Commit c7146dd0 authored by Linus Torvalds's avatar Linus Torvalds

Merge branches 'stable/p2m-identity.v4.9.1' and 'stable/e820' of...

Merge branches 'stable/p2m-identity.v4.9.1' and 'stable/e820' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen

* 'stable/p2m-identity.v4.9.1' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
  xen/m2p: Check whether the MFN has IDENTITY_FRAME bit set..
  xen/m2p: No need to catch exceptions when we know that there is no RAM
  xen/debug: WARN_ON when identity PFN has no _PAGE_IOMAP flag set.
  xen/debugfs: Add 'p2m' file for printing out the P2M layout.
  xen/setup: Set identity mapping for non-RAM E820 and E820 gaps.
  xen/mmu: WARN_ON when racing to swap middle leaf.
  xen/mmu: Set _PAGE_IOMAP if PFN is an identity PFN.
  xen/mmu: Add the notion of identity (1-1) mapping.
  xen: Mark all initial reserved pages for the balloon as INVALID_P2M_ENTRY.

* 'stable/e820' of git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen:
  xen/e820: Don't mark balloon memory as E820_UNUSABLE when running as guest and fix overflow.
  xen/setup: Inhibit resource API from using System RAM E820 gaps as PCI mem gaps.
...@@ -29,8 +29,10 @@ typedef struct xpaddr { ...@@ -29,8 +29,10 @@ typedef struct xpaddr {
/**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/ /**** MACHINE <-> PHYSICAL CONVERSION MACROS ****/
#define INVALID_P2M_ENTRY (~0UL) #define INVALID_P2M_ENTRY (~0UL)
#define FOREIGN_FRAME_BIT (1UL<<31) #define FOREIGN_FRAME_BIT (1UL<<(BITS_PER_LONG-1))
#define IDENTITY_FRAME_BIT (1UL<<(BITS_PER_LONG-2))
#define FOREIGN_FRAME(m) ((m) | FOREIGN_FRAME_BIT) #define FOREIGN_FRAME(m) ((m) | FOREIGN_FRAME_BIT)
#define IDENTITY_FRAME(m) ((m) | IDENTITY_FRAME_BIT)
/* Maximum amount of memory we can handle in a domain in pages */ /* Maximum amount of memory we can handle in a domain in pages */
#define MAX_DOMAIN_PAGES \ #define MAX_DOMAIN_PAGES \
...@@ -41,12 +43,18 @@ extern unsigned int machine_to_phys_order; ...@@ -41,12 +43,18 @@ extern unsigned int machine_to_phys_order;
extern unsigned long get_phys_to_machine(unsigned long pfn); extern unsigned long get_phys_to_machine(unsigned long pfn);
extern bool set_phys_to_machine(unsigned long pfn, unsigned long mfn); extern bool set_phys_to_machine(unsigned long pfn, unsigned long mfn);
extern bool __set_phys_to_machine(unsigned long pfn, unsigned long mfn);
extern unsigned long set_phys_range_identity(unsigned long pfn_s,
unsigned long pfn_e);
extern int m2p_add_override(unsigned long mfn, struct page *page); extern int m2p_add_override(unsigned long mfn, struct page *page);
extern int m2p_remove_override(struct page *page); extern int m2p_remove_override(struct page *page);
extern struct page *m2p_find_override(unsigned long mfn); extern struct page *m2p_find_override(unsigned long mfn);
extern unsigned long m2p_find_override_pfn(unsigned long mfn, unsigned long pfn); extern unsigned long m2p_find_override_pfn(unsigned long mfn, unsigned long pfn);
#ifdef CONFIG_XEN_DEBUG_FS
extern int p2m_dump_show(struct seq_file *m, void *v);
#endif
static inline unsigned long pfn_to_mfn(unsigned long pfn) static inline unsigned long pfn_to_mfn(unsigned long pfn)
{ {
unsigned long mfn; unsigned long mfn;
...@@ -57,7 +65,7 @@ static inline unsigned long pfn_to_mfn(unsigned long pfn) ...@@ -57,7 +65,7 @@ static inline unsigned long pfn_to_mfn(unsigned long pfn)
mfn = get_phys_to_machine(pfn); mfn = get_phys_to_machine(pfn);
if (mfn != INVALID_P2M_ENTRY) if (mfn != INVALID_P2M_ENTRY)
mfn &= ~FOREIGN_FRAME_BIT; mfn &= ~(FOREIGN_FRAME_BIT | IDENTITY_FRAME_BIT);
return mfn; return mfn;
} }
...@@ -73,25 +81,44 @@ static inline int phys_to_machine_mapping_valid(unsigned long pfn) ...@@ -73,25 +81,44 @@ static inline int phys_to_machine_mapping_valid(unsigned long pfn)
static inline unsigned long mfn_to_pfn(unsigned long mfn) static inline unsigned long mfn_to_pfn(unsigned long mfn)
{ {
unsigned long pfn; unsigned long pfn;
int ret = 0;
if (xen_feature(XENFEAT_auto_translated_physmap)) if (xen_feature(XENFEAT_auto_translated_physmap))
return mfn; return mfn;
if (unlikely((mfn >> machine_to_phys_order) != 0)) {
pfn = ~0;
goto try_override;
}
pfn = 0; pfn = 0;
/* /*
* The array access can fail (e.g., device space beyond end of RAM). * The array access can fail (e.g., device space beyond end of RAM).
* In such cases it doesn't matter what we return (we return garbage), * In such cases it doesn't matter what we return (we return garbage),
* but we must handle the fault without crashing! * but we must handle the fault without crashing!
*/ */
__get_user(pfn, &machine_to_phys_mapping[mfn]); ret = __get_user(pfn, &machine_to_phys_mapping[mfn]);
try_override:
/* /* ret might be < 0 if there are no entries in the m2p for mfn */
* If this appears to be a foreign mfn (because the pfn if (ret < 0)
* doesn't map back to the mfn), then check the local override pfn = ~0;
* table to see if there's a better pfn to use. else if (get_phys_to_machine(pfn) != mfn)
/*
* If this appears to be a foreign mfn (because the pfn
* doesn't map back to the mfn), then check the local override
* table to see if there's a better pfn to use.
*
* m2p_find_override_pfn returns ~0 if it doesn't find anything.
*/
pfn = m2p_find_override_pfn(mfn, ~0);
/*
* pfn is ~0 if there are no entries in the m2p for mfn or if the
* entry doesn't map back to the mfn and m2p_override doesn't have a
* valid entry for it.
*/ */
if (get_phys_to_machine(pfn) != mfn) if (pfn == ~0 &&
pfn = m2p_find_override_pfn(mfn, pfn); get_phys_to_machine(mfn) == IDENTITY_FRAME(mfn))
pfn = mfn;
return pfn; return pfn;
} }
......
...@@ -48,3 +48,11 @@ config XEN_DEBUG_FS ...@@ -48,3 +48,11 @@ config XEN_DEBUG_FS
help help
Enable statistics output and various tuning options in debugfs. Enable statistics output and various tuning options in debugfs.
Enabling this option may incur a significant performance overhead. Enabling this option may incur a significant performance overhead.
config XEN_DEBUG
bool "Enable Xen debug checks"
depends on XEN
default n
help
Enable various WARN_ON checks in the Xen MMU code.
Enabling this option WILL incur a significant performance overhead.
...@@ -46,6 +46,7 @@ ...@@ -46,6 +46,7 @@
#include <linux/module.h> #include <linux/module.h>
#include <linux/gfp.h> #include <linux/gfp.h>
#include <linux/memblock.h> #include <linux/memblock.h>
#include <linux/seq_file.h>
#include <asm/pgtable.h> #include <asm/pgtable.h>
#include <asm/tlbflush.h> #include <asm/tlbflush.h>
...@@ -416,8 +417,12 @@ static pteval_t pte_pfn_to_mfn(pteval_t val) ...@@ -416,8 +417,12 @@ static pteval_t pte_pfn_to_mfn(pteval_t val)
if (val & _PAGE_PRESENT) { if (val & _PAGE_PRESENT) {
unsigned long pfn = (val & PTE_PFN_MASK) >> PAGE_SHIFT; unsigned long pfn = (val & PTE_PFN_MASK) >> PAGE_SHIFT;
pteval_t flags = val & PTE_FLAGS_MASK; pteval_t flags = val & PTE_FLAGS_MASK;
unsigned long mfn = pfn_to_mfn(pfn); unsigned long mfn;
if (!xen_feature(XENFEAT_auto_translated_physmap))
mfn = get_phys_to_machine(pfn);
else
mfn = pfn;
/* /*
* If there's no mfn for the pfn, then just create an * If there's no mfn for the pfn, then just create an
* empty non-present pte. Unfortunately this loses * empty non-present pte. Unfortunately this loses
...@@ -427,8 +432,18 @@ static pteval_t pte_pfn_to_mfn(pteval_t val) ...@@ -427,8 +432,18 @@ static pteval_t pte_pfn_to_mfn(pteval_t val)
if (unlikely(mfn == INVALID_P2M_ENTRY)) { if (unlikely(mfn == INVALID_P2M_ENTRY)) {
mfn = 0; mfn = 0;
flags = 0; flags = 0;
} else {
/*
* Paramount to do this test _after_ the
* INVALID_P2M_ENTRY as INVALID_P2M_ENTRY &
* IDENTITY_FRAME_BIT resolves to true.
*/
mfn &= ~FOREIGN_FRAME_BIT;
if (mfn & IDENTITY_FRAME_BIT) {
mfn &= ~IDENTITY_FRAME_BIT;
flags |= _PAGE_IOMAP;
}
} }
val = ((pteval_t)mfn << PAGE_SHIFT) | flags; val = ((pteval_t)mfn << PAGE_SHIFT) | flags;
} }
...@@ -532,6 +547,41 @@ pte_t xen_make_pte(pteval_t pte) ...@@ -532,6 +547,41 @@ pte_t xen_make_pte(pteval_t pte)
} }
PV_CALLEE_SAVE_REGS_THUNK(xen_make_pte); PV_CALLEE_SAVE_REGS_THUNK(xen_make_pte);
#ifdef CONFIG_XEN_DEBUG
pte_t xen_make_pte_debug(pteval_t pte)
{
phys_addr_t addr = (pte & PTE_PFN_MASK);
phys_addr_t other_addr;
bool io_page = false;
pte_t _pte;
if (pte & _PAGE_IOMAP)
io_page = true;
_pte = xen_make_pte(pte);
if (!addr)
return _pte;
if (io_page &&
(xen_initial_domain() || addr >= ISA_END_ADDRESS)) {
other_addr = pfn_to_mfn(addr >> PAGE_SHIFT) << PAGE_SHIFT;
WARN(addr != other_addr,
"0x%lx is using VM_IO, but it is 0x%lx!\n",
(unsigned long)addr, (unsigned long)other_addr);
} else {
pteval_t iomap_set = (_pte.pte & PTE_FLAGS_MASK) & _PAGE_IOMAP;
other_addr = (_pte.pte & PTE_PFN_MASK);
WARN((addr == other_addr) && (!io_page) && (!iomap_set),
"0x%lx is missing VM_IO (and wasn't fixed)!\n",
(unsigned long)addr);
}
return _pte;
}
PV_CALLEE_SAVE_REGS_THUNK(xen_make_pte_debug);
#endif
pgd_t xen_make_pgd(pgdval_t pgd) pgd_t xen_make_pgd(pgdval_t pgd)
{ {
pgd = pte_pfn_to_mfn(pgd); pgd = pte_pfn_to_mfn(pgd);
...@@ -1940,6 +1990,9 @@ __init void xen_ident_map_ISA(void) ...@@ -1940,6 +1990,9 @@ __init void xen_ident_map_ISA(void)
static __init void xen_post_allocator_init(void) static __init void xen_post_allocator_init(void)
{ {
#ifdef CONFIG_XEN_DEBUG
pv_mmu_ops.make_pte = PV_CALLEE_SAVE(xen_make_pte_debug);
#endif
pv_mmu_ops.set_pte = xen_set_pte; pv_mmu_ops.set_pte = xen_set_pte;
pv_mmu_ops.set_pmd = xen_set_pmd; pv_mmu_ops.set_pmd = xen_set_pmd;
pv_mmu_ops.set_pud = xen_set_pud; pv_mmu_ops.set_pud = xen_set_pud;
...@@ -2072,7 +2125,7 @@ static void xen_zap_pfn_range(unsigned long vaddr, unsigned int order, ...@@ -2072,7 +2125,7 @@ static void xen_zap_pfn_range(unsigned long vaddr, unsigned int order,
in_frames[i] = virt_to_mfn(vaddr); in_frames[i] = virt_to_mfn(vaddr);
MULTI_update_va_mapping(mcs.mc, vaddr, VOID_PTE, 0); MULTI_update_va_mapping(mcs.mc, vaddr, VOID_PTE, 0);
set_phys_to_machine(virt_to_pfn(vaddr), INVALID_P2M_ENTRY); __set_phys_to_machine(virt_to_pfn(vaddr), INVALID_P2M_ENTRY);
if (out_frames) if (out_frames)
out_frames[i] = virt_to_pfn(vaddr); out_frames[i] = virt_to_pfn(vaddr);
...@@ -2351,6 +2404,18 @@ EXPORT_SYMBOL_GPL(xen_remap_domain_mfn_range); ...@@ -2351,6 +2404,18 @@ EXPORT_SYMBOL_GPL(xen_remap_domain_mfn_range);
#ifdef CONFIG_XEN_DEBUG_FS #ifdef CONFIG_XEN_DEBUG_FS
static int p2m_dump_open(struct inode *inode, struct file *filp)
{
return single_open(filp, p2m_dump_show, NULL);
}
static const struct file_operations p2m_dump_fops = {
.open = p2m_dump_open,
.read = seq_read,
.llseek = seq_lseek,
.release = single_release,
};
static struct dentry *d_mmu_debug; static struct dentry *d_mmu_debug;
static int __init xen_mmu_debugfs(void) static int __init xen_mmu_debugfs(void)
...@@ -2406,6 +2471,7 @@ static int __init xen_mmu_debugfs(void) ...@@ -2406,6 +2471,7 @@ static int __init xen_mmu_debugfs(void)
debugfs_create_u32("prot_commit_batched", 0444, d_mmu_debug, debugfs_create_u32("prot_commit_batched", 0444, d_mmu_debug,
&mmu_stats.prot_commit_batched); &mmu_stats.prot_commit_batched);
debugfs_create_file("p2m", 0600, d_mmu_debug, NULL, &p2m_dump_fops);
return 0; return 0;
} }
fs_initcall(xen_mmu_debugfs); fs_initcall(xen_mmu_debugfs);
......
This diff is collapsed.
...@@ -52,6 +52,8 @@ phys_addr_t xen_extra_mem_start, xen_extra_mem_size; ...@@ -52,6 +52,8 @@ phys_addr_t xen_extra_mem_start, xen_extra_mem_size;
static __init void xen_add_extra_mem(unsigned long pages) static __init void xen_add_extra_mem(unsigned long pages)
{ {
unsigned long pfn;
u64 size = (u64)pages * PAGE_SIZE; u64 size = (u64)pages * PAGE_SIZE;
u64 extra_start = xen_extra_mem_start + xen_extra_mem_size; u64 extra_start = xen_extra_mem_start + xen_extra_mem_size;
...@@ -66,6 +68,9 @@ static __init void xen_add_extra_mem(unsigned long pages) ...@@ -66,6 +68,9 @@ static __init void xen_add_extra_mem(unsigned long pages)
xen_extra_mem_size += size; xen_extra_mem_size += size;
xen_max_p2m_pfn = PFN_DOWN(extra_start + size); xen_max_p2m_pfn = PFN_DOWN(extra_start + size);
for (pfn = PFN_DOWN(extra_start); pfn <= xen_max_p2m_pfn; pfn++)
__set_phys_to_machine(pfn, INVALID_P2M_ENTRY);
} }
static unsigned long __init xen_release_chunk(phys_addr_t start_addr, static unsigned long __init xen_release_chunk(phys_addr_t start_addr,
...@@ -104,7 +109,7 @@ static unsigned long __init xen_release_chunk(phys_addr_t start_addr, ...@@ -104,7 +109,7 @@ static unsigned long __init xen_release_chunk(phys_addr_t start_addr,
WARN(ret != 1, "Failed to release memory %lx-%lx err=%d\n", WARN(ret != 1, "Failed to release memory %lx-%lx err=%d\n",
start, end, ret); start, end, ret);
if (ret == 1) { if (ret == 1) {
set_phys_to_machine(pfn, INVALID_P2M_ENTRY); __set_phys_to_machine(pfn, INVALID_P2M_ENTRY);
len++; len++;
} }
} }
...@@ -138,12 +143,55 @@ static unsigned long __init xen_return_unused_memory(unsigned long max_pfn, ...@@ -138,12 +143,55 @@ static unsigned long __init xen_return_unused_memory(unsigned long max_pfn,
return released; return released;
} }
static unsigned long __init xen_set_identity(const struct e820entry *list,
ssize_t map_size)
{
phys_addr_t last = xen_initial_domain() ? 0 : ISA_END_ADDRESS;
phys_addr_t start_pci = last;
const struct e820entry *entry;
unsigned long identity = 0;
int i;
for (i = 0, entry = list; i < map_size; i++, entry++) {
phys_addr_t start = entry->addr;
phys_addr_t end = start + entry->size;
if (start < last)
start = last;
if (end <= start)
continue;
/* Skip over the 1MB region. */
if (last > end)
continue;
if (entry->type == E820_RAM) {
if (start > start_pci)
identity += set_phys_range_identity(
PFN_UP(start_pci), PFN_DOWN(start));
/* Without saving 'last' we would gooble RAM too
* at the end of the loop. */
last = end;
start_pci = end;
continue;
}
start_pci = min(start, start_pci);
last = end;
}
if (last > start_pci)
identity += set_phys_range_identity(
PFN_UP(start_pci), PFN_DOWN(last));
return identity;
}
/** /**
* machine_specific_memory_setup - Hook for machine specific memory setup. * machine_specific_memory_setup - Hook for machine specific memory setup.
**/ **/
char * __init xen_memory_setup(void) char * __init xen_memory_setup(void)
{ {
static struct e820entry map[E820MAX] __initdata; static struct e820entry map[E820MAX] __initdata;
static struct e820entry map_raw[E820MAX] __initdata;
unsigned long max_pfn = xen_start_info->nr_pages; unsigned long max_pfn = xen_start_info->nr_pages;
unsigned long long mem_end; unsigned long long mem_end;
...@@ -151,6 +199,7 @@ char * __init xen_memory_setup(void) ...@@ -151,6 +199,7 @@ char * __init xen_memory_setup(void)
struct xen_memory_map memmap; struct xen_memory_map memmap;
unsigned long extra_pages = 0; unsigned long extra_pages = 0;
unsigned long extra_limit; unsigned long extra_limit;
unsigned long identity_pages = 0;
int i; int i;
int op; int op;
...@@ -176,6 +225,7 @@ char * __init xen_memory_setup(void) ...@@ -176,6 +225,7 @@ char * __init xen_memory_setup(void)
} }
BUG_ON(rc); BUG_ON(rc);
memcpy(map_raw, map, sizeof(map));
e820.nr_map = 0; e820.nr_map = 0;
xen_extra_mem_start = mem_end; xen_extra_mem_start = mem_end;
for (i = 0; i < memmap.nr_entries; i++) { for (i = 0; i < memmap.nr_entries; i++) {
...@@ -194,6 +244,15 @@ char * __init xen_memory_setup(void) ...@@ -194,6 +244,15 @@ char * __init xen_memory_setup(void)
end -= delta; end -= delta;
extra_pages += PFN_DOWN(delta); extra_pages += PFN_DOWN(delta);
/*
* Set RAM below 4GB that is not for us to be unusable.
* This prevents "System RAM" address space from being
* used as potential resource for I/O address (happens
* when 'allocate_resource' is called).
*/
if (delta &&
(xen_initial_domain() && end < 0x100000000ULL))
e820_add_region(end, delta, E820_UNUSABLE);
} }
if (map[i].size > 0 && end > xen_extra_mem_start) if (map[i].size > 0 && end > xen_extra_mem_start)
...@@ -251,6 +310,13 @@ char * __init xen_memory_setup(void) ...@@ -251,6 +310,13 @@ char * __init xen_memory_setup(void)
xen_add_extra_mem(extra_pages); xen_add_extra_mem(extra_pages);
/*
* Set P2M for all non-RAM pages and E820 gaps to be identity
* type PFNs. We supply it with the non-sanitized version
* of the E820.
*/
identity_pages = xen_set_identity(map_raw, memmap.nr_entries);
printk(KERN_INFO "Set %ld page(s) to 1-1 mapping.\n", identity_pages);
return "Xen"; return "Xen";
} }
......
...@@ -296,7 +296,7 @@ static int decrease_reservation(unsigned long nr_pages) ...@@ -296,7 +296,7 @@ static int decrease_reservation(unsigned long nr_pages)
/* No more mappings: invalidate P2M and add to balloon. */ /* No more mappings: invalidate P2M and add to balloon. */
for (i = 0; i < nr_pages; i++) { for (i = 0; i < nr_pages; i++) {
pfn = mfn_to_pfn(frame_list[i]); pfn = mfn_to_pfn(frame_list[i]);
set_phys_to_machine(pfn, INVALID_P2M_ENTRY); __set_phys_to_machine(pfn, INVALID_P2M_ENTRY);
balloon_append(pfn_to_page(pfn)); balloon_append(pfn_to_page(pfn));
} }
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment