32_t devfn, bar_data;
uint16_t vendor_id, device_id;
+ uint16_t xen_pfiob = 0;
for ( devfn = 0; devfn < 128; devfn++ )
{
@@ -443,14 +444,17 @@ static void init_xen_platform_io_base(vo
if ( (vendor_id != 0x5853) || (device_id != 0x0001) )
continue;
bar_data = pci_readl(devfn, PCI_BASE_ADDRESS_0);
- bios_info->xen_pfiob = bar_data & PCI_BASE_ADDRESS_IO_MASK;
+ bios_info->xen_pfiob = xen_pfiob = bar_data &
PCI_BASE_ADDRESS_IO_MASK;
}
+ return xen_pfiob;
}
int main(void)
{
int vgabios_sz = 0, etherboot_sz = 0, rombios_sz, smbios_sz;
int extboot_sz = 0;
+ uint32_t vga_ram = ~0;
+ uint16_t xen_pfiob;
printf("HVM Loader\n");
@@ -495,6 +499,11 @@ int main(void)
break;
}
+ if ( virtual_vga != VGA_none ) {
+ vga_ram = e820_malloc(8 << 20);
+ printf("VGA RAM at %x\n", vga_ram);
+ }
+
etherboot_sz = scan_etherboot_nic((void*)ETHERBOOT_PHYSICAL_ADDRESS);
if ( must_load_extboot() )
@@ -535,7 +544,10 @@ int main(void)
ROMBIOS_PHYSICAL_ADDRESS,
ROMBIOS_PHYSICAL_ADDRESS + rombios_sz - 1);
- init_xen_platform_io_base();
+ xen_pfiob = init_xen_platform_io_base();
+
+ if (xen_pfiob && vga_ram != ~0)
+ outl(xen_pfiob + 4, vga_ram);
printf("Invoking ROMBIOS ...\n");
return 0;
diff -r 83b3a8a2a4f3 tools/firmware/hvmloader/util.c
--- a/tools/firmware/hvmloader/util.c Thu Jul 31 12:58:07 2008 +0100
+++ b/tools/firmware/hvmloader/util.c Thu Jul 31 13:20:02 2008 +0100
@@ -331,8 +331,8 @@ uint32_t e820_malloc(uint32_t size)
int i;
struct e820entry *ent = (struct e820entry *)HVM_E820;
- /* Align allocation request to a reasonable boundary (1kB). */
- size = (size + 1023) & ~1023;
+ /* Align allocation request to a reasonable boundary (4kB). */
+ size = (size + 4095) & ~4095;
for ( i = *HVM_E820_NR - 1; i >= 0; i-- )
{
diff -r 83b3a8a2a4f3 tools/ioemu/hw/cirrus_vga.c
--- a/tools/ioemu/hw/cirrus_vga.c Thu Jul 31 12:58:07 2008 +0100
+++ b/tools/ioemu/hw/cirrus_vga.c Thu Jul 31 13:20:02 2008 +0100
@@ -2543,34 +2543,28 @@ static CPUWriteMemoryFunc *cirrus_linear
cirrus_linear_bitblt_writel,
};
-static void *set_vram_mapping(unsigned long begin, unsigned long end)
+static void set_vram_mapping(CirrusVGAState *s, unsigned long begin, unsigned
long end)
{
- xen_pfn_t *extent_start = NULL;
- unsigned long nr_extents;
- void *vram_pointer = NULL;
- int i;
+ unsigned long i;
+ struct xen_add_to_physmap xatp;
+ int rc;
- /* align begin and end address */
- begin = begin & TARGET_PAGE_MASK;
- end = begin + VGA_RAM_SIZE;
- end = (end + TARGET_PAGE_SIZE -1 ) & TARGET_PAGE_MASK;
- nr_extents = (end - begin) >> TARGET_PAGE_BITS;
+ if (end > begin + VGA_RAM_SIZE)
+ end = begin + VGA_RAM_SIZE;
- extent_start = malloc(sizeof(xen_pfn_t) * nr_extents);
- if (extent_start == NULL) {
- fprintf(stderr, "Failed malloc on set_vram_mapping\n");
- return NULL;
- }
+ fprintf(logfile,"mapping vram to %lx - %lx\n", begin, end);
- memset(extent_start, 0, sizeof(xen_pfn_t) * nr_extents);
+ xatp.domid = domid;
+ xatp.space = XENMAPSPACE_mfn;
- for (i = 0; i < nr_extents; i++)
- extent_start[i] = (begin + i * TARGET_PAGE_SIZE) >>
TARGET_PAGE_BITS;
-
- if (set_mm_mapping(xc_handle, domid, nr_extents, 0, extent_start) < 0) {
- fprintf(logfile, "Failed set_mm_mapping\n");
- free(extent_start);
- return NULL;
+ for (i = 0; i < (end - begin) >> TARGET_PAGE_BITS; i++) {
+ xatp.idx = s->vram_mfns[i];
+ xatp.gpfn = (begin >> TARGET_PAGE_BITS) + i;
+ rc = xc_memory_op(xc_handle, XENMEM_add_to_physmap, &xatp);
+ if (rc) {
+ fprintf(stderr, "add_to_physmap MFN %"PRI_xen_pfn"
to PFN %"PRI_xen_pfn" failed: %d\n", xatp.idx, xatp.gpfn, rc);
+ return;
+ }
}
(void)xc_domain_pin_memory_cacheattr(
@@ -2578,61 +2572,42 @@ static void *set_vram_mapping(unsigned l
begin >> TARGET_PAGE_BITS,
end >> TARGET_PAGE_BITS,
XEN_DOMCTL_MEM_CACHEATTR_WB);
-
- vram_pointer = xc_map_foreign_pages(xc_handle, domid,
- PROT_READ|PROT_WRITE,
- extent_start, nr_extents);
- if (vram_pointer == NULL) {
- fprintf(logfile, "xc_map_foreign_batch vgaram returned error
%d\n",
- errno);
- free(extent_start);
- return NULL;
- }
-
- memset(vram_pointer, 0, nr_extents * TARGET_PAGE_SIZE);
-
-#ifdef CONFIG_STUBDOM
- xenfb_pv_display_start(vram_pointer);
-#endif
-
- free(extent_start);
-
- return vram_pointer;
}
-static int unset_vram_mapping(unsigned long begin, unsigned long end,
- void *mapping)
+static void unset_vram_mapping(CirrusVGAState *s, unsigned long begin, unsigned
long end)
{
- xen_pfn_t *extent_start = NULL;
- unsigned long nr_extents;
- int i;
+ if (s->stolen_vram_addr) {
+ /* We can put it there for xend to save it efficiently */
+ set_vram_mapping(s, s->stolen_vram_addr, s->stolen_vram_addr +
VGA_RAM_SIZE);
+ } else {
+ /* Old image, we have to unmap them completely */
+ struct xen_remove_from_physmap xrfp;
+ unsigned long i;
+ int rc;
- /* align begin and end address */
+ if (end > begin + VGA_RAM_SIZE)
+ end = begin + VGA_RAM_SIZE;
- end = begin + VGA_RAM_SIZE;
- begin = begin & TARGET_PAGE_MASK;
- end = (end + TARGET_PAGE_SIZE -1 ) & TARGET_PAGE_MASK;
- nr_extents = (end - begin) >> TARGET_PAGE_BITS;
+ fprintf(logfile,"unmapping vram from %lx - %lx\n", begin,
end);
- extent_start = malloc(sizeof(xen_pfn_t) * nr_extents);
+ xrfp.domid = domid;
- if (extent_start == NULL) {
- fprintf(stderr, "Failed malloc on set_mm_mapping\n");
- return -1;
+ for (i = 0; i < (end - begin) >> TARGET_PAGE_BITS; i++) {
+ xrfp.gpfn = (begin >> TARGET_PAGE_BITS) + i;
+ rc = xc_memory_op(xc_handle, XENMEM_remove_from_physmap,
&xrfp);
+ if (rc) {
+ fprintf(stderr, "remove_from_physmap PFN
%"PRI_xen_pfn" failed: %d\n", xrfp.gpfn, rc);
+ return;
+ }
+ }
}
+}
- /* Drop our own references to the vram pages */
- munmap(mapping, nr_extents * TARGET_PAGE_SIZE);
-
- /* Now drop the guest''s mappings */
- memset(extent_start, 0, sizeof(xen_pfn_t) * nr_extents);
- for (i = 0; i < nr_extents; i++)
- extent_start[i] = (begin + (i * TARGET_PAGE_SIZE)) >>
TARGET_PAGE_BITS;
- unset_mm_mapping(xc_handle, domid, nr_extents, 0, extent_start);
-
- free(extent_start);
-
- return 0;
+void cirrus_restart_acc(CirrusVGAState *s)
+{
+ set_vram_mapping(s, s->lfb_addr, s->lfb_end);
+ s->map_addr = s->lfb_addr;
+ s->map_end = s->lfb_end;
}
/* Compute the memory access functions */
@@ -2654,17 +2629,7 @@ static void cirrus_update_memory_access(
mode = s->gr[0x05] & 0x7;
if (mode < 4 || mode > 5 || ((s->gr[0x0B] & 0x4) == 0)) {
if (s->lfb_addr && s->lfb_end &&
!s->map_addr) {
- void *vram_pointer, *old_vram;
-
- vram_pointer = set_vram_mapping(s->lfb_addr,
- s->lfb_end);
- if (!vram_pointer)
- fprintf(stderr, "NULL vram_pointer\n");
- else {
- old_vram = vga_update_vram((VGAState *)s, vram_pointer,
- VGA_RAM_SIZE);
- qemu_free(old_vram);
- }
+ set_vram_mapping(s, s->lfb_addr, s->lfb_end);
s->map_addr = s->lfb_addr;
s->map_end = s->lfb_end;
}
@@ -2674,14 +2639,7 @@ static void cirrus_update_memory_access(
} else {
generic_io:
if (s->lfb_addr && s->lfb_end &&
s->map_addr) {
- void *old_vram;
-
- old_vram = vga_update_vram((VGAState *)s, NULL, VGA_RAM_SIZE);
-
- unset_vram_mapping(s->lfb_addr,
- s->lfb_end,
- old_vram);
-
+ unset_vram_mapping(s, s->map_addr, s->map_end);
s->map_addr = s->map_end = 0;
}
s->cirrus_linear_write[0] = cirrus_linear_writeb;
@@ -3040,36 +2998,6 @@ static CPUWriteMemoryFunc *cirrus_mmio_w
cirrus_mmio_writel,
};
-void cirrus_stop_acc(CirrusVGAState *s)
-{
- if (s->map_addr){
- int error;
- s->map_addr = 0;
- error = unset_vram_mapping(s->lfb_addr,
- s->lfb_end, s->vram_ptr);
- fprintf(stderr, "cirrus_stop_acc:unset_vram_mapping.\n");
- }
-}
-
-void cirrus_restart_acc(CirrusVGAState *s)
-{
- if (s->lfb_addr && s->lfb_end) {
- void *vram_pointer, *old_vram;
- fprintf(stderr, "cirrus_vga_load:re-enable vga acc.lfb_addr=0x%lx,
lfb_end=0x%lx.\n",
- s->lfb_addr, s->lfb_end);
- vram_pointer = set_vram_mapping(s->lfb_addr ,s->lfb_end);
- if (!vram_pointer){
- fprintf(stderr, "cirrus_vga_load:NULL vram_pointer\n");
- } else {
- old_vram = vga_update_vram((VGAState *)s, vram_pointer,
- VGA_RAM_SIZE);
- qemu_free(old_vram);
- s->map_addr = s->lfb_addr;
- s->map_end = s->lfb_end;
- }
- }
-}
-
/* load/save state */
static void cirrus_vga_save(QEMUFile *f, void *opaque)
@@ -3118,7 +3046,10 @@ static void cirrus_vga_save(QEMUFile *f,
qemu_put_8s(f, &vga_acc);
qemu_put_be64s(f, (uint64_t*)&s->lfb_addr);
qemu_put_be64s(f, (uint64_t*)&s->lfb_end);
- qemu_put_buffer(f, s->vram_ptr, VGA_RAM_SIZE);
+ qemu_put_be64s(f, &s->stolen_vram_addr);
+ if (!s->stolen_vram_addr && !vga_acc)
+ /* Old guest: VRAM is not mapped, we have to save it ourselves */
+ qemu_put_buffer(f, s->vram_ptr, VGA_RAM_SIZE);
}
static int cirrus_vga_load(QEMUFile *f, void *opaque, int version_id)
@@ -3127,7 +3058,7 @@ static int cirrus_vga_load(QEMUFile *f,
uint8_t vga_acc = 0;
int ret;
- if (version_id > 2)
+ if (version_id > 3)
return -EINVAL;
if (s->pci_dev && version_id >= 2) {
@@ -3173,9 +3104,20 @@ static int cirrus_vga_load(QEMUFile *f,
qemu_get_8s(f, &vga_acc);
qemu_get_be64s(f, (uint64_t*)&s->lfb_addr);
qemu_get_be64s(f, (uint64_t*)&s->lfb_end);
- qemu_get_buffer(f, s->vram_ptr, VGA_RAM_SIZE);
- if (vga_acc){
- cirrus_restart_acc(s);
+ if (version_id >= 3) {
+ qemu_get_be64s(f, &s->stolen_vram_addr);
+ if (!s->stolen_vram_addr && !vga_acc) {
+ /* Old guest, VRAM is not mapped, we have to restore it ourselves
*/
+ qemu_get_buffer(f, s->vram_ptr, VGA_RAM_SIZE);
+ xen_vga_populate_vram(s->lfb_addr);
+ } else
+ xen_vga_vram_map(vga_acc ? s->lfb_addr : s->stolen_vram_addr,
0);
+ } else {
+ /* Old image, we have to populate and restore VRAM ourselves */
+ xen_vga_populate_vram(s->lfb_addr);
+ qemu_get_buffer(f, s->vram_ptr, VGA_RAM_SIZE);
+ if (vga_acc)
+ cirrus_restart_acc(s);
}
/* force refresh */
@@ -3297,7 +3239,7 @@ static void cirrus_init_common(CirrusVGA
s->cursor_invalidate = cirrus_cursor_invalidate;
s->cursor_draw_line = cirrus_cursor_draw_line;
- register_savevm("cirrus_vga", 0, 2, cirrus_vga_save,
cirrus_vga_load, s);
+ register_savevm("cirrus_vga", 0, 3, cirrus_vga_save,
cirrus_vga_load, s);
}
/***************************************
diff -r 83b3a8a2a4f3 tools/ioemu/hw/vga.c
--- a/tools/ioemu/hw/vga.c Thu Jul 31 12:58:07 2008 +0100
+++ b/tools/ioemu/hw/vga.c Thu Jul 31 13:20:02 2008 +0100
@@ -23,6 +23,7 @@
*/
#include "vl.h"
#include "vga_int.h"
+#include <sys/mman.h>
//#define DEBUG_VGA
//#define DEBUG_VGA_MEM
@@ -1776,7 +1777,10 @@ static void vga_save(QEMUFile *f, void *
#endif
vram_size = s->vram_size;
qemu_put_be32s(f, &vram_size);
- qemu_put_buffer(f, s->vram_ptr, s->vram_size);
+ qemu_put_be64s(f, &s->stolen_vram_addr);
+ if (!s->stolen_vram_addr)
+ /* Old guest: VRAM is not mapped, we have to save it ourselves */
+ qemu_put_buffer(f, s->vram_ptr, VGA_RAM_SIZE);
}
static int vga_load(QEMUFile *f, void *opaque, int version_id)
@@ -1788,7 +1792,7 @@ static int vga_load(QEMUFile *f, void *o
int i;
#endif
- if (version_id > 3)
+ if (version_id > 4)
return -EINVAL;
if (s->pci_dev && version_id >= 2) {
@@ -1839,7 +1843,14 @@ static int vga_load(QEMUFile *f, void *o
qemu_get_be32s(f, &vram_size);
if (vram_size != s->vram_size)
return -EINVAL;
- qemu_get_buffer(f, s->vram_ptr, s->vram_size);
+ if (version_id >= 4) {
+ qemu_get_be64s(f, &s->stolen_vram_addr);
+ if (s->stolen_vram_addr)
+ xen_vga_vram_map(s->stolen_vram_addr, 0);
+ }
+ /* Old guest, VRAM is not mapped, we have to restore it ourselves */
+ if (!s->stolen_vram_addr)
+ qemu_get_buffer(f, s->vram_ptr, s->vram_size);
}
/* force refresh */
@@ -1994,6 +2005,100 @@ void vga_bios_init(VGAState *s)
/* TODO: add vbe support if enabled */
}
+
+static VGAState *xen_vga_state;
+
+/* When loading old images we have to populate the video ram ourselves */
+void xen_vga_populate_vram(uint64_t vram_addr)
+{
+ unsigned long nr_pfn;
+ struct xen_remove_from_physmap xrfp;
+ xen_pfn_t *pfn_list;
+ int i;
+ int rc;
+
+ fprintf(logfile, "populating video RAM at %lx\n", vram_addr);
+
+ nr_pfn = VGA_RAM_SIZE >> TARGET_PAGE_BITS;
+
+ pfn_list = malloc(sizeof(*pfn_list) * nr_pfn);
+
+ for (i = 0; i < nr_pfn; i++)
+ pfn_list[i] = (vram_addr >> TARGET_PAGE_BITS) + i;
+
+ if (xc_domain_memory_populate_physmap(xc_handle, domid, nr_pfn, 0, 0,
pfn_list)) {
+ fprintf(stderr, "Failed to populate video ram\n");
+ exit(1);
+ }
+ free(pfn_list);
+
+ xen_vga_vram_map(vram_addr, 0);
+
+ /* Unmap them from the guest for now. */
+ xrfp.domid = domid;
+ for (i = 0; i < nr_pfn; i++) {
+ xrfp.gpfn = (vram_addr >> TARGET_PAGE_BITS) + i;
+ rc = xc_memory_op(xc_handle, XENMEM_remove_from_physmap, &xrfp);
+ if (rc) {
+ fprintf(stderr, "remove_from_physmap PFN
%"PRI_xen_pfn" failed: %d\n", xrfp.gpfn, rc);
+ break;
+ }
+ }
+}
+
+/* Called once video memory has been allocated in the GPFN space */
+void xen_vga_vram_map(uint64_t vram_addr, int copy)
+{
+ unsigned long nr_pfn;
+ xen_pfn_t *pfn_list;
+ int i;
+ void *vram;
+
+ fprintf(logfile, "mapping video RAM from %lx\n", vram_addr);
+
+ nr_pfn = VGA_RAM_SIZE >> TARGET_PAGE_BITS;
+
+ pfn_list = malloc(sizeof(*pfn_list) * nr_pfn);
+
+ for (i = 0; i < nr_pfn; i++)
+ pfn_list[i] = (vram_addr >> TARGET_PAGE_BITS) + i;
+
+ vram = xc_map_foreign_pages(xc_handle, domid,
+ PROT_READ|PROT_WRITE,
+ pfn_list, nr_pfn);
+
+ if (!vram) {
+ fprintf(stderr, "Failed to map vram\n");
+ exit(1);
+ }
+
+ if (xc_domain_memory_translate_gpfn_list(xc_handle, domid, nr_pfn,
+ pfn_list, pfn_list)) {
+ fprintf(stderr, "Failed translation in xen_vga_vram_addr\n");
+ exit(1);
+ }
+
+ if (copy)
+ memcpy(vram, xen_vga_state->vram_ptr, VGA_RAM_SIZE);
+ qemu_free(xen_vga_state->vram_ptr);
+ xen_vga_state->vram_ptr = vram;
+ xen_vga_state->vram_mfns = pfn_list;
+#ifdef CONFIG_STUBDOM
+ xenfb_pv_display_start(vram);
+#endif
+}
+
+/* Called at boot time when the BIOS has allocated video RAM */
+void xen_vga_stolen_vram_addr(uint64_t stolen_vram_addr)
+{
+ fprintf(logfile, "stolen video RAM at %lx\n", stolen_vram_addr);
+
+ xen_vga_state->stolen_vram_addr = stolen_vram_addr;
+
+ /* And copy from the initialization value */
+ xen_vga_vram_map(stolen_vram_addr, 1);
+}
+
/* when used on xen environment, the vga_ram_base is not used */
void vga_common_init(VGAState *s, DisplayState *ds, uint8_t *vga_ram_base,
unsigned long vga_ram_offset, int vga_ram_size)
@@ -2025,13 +2130,9 @@ void vga_common_init(VGAState *s, Displa
vga_reset(s);
- /* Video RAM must be page-aligned for PVFB memory sharing */
- s->vram_ptr = s->vram_alloc = qemu_memalign(TARGET_PAGE_SIZE,
vga_ram_size);
-
-#ifdef CONFIG_STUBDOM
- if (!cirrus_vga_enabled)
- xenfb_pv_display_start(s->vram_ptr);
-#endif
+ s->vram_ptr = qemu_malloc(vga_ram_size);
+ s->vram_mfns = NULL;
+ xen_vga_state = s;
s->vram_offset = vga_ram_offset;
s->vram_size = vga_ram_size;
@@ -2051,7 +2152,7 @@ static void vga_init(VGAState *s)
{
int vga_io_memory;
- register_savevm("vga", 0, 3, vga_save, vga_load, s);
+ register_savevm("vga", 0, 4, vga_save, vga_load, s);
register_ioport_write(0x3c0, 16, 1, vga_ioport_write, s);
@@ -2163,33 +2264,6 @@ int pci_vga_init(PCIBus *bus, DisplaySta
return 0;
}
-void *vga_update_vram(VGAState *s, void *vga_ram_base, int vga_ram_size)
-{
- uint8_t *old_pointer;
-
- if (s->vram_size != vga_ram_size) {
- fprintf(stderr, "No support to change vga_ram_size\n");
- return NULL;
- }
-
- if (!vga_ram_base) {
- vga_ram_base = qemu_memalign(TARGET_PAGE_SIZE, vga_ram_size +
TARGET_PAGE_SIZE + 1);
- if (!vga_ram_base) {
- fprintf(stderr, "reallocate error\n");
- return NULL;
- }
- }
-
- /* XXX lock needed? */
- old_pointer = s->vram_alloc;
- s->vram_alloc = vga_ram_base;
- vga_ram_base = (uint8_t *)((long)(vga_ram_base + 15) & ~15L);
- memcpy(vga_ram_base, s->vram_ptr, vga_ram_size);
- s->vram_ptr = vga_ram_base;
-
- return old_pointer;
-}
-
/********************************************************/
/* vga screen dump */
diff -r 83b3a8a2a4f3 tools/ioemu/hw/vga_int.h
--- a/tools/ioemu/hw/vga_int.h Thu Jul 31 12:58:07 2008 +0100
+++ b/tools/ioemu/hw/vga_int.h Thu Jul 31 13:20:02 2008 +0100
@@ -80,9 +80,9 @@
#define VGA_MAX_HEIGHT 2048
#define VGA_STATE_COMMON \
- uint8_t *vram_alloc; \
uint8_t *vram_ptr; \
- uint8_t *vram_shadow; \
+ xen_pfn_t *vram_mfns; \
+ uint64_t stolen_vram_addr; /* Address of stolen RAM */ \
unsigned long vram_offset; \
unsigned int vram_size; \
unsigned long bios_offset; \
diff -r 83b3a8a2a4f3 tools/ioemu/hw/xen_platform.c
--- a/tools/ioemu/hw/xen_platform.c Thu Jul 31 12:58:07 2008 +0100
+++ b/tools/ioemu/hw/xen_platform.c Thu Jul 31 13:20:02 2008 +0100
@@ -34,6 +34,7 @@ typedef struct PCIXenPlatformState
{
PCIDevice pci_dev;
uint8_t platform_flags;
+ uint64_t vga_stolen_ram;
} PCIXenPlatformState;
static uint32_t xen_platform_ioport_readb(void *opaque, uint32_t addr)
@@ -69,11 +70,46 @@ static void xen_platform_ioport_writeb(v
}
+static uint32_t xen_platform_ioport_readl(void *opaque, uint32_t addr)
+{
+ PCIXenPlatformState *d = opaque;
+
+ addr &= 0xff;
+
+ switch (addr) {
+ case 4: /* VGA stolen memory address */
+ return d->vga_stolen_ram;
+ default:
+ return ~0u;
+ }
+}
+
+static void xen_platform_ioport_writel(void *opaque, uint32_t addr, uint32_t
val)
+{
+ PCIXenPlatformState *d = opaque;
+
+ addr &= 0xff;
+ val &= 0xffffffff;
+
+ switch (addr) {
+ case 4: /* VGA stolen memory address */
+ d->vga_stolen_ram = val;
+ xen_vga_stolen_vram_addr(val);
+ break;
+ default:
+ break;
+ }
+}
+
+
+
static void platform_ioport_map(PCIDevice *pci_dev, int region_num, uint32_t
addr, uint32_t size, int type)
{
PCIXenPlatformState *d = (PCIXenPlatformState *)pci_dev;
register_ioport_write(addr, size, 1, xen_platform_ioport_writeb, d);
+ register_ioport_write(addr, size, 4, xen_platform_ioport_writel, d);
register_ioport_read(addr, size, 1, xen_platform_ioport_readb, d);
+ register_ioport_read(addr, size, 4, xen_platform_ioport_readl, d);
}
static uint32_t platform_mmio_read(void *opaque, target_phys_addr_t addr)
@@ -155,6 +191,7 @@ void xen_pci_save(QEMUFile *f, void *opa
pci_device_save(&d->pci_dev, f);
qemu_put_8s(f, &d->platform_flags);
+ qemu_put_be64s(f, &d->vga_stolen_ram);
}
int xen_pci_load(QEMUFile *f, void *opaque, int version_id)
@@ -173,6 +210,7 @@ int xen_pci_load(QEMUFile *f, void *opaq
uint8_t flags;
qemu_get_8s(f, &flags);
xen_platform_ioport_writeb(d, 0, flags);
+ qemu_get_be64s(f, &d->vga_stolen_ram);
}
return 0;
diff -r 83b3a8a2a4f3 tools/ioemu/vl.c
--- a/tools/ioemu/vl.c Thu Jul 31 12:58:07 2008 +0100
+++ b/tools/ioemu/vl.c Thu Jul 31 13:20:02 2008 +0100
@@ -7022,38 +7022,6 @@ static BOOL WINAPI qemu_ctrl_handler(DWO
#define MAX_NET_CLIENTS 32
#include <xg_private.h>
-
-/* FIXME Flush the shadow page */
-int unset_mm_mapping(int xc_handle, uint32_t domid,
- unsigned long nr_pages, unsigned int address_bits,
- xen_pfn_t *extent_start)
-{
- int err = 0;
-
- err = xc_domain_memory_decrease_reservation(xc_handle, domid,
- nr_pages, 0, extent_start);
- if (err)
- fprintf(stderr, "Failed to decrease physmap\n");
-
- return err;
-}
-
-int set_mm_mapping(int xc_handle, uint32_t domid,
- unsigned long nr_pages, unsigned int address_bits,
- xen_pfn_t *extent_start)
-{
- int err = 0;
-
- err = xc_domain_memory_populate_physmap(
- xc_handle, domid, nr_pages, 0,
- XENMEMF_address_bits(address_bits), extent_start);
- if (err) {
- fprintf(stderr, "Failed to populate physmap\n");
- return -1;
- }
-
- return 0;
-}
int main(int argc, char **argv)
diff -r 83b3a8a2a4f3 tools/ioemu/vl.h
--- a/tools/ioemu/vl.h Thu Jul 31 12:58:07 2008 +0100
+++ b/tools/ioemu/vl.h Thu Jul 31 13:20:02 2008 +0100
@@ -1560,6 +1560,9 @@ void timeoffset_get(void);
/* xen_platform.c */
#ifndef QEMU_TOOL
void pci_xen_platform_init(PCIBus *bus);
+void xen_vga_stolen_vram_addr(uint64_t vram_addr);
+void xen_vga_populate_vram(uint64_t vram_addr);
+void xen_vga_vram_map(uint64_t vram_addr, int copy);
#endif
/* pci_emulation.c */
diff -r 83b3a8a2a4f3 tools/libxc/xc_domain.c
--- a/tools/libxc/xc_domain.c Thu Jul 31 12:58:07 2008 +0100
+++ b/tools/libxc/xc_domain.c Thu Jul 31 13:20:02 2008 +0100
@@ -531,6 +531,33 @@ int xc_domain_memory_populate_physmap(in
DPRINTF("Failed allocation for dom %d: %ld extents of order
%d\n",
domid, nr_extents, extent_order);
errno = EBUSY;
+ err = -1;
+ }
+
+ return err;
+}
+
+int xc_domain_memory_translate_gpfn_list(int xc_handle,
+ uint32_t domid,
+ unsigned long nr_gpfns,
+ xen_pfn_t *gpfn_list,
+ xen_pfn_t *mfn_list)
+{
+ int err;
+ struct xen_translate_gpfn_list translate_gpfn_list = {
+ .domid = domid,
+ .nr_gpfns = nr_gpfns,
+ };
+ set_xen_guest_handle(translate_gpfn_list.gpfn_list, gpfn_list);
+ set_xen_guest_handle(translate_gpfn_list.mfn_list, mfn_list);
+
+ err = xc_memory_op(xc_handle, XENMEM_translate_gpfn_list,
&translate_gpfn_list);
+
+ if ( err != 0 )
+ {
+ DPRINTF("Failed translation for dom %d (%ld PFNs)\n",
+ domid, nr_gpfns);
+ errno = -err;
err = -1;
}
diff -r 83b3a8a2a4f3 tools/libxc/xc_domain_save.c
--- a/tools/libxc/xc_domain_save.c Thu Jul 31 12:58:07 2008 +0100
+++ b/tools/libxc/xc_domain_save.c Thu Jul 31 13:20:02 2008 +0100
@@ -1112,9 +1112,7 @@ int xc_domain_save(int xc_handle, int io
continue;
/* Skip PFNs that aren''t really there */
- if ( hvm && ((n >= 0xa0 && n < 0xc0) /*
VGA hole */
- || (n >= (HVM_BELOW_4G_MMIO_START >>
PAGE_SHIFT)
- && n < (1ULL<<32) >>
PAGE_SHIFT)) /* MMIO */ )
+ if ( hvm && n == shared_info_frame )
continue;
/*
diff -r 83b3a8a2a4f3 tools/libxc/xenctrl.h
--- a/tools/libxc/xenctrl.h Thu Jul 31 12:58:07 2008 +0100
+++ b/tools/libxc/xenctrl.h Thu Jul 31 13:20:02 2008 +0100
@@ -628,6 +628,12 @@ int xc_domain_memory_populate_physmap(in
unsigned int mem_flags,
xen_pfn_t *extent_start);
+int xc_domain_memory_translate_gpfn_list(int xc_handle,
+ uint32_t domid,
+ unsigned long nr_gpfns,
+ xen_pfn_t *gpfn_list,
+ xen_pfn_t *mfn_list);
+
int xc_domain_ioport_permission(int xc_handle,
uint32_t domid,
uint32_t first_port,
diff -r 83b3a8a2a4f3 xen/arch/ia64/xen/mm.c
--- a/xen/arch/ia64/xen/mm.c Thu Jul 31 12:58:07 2008 +0100
+++ b/xen/arch/ia64/xen/mm.c Thu Jul 31 13:20:02 2008 +0100
@@ -2698,6 +2698,20 @@ void put_page_type(struct page_info *pag
}
+static int get_page_from_pagenr(unsigned long page_nr, struct domain *d)
+{
+ struct page_info *page = mfn_to_page(page_nr);
+
+ if ( unlikely(!mfn_valid(page_nr)) || unlikely(!get_page(page, d)) )
+ {
+ MEM_LOG("Could not get page ref for pfn %lx", page_nr);
+ return 0;
+ }
+
+ return 1;
+}
+
+
int get_page_type(struct page_info *page, u32 type)
{
u64 nx, x, y = page->u.inuse.type_info;
@@ -2792,6 +2806,8 @@ long
long
arch_memory_op(int op, XEN_GUEST_HANDLE(void) arg)
{
+ struct page_info *page = NULL;
+
switch (op) {
case XENMEM_add_to_physmap:
{
@@ -2836,11 +2852,21 @@ arch_memory_op(int op, XEN_GUEST_HANDLE(
spin_unlock(&d->grant_table->lock);
break;
+ case XENMAPSPACE_mfn:
+ {
+ if ( get_page_from_pagenr(xatp.idx, d) ) {
+ mfn = xatp.idx;
+ page = mfn_to_page(mfn);
+ }
+ break;
+ }
default:
break;
}
if (mfn == 0) {
+ if ( page )
+ put_page(page);
rcu_unlock_domain(d);
return -EINVAL;
}
@@ -2872,11 +2898,53 @@ arch_memory_op(int op, XEN_GUEST_HANDLE(
out:
domain_unlock(d);
-
+
+ if ( page )
+ put_page(page);
+
rcu_unlock_domain(d);
break;
}
+
+ case XENMEM_remove_from_physmap:
+ {
+ struct xen_remove_from_physmap xrfp;
+ unsigned long mfn;
+ struct domain *d;
+
+ if ( copy_from_guest(&xrfp, arg, 1) )
+ return -EFAULT;
+
+ if ( xrfp.domid == DOMID_SELF )
+ {
+ d = rcu_lock_current_domain();
+ }
+ else
+ {
+ if ( (d = rcu_lock_domain_by_id(xrfp.domid)) == NULL )
+ return -ESRCH;
+ if ( !IS_PRIV_FOR(current->domain, d) )
+ {
+ rcu_unlock_domain(d);
+ return -EPERM;
+ }
+ }
+
+ domain_lock(d);
+
+ mfn = gmfn_to_mfn(d, xrfp.gpfn);
+
+ if ( mfn_valid(mfn) )
+ guest_physmap_remove_page(d, xrfp.gpfn, mfn, 0);
+
+ domain_unlock(d);
+
+ rcu_unlock_domain(d);
+
+ break;
+ }
+
case XENMEM_machine_memory_map:
{
diff -r 83b3a8a2a4f3 xen/arch/x86/mm.c
--- a/xen/arch/x86/mm.c Thu Jul 31 12:58:07 2008 +0100
+++ b/xen/arch/x86/mm.c Thu Jul 31 13:20:02 2008 +0100
@@ -3332,6 +3332,7 @@ DEFINE_XEN_GUEST_HANDLE(e820entry_t);
long arch_memory_op(int op, XEN_GUEST_HANDLE(void) arg)
{
+ struct page_info *page = NULL;
switch ( op )
{
case XENMEM_add_to_physmap:
@@ -3382,12 +3383,22 @@ long arch_memory_op(int op, XEN_GUEST_HA
spin_unlock(&d->grant_table->lock);
break;
+ case XENMAPSPACE_mfn:
+ {
+ if ( get_page_from_pagenr(xatp.idx, d) ) {
+ mfn = xatp.idx;
+ page = mfn_to_page(mfn);
+ }
+ break;
+ }
default:
break;
}
if ( !paging_mode_translate(d) || (mfn == 0) )
{
+ if ( page )
+ put_page(page);
rcu_unlock_domain(d);
return -EINVAL;
}
@@ -3413,6 +3424,53 @@ long arch_memory_op(int op, XEN_GUEST_HA
/* Map at new location. */
guest_physmap_add_page(d, xatp.gpfn, mfn, 0);
+
+ domain_unlock(d);
+
+ if ( page )
+ put_page(page);
+
+ rcu_unlock_domain(d);
+
+ break;
+ }
+
+ case XENMEM_remove_from_physmap:
+ {
+ struct xen_remove_from_physmap xrfp;
+ unsigned long mfn;
+ struct domain *d;
+
+ if ( copy_from_guest(&xrfp, arg, 1) )
+ return -EFAULT;
+
+ if ( xrfp.domid == DOMID_SELF )
+ {
+ d = rcu_lock_current_domain();
+ }
+ else
+ {
+ if ( (d = rcu_lock_domain_by_id(xrfp.domid)) == NULL )
+ return -ESRCH;
+ if ( !IS_PRIV_FOR(current->domain, d) )
+ {
+ rcu_unlock_domain(d);
+ return -EPERM;
+ }
+ }
+
+ if ( xsm_remove_from_physmap(current->domain, d) )
+ {
+ rcu_unlock_domain(d);
+ return -EPERM;
+ }
+
+ domain_lock(d);
+
+ mfn = gmfn_to_mfn(d, xrfp.gpfn);
+
+ if ( mfn_valid(mfn) )
+ guest_physmap_remove_page(d, xrfp.gpfn, mfn, 0);
domain_unlock(d);
diff -r 83b3a8a2a4f3 xen/arch/x86/x86_64/compat/mm.c
--- a/xen/arch/x86/x86_64/compat/mm.c Thu Jul 31 12:58:07 2008 +0100
+++ b/xen/arch/x86/x86_64/compat/mm.c Thu Jul 31 13:20:02 2008 +0100
@@ -64,6 +64,20 @@ int compat_arch_memory_op(int op, XEN_GU
return -EFAULT;
XLAT_add_to_physmap(nat, &cmp);
+ rc = arch_memory_op(op, guest_handle_from_ptr(nat, void));
+
+ break;
+ }
+
+ case XENMEM_remove_from_physmap:
+ {
+ struct compat_remove_from_physmap cmp;
+ struct xen_remove_from_physmap *nat = (void
*)COMPAT_ARG_XLAT_VIRT_BASE;
+
+ if ( copy_from_guest(&cmp, arg, 1) )
+ return -EFAULT;
+
+ XLAT_remove_from_physmap(nat, &cmp);
rc = arch_memory_op(op, guest_handle_from_ptr(nat, void));
break;
diff -r 83b3a8a2a4f3 xen/include/public/memory.h
--- a/xen/include/public/memory.h Thu Jul 31 12:58:07 2008 +0100
+++ b/xen/include/public/memory.h Thu Jul 31 13:20:02 2008 +0100
@@ -204,6 +204,7 @@ struct xen_add_to_physmap {
/* Source mapping space. */
#define XENMAPSPACE_shared_info 0 /* shared info page */
#define XENMAPSPACE_grant_table 1 /* grant table page */
+#define XENMAPSPACE_mfn 2 /* usual MFN */
unsigned int space;
/* Index into source mapping space. */
@@ -214,6 +215,22 @@ struct xen_add_to_physmap {
};
typedef struct xen_add_to_physmap xen_add_to_physmap_t;
DEFINE_XEN_GUEST_HANDLE(xen_add_to_physmap_t);
+
+/*
+ * Unmaps the page appearing at a particular GPFN from the specified
guest''s
+ * pseudophysical address space.
+ * arg == addr of xen_remove_from_physmap_t.
+ */
+#define XENMEM_remove_from_physmap 15
+struct xen_remove_from_physmap {
+ /* Which domain to change the mapping for. */
+ domid_t domid;
+
+ /* GPFN of the current mapping of the page. */
+ xen_pfn_t gpfn;
+};
+typedef struct xen_remove_from_physmap xen_remove_from_physmap_t;
+DEFINE_XEN_GUEST_HANDLE(xen_remove_from_physmap_t);
/*
* Translates a list of domain-specific GPFNs into MFNs. Returns a -ve error
diff -r 83b3a8a2a4f3 xen/include/xlat.lst
--- a/xen/include/xlat.lst Thu Jul 31 12:58:07 2008 +0100
+++ b/xen/include/xlat.lst Thu Jul 31 13:20:02 2008 +0100
@@ -33,6 +33,7 @@
! kexec_image kexec.h
! kexec_range kexec.h
! add_to_physmap memory.h
+! remove_from_physmap memory.h
! foreign_memory_map memory.h
! memory_exchange memory.h
! memory_map memory.h
diff -r 83b3a8a2a4f3 xen/include/xsm/xsm.h
--- a/xen/include/xsm/xsm.h Thu Jul 31 12:58:07 2008 +0100
+++ b/xen/include/xsm/xsm.h Thu Jul 31 13:20:02 2008 +0100
@@ -136,6 +136,7 @@ struct xsm_operations {
int (*mmu_machphys_update) (struct domain *d, unsigned long mfn);
int (*update_va_mapping) (struct domain *d, l1_pgentry_t pte);
int (*add_to_physmap) (struct domain *d1, struct domain *d2);
+ int (*remove_from_physmap) (struct domain *d1, struct domain *d2);
#endif
};
@@ -532,6 +533,11 @@ static inline int xsm_add_to_physmap(str
{
return xsm_call(add_to_physmap(d1, d2));
}
+
+static inline int xsm_remove_from_physmap(struct domain *d1, struct domain *d2)
+{
+ return xsm_call(remove_from_physmap(d1, d2));
+}
#endif /* CONFIG_X86 */
#endif /* __XSM_H */
diff -r 83b3a8a2a4f3 xen/xsm/dummy.c
--- a/xen/xsm/dummy.c Thu Jul 31 12:58:07 2008 +0100
+++ b/xen/xsm/dummy.c Thu Jul 31 13:20:02 2008 +0100
@@ -385,6 +385,11 @@ static int dummy_add_to_physmap (struct
{
return 0;
}
+
+static int dummy_remove_from_physmap (struct domain *d1, struct domain *d2)
+{
+ return 0;
+}
#endif
struct xsm_operations dummy_xsm_ops;
@@ -484,5 +489,6 @@ void xsm_fixup_ops (struct xsm_operation
set_to_dummy_if_null(ops, mmu_machphys_update);
set_to_dummy_if_null(ops, update_va_mapping);
set_to_dummy_if_null(ops, add_to_physmap);
+ set_to_dummy_if_null(ops, remove_from_physmap);
#endif
}
diff -r 83b3a8a2a4f3 xen/xsm/flask/hooks.c
--- a/xen/xsm/flask/hooks.c Thu Jul 31 12:58:07 2008 +0100
+++ b/xen/xsm/flask/hooks.c Thu Jul 31 13:20:02 2008 +0100
@@ -1028,6 +1028,11 @@ static int flask_add_to_physmap(struct d
{
return domain_has_perm(d1, d2, SECCLASS_MMU, MMU__PHYSMAP);
}
+
+static int flask_remove_from_physmap(struct domain *d1, struct domain *d2)
+{
+ return domain_has_perm(d1, d2, SECCLASS_MMU, MMU__PHYSMAP);
+}
#endif
long do_flask_op(XEN_GUEST_HANDLE(xsm_op_t) u_flask_op);
@@ -1115,6 +1120,7 @@ static struct xsm_operations flask_ops
.mmu_machphys_update = flask_mmu_machphys_update,
.update_va_mapping = flask_update_va_mapping,
.add_to_physmap = flask_add_to_physmap,
+ .remove_from_physmap = flask_remove_from_physmap,
#endif
};
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel
Keir Fraser
2008-Jul-31 13:22 UTC
Re: [Xen-devel] [PATCH] hvm: Use main memory for video memory. [Was: xend, 8MB video memory and ballooning]
On 31/7/08 14:18, "Samuel Thibault" <samuel.thibault@eu.citrix.com> wrote:> hvm: Use main memory for video memory. > > When creating an HVM domain, if e.g. another domain is created before > qemu allocates video memory, the extra 8MB memory ballooning is not > available any more, because it got consumed by the other domain.Really too big for Xen 3.3 at this point, unfortunately. I''ll queue it up. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Samuel Thibault
2008-Jul-31 13:27 UTC
Re: [Xen-devel] [PATCH] hvm: Use main memory for video memory. [Was: xend, 8MB video memory and ballooning]
Keir Fraser, le Thu 31 Jul 2008 14:22:30 +0100, a écrit :> On 31/7/08 14:18, "Samuel Thibault" <samuel.thibault@eu.citrix.com> wrote: > > > hvm: Use main memory for video memory. > > > > When creating an HVM domain, if e.g. another domain is created before > > qemu allocates video memory, the extra 8MB memory ballooning is not > > available any more, because it got consumed by the other domain. > > Really too big for Xen 3.3 at this point, unfortunately.Ok, then I guess I should document that for stub domains to work one shouldn''t use the ballooner. Or a quick&dirty fix would be to add the 8MB extra ballooning for PV guests as well. Samuel _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2008-Jul-31 13:39 UTC
Re: [Xen-devel] [PATCH] hvm: Use main memory for video memory. [Was: xend, 8MB video memory and ballooning]
On 31/7/08 14:18, "Samuel Thibault" <samuel.thibault@eu.citrix.com> wrote:> diff -r 83b3a8a2a4f3 tools/libxc/xc_domain_save.c > --- a/tools/libxc/xc_domain_save.c Thu Jul 31 12:58:07 2008 +0100 > +++ b/tools/libxc/xc_domain_save.c Thu Jul 31 13:20:02 2008 +0100 > @@ -1112,9 +1112,7 @@ int xc_domain_save(int xc_handle, int io > continue; > > /* Skip PFNs that aren''t really there */ > - if ( hvm && ((n >= 0xa0 && n < 0xc0) /* VGA hole */ > - || (n >= (HVM_BELOW_4G_MMIO_START >> PAGE_SHIFT) > - && n < (1ULL<<32) >> PAGE_SHIFT)) /* MMIO */ > ) > + if ( hvm && n == shared_info_frame ) > continue;What happens if you don''t special-case shared_info_frame? I''d''ve hoped it would just work... does the extra page of allocation during restore fail? I''m thinking that the grant tables will be mapped in the I/O hole too, and that is from ''Xen heap'' just like shared_info is. So special-casing one and not the other sure is odd. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2008-Jul-31 13:41 UTC
Re: [Xen-devel] [PATCH] hvm: Use main memory for video memory. [Was: xend, 8MB video memory and ballooning]
On 31/7/08 14:27, "Samuel Thibault" <samuel.thibault@eu.citrix.com> wrote:>> Really too big for Xen 3.3 at this point, unfortunately. > > Ok, then I guess I should document that for stub domains to work one > shouldn''t use the ballooner. Or a quick&dirty fix would be to add the > 8MB extra ballooning for PV guests as well.I think documenting the mutual incompatibility is fine for 3.3. But it''s up to you. -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Samuel Thibault
2008-Jul-31 14:57 UTC
Re: [Xen-devel] [PATCH] hvm: Use main memory for video memory. [Was: xend, 8MB video memory and ballooning]
Keir Fraser, le Thu 31 Jul 2008 14:39:59 +0100, a écrit :> On 31/7/08 14:18, "Samuel Thibault" <samuel.thibault@eu.citrix.com> wrote: > > > diff -r 83b3a8a2a4f3 tools/libxc/xc_domain_save.c > > --- a/tools/libxc/xc_domain_save.c Thu Jul 31 12:58:07 2008 +0100 > > +++ b/tools/libxc/xc_domain_save.c Thu Jul 31 13:20:02 2008 +0100 > > @@ -1112,9 +1112,7 @@ int xc_domain_save(int xc_handle, int io > > continue; > > > > /* Skip PFNs that aren''t really there */ > > - if ( hvm && ((n >= 0xa0 && n < 0xc0) /* VGA hole */ > > - || (n >= (HVM_BELOW_4G_MMIO_START >> PAGE_SHIFT) > > - && n < (1ULL<<32) >> PAGE_SHIFT)) /* MMIO */ > > ) > > + if ( hvm && n == shared_info_frame ) > > continue; > > What happens if you don''t special-case shared_info_frame?I actually hadn''t even tried.> I''d''ve hoped it would just work... does the extra page of allocation > during restore fail?It does work, but I guess it replaces the mapping of the "magic" page with the mapping of the allocated page, so that shared_info wouldn''t work any more.> I''m thinking that the grant tables will be mapped in the I/O hole too, and > that is from ''Xen heap'' just like shared_info is.Urgl, hadn''t thought about those, I guess the PV drivers would re-etablish them on resume anyway?> So special-casing one and not the other sure is odd.Actually, shouldn''t it be up to the hypervisor to not return all these from the SHADOW_OP_PEEK operation? Samuel _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2008-Jul-31 15:03 UTC
Re: [Xen-devel] [PATCH] hvm: Use main memory for video memory. [Was: xend, 8MB video memory and ballooning]
On 31/7/08 15:57, "Samuel Thibault" <samuel.thibault@eu.citrix.com> wrote:>> I''d''ve hoped it would just work... does the extra page of allocation >> during restore fail? > > It does work, but I guess it replaces the mapping of the "magic" page > with the mapping of the allocated page, so that shared_info wouldn''t > work any more.PV drivers will add_to_physmap() these pages again during resume, which will map the correct pages (shared_info and grant_table, notably) on top of the RAM pages that save/restore will leave behind. things should work fine then and the displaced RAM pages *should* get freed immediately. Even if they don''t they will on domain destruction and it''s only a few pages...>> I''m thinking that the grant tables will be mapped in the I/O hole too, and >> that is from ''Xen heap'' just like shared_info is. > > Urgl, hadn''t thought about those, I guess the PV drivers would > re-etablish them on resume anyway?Yep, as above.>> So special-casing one and not the other sure is odd. > > Actually, shouldn''t it be up to the hypervisor to not return all these > from the SHADOW_OP_PEEK operation?Well, maybe. :-) But shadow code doesn''t special-case these Xen-heap pages I think. And doing so could be a pain, and maybe cause other issues? -- Keir _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Samuel Thibault
2008-Jul-31 15:10 UTC
Re: [Xen-devel] [PATCH] hvm: Use main memory for video memory. [Was: xend, 8MB video memory and ballooning]
Keir Fraser, le Thu 31 Jul 2008 16:03:12 +0100, a écrit :> On 31/7/08 15:57, "Samuel Thibault" <samuel.thibault@eu.citrix.com> wrote: > > >> I''d''ve hoped it would just work... does the extra page of allocation > >> during restore fail? > > > > It does work, but I guess it replaces the mapping of the "magic" page > > with the mapping of the allocated page, so that shared_info wouldn''t > > work any more. > > PV drivers will add_to_physmap() these pages again during resume,Ok, so it should be safe indeed. Samuel _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Samuel Thibault
2008-Jul-31 16:29 UTC
Re: [Xen-devel] [PATCH] hvm: Use main memory for video memory. [Was: xend, 8MB video memory and ballooning]
Keir Fraser, le Thu 31 Jul 2008 14:41:01 +0100, a écrit :> On 31/7/08 14:27, "Samuel Thibault" <samuel.thibault@eu.citrix.com> wrote: > >> Really too big for Xen 3.3 at this point, unfortunately. > > > > Ok, then I guess I should document that for stub domains to work one > > shouldn''t use the ballooner. Or a quick&dirty fix would be to add the > > 8MB extra ballooning for PV guests as well. > > I think documenting the mutual incompatibility is fine for 3.3. But it''s up > to you.The sad thing is that although it would fix HVM+ioemu-dom boot, this quick&dirty fix of course fixes the underlying race only for one HVM domain lagging behind. If there are several, 8MB won''t be enough, etc. I guess it''s simpler to just document. Samuel stubdom: update documentation stubdom/ is now compiled and installed by default HVM+IOEMU-stubdom can not boot if dom0 has to be ballooned. diff -r e75cd1659ea1 stubdom/README --- a/stubdom/README Thu Jul 31 16:48:11 2008 +0100 +++ b/stubdom/README Thu Jul 31 17:24:07 2008 +0100 @@ -1,13 +1,3 @@ To compile -To compile -=========- -Just run make -j 4, that will download / patch / compile -Then make install to install the result. - -Also, run make and make install in $XEN_ROOT/tools/fs-back - - - IOEMU stubdom ============ @@ -15,6 +5,11 @@ Also, run make and make install in $XEN_ General Configuration ====================+ +Due to a race between the creation of the IOEMU stubdomain itself and allocation +of video memory for the HVM domain, you need to avoid the need for ballooning, +by using the hypervisor dom0_mem= option for instance. + In your HVM config "hvmconfig", _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel