| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
This finally resolve crash if loaded under qemu + haxm. Haitao Shan pointed
out that the reason of that crash is that NX bit get set for page tables.
It seems we missed checking if _PAGE_NX is supported in kaiser_add_user_map
Link: https://www.spinics.net/lists/kernel/msg2689835.html
Reviewed-by: Guenter Roeck <groeck@chromium.org>
Signed-off-by: Lepton Wu <ytht.net@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
(backported from Greg K-H's 4.4 stable-queue)
Signed-off-by: Juerg Haefliger <juerg.haefliger@canonical.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
This resolves a crash if loaded under qemu + haxm under windows.
See https://www.spinics.net/lists/kernel/msg2689835.html for details.
Here is a boot log (the log is from chromeos-4.4, but Tao Wu says that
the same log is also seen with vanilla v4.4.110-rc1).
[ 0.712750] Freeing unused kernel memory: 552K
[ 0.721821] init: Corrupted page table at address 57b029b332e0
[ 0.722761] PGD 80000000bb238067 PUD bc36a067 PMD bc369067 PTE 45d2067
[ 0.722761] Bad pagetable: 000b [#1] PREEMPT SMP
[ 0.722761] Modules linked in:
[ 0.722761] CPU: 1 PID: 1 Comm: init Not tainted 4.4.96 #31
[ 0.722761] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS
rel-1.7.5.1-0-g8936dbb-20141113_115728-nilsson.home.kraxel.org 04/01/2014
[ 0.722761] task: ffff8800bc290000 ti: ffff8800bc28c000 task.ti: ffff8800bc28c000
[ 0.722761] RIP: 0010:[<ffffffff83f4129e>] [<ffffffff83f4129e>] __clear_user+0x42/0x67
[ 0.722761] RSP: 0000:ffff8800bc28fcf8 EFLAGS: 00010202
[ 0.722761] RAX: 0000000000000000 RBX: 00000000000001a4 RCX: 00000000000001a4
[ 0.722761] RDX: 0000000000000000 RSI: 0000000000000008 RDI: 000057b029b332e0
[ 0.722761] RBP: ffff8800bc28fd08 R08: ffff8800bc290000 R09: ffff8800bb2f4000
[ 0.722761] R10: ffff8800bc290000 R11: ffff8800bb2f4000 R12: 000057b029b332e0
[ 0.722761] R13: 0000000000000000 R14: 000057b029b33340 R15: ffff8800bb1e2a00
[ 0.722761] FS: 0000000000000000(0000) GS:ffff8800bfb00000(0000) knlGS:0000000000000000
[ 0.722761] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[ 0.722761] CR2: 000057b029b332e0 CR3: 00000000bb2f8000 CR4: 00000000000006e0
[ 0.722761] Stack:
[ 0.722761] 000057b029b332e0 ffff8800bb95fa80 ffff8800bc28fd18 ffffffff83f4120c
[ 0.722761] ffff8800bc28fe18 ffffffff83e9e7a1 ffff8800bc28fd68 0000000000000000
[ 0.722761] ffff8800bc290000 ffff8800bc290000 ffff8800bc290000 ffff8800bc290000
[ 0.722761] Call Trace:
[ 0.722761] [<ffffffff83f4120c>] clear_user+0x2e/0x30
[ 0.722761] [<ffffffff83e9e7a1>] load_elf_binary+0xa7f/0x18f7
[ 0.722761] [<ffffffff83de2088>] search_binary_handler+0x86/0x19c
[ 0.722761] [<ffffffff83de389e>] do_execveat_common.isra.26+0x909/0xf98
[ 0.722761] [<ffffffff844febe0>] ? rest_init+0x87/0x87
[ 0.722761] [<ffffffff83de40be>] do_execve+0x23/0x25
[ 0.722761] [<ffffffff83c002e3>] run_init_process+0x2b/0x2d
[ 0.722761] [<ffffffff844fec4d>] kernel_init+0x6d/0xda
[ 0.722761] [<ffffffff84505b2f>] ret_from_fork+0x3f/0x70
[ 0.722761] [<ffffffff844febe0>] ? rest_init+0x87/0x87
[ 0.722761] Code: 86 84 be 12 00 00 00 e8 87 0d e8 ff 66 66 90 48 89 d8 48 c1
eb 03 4c 89 e7 83 e0 07 48 89 d9 be 08 00 00 00 31 d2 48 85 c9 74 0a <48> 89 17
48 01 f7 ff c9 75 f6 48 89 c1 85 c9 74 09 88 17 48 ff
[ 0.722761] RIP [<ffffffff83f4129e>] __clear_user+0x42/0x67
[ 0.722761] RSP <ffff8800bc28fcf8>
[ 0.722761] ---[ end trace def703879b4ff090 ]---
[ 0.722761] BUG: sleeping function called from invalid context at /mnt/host/source/src/third_party/kernel/v4.4/kernel/locking/rwsem.c:21
[ 0.722761] in_atomic(): 0, irqs_disabled(): 1, pid: 1, name: init
[ 0.722761] CPU: 1 PID: 1 Comm: init Tainted: G D 4.4.96 #31
[ 0.722761] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.7.5.1-0-g8936dbb-20141113_115728-nilsson.home.kraxel.org 04/01/2014
[ 0.722761] 0000000000000086 dcb5d76098c89836 ffff8800bc28fa30 ffffffff83f34004
[ 0.722761] ffffffff84839dc2 0000000000000015 ffff8800bc28fa40 ffffffff83d57dc9
[ 0.722761] ffff8800bc28fa68 ffffffff83d57e6a ffffffff84a53640 0000000000000000
[ 0.722761] Call Trace:
[ 0.722761] [<ffffffff83f34004>] dump_stack+0x4d/0x63
[ 0.722761] [<ffffffff83d57dc9>] ___might_sleep+0x13a/0x13c
[ 0.722761] [<ffffffff83d57e6a>] __might_sleep+0x9f/0xa6
[ 0.722761] [<ffffffff84502788>] down_read+0x20/0x31
[ 0.722761] [<ffffffff83cc5d9b>] __blocking_notifier_call_chain+0x35/0x63
[ 0.722761] [<ffffffff83cc5ddd>] blocking_notifier_call_chain+0x14/0x16
[ 0.800374] usb 1-1: new full-speed USB device number 2 using uhci_hcd
[ 0.722761] [<ffffffff83cefe97>] profile_task_exit+0x1a/0x1c
[ 0.802309] [<ffffffff83cac84e>] do_exit+0x39/0xe7f
[ 0.802309] [<ffffffff83ce5938>] ? vprintk_default+0x1d/0x1f
[ 0.802309] [<ffffffff83d7bb95>] ? printk+0x57/0x73
[ 0.802309] [<ffffffff83c46e25>] oops_end+0x80/0x85
[ 0.802309] [<ffffffff83c7b747>] pgtable_bad+0x8a/0x95
[ 0.802309] [<ffffffff83ca7f4a>] __do_page_fault+0x8c/0x352
[ 0.802309] [<ffffffff83eefba5>] ? file_has_perm+0xc4/0xe5
[ 0.802309] [<ffffffff83ca821c>] do_page_fault+0xc/0xe
[ 0.802309] [<ffffffff84507682>] page_fault+0x22/0x30
[ 0.802309] [<ffffffff83f4129e>] ? __clear_user+0x42/0x67
[ 0.802309] [<ffffffff83f4127f>] ? __clear_user+0x23/0x67
[ 0.802309] [<ffffffff83f4120c>] clear_user+0x2e/0x30
[ 0.802309] [<ffffffff83e9e7a1>] load_elf_binary+0xa7f/0x18f7
[ 0.802309] [<ffffffff83de2088>] search_binary_handler+0x86/0x19c
[ 0.802309] [<ffffffff83de389e>] do_execveat_common.isra.26+0x909/0xf98
[ 0.802309] [<ffffffff844febe0>] ? rest_init+0x87/0x87
[ 0.802309] [<ffffffff83de40be>] do_execve+0x23/0x25
[ 0.802309] [<ffffffff83c002e3>] run_init_process+0x2b/0x2d
[ 0.802309] [<ffffffff844fec4d>] kernel_init+0x6d/0xda
[ 0.802309] [<ffffffff84505b2f>] ret_from_fork+0x3f/0x70
[ 0.802309] [<ffffffff844febe0>] ? rest_init+0x87/0x87
[ 0.830559] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000009
[ 0.830559]
[ 0.831305] Kernel Offset: 0x2c00000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff)
[ 0.831305] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000009
The crash part of this problem may be solved with the following patch
(thanks to Hugh for the hint). There is still another problem, though -
with this patch applied, the qemu session aborts with "VCPU Shutdown
request", whatever that means.
Cc: lepton <ytht.net@gmail.com>
Signed-off-by: Guenter Roeck <groeck@chromium.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
(cherry picked from commit b33c3c64c4786cd724ccde6fa97c87ada49f6a73 linux-4.4.y)
Signed-off-by: Juerg Haefliger <juerg.haefliger@canonical.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
Make sure dmesg reports when KPTI is enabled.
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
This renames CONFIG_KAISER to CONFIG_PAGE_TABLE_ISOLATION.
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[bwh: Backported to 3.2]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
... before the first use of kaiser_enabled as otherwise funky
things happen:
about to get started...
(XEN) d0v0 Unhandled page fault fault/trap [#14, ec=0000]
(XEN) Pagetable walk from ffff88022a449090:
(XEN) L4[0x110] = 0000000229e0e067 0000000000001e0e
(XEN) L3[0x008] = 0000000000000000 ffffffffffffffff
(XEN) domain_crash_sync called from entry.S: fault at ffff82d08033fd08
entry.o#create_bounce_frame+0x135/0x14d
(XEN) Domain 0 (vcpu#0) crashed on cpu#0:
(XEN) ----[ Xen-4.9.1_02-3.21 x86_64 debug=n Not tainted ]----
(XEN) CPU: 0
(XEN) RIP: e033:[<ffffffff81007460>]
(XEN) RFLAGS: 0000000000000286 EM: 1 CONTEXT: pv guest (d0v0)
Signed-off-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
Kaiser cannot be used on paravirtualized MMUs (namely reading and writing CR3).
This does not work with KAISER as the CR3 switch from and to user space PGD
would require to map the whole XEN_PV machinery into both.
More importantly, enabling KAISER on Xen PV doesn't make too much sense, as PV
guests use distinct %cr3 values for kernel and user already.
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[bwh: Backported to 3.2: use xen_pv_domain()]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
Let kaiser_flush_tlb_on_return_to_user() do the X86_FEATURE_PCID
check, instead of each caller doing it inline first: nobody needs
to optimize for the noPCID case, it's clearer this way, and better
suits later changes. Replace those no-op X86_CR3_PCID_KERN_FLUSH lines
by a BUILD_BUG_ON() in load_new_mm_cr3(), in case something changes.
(cherry picked from Change-Id: I9b528ed9d7c1ae4a3b4738c2894ee1740b6fb0b9)
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
Now that we're playing the ALTERNATIVE game, use that more efficient
method: instead of user-mapping an extra page, and reading an extra
cacheline each time for x86_cr3_pcid_noflush.
Neel has found that __stringify(bts $X86_CR3_PCID_NOFLUSH_BIT, %rax)
is a working substitute for the "bts $63, %rax" in these ALTERNATIVEs;
but the one line with $63 in looks clearer, so let's stick with that.
Worried about what happens with an ALTERNATIVE between the jump and
jump label in another ALTERNATIVE? I was, but have checked the
combinations in SWITCH_KERNEL_CR3_NO_STACK at entry_SYSCALL_64,
and it does a good job.
(cherry picked from Change-Id: I46d06167615aa8d628eed9972125ab2faca93f05)
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
AMD (and possibly other vendors) are not affected by the leak
KAISER is protecting against.
Keep the "nopti" for traditional reasons and add pti=<on|off|auto>
like upstream.
Signed-off-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
Concentrate it in arch/x86/mm/kaiser.c and use the upstream string "nopti".
Signed-off-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
Added "nokaiser" boot option: an early param like "noinvpcid".
Most places now check int kaiser_enabled (#defined 0 when not
CONFIG_KAISER) instead of #ifdef CONFIG_KAISER; but entry_64.S
and entry_64_compat.S are using the ALTERNATIVE technique, which
patches in the preferred instructions at runtime. That technique
is tied to x86 cpu features, so X86_FEATURE_KAISER fabricated
("" in its comment so "kaiser" not magicked into /proc/cpuinfo).
Prior to "nokaiser", Kaiser #defined _PAGE_GLOBAL 0: revert that,
but be careful with both _PAGE_GLOBAL and CR4.PGE: setting them when
nokaiser like when !CONFIG_KAISER, but not setting either when kaiser -
neither matters on its own, but it's hard to be sure that _PAGE_GLOBAL
won't get set in some obscure corner, or something add PGE into CR4.
By omitting _PAGE_GLOBAL from __supported_pte_mask when kaiser_enabled,
all page table setup which uses pte_pfn() masks it out of the ptes.
It's slightly shameful that the same declaration versus definition of
kaiser_enabled appears in not one, not two, but in three header files
(asm/kaiser.h, asm/pgtable.h, asm/tlbflush.h). I felt safer that way,
than with #including any of those in any of the others; and did not
feel it worth an asm/kaiser_enabled.h - kernel/cpu/common.c includes
them all, so we shall hear about it if they get out of synch.
Cleanups while in the area: removed the silly #ifdef CONFIG_KAISER
from kaiser.c; removed the unused native_get_normal_pgd(); removed
the spurious reg clutter from SWITCH_*_CR3 macro stubs; corrected some
comments. But more interestingly, set CR4.PSE in secondary_startup_64:
the manual is clear that it does not matter whether it's 0 or 1 when
4-level-pts are enabled, but I was distracted to find cr4 different on
BSP and auxiliaries - BSP alone was adding PSE, in init_memory_mapping().
(cherry picked from Change-Id: I8e5bec716944444359cbd19f6729311eff943e9a)
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
In 3.2 (and earlier, and up to 3.15) Kaiser needs to user_map the
__kprobes_text as well as the __entry_text: entry_64.S places some
vital functions there, so without this you very soon triple-fault.
Many thanks to Jiri Kosina for pointing me in this direction.
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
To avoid breaking the kernel ABI.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
[Hugh Dickins: Backported to 3.2:
- Leave out the PVCLOCK_FIXMAP user mapping, which does not apply to
this tree
- For safety added vsyscall_pgprot, and a BUG_ON if _PAGE_USER
outside of FIXMAP.]
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
This patch introduces our implementation of KAISER (Kernel Address
Isolation to have Side-channels Efficiently Removed), a kernel isolation
technique to close hardware side channels on kernel address information.
More information about the original patch can be found at:
https://github.com/IAIK/KAISER
http://marc.info/?l=linux-kernel&m=149390087310405&w=2
Daniel Gruss <daniel.gruss@iaik.tugraz.at>
Richard Fellner <richard.fellner@student.tugraz.at>
Michael Schwarz <michael.schwarz@iaik.tugraz.at>
<clementine.maurice@iaik.tugraz.at>
<moritz.lipp@iaik.tugraz.at>
That original was then developed further by
Dave Hansen <dave.hansen@intel.com>
Hugh Dickins <hughd@google.com>
then others after this snapshot.
This combined patch for 3.2.96 was derived from hughd's patches below
for 3.18.72, in 2017-12-04's kaiser-3.18.72.tar; except for the last,
which was sent in 2017-12-09's nokaiser-3.18.72.tar. They have been
combined in order to minimize the effort of rebasing: most of the
patches in the 3.18.72 series were small fixes and cleanups and
enhancements to three large patches. About the only new work in this
backport is a simple reimplementation of kaiser_remove_mapping():
since mm/pageattr.c changed a lot between 3.2 and 3.18, and the
mods there for Kaiser never seemed necessary.
KAISER: Kernel Address Isolation
kaiser: merged update
kaiser: do not set _PAGE_NX on pgd_none
kaiser: stack map PAGE_SIZE at THREAD_SIZE-PAGE_SIZE
kaiser: fix build and FIXME in alloc_ldt_struct()
kaiser: KAISER depends on SMP
kaiser: fix regs to do_nmi() ifndef CONFIG_KAISER
kaiser: fix perf crashes
kaiser: ENOMEM if kaiser_pagetable_walk() NULL
kaiser: tidied up asm/kaiser.h somewhat
kaiser: tidied up kaiser_add/remove_mapping slightly
kaiser: kaiser_remove_mapping() move along the pgd
kaiser: align addition to x86/mm/Makefile
kaiser: cleanups while trying for gold link
kaiser: name that 0x1000 KAISER_SHADOW_PGD_OFFSET
kaiser: delete KAISER_REAL_SWITCH option
kaiser: vmstat show NR_KAISERTABLE as nr_overhead
kaiser: enhanced by kernel and user PCIDs
kaiser: load_new_mm_cr3() let SWITCH_USER_CR3 flush user
kaiser: PCID 0 for kernel and 128 for user
kaiser: x86_cr3_pcid_noflush and x86_cr3_pcid_user
kaiser: paranoid_entry pass cr3 need to paranoid_exit
kaiser: _pgd_alloc() without __GFP_REPEAT to avoid stalls
kaiser: fix unlikely error in alloc_ldt_struct()
kaiser: drop is_atomic arg to kaiser_pagetable_walk()
Signed-off-by: Hugh Dickins <hughd@google.com>
[bwh:
- Fixed the #undef in arch/x86/boot/compressed/misc.h
- Add missing #include in arch/x86/mm/kaiser.c]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit ce4a4e565f5264909a18c733b864c3f74467f69e upstream.
The UP asm/tlbflush.h generates somewhat nicer code than the SMP version.
Aside from that, it's fallen quite a bit behind the SMP code:
- flush_tlb_mm_range() didn't flush individual pages if the range
was small.
- The lazy TLB code was much weaker. This usually wouldn't matter,
but, if a kernel thread flushed its lazy "active_mm" more than
once (due to reclaim or similar), it wouldn't be unlazied and
would instead pointlessly flush repeatedly.
- Tracepoints were missing.
Aside from that, simply having the UP code around was a maintanence
burden, since it means that any change to the TLB flush code had to
make sure not to break it.
Simplify everything by deleting the UP code.
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Borislav Petkov <bpetkov@suse.de>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Nadav Amit <nadav.amit@gmail.com>
Cc: Nadav Amit <namit@vmware.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-mm@kvack.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
[Hugh Dickins: Backported to 3.2]
Signed-off-by: Hugh Dickins <hughd@google.com>
[bwh: Fix allnoconfig build failure due to direct use of 'apic' in
flush_tlb_others_ipi()]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 078194f8e9fe3cf54c8fd8bded48a1db5bd8eb8a upstream.
Potential races between switch_mm() and TLB-flush or LDT-flush IPIs
could be very messy. AFAICT the code is currently okay, whether by
accident or by careful design, but enabling PCID will make it
considerably more complicated and will no longer be obviously safe.
Fix it with a big hammer: run switch_mm() with IRQs off.
To avoid a performance hit in the scheduler, we take advantage of
our knowledge that the scheduler already has IRQs disabled when it
calls switch_mm().
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Reviewed-by: Borislav Petkov <bp@suse.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/f19baf759693c9dcae64bbff76189db77cb13398.1461688545.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 69c0319aabba45bcf33178916a2f06967b4adede upstream.
It's fairly large and it has quite a few callers. This may also
help untangle some headers down the road.
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Reviewed-by: Borislav Petkov <bp@suse.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/54f3367803e7f80b2be62c8a21879aa74b1a5f57.1461688545.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
[Hugh Dickins: Backported to 3.2]
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit e1074888c326038340a1ada9129d679e661f2ea6 upstream.
Currently all of the functions that live in tlb.c are inlined on
!SMP builds. One can debate whether this is a good idea (in many
respects the code in tlb.c is better than the inlined UP code).
Regardless, I want to add code that needs to be built on UP and SMP
kernels and relates to tlb flushing, so arrange for tlb.c to be
compiled unconditionally.
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Reviewed-by: Borislav Petkov <bp@suse.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/f0d778f0d828fc46e5d1946bca80f0aaf9abf032.1461688545.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 861ce4a3244c21b0af64f880d5bfe5e6e2fb9e4a upstream.
'__vmalloc_start_set' currently only gets set in initmem_init() when
!CONFIG_NEED_MULTIPLE_NODES. This breaks detection of vmalloc address
with virt_addr_valid() with CONFIG_NEED_MULTIPLE_NODES=y, causing
a kernel crash:
[mm/usercopy] 517e1fbeb6: kernel BUG at arch/x86/mm/physaddr.c:78!
Set '__vmalloc_start_set' appropriately for that case as well.
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Laura Abbott <labbott@redhat.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: dc16ecf7fd1f ("x86-32: use specific __vmalloc_start_set flag in __virt_addr_valid")
Link: http://lkml.kernel.org/r/1494278596-30373-1-git-send-email-labbott@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit a4866aa812518ed1a37d8ea0c881dc946409de94 upstream.
Under CONFIG_STRICT_DEVMEM, reading System RAM through /dev/mem is
disallowed. However, on x86, the first 1MB was always allowed for BIOS
and similar things, regardless of it actually being System RAM. It was
possible for heap to end up getting allocated in low 1MB RAM, and then
read by things like x86info or dd, which would trip hardened usercopy:
usercopy: kernel memory exposure attempt detected from ffff880000090000 (dma-kmalloc-256) (4096 bytes)
This changes the x86 exception for the low 1MB by reading back zeros for
System RAM areas instead of blindly allowing them. More work is needed to
extend this to mmap, but currently mmap doesn't go through usercopy, so
hardened usercopy won't Oops the kernel.
Reported-by: Tommi Rantala <tommi.t.rantala@nokia.com>
Tested-by: Tommi Rantala <tommi.t.rantala@nokia.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 73e8f3d7e2cb23614d5115703d76d8e54764b641 upstream.
Fixing an off-by-one error in devmem_is_allowed(), which allows
accesses to physical addresses 0x100000-0x100fff, an extra page
past 1MB.
Signed-off-by: T Makphaibulchoke <tmac@hp.com>
Acked-by: H. Peter Anvin <hpa@zytor.com>
Cc: yinghai@kernel.org
Cc: tiwai@suse.de
Cc: dhowells@redhat.com
Link: http://lkml.kernel.org/r/1346210503-14276-1-git-send-email-tmac@hp.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 1be7107fbe18eed3e319a6c3e83c78254b693acb upstream.
Stack guard page is a useful feature to reduce a risk of stack smashing
into a different mapping. We have been using a single page gap which
is sufficient to prevent having stack adjacent to a different mapping.
But this seems to be insufficient in the light of the stack usage in
userspace. E.g. glibc uses as large as 64kB alloca() in many commonly
used functions. Others use constructs liks gid_t buffer[NGROUPS_MAX]
which is 256kB or stack strings with MAX_ARG_STRLEN.
This will become especially dangerous for suid binaries and the default
no limit for the stack size limit because those applications can be
tricked to consume a large portion of the stack and a single glibc call
could jump over the guard page. These attacks are not theoretical,
unfortunatelly.
Make those attacks less probable by increasing the stack guard gap
to 1MB (on systems with 4k pages; but make it depend on the page size
because systems with larger base pages might cap stack allocations in
the PAGE_SIZE units) which should cover larger alloca() and VLA stack
allocations. It is obviously not a full fix because the problem is
somehow inherent, but it should reduce attack space a lot.
One could argue that the gap size should be configurable from userspace,
but that can be done later when somebody finds that the new 1MB is wrong
for some special case applications. For now, add a kernel command line
option (stack_guard_gap) to specify the stack gap size (in page units).
Implementation wise, first delete all the old code for stack guard page:
because although we could get away with accounting one extra page in a
stack vma, accounting a larger gap can break userspace - case in point,
a program run with "ulimit -S -v 20000" failed when the 1MB gap was
counted for RLIMIT_AS; similar problems could come with RLIMIT_MLOCK
and strict non-overcommit mode.
Instead of keeping gap inside the stack vma, maintain the stack guard
gap as a gap between vmas: using vm_start_gap() in place of vm_start
(or vm_end_gap() in place of vm_end if VM_GROWSUP) in just those few
places which need to respect the gap - mainly arch_get_unmapped_area(),
and and the vma tree's subtree_gap support for that.
Original-patch-by: Oleg Nesterov <oleg@redhat.com>
Original-patch-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Hugh Dickins <hughd@google.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Tested-by: Helge Deller <deller@gmx.de> # parisc
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
[Hugh Dickins: Backported to 3.2]
[bwh: Fix more instances of vma->vm_start in sparc64 impl. of
arch_get_unmapped_area_topdown() and generic impl. of
hugetlb_get_unmapped_area()]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 8b8addf891de8a00e4d39fc32f93f7c5eb8feceb upstream.
Currently on i386 and on X86_64 when emulating X86_32 in legacy mode, only
the stack and the executable are randomized but not other mmapped files
(libraries, vDSO, etc.). This patch enables randomization for the
libraries, vDSO and mmap requests on i386 and in X86_32 in legacy mode.
By default on i386 there are 8 bits for the randomization of the libraries,
vDSO and mmaps which only uses 1MB of VA.
This patch preserves the original randomness, using 1MB of VA out of 3GB or
4GB. We think that 1MB out of 3GB is not a big cost for having the ASLR.
The first obvious security benefit is that all objects are randomized (not
only the stack and the executable) in legacy mode which highly increases
the ASLR effectiveness, otherwise the attackers may use these
non-randomized areas. But also sensitive setuid/setgid applications are
more secure because currently, attackers can disable the randomization of
these applications by setting the ulimit stack to "unlimited". This is a
very old and widely known trick to disable the ASLR in i386 which has been
allowed for too long.
Another trick used to disable the ASLR was to set the ADDR_NO_RANDOMIZE
personality flag, but fortunately this doesn't work on setuid/setgid
applications because there is security checks which clear Security-relevant
flags.
This patch always randomizes the mmap_legacy_base address, removing the
possibility to disable the ASLR by setting the stack to "unlimited".
Signed-off-by: Hector Marco-Gisbert <hecmargi@upv.es>
Acked-by: Ismael Ripoll Ripoll <iripoll@upv.es>
Acked-by: Kees Cook <keescook@chromium.org>
Acked-by: Arjan van de Ven <arjan@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: akpm@linux-foundation.org
Cc: kees Cook <keescook@chromium.org>
Link: http://lkml.kernel.org/r/1457639460-5242-1-git-send-email-hecmargi@upv.es
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 82168140bc4cec7ec9bad39705518541149ff8b7 upstream.
In preparation for splitting out ET_DYN ASLR, this refactors the use of
mmap_rnd() to be used similarly to arm, and extracts the checking of
PF_RANDOMIZE.
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Ingo Molnar <mingo@kernel.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Andy Lutomirski <luto@amacapital.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 742563777e8da62197d6cb4b99f4027f59454735 upstream.
There are a couple of nasty truncation bugs lurking in the pageattr
code that can be triggered when mapping EFI regions, e.g. when we pass
a cpa->pgd pointer. Because cpa->numpages is a 32-bit value, shifting
left by PAGE_SHIFT will truncate the resultant address to 32-bits.
Viorel-Cătălin managed to trigger this bug on his Dell machine that
provides a ~5GB EFI region which requires 1236992 pages to be mapped.
When calling populate_pud() the end of the region gets calculated
incorrectly in the following buggy expression,
end = start + (cpa->numpages << PAGE_SHIFT);
And only 188416 pages are mapped. Next, populate_pud() gets invoked
for a second time because of the loop in __change_page_attr_set_clr(),
only this time no pages get mapped because shifting the remaining
number of pages (1048576) by PAGE_SHIFT is zero. At which point the
loop in __change_page_attr_set_clr() spins forever because we fail to
map progress.
Hitting this bug depends very much on the virtual address we pick to
map the large region at and how many pages we map on the initial run
through the loop. This explains why this issue was only recently hit
with the introduction of commit
a5caa209ba9c ("x86/efi: Fix boot crash by mapping EFI memmap
entries bottom-up at runtime, instead of top-down")
It's interesting to note that safe uses of cpa->numpages do exist in
the pageattr code. If instead of shifting ->numpages we multiply by
PAGE_SIZE, no truncation occurs because PAGE_SIZE is a UL value, and
so the result is unsigned long.
To avoid surprises when users try to convert very large cpa->numpages
values to addresses, change the data type from 'int' to 'unsigned
long', thereby making it suitable for shifting by PAGE_SHIFT without
any type casting.
The alternative would be to make liberal use of casting, but that is
far more likely to cause problems in the future when someone adds more
code and fails to cast properly; this bug was difficult enough to
track down in the first place.
Reported-and-tested-by: Viorel-Cătălin Răpițeanu <rapiteanu.catalin@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Sai Praneeth Prakhya <sai.praneeth.prakhya@intel.com>
Signed-off-by: Matt Fleming <matt@codeblueprint.co.uk>
Link: https://bugzilla.kernel.org/show_bug.cgi?id=110131
Link: http://lkml.kernel.org/r/1454067370-10374-1-git-send-email-matt@codeblueprint.co.uk
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 71b3c126e61177eb693423f2e18a1914205b165e upstream.
When switch_mm() activates a new PGD, it also sets a bit that
tells other CPUs that the PGD is in use so that TLB flush IPIs
will be sent. In order for that to work correctly, the bit
needs to be visible prior to loading the PGD and therefore
starting to fill the local TLB.
Document all the barriers that make this work correctly and add
a couple that were missing.
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-mm@kvack.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
[bwh: Backported to 3.2:
- There's no flush_tlb_mm_range(), only flush_tlb_mm() which does not use
INVLPG
- Adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 4e7c22d447bb6d7e37bfe39ff658486ae78e8d77 upstream.
The issue is that the stack for processes is not properly randomized on
64 bit architectures due to an integer overflow.
The affected function is randomize_stack_top() in file
"fs/binfmt_elf.c":
static unsigned long randomize_stack_top(unsigned long stack_top)
{
unsigned int random_variable = 0;
if ((current->flags & PF_RANDOMIZE) &&
!(current->personality & ADDR_NO_RANDOMIZE)) {
random_variable = get_random_int() & STACK_RND_MASK;
random_variable <<= PAGE_SHIFT;
}
return PAGE_ALIGN(stack_top) + random_variable;
return PAGE_ALIGN(stack_top) - random_variable;
}
Note that, it declares the "random_variable" variable as "unsigned int".
Since the result of the shifting operation between STACK_RND_MASK (which
is 0x3fffff on x86_64, 22 bits) and PAGE_SHIFT (which is 12 on x86_64):
random_variable <<= PAGE_SHIFT;
then the two leftmost bits are dropped when storing the result in the
"random_variable". This variable shall be at least 34 bits long to hold
the (22+12) result.
These two dropped bits have an impact on the entropy of process stack.
Concretely, the total stack entropy is reduced by four: from 2^28 to
2^30 (One fourth of expected entropy).
This patch restores back the entropy by correcting the types involved
in the operations in the functions randomize_stack_top() and
stack_maxrandom_size().
The successful fix can be tested with:
$ for i in `seq 1 10`; do cat /proc/self/maps | grep stack; done
7ffeda566000-7ffeda587000 rw-p 00000000 00:00 0 [stack]
7fff5a332000-7fff5a353000 rw-p 00000000 00:00 0 [stack]
7ffcdb7a1000-7ffcdb7c2000 rw-p 00000000 00:00 0 [stack]
7ffd5e2c4000-7ffd5e2e5000 rw-p 00000000 00:00 0 [stack]
...
Once corrected, the leading bytes should be between 7ffc and 7fff,
rather than always being 7fff.
Signed-off-by: Hector Marco-Gisbert <hecmargi@upv.es>
Signed-off-by: Ismael Ripoll <iripoll@upv.es>
[ Rebased, fixed 80 char bugs, cleaned up commit message, added test example and CVE ]
Signed-off-by: Kees Cook <keescook@chromium.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Fixes: CVE-2015-1593
Link: http://lkml.kernel.org/r/20150214173350.GA18393@www.outflux.net
Signed-off-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
When backporting commit 33692f27597f ('vm: add VM_FAULT_SIGSEGV
handling support') I didn't notice that it depended on a recent change
to the locking context of mm_fault_error() (commit 7fb08eca4527,
'x86: mm: move mmap_sem unlock from mm_fault_error() to caller').
That isn't easily applicable to 3.2, so instead make sure we drop
mm->mmap_sem on the new branch of mm_fault_error().
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
This reverts commit e105c8187b7101e8a8a54ac0218c9d9c9463c636 which
was commit 72212675d1c96f5db8ec6fb35701879911193158 upstream.
This caused suspend/resume to stop working on at least some systems -
specifically, the system would reboot when woken.
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Cc: Steven Rostedt <rostedt@goodmis.org>
|
|
This reverts commit a5c187d92d2ce30315f333b9dff33af832e8b443 which
was commit 45e2a9d4701d8c624d4a4bcdd1084eae31e92f58 upstream.
The previous commit caused suspend/resume to stop working on at least
some systems - specifically, the system would reboot when woken.
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Cc: Steven Rostedt <rostedt@goodmis.org>
|
|
commit 33692f27597fcab536d7cbbcc8f52905133e4aa7 upstream.
The core VM already knows about VM_FAULT_SIGBUS, but cannot return a
"you should SIGSEGV" error, because the SIGSEGV case was generally
handled by the caller - usually the architecture fault handler.
That results in lots of duplication - all the architecture fault
handlers end up doing very similar "look up vma, check permissions, do
retries etc" - but it generally works. However, there are cases where
the VM actually wants to SIGSEGV, and applications _expect_ SIGSEGV.
In particular, when accessing the stack guard page, libsigsegv expects a
SIGSEGV. And it usually got one, because the stack growth is handled by
that duplicated architecture fault handler.
However, when the generic VM layer started propagating the error return
from the stack expansion in commit fee7e49d4514 ("mm: propagate error
from stack expansion even for guard page"), that now exposed the
existing VM_FAULT_SIGBUS result to user space. And user space really
expected SIGSEGV, not SIGBUS.
To fix that case, we need to add a VM_FAULT_SIGSEGV, and teach all those
duplicate architecture fault handlers about it. They all already have
the code to handle SIGSEGV, so it's about just tying that new return
value to the existing code, but it's all a bit annoying.
This is the mindless minimal patch to do this. A more extensive patch
would be to try to gather up the mostly shared fault handling logic into
one generic helper routine, and long-term we really should do that
cleanup.
Just from this patch, you can generally see that most architectures just
copied (directly or indirectly) the old x86 way of doing things, but in
the meantime that original x86 model has been improved to hold the VM
semaphore for shorter times etc and to handle VM_FAULT_RETRY and other
"newer" things, so it would be a good idea to bring all those
improvements to the generic case and teach other architectures about
them too.
Reported-and-tested-by: Takashi Iwai <tiwai@suse.de>
Tested-by: Jan Engelhardt <jengelh@inai.de>
Acked-by: Heiko Carstens <heiko.carstens@de.ibm.com> # "s390 still compiles and boots"
Cc: linux-arch@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
[bwh: Backported to 3.2:
- Adjust filenames, context
- Drop arc, metag, nios2 and lustre changes
- For sh, patch both 32-bit and 64-bit implementations to use goto bad_area
- For s390, pass int_code and trans_exc_code as arguments to do_no_context()
and do_sigsegv()]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 45e2a9d4701d8c624d4a4bcdd1084eae31e92f58 upstream.
When setting up permissions on kernel memory at boot, the end of the
PMD that was split from bss remained executable. It should be NX like
the rest. This performs a PMD alignment instead of a PAGE alignment to
get the correct span of memory.
Before:
---[ High Kernel Mapping ]---
...
0xffffffff8202d000-0xffffffff82200000 1868K RW GLB NX pte
0xffffffff82200000-0xffffffff82c00000 10M RW PSE GLB NX pmd
0xffffffff82c00000-0xffffffff82df5000 2004K RW GLB NX pte
0xffffffff82df5000-0xffffffff82e00000 44K RW GLB x pte
0xffffffff82e00000-0xffffffffc0000000 978M pmd
After:
---[ High Kernel Mapping ]---
...
0xffffffff8202d000-0xffffffff82200000 1868K RW GLB NX pte
0xffffffff82200000-0xffffffff82e00000 12M RW PSE GLB NX pmd
0xffffffff82e00000-0xffffffffc0000000 978M pmd
[ tglx: Changed it to roundup(_brk_end, PMD_SIZE) and added a comment.
We really should unmap the reminder along with the holes
caused by init,initdata etc. but thats a different issue ]
Signed-off-by: Kees Cook <keescook@chromium.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Toshi Kani <toshi.kani@hp.com>
Cc: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Wang Nan <wangnan0@huawei.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Link: http://lkml.kernel.org/r/20141114194737.GA3091@www.outflux.net
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
[bwh: BAckported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 72212675d1c96f5db8ec6fb35701879911193158 upstream.
HPA said, we should not have RW and +x set at the time.
for kernel layout:
[ 0.000000] Kernel Layout:
[ 0.000000] .text: [0x01000000-0x021434f8]
[ 0.000000] .rodata: [0x02200000-0x02a13fff]
[ 0.000000] .data: [0x02c00000-0x02dc763f]
[ 0.000000] .init: [0x02dc9000-0x0312cfff]
[ 0.000000] .bss: [0x0313b000-0x03dd6fff]
[ 0.000000] .brk: [0x03dd7000-0x03dfffff]
before the patch, we have
---[ High Kernel Mapping ]---
0xffffffff80000000-0xffffffff81000000 16M pmd
0xffffffff81000000-0xffffffff82200000 18M ro PSE GLB x pmd
0xffffffff82200000-0xffffffff82c00000 10M ro PSE GLB NX pmd
0xffffffff82c00000-0xffffffff82dc9000 1828K RW GLB x pte
0xffffffff82dc9000-0xffffffff82e00000 220K RW GLB NX pte
0xffffffff82e00000-0xffffffff83000000 2M RW PSE GLB NX pmd
0xffffffff83000000-0xffffffff8313a000 1256K RW GLB NX pte
0xffffffff8313a000-0xffffffff83200000 792K RW GLB x pte
0xffffffff83200000-0xffffffff83e00000 12M RW PSE GLB x pmd
0xffffffff83e00000-0xffffffffa0000000 450M pmd
after patch,, we get
---[ High Kernel Mapping ]---
0xffffffff80000000-0xffffffff81000000 16M pmd
0xffffffff81000000-0xffffffff82200000 18M ro PSE GLB x pmd
0xffffffff82200000-0xffffffff82c00000 10M ro PSE GLB NX pmd
0xffffffff82c00000-0xffffffff82e00000 2M RW GLB NX pte
0xffffffff82e00000-0xffffffff83000000 2M RW PSE GLB NX pmd
0xffffffff83000000-0xffffffff83200000 2M RW GLB NX pte
0xffffffff83200000-0xffffffff83e00000 12M RW PSE GLB NX pmd
0xffffffff83e00000-0xffffffffa0000000 450M pmd
so data, bss, brk get NX ...
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Link: http://lkml.kernel.org/r/1359058816-7615-33-git-send-email-yinghai@kernel.org
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 3891a04aafd668686239349ea58f3314ea2af86b upstream.
The IRET instruction, when returning to a 16-bit segment, only
restores the bottom 16 bits of the user space stack pointer. This
causes some 16-bit software to break, but it also leaks kernel state
to user space. We have a software workaround for that ("espfix") for
the 32-bit kernel, but it relies on a nonzero stack segment base which
is not available in 64-bit mode.
In checkin:
b3b42ac2cbae x86-64, modify_ldt: Ban 16-bit segments on 64-bit kernels
we "solved" this by forbidding 16-bit segments on 64-bit kernels, with
the logic that 16-bit support is crippled on 64-bit kernels anyway (no
V86 support), but it turns out that people are doing stuff like
running old Win16 binaries under Wine and expect it to work.
This works around this by creating percpu "ministacks", each of which
is mapped 2^16 times 64K apart. When we detect that the return SS is
on the LDT, we copy the IRET frame to the ministack and use the
relevant alias to return to userspace. The ministacks are mapped
readonly, so if IRET faults we promote #GP to #DF which is an IST
vector and thus has its own stack; we then do the fixup in the #DF
handler.
(Making #GP an IST exception would make the msr_safe functions unsafe
in NMI/MC context, and quite possibly have other effects.)
Special thanks to:
- Andy Lutomirski, for the suggestion of using very small stack slots
and copy (as opposed to map) the IRET frame there, and for the
suggestion to mark them readonly and let the fault promote to #DF.
- Konrad Wilk for paravirt fixup and testing.
- Borislav Petkov for testing help and useful comments.
Reported-by: Brian Gerst <brgerst@gmail.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Link: http://lkml.kernel.org/r/1398816946-3351-1-git-send-email-hpa@linux.intel.com
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Andrew Lutomriski <amluto@gmail.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Dirk Hohndel <dirk@hohndel.org>
Cc: Arjan van de Ven <arjan.van.de.ven@intel.com>
Cc: comex <comexk@gmail.com>
Cc: Alexander van Heukelum <heukelum@fastmail.fm>
Cc: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit c81c8a1eeede61e92a15103748c23d100880cc8a upstream.
In __ioremap_caller() (the guts of ioremap), we loop over the range of
pfns being remapped and checks each one individually with page_is_ram().
For large ioremaps, this can be very slow. For example, we have a
device with a 256 GiB PCI BAR, and ioremapping this BAR can take 20+
seconds -- sometimes long enough to trigger the soft lockup detector!
Internally, page_is_ram() calls walk_system_ram_range() on a single
page. Instead, we can make a single call to walk_system_ram_range()
from __ioremap_caller(), and do our further checks only for any RAM
pages that we find. For the common case of MMIO, this saves an enormous
amount of work, since the range being ioremapped doesn't intersect
system RAM at all.
With this change, ioremap on our 256 GiB BAR takes less than 1 second.
Signed-off-by: Roland Dreier <roland@purestorage.com>
Link: http://lkml.kernel.org/r/1399054721-1331-1-git-send-email-roland@kernel.org
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 41aacc1eea645c99edbe8fbcf78a97dc9b862adc upstream.
This is the updated version of df54d6fa5427 ("x86 get_unmapped_area():
use proper mmap base for bottom-up direction") that only randomizes the
mmap base address once.
Signed-off-by: Radu Caragea <sinaelgl@gmail.com>
Reported-and-tested-by: Jeff Shorey <shoreyjeff@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Michel Lespinasse <walken@google.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Adrian Sendroiu <molecula2788@gmail.com>
Cc: Greg KH <greg@kroah.com>
Cc: Kamal Mostafa <kamal@canonical.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
Patch for 3.0-stable. Function find_early_table_space removed upstream.
Fixes panic in alloc_low_page due to pgt_buf overflow during
init_memory_mapping.
find_early_table_space sizes pgt_buf based upon the size of the
memory being mapped, but it does not take into account the alignment
of the memory. When the region being mapped spans a 512GB (PGDIR_SIZE)
alignment, a panic from alloc_low_pages occurs.
kernel_physical_mapping_init takes into account PGDIR_SIZE alignment.
This causes an extra call to alloc_low_page to be made. This extra call
isn't accounted for by find_early_table_space and causes a kernel panic.
Change is to take into account PGDIR_SIZE alignment in find_early_table_space.
Signed-off-by: Jerry Hoemann <jerry.hoemann@hp.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 1160c2779b826c6f5c08e5cc542de58fd1f667d5 upstream.
In paravirtualized x86_64 kernels, vmalloc_fault may cause an oops
when lazy MMU updates are enabled, because set_pgd effects are being
deferred.
One instance of this problem is during process mm cleanup with memory
cgroups enabled. The chain of events is as follows:
- zap_pte_range enables lazy MMU updates
- zap_pte_range eventually calls mem_cgroup_charge_statistics,
which accesses the vmalloc'd mem_cgroup per-cpu stat area
- vmalloc_fault is triggered which tries to sync the corresponding
PGD entry with set_pgd, but the update is deferred
- vmalloc_fault oopses due to a mismatch in the PUD entries
The OOPs usually looks as so:
------------[ cut here ]------------
kernel BUG at arch/x86/mm/fault.c:396!
invalid opcode: 0000 [#1] SMP
.. snip ..
CPU 1
Pid: 10866, comm: httpd Not tainted 3.6.10-4.fc18.x86_64 #1
RIP: e030:[<ffffffff816271bf>] [<ffffffff816271bf>] vmalloc_fault+0x11f/0x208
.. snip ..
Call Trace:
[<ffffffff81627759>] do_page_fault+0x399/0x4b0
[<ffffffff81004f4c>] ? xen_mc_extend_args+0xec/0x110
[<ffffffff81624065>] page_fault+0x25/0x30
[<ffffffff81184d03>] ? mem_cgroup_charge_statistics.isra.13+0x13/0x50
[<ffffffff81186f78>] __mem_cgroup_uncharge_common+0xd8/0x350
[<ffffffff8118aac7>] mem_cgroup_uncharge_page+0x57/0x60
[<ffffffff8115fbc0>] page_remove_rmap+0xe0/0x150
[<ffffffff8115311a>] ? vm_normal_page+0x1a/0x80
[<ffffffff81153e61>] unmap_single_vma+0x531/0x870
[<ffffffff81154962>] unmap_vmas+0x52/0xa0
[<ffffffff81007442>] ? pte_mfn_to_pfn+0x72/0x100
[<ffffffff8115c8f8>] exit_mmap+0x98/0x170
[<ffffffff810050d9>] ? __raw_callee_save_xen_pmd_val+0x11/0x1e
[<ffffffff81059ce3>] mmput+0x83/0xf0
[<ffffffff810624c4>] exit_mm+0x104/0x130
[<ffffffff8106264a>] do_exit+0x15a/0x8c0
[<ffffffff810630ff>] do_group_exit+0x3f/0xa0
[<ffffffff81063177>] sys_exit_group+0x17/0x20
[<ffffffff8162bae9>] system_call_fastpath+0x16/0x1b
Calling arch_flush_lazy_mmu_mode immediately after set_pgd makes the
changes visible to the consistency checks.
RedHat-Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=914737
Tested-by: Josh Boyer <jwboyer@redhat.com>
Reported-and-Tested-by: Krishna Raman <kraman@redhat.com>
Signed-off-by: Samu Kallio <samu.kallio@aberdeencloud.com>
Link: http://lkml.kernel.org/r/1364045796-10720-1-git-send-email-konrad.wilk@oracle.com
Tested-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 0ee364eb316348ddf3e0dfcd986f5f13f528f821 upstream.
A user reported the following oops when a backup process reads
/proc/kcore:
BUG: unable to handle kernel paging request at ffffbb00ff33b000
IP: [<ffffffff8103157e>] kern_addr_valid+0xbe/0x110
[...]
Call Trace:
[<ffffffff811b8aaa>] read_kcore+0x17a/0x370
[<ffffffff811ad847>] proc_reg_read+0x77/0xc0
[<ffffffff81151687>] vfs_read+0xc7/0x130
[<ffffffff811517f3>] sys_read+0x53/0xa0
[<ffffffff81449692>] system_call_fastpath+0x16/0x1b
Investigation determined that the bug triggered when reading
system RAM at the 4G mark. On this system, that was the first
address using 1G pages for the virt->phys direct mapping so the
PUD is pointing to a physical address, not a PMD page.
The problem is that the page table walker in kern_addr_valid() is
not checking pud_large() and treats the physical address as if
it was a PMD. If it happens to look like pmd_none then it'll
silently fail, probably returning zeros instead of real data. If
the data happens to look like a present PMD though, it will be
walked resulting in the oops above.
This patch adds the necessary pud_large() check.
Unfortunately the problem was not readily reproducible and now
they are running the backup program without accessing
/proc/kcore so the patch has not been validated but I think it
makes sense.
Signed-off-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Rik van Riel <riel@redhat.coM>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: linux-mm@kvack.org
Link: http://lkml.kernel.org/r/20130211145236.GX21389@suse.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit e575a86fdc50d013bf3ad3aa81d9100e8e6cc60d upstream.
Without this patch, it is trivial to determine kernel page
mappings by examining the error code reported to dmesg[1].
Instead, declare the entire kernel memory space as a violation
of a present page.
Additionally, since show_unhandled_signals is enabled by
default, switch branch hinting to the more realistic
expectation, and unobfuscate the setting of the PF_PROT bit to
improve readability.
[1] http://vulnfactory.org/blog/2013/02/06/a-linux-memory-trick/
Reported-by: Dan Rosenberg <dan.j.rosenberg@gmail.com>
Suggested-by: Brad Spengler <spender@grsecurity.net>
Signed-off-by: Kees Cook <keescook@chromium.org>
Acked-by: H. Peter Anvin <hpa@zytor.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20130207174413.GA12485@www.outflux.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit f82f64dd9f485e13f29f369772d4a0e868e5633a upstream.
Commit
844ab6f9 x86, mm: Find_early_table_space based on ranges that are actually being mapped
added back some lines back wrongly that has been removed in commit
7b16bbf97 Revert "x86/mm: Fix the size calculation of mapping tables"
remove them again.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Link: http://lkml.kernel.org/r/CAE9FiQW_vuaYQbmagVnxT2DGsYc=9tNeAbdBq53sYkitPOwxSQ@mail.gmail.com
Acked-by: Jacob Shin <jacob.shin@amd.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit 844ab6f993b1d32eb40512503d35ff6ad0c57030 upstream.
Current logic finds enough space for direct mapping page tables from 0
to end. Instead, we only need to find enough space to cover mr[0].start
to mr[nr_range].end -- the range that is actually being mapped by
init_memory_mapping()
This is needed after 1bbbbe779aabe1f0768c2bf8f8c0a5583679b54a, to address
the panic reported here:
https://lkml.org/lkml/2012/10/20/160
https://lkml.org/lkml/2012/10/21/157
Signed-off-by: Jacob Shin <jacob.shin@amd.com>
Link: http://lkml.kernel.org/r/20121024195311.GB11779@jshin-Toonie
Tested-by: Tom Rini <trini@ti.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
[bwh: Backported to 3.2:
- Adjust context
- The log message format is a bit different]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit eb48c071464757414538c68a6033c8f8c15196f8 upstream.
Each page mapped in a process's address space must be correctly
accounted for in _mapcount. Normally the rules for this are
straightforward but hugetlbfs page table sharing is different. The page
table pages at the PMD level are reference counted while the mapcount
remains the same.
If this accounting is wrong, it causes bugs like this one reported by
Larry Woodman:
kernel BUG at mm/filemap.c:135!
invalid opcode: 0000 [#1] SMP
CPU 22
Modules linked in: bridge stp llc sunrpc binfmt_misc dcdbas microcode pcspkr acpi_pad acpi]
Pid: 18001, comm: mpitest Tainted: G W 3.3.0+ #4 Dell Inc. PowerEdge R620/07NDJ2
RIP: 0010:[<ffffffff8112cfed>] [<ffffffff8112cfed>] __delete_from_page_cache+0x15d/0x170
Process mpitest (pid: 18001, threadinfo ffff880428972000, task ffff880428b5cc20)
Call Trace:
delete_from_page_cache+0x40/0x80
truncate_hugepages+0x115/0x1f0
hugetlbfs_evict_inode+0x18/0x30
evict+0x9f/0x1b0
iput_final+0xe3/0x1e0
iput+0x3e/0x50
d_kill+0xf8/0x110
dput+0xe2/0x1b0
__fput+0x162/0x240
During fork(), copy_hugetlb_page_range() detects if huge_pte_alloc()
shared page tables with the check dst_pte == src_pte. The logic is if
the PMD page is the same, they must be shared. This assumes that the
sharing is between the parent and child. However, if the sharing is
with a different process entirely then this check fails as in this
diagram:
parent
|
------------>pmd
src_pte----------> data page
^
other--------->pmd--------------------|
^
child-----------|
dst_pte
For this situation to occur, it must be possible for Parent and Other to
have faulted and failed to share page tables with each other. This is
possible due to the following style of race.
PROC A PROC B
copy_hugetlb_page_range copy_hugetlb_page_range
src_pte == huge_pte_offset src_pte == huge_pte_offset
!src_pte so no sharing !src_pte so no sharing
(time passes)
hugetlb_fault hugetlb_fault
huge_pte_alloc huge_pte_alloc
huge_pmd_share huge_pmd_share
LOCK(i_mmap_mutex)
find nothing, no sharing
UNLOCK(i_mmap_mutex)
LOCK(i_mmap_mutex)
find nothing, no sharing
UNLOCK(i_mmap_mutex)
pmd_alloc pmd_alloc
LOCK(instantiation_mutex)
fault
UNLOCK(instantiation_mutex)
LOCK(instantiation_mutex)
fault
UNLOCK(instantiation_mutex)
These two processes are not poing to the same data page but are not
sharing page tables because the opportunity was missed. When either
process later forks, the src_pte == dst pte is potentially insufficient.
As the check falls through, the wrong PTE information is copied in
(harmless but wrong) and the mapcount is bumped for a page mapped by a
shared page table leading to the BUG_ON.
This patch addresses the issue by moving pmd_alloc into huge_pmd_share
which guarantees that the shared pud is populated in the same critical
section as pmd. This also means that huge_pte_offset test in
huge_pmd_share is serialized correctly now which in turn means that the
success of the sharing will be higher as the racing tasks see the pud
and pmd populated together.
Race identified and changelog written mostly by Mel Gorman.
{akpm@linux-foundation.org: attempt to make the huge_pmd_share() comment comprehensible, clean up coding style]
Reported-by: Larry Woodman <lwoodman@redhat.com>
Tested-by: Larry Woodman <lwoodman@redhat.com>
Reviewed-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Reviewed-by: Rik van Riel <riel@redhat.com>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: Ken Chen <kenchen@google.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Hillf Danton <dhillf@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
|
|
commit cd298f60a2451a16e0f077404bf69b62ec868733 upstream.
In SRAT v1, we had 8bit proximity domain (PXM) fields; SRAT v2 provides
32bits for these. The new fields were reserved before.
According to the ACPI spec, the OS must disregrard reserved fields.
x86/x86-64 was rather inconsistent prior to this patch; it used 8 bits
for the pxm field in cpu_affinity, but 32 bits in mem_affinity.
This patch makes it consistent: Either use 8 bits consistently (SRAT
rev 1 or lower) or 32 bits (SRAT rev 2 or higher).
cc: x86@kernel.org
Signed-off-by: Kurt Garloff <kurt@garloff.de>
Signed-off-by: Len Brown <len.brown@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
|
|
commit 9af0c7a6fa860698d080481f24a342ba74b68982 upstream.
On x86_32 casting the unsigned int result of get_random_int() to
long may result in a negative value. On x86_32 the range of
mmap_rnd() therefore was -255 to 255. The 32bit mode on x86_64
used 0 to 255 as intended.
The bug was introduced by 675a081 ("x86: unify mmap_{32|64}.c")
in January 2008.
Signed-off-by: Ludwig Nussel <ludwig.nussel@suse.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: harvey.harrison@gmail.com
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/201111152246.pAFMklOB028527@wpaz5.hot.corp.google.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
|
|
With the 3.2-rc kernel, IOMMU 2M pages in KVM works. But when I tried
to use IOMMU 1GB pages in KVM, I encountered an oops and the 1GB page
failed to be used.
The root cause is that 1GB page allocation calls gup_huge_pud() while 2M
page calls gup_huge_pmd. If compound pages are used and the page is a
tail page, gup_huge_pmd() increases _mapcount to record tail page are
mapped while gup_huge_pud does not do that.
So when the mapped page is relesed, it will result in kernel oops
because the page is not marked mapped.
This patch add tail process for compound page in 1GB huge page which
keeps the same process as 2M page.
Reproduce like:
1. Add grub boot option: hugepagesz=1G hugepages=8
2. mount -t hugetlbfs -o pagesize=1G hugetlbfs /dev/hugepages
3. qemu-kvm -m 2048 -hda os-kvm.img -cpu kvm64 -smp 4 -mem-path /dev/hugepages
-net none -device pci-assign,host=07:00.1
kernel BUG at mm/swap.c:114!
invalid opcode: 0000 [#1] SMP
Call Trace:
put_page+0x15/0x37
kvm_release_pfn_clean+0x31/0x36
kvm_iommu_put_pages+0x94/0xb1
kvm_iommu_unmap_memslots+0x80/0xb6
kvm_assign_device+0xba/0x117
kvm_vm_ioctl_assigned_device+0x301/0xa47
kvm_vm_ioctl+0x36c/0x3a2
do_vfs_ioctl+0x49e/0x4e4
sys_ioctl+0x5a/0x7c
system_call_fastpath+0x16/0x1b
RIP put_compound_page+0xd4/0x168
Signed-off-by: Youquan Song <youquan.song@intel.com>
Reviewed-by: Andrea Arcangeli <aarcange@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
regardless of lazy_mmu mode
Fix an outstanding issue that has been reported since 2.6.37.
Under a heavy loaded machine processing "fork()" calls could
crash with:
BUG: unable to handle kernel paging request at f573fc8c
IP: [<c01abc54>] swap_count_continued+0x104/0x180
*pdpt = 000000002a3b9027 *pde = 0000000001bed067 *pte = 0000000000000000 Oops: 0000 [#1] SMP
Modules linked in:
Pid: 1638, comm: apache2 Not tainted 3.0.4-linode37 #1
EIP: 0061:[<c01abc54>] EFLAGS: 00210246 CPU: 3
EIP is at swap_count_continued+0x104/0x180
.. snip..
Call Trace:
[<c01ac222>] ? __swap_duplicate+0xc2/0x160
[<c01040f7>] ? pte_mfn_to_pfn+0x87/0xe0
[<c01ac2e4>] ? swap_duplicate+0x14/0x40
[<c01a0a6b>] ? copy_pte_range+0x45b/0x500
[<c01a0ca5>] ? copy_page_range+0x195/0x200
[<c01328c6>] ? dup_mmap+0x1c6/0x2c0
[<c0132cf8>] ? dup_mm+0xa8/0x130
[<c013376a>] ? copy_process+0x98a/0xb30
[<c013395f>] ? do_fork+0x4f/0x280
[<c01573b3>] ? getnstimeofday+0x43/0x100
[<c010f770>] ? sys_clone+0x30/0x40
[<c06c048d>] ? ptregs_clone+0x15/0x48
[<c06bfb71>] ? syscall_call+0x7/0xb
The problem is that in copy_page_range() we turn lazy mode on,
and then in swap_entry_free() we call swap_count_continued()
which ends up in:
map = kmap_atomic(page, KM_USER0) + offset;
and then later we touch *map.
Since we are running in batched mode (lazy) we don't actually
set up the PTE mappings and the kmap_atomic is not done
synchronously and ends up trying to dereference a page that has
not been set.
Looking at kmap_atomic_prot_pfn(), it uses
'arch_flush_lazy_mmu_mode' and doing the same in
kmap_atomic_prot() and __kunmap_atomic() makes the problem go
away.
Interestingly, commit b8bcfe997e4615 ("x86/paravirt: remove lazy
mode in interrupts") removed part of this to fix an interrupt
issue - but it went to far and did not consider this scenario.
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
|
|
This avoids duplicating the function in every arch gup_fast.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Hugh Dickins <hughd@google.com>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: David Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
Michel while working on the working set estimation code, noticed that
calling get_page_unless_zero() on a random pfn_to_page(random_pfn)
wasn't safe, if the pfn ended up being a tail page of a transparent
hugepage under splitting by __split_huge_page_refcount().
He then found the problem could also theoretically materialize with
page_cache_get_speculative() during the speculative radix tree lookups
that uses get_page_unless_zero() in SMP if the radix tree page is freed
and reallocated and get_user_pages is called on it before
page_cache_get_speculative has a chance to call get_page_unless_zero().
So the best way to fix the problem is to keep page_tail->_count zero at
all times. This will guarantee that get_page_unless_zero() can never
succeed on any tail page. page_tail->_mapcount is guaranteed zero and
is unused for all tail pages of a compound page, so we can simply
account the tail page references there and transfer them to
tail_page->_count in __split_huge_page_refcount() (in addition to the
head_page->_mapcount).
While debugging this s/_count/_mapcount/ change I also noticed get_page is
called by direct-io.c on pages returned by get_user_pages. That wasn't
entirely safe because the two atomic_inc in get_page weren't atomic. As
opposed to other get_user_page users like secondary-MMU page fault to
establish the shadow pagetables would never call any superflous get_page
after get_user_page returns. It's safer to make get_page universally safe
for tail pages and to use get_page_foll() within follow_page (inside
get_user_pages()). get_page_foll() is safe to do the refcounting for tail
pages without taking any locks because it is run within PT lock protected
critical sections (PT lock for pte and page_table_lock for
pmd_trans_huge).
The standard get_page() as invoked by direct-io instead will now take
the compound_lock but still only for tail pages. The direct-io paths
are usually I/O bound and the compound_lock is per THP so very
finegrined, so there's no risk of scalability issues with it. A simple
direct-io benchmarks with all lockdep prove locking and spinlock
debugging infrastructure enabled shows identical performance and no
overhead. So it's worth it. Ideally direct-io should stop calling
get_page() on pages returned by get_user_pages(). The spinlock in
get_page() is already optimized away for no-THP builds but doing
get_page() on tail pages returned by GUP is generally a rare operation
and usually only run in I/O paths.
This new refcounting on page_tail->_mapcount in addition to avoiding new
RCU critical sections will also allow the working set estimation code to
work without any further complexity associated to the tail page
refcounting with THP.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reported-by: Michel Lespinasse <walken@google.com>
Reviewed-by: Michel Lespinasse <walken@google.com>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Hugh Dickins <hughd@google.com>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: <stable@kernel.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
* 'x86-vdso-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86-64, doc: Remove int 0xcc from entry_64.S documentation
x86, vsyscall: Add missing <asm/fixmap.h> to arch/x86/mm/fault.c
Fix up trivial conflicts in arch/x86/mm/fault.c (asm/fixmap.h vs
asm/vsyscall.h: both work, which to use? Whatever..)
|
