[PATCH 4.14 48/62] x86/mm: Move LDT remap out of KASLR region on 5-level paging

[PATCH 4.14 48/62] x86/mm: Move LDT remap out of KASLR region on 5-level paging

[Greg Kroah-Hartman]

4.14-stable review patch.  If anyone has any objections, please let me know.

------------------

commit d52888aa2753e3063a9d3a0c9f72f94aa9809c15 upstream

On 5-level paging the LDT remap area is placed in the middle of the KASLR
randomization region and it can overlap with the direct mapping, the
vmalloc or the vmap area.

The LDT mapping is per mm, so it cannot be moved into the P4D page table
next to the CPU_ENTRY_AREA without complicating PGD table allocation for
5-level paging.

The 4 PGD slot gap just before the direct mapping is reserved for
hypervisors, so it cannot be used.

Move the direct mapping one slot deeper and use the resulting gap for the
LDT remap area. The resulting layout is the same for 4 and 5 level paging.

[ tglx: Massaged changelog ]

Fixes: f55f0501cbf6 ("x86/pti: Put the LDT in its own PGD if PTI is on")
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Andy Lutomirski <luto@kernel.org>
Cc: bp@alien8.de
Cc: hpa@zytor.com
Cc: dave.hansen@linux.intel.com
Cc: peterz@infradead.org
Cc: boris.ostrovsky@oracle.com
Cc: jgross@suse.com
Cc: bhe@redhat.com
Cc: willy@infradead.org
Cc: linux-mm@kvack.org
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20181026122856.66224-2-kirill.shutemov@linux.intel.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 Documentation/x86/x86_64/mm.txt         | 10 ++++++----
 arch/x86/include/asm/page_64_types.h    | 12 +++++++-----
 arch/x86/include/asm/pgtable_64_types.h |  7 +++----
 arch/x86/xen/mmu_pv.c                   |  6 +++---
 4 files changed, 19 insertions(+), 16 deletions(-)

diff --git a/Documentation/x86/x86_64/mm.txt b/Documentation/x86/x86_64/mm.txt
index ea91cb61a602..43f066cde67d 100644
--- a/Documentation/x86/x86_64/mm.txt
+++ b/Documentation/x86/x86_64/mm.txt
@@ -4,8 +4,9 @@ Virtual memory map with 4 level page tables:
 0000000000000000 - 00007fffffffffff (=47 bits) user space, different per mm
 hole caused by [47:63] sign extension
 ffff800000000000 - ffff87ffffffffff (=43 bits) guard hole, reserved for hypervisor
-ffff880000000000 - ffffc7ffffffffff (=64 TB) direct mapping of all phys. memory
-ffffc80000000000 - ffffc8ffffffffff (=40 bits) hole
+ffff880000000000 - ffff887fffffffff (=39 bits) LDT remap for PTI
+ffff888000000000 - ffffc87fffffffff (=64 TB) direct mapping of all phys. memory
+ffffc88000000000 - ffffc8ffffffffff (=39 bits) hole
 ffffc90000000000 - ffffe8ffffffffff (=45 bits) vmalloc/ioremap space
 ffffe90000000000 - ffffe9ffffffffff (=40 bits) hole
 ffffea0000000000 - ffffeaffffffffff (=40 bits) virtual memory map (1TB)
@@ -30,8 +31,9 @@ Virtual memory map with 5 level page tables:
 0000000000000000 - 00ffffffffffffff (=56 bits) user space, different per mm
 hole caused by [56:63] sign extension
 ff00000000000000 - ff0fffffffffffff (=52 bits) guard hole, reserved for hypervisor
-ff10000000000000 - ff8fffffffffffff (=55 bits) direct mapping of all phys. memory
-ff90000000000000 - ff9fffffffffffff (=52 bits) LDT remap for PTI
+ff10000000000000 - ff10ffffffffffff (=48 bits) LDT remap for PTI
+ff11000000000000 - ff90ffffffffffff (=55 bits) direct mapping of all phys. memory
+ff91000000000000 - ff9fffffffffffff (=3840 TB) hole
 ffa0000000000000 - ffd1ffffffffffff (=54 bits) vmalloc/ioremap space (12800 TB)
 ffd2000000000000 - ffd3ffffffffffff (=49 bits) hole
 ffd4000000000000 - ffd5ffffffffffff (=49 bits) virtual memory map (512TB)
diff --git a/arch/x86/include/asm/page_64_types.h b/arch/x86/include/asm/page_64_types.h
index e1407312c412..74d531f6d518 100644
--- a/arch/x86/include/asm/page_64_types.h
+++ b/arch/x86/include/asm/page_64_types.h
@@ -33,14 +33,16 @@
 
 /*
  * Set __PAGE_OFFSET to the most negative possible address +
- * PGDIR_SIZE*16 (pgd slot 272).  The gap is to allow a space for a
- * hypervisor to fit.  Choosing 16 slots here is arbitrary, but it's
- * what Xen requires.
+ * PGDIR_SIZE*17 (pgd slot 273).
+ *
+ * The gap is to allow a space for LDT remap for PTI (1 pgd slot) and space for
+ * a hypervisor (16 slots). Choosing 16 slots for a hypervisor is arbitrary,
+ * but it's what Xen requires.
  */
 #ifdef CONFIG_X86_5LEVEL
-#define __PAGE_OFFSET_BASE      _AC(0xff10000000000000, UL)
+#define __PAGE_OFFSET_BASE	_AC(0xff11000000000000, UL)
 #else
-#define __PAGE_OFFSET_BASE      _AC(0xffff880000000000, UL)
+#define __PAGE_OFFSET_BASE	_AC(0xffff888000000000, UL)
 #endif
 
 #ifdef CONFIG_RANDOMIZE_MEMORY
diff --git a/arch/x86/include/asm/pgtable_64_types.h b/arch/x86/include/asm/pgtable_64_types.h
index 6b8f73dcbc2c..7764617b8f9c 100644
--- a/arch/x86/include/asm/pgtable_64_types.h
+++ b/arch/x86/include/asm/pgtable_64_types.h
@@ -88,16 +88,15 @@ typedef struct { pteval_t pte; } pte_t;
 # define VMALLOC_SIZE_TB	_AC(12800, UL)
 # define __VMALLOC_BASE		_AC(0xffa0000000000000, UL)
 # define __VMEMMAP_BASE		_AC(0xffd4000000000000, UL)
-# define LDT_PGD_ENTRY		_AC(-112, UL)
-# define LDT_BASE_ADDR		(LDT_PGD_ENTRY << PGDIR_SHIFT)
 #else
 # define VMALLOC_SIZE_TB	_AC(32, UL)
 # define __VMALLOC_BASE		_AC(0xffffc90000000000, UL)
 # define __VMEMMAP_BASE		_AC(0xffffea0000000000, UL)
-# define LDT_PGD_ENTRY		_AC(-3, UL)
-# define LDT_BASE_ADDR		(LDT_PGD_ENTRY << PGDIR_SHIFT)
 #endif
 
+#define LDT_PGD_ENTRY		-240UL
+#define LDT_BASE_ADDR		(LDT_PGD_ENTRY << PGDIR_SHIFT)
+
 #ifdef CONFIG_RANDOMIZE_MEMORY
 # define VMALLOC_START		vmalloc_base
 # define VMEMMAP_START		vmemmap_base
diff --git a/arch/x86/xen/mmu_pv.c b/arch/x86/xen/mmu_pv.c
index 8ed11a5b1a9d..b33fa127a613 100644
--- a/arch/x86/xen/mmu_pv.c
+++ b/arch/x86/xen/mmu_pv.c
@@ -1869,7 +1869,7 @@ void __init xen_setup_kernel_pagetable(pgd_t *pgd, unsigned long max_pfn)
 	init_top_pgt[0] = __pgd(0);
 
 	/* Pre-constructed entries are in pfn, so convert to mfn */
-	/* L4[272] -> level3_ident_pgt  */
+	/* L4[273] -> level3_ident_pgt  */
 	/* L4[511] -> level3_kernel_pgt */
 	convert_pfn_mfn(init_top_pgt);
 
@@ -1889,8 +1889,8 @@ void __init xen_setup_kernel_pagetable(pgd_t *pgd, unsigned long max_pfn)
 	addr[0] = (unsigned long)pgd;
 	addr[1] = (unsigned long)l3;
 	addr[2] = (unsigned long)l2;
-	/* Graft it onto L4[272][0]. Note that we creating an aliasing problem:
-	 * Both L4[272][0] and L4[511][510] have entries that point to the same
+	/* Graft it onto L4[273][0]. Note that we creating an aliasing problem:
+	 * Both L4[273][0] and L4[511][510] have entries that point to the same
 	 * L2 (PMD) tables. Meaning that if you modify it in __va space
 	 * it will be also modified in the __ka space! (But if you just
 	 * modify the PMD table to point to other PTE's or none, then you
-- 
2.17.1

[PATCH 4.14 49/62] x86/ldt: Unmap PTEs for the slot before freeing LDT pages

[Greg Kroah-Hartman]

4.14-stable review patch.  If anyone has any objections, please let me know.

------------------

commit a0e6e0831c516860fc7f9be1db6c081fe902ebcf upstream

modify_ldt(2) leaves the old LDT mapped after switching over to the new
one. The old LDT gets freed and the pages can be re-used.

Leaving the mapping in place can have security implications. The mapping is
present in the userspace page tables and Meltdown-like attacks can read
these freed and possibly reused pages.

It's relatively simple to fix: unmap the old LDT and flush TLB before
freeing the old LDT memory.

This further allows to avoid flushing the TLB in map_ldt_struct() as the
slot is unmapped and flushed by unmap_ldt_struct() or has never been mapped
at all.

[ tglx: Massaged changelog and removed the needless line breaks ]

Fixes: f55f0501cbf6 ("x86/pti: Put the LDT in its own PGD if PTI is on")
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: bp@alien8.de
Cc: hpa@zytor.com
Cc: dave.hansen@linux.intel.com
Cc: luto@kernel.org
Cc: peterz@infradead.org
Cc: boris.ostrovsky@oracle.com
Cc: jgross@suse.com
Cc: bhe@redhat.com
Cc: willy@infradead.org
Cc: linux-mm@kvack.org
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20181026122856.66224-3-kirill.shutemov@linux.intel.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 arch/x86/kernel/ldt.c | 49 +++++++++++++++++++++++++++++++------------
 1 file changed, 36 insertions(+), 13 deletions(-)

diff --git a/arch/x86/kernel/ldt.c b/arch/x86/kernel/ldt.c
index 26d713ecad34..65df298d4e9e 100644
--- a/arch/x86/kernel/ldt.c
+++ b/arch/x86/kernel/ldt.c
@@ -103,14 +103,6 @@ static struct ldt_struct *alloc_ldt_struct(unsigned int num_entries)
 /*
  * If PTI is enabled, this maps the LDT into the kernelmode and
  * usermode tables for the given mm.
- *
- * There is no corresponding unmap function.  Even if the LDT is freed, we
- * leave the PTEs around until the slot is reused or the mm is destroyed.
- * This is harmless: the LDT is always in ordinary memory, and no one will
- * access the freed slot.
- *
- * If we wanted to unmap freed LDTs, we'd also need to do a flush to make
- * it useful, and the flush would slow down modify_ldt().
  */
 static int
 map_ldt_struct(struct mm_struct *mm, struct ldt_struct *ldt, int slot)
@@ -119,8 +111,8 @@ map_ldt_struct(struct mm_struct *mm, struct ldt_struct *ldt, int slot)
 	bool is_vmalloc, had_top_level_entry;
 	unsigned long va;
 	spinlock_t *ptl;
+	int i, nr_pages;
 	pgd_t *pgd;
-	int i;
 
 	if (!static_cpu_has(X86_FEATURE_PTI))
 		return 0;
@@ -141,7 +133,9 @@ map_ldt_struct(struct mm_struct *mm, struct ldt_struct *ldt, int slot)
 
 	is_vmalloc = is_vmalloc_addr(ldt->entries);
 
-	for (i = 0; i * PAGE_SIZE < ldt->nr_entries * LDT_ENTRY_SIZE; i++) {
+	nr_pages = DIV_ROUND_UP(ldt->nr_entries * LDT_ENTRY_SIZE, PAGE_SIZE);
+
+	for (i = 0; i < nr_pages; i++) {
 		unsigned long offset = i << PAGE_SHIFT;
 		const void *src = (char *)ldt->entries + offset;
 		unsigned long pfn;
@@ -189,14 +183,42 @@ map_ldt_struct(struct mm_struct *mm, struct ldt_struct *ldt, int slot)
 		}
 	}
 
-	va = (unsigned long)ldt_slot_va(slot);
-	flush_tlb_mm_range(mm, va, va + LDT_SLOT_STRIDE, 0);
-
 	ldt->slot = slot;
 #endif
 	return 0;
 }
 
+static void unmap_ldt_struct(struct mm_struct *mm, struct ldt_struct *ldt)
+{
+#ifdef CONFIG_PAGE_TABLE_ISOLATION
+	unsigned long va;
+	int i, nr_pages;
+
+	if (!ldt)
+		return;
+
+	/* LDT map/unmap is only required for PTI */
+	if (!static_cpu_has(X86_FEATURE_PTI))
+		return;
+
+	nr_pages = DIV_ROUND_UP(ldt->nr_entries * LDT_ENTRY_SIZE, PAGE_SIZE);
+
+	for (i = 0; i < nr_pages; i++) {
+		unsigned long offset = i << PAGE_SHIFT;
+		spinlock_t *ptl;
+		pte_t *ptep;
+
+		va = (unsigned long)ldt_slot_va(ldt->slot) + offset;
+		ptep = get_locked_pte(mm, va, &ptl);
+		pte_clear(mm, va, ptep);
+		pte_unmap_unlock(ptep, ptl);
+	}
+
+	va = (unsigned long)ldt_slot_va(ldt->slot);
+	flush_tlb_mm_range(mm, va, va + nr_pages * PAGE_SIZE, 0);
+#endif /* CONFIG_PAGE_TABLE_ISOLATION */
+}
+
 static void free_ldt_pgtables(struct mm_struct *mm)
 {
 #ifdef CONFIG_PAGE_TABLE_ISOLATION
@@ -433,6 +455,7 @@ static int write_ldt(void __user *ptr, unsigned long bytecount, int oldmode)
 	}
 
 	install_ldt(mm, new_ldt);
+	unmap_ldt_struct(mm, old_ldt);
 	free_ldt_struct(old_ldt);
 	error = 0;
 
-- 
2.17.1