A recent patch changed the format of a swap PTE.
The comment explaining the format of the swap PTE is wrong about
the bits used for the swap type field. Amusingly, the ASCII art
and the patch description are correct, but the comment itself
is wrong.
As I was looking at this, I also noticed that the
SWP_OFFSET_FIRST_BIT has an off-by-one error. This does not
really hurt anything. It just wasted a bit of space in the PTE,
giving us 2^59 bytes of addressable space in our swapfiles
instead of 2^60. But, it doesn't match with the comments, and it
wastes a bit of space, so fix it.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dave Hansen <dave@sr71.net>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Luis R. Rodriguez <mcgrof@suse.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Toshi Kani <toshi.kani@hp.com>
Fixes: 00839ee3b2 ("x86/mm: Move swap offset/type up in PTE to work around erratum")
Link: http://lkml.kernel.org/r/20160810172325.E56AD7DA@viggo.jf.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This erratum can result in Accessed/Dirty getting set by the hardware
when we do not expect them to be (on !Present PTEs).
Instead of trying to fix them up after this happens, we just
allow the bits to get set and try to ignore them. We do this by
shifting the layout of the bits we use for swap offset/type in
our 64-bit PTEs.
It looks like this:
bitnrs: | ... | 11| 10| 9|8|7|6|5| 4| 3|2|1|0|
names: | ... |SW3|SW2|SW1|G|L|D|A|CD|WT|U|W|P|
before: | OFFSET (9-63) |0|X|X| TYPE(1-5) |0|
after: | OFFSET (14-63) | TYPE (9-13) |0|X|X|X| X| X|X|X|0|
Note that D was already a don't care (X) even before. We just
move TYPE up and turn its old spot (which could be hit by the
A bit) into all don't cares.
We take 5 bits away from the offset, but that still leaves us
with 50 bits which lets us index into a 62-bit swapfile (4 EiB).
I think that's probably fine for the moment. We could
theoretically reclaim 5 of the bits (1, 2, 3, 4, 7) but it
doesn't gain us anything.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dave Hansen <dave@sr71.net>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Luis R. Rodriguez <mcgrof@suse.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Toshi Kani <toshi.kani@hp.com>
Cc: dave.hansen@intel.com
Cc: linux-mm@kvack.org
Cc: mhocko@suse.com
Link: http://lkml.kernel.org/r/20160708001911.9A3FD2B6@viggo.jf.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This patch removes the NUMA PTE bits and associated helpers. As a
side-effect it increases the maximum possible swap space on x86-64.
One potential source of problems is races between the marking of PTEs
PROT_NONE, NUMA hinting faults and migration. It must be guaranteed that
a PTE being protected is not faulted in parallel, seen as a pte_none and
corrupting memory. The base case is safe but transhuge has problems in
the past due to an different migration mechanism and a dependance on page
lock to serialise migrations and warrants a closer look.
task_work hinting update parallel fault
------------------------ --------------
change_pmd_range
change_huge_pmd
__pmd_trans_huge_lock
pmdp_get_and_clear
__handle_mm_fault
pmd_none
do_huge_pmd_anonymous_page
read? pmd_lock blocks until hinting complete, fail !pmd_none test
write? __do_huge_pmd_anonymous_page acquires pmd_lock, checks pmd_none
pmd_modify
set_pmd_at
task_work hinting update parallel migration
------------------------ ------------------
change_pmd_range
change_huge_pmd
__pmd_trans_huge_lock
pmdp_get_and_clear
__handle_mm_fault
do_huge_pmd_numa_page
migrate_misplaced_transhuge_page
pmd_lock waits for updates to complete, recheck pmd_same
pmd_modify
set_pmd_at
Both of those are safe and the case where a transhuge page is inserted
during a protection update is unchanged. The case where two processes try
migrating at the same time is unchanged by this series so should still be
ok. I could not find a case where we are accidentally depending on the
PTE not being cleared and flushed. If one is missed, it'll manifest as
corruption problems that start triggering shortly after this series is
merged and only happen when NUMA balancing is enabled.
Signed-off-by: Mel Gorman <mgorman@suse.de>
Tested-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Dave Jones <davej@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Kirill Shutemov <kirill.shutemov@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mark Brown <broonie@kernel.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull x86 mm updates from Ingo Molnar:
"This tree includes the following changes:
- fix memory hotplug
- fix hibernation bootup memory layout assumptions
- fix hyperv numa guest kernel messages
- remove dead code
- update documentation"
* 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/mm: Update memory map description to list hypervisor-reserved area
x86/mm, hibernate: Do not assume the first e820 area to be RAM
x86/mm/numa: Drop dead code and rename setup_node_data() to setup_alloc_data()
x86/mm/hotplug: Modify PGD entry when removing memory
x86/mm/hotplug: Pass sync_global_pgds() a correct argument in remove_pagetable()
x86: Remove set_pmd_pfn
When hot-adding/removing memory, sync_global_pgds() is called
for synchronizing PGD to PGD entries of all processes MM. But
when hot-removing memory, sync_global_pgds() does not work
correctly.
At first, sync_global_pgds() checks whether target PGD is none
or not. And if PGD is none, the PGD is skipped. But when
hot-removing memory, PGD may be none since PGD may be cleared by
free_pud_table(). So when sync_global_pgds() is called after
hot-removing memory, sync_global_pgds() should not skip PGD even
if the PGD is none. And sync_global_pgds() must clear PGD
entries of all processes MM.
Currently sync_global_pgds() does not clear PGD entries of all
processes MM when hot-removing memory. So when hot adding
memory which is same memory range as removed memory after
hot-removing memory, following call traces are shown:
kernel BUG at arch/x86/mm/init_64.c:206!
...
[<ffffffff815e0c80>] kernel_physical_mapping_init+0x1b2/0x1d2
[<ffffffff815ced94>] init_memory_mapping+0x1d4/0x380
[<ffffffff8104aebd>] arch_add_memory+0x3d/0xd0
[<ffffffff815d03d9>] add_memory+0xb9/0x1b0
[<ffffffff81352415>] acpi_memory_device_add+0x1af/0x28e
[<ffffffff81325dc4>] acpi_bus_device_attach+0x8c/0xf0
[<ffffffff813413b9>] acpi_ns_walk_namespace+0xc8/0x17f
[<ffffffff81325d38>] ? acpi_bus_type_and_status+0xb7/0xb7
[<ffffffff81325d38>] ? acpi_bus_type_and_status+0xb7/0xb7
[<ffffffff813418ed>] acpi_walk_namespace+0x95/0xc5
[<ffffffff81326b4c>] acpi_bus_scan+0x9a/0xc2
[<ffffffff81326bff>] acpi_scan_bus_device_check+0x8b/0x12e
[<ffffffff81326cb5>] acpi_scan_device_check+0x13/0x15
[<ffffffff81320122>] acpi_os_execute_deferred+0x25/0x32
[<ffffffff8107e02b>] process_one_work+0x17b/0x460
[<ffffffff8107edfb>] worker_thread+0x11b/0x400
[<ffffffff8107ece0>] ? rescuer_thread+0x400/0x400
[<ffffffff81085aef>] kthread+0xcf/0xe0
[<ffffffff81085a20>] ? kthread_create_on_node+0x140/0x140
[<ffffffff815fc76c>] ret_from_fork+0x7c/0xb0
[<ffffffff81085a20>] ? kthread_create_on_node+0x140/0x140
This patch clears PGD entries of all processes MM when
sync_global_pgds() is called after hot-removing memory
Signed-off-by: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Acked-by: Toshi Kani <toshi.kani@hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Tang Chen <tangchen@cn.fujitsu.com>
Cc: Gu Zheng <guz.fnst@cn.fujitsu.com>
Cc: Zhang Yanfei <zhangyanfei@cn.fujitsu.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
When RANDOMIZE_BASE (KASLR) is enabled; or the sum of all loaded
modules exceeds 512 MiB, then loading modules fails with a warning
(and hence a vmalloc allocation failure) because the PTEs for the
newly-allocated vmalloc address space are not zero.
WARNING: CPU: 0 PID: 494 at linux/mm/vmalloc.c:128
vmap_page_range_noflush+0x2a1/0x360()
This is caused by xen_setup_kernel_pagetables() copying
level2_kernel_pgt into level2_fixmap_pgt, overwriting many non-present
entries.
Without KASLR, the normal kernel image size only covers the first half
of level2_kernel_pgt and module space starts after that.
L4[511]->level3_kernel_pgt[510]->level2_kernel_pgt[ 0..255]->kernel
[256..511]->module
[511]->level2_fixmap_pgt[ 0..505]->module
This allows 512 MiB of of module vmalloc space to be used before
having to use the corrupted level2_fixmap_pgt entries.
With KASLR enabled, the kernel image uses the full PUD range of 1G and
module space starts in the level2_fixmap_pgt. So basically:
L4[511]->level3_kernel_pgt[510]->level2_kernel_pgt[0..511]->kernel
[511]->level2_fixmap_pgt[0..505]->module
And now no module vmalloc space can be used without using the corrupt
level2_fixmap_pgt entries.
Fix this by properly converting the level2_fixmap_pgt entries to MFNs,
and setting level1_fixmap_pgt as read-only.
A number of comments were also using the the wrong L3 offset for
level2_kernel_pgt. These have been corrected.
Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: David Vrabel <david.vrabel@citrix.com>
Reviewed-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: stable@vger.kernel.org
_PAGE_NUMA is currently an alias of _PROT_PROTNONE to trap NUMA hinting
faults on x86. Care is taken such that _PAGE_NUMA is used only in
situations where the VMA flags distinguish between NUMA hinting faults
and prot_none faults. This decision was x86-specific and conceptually
it is difficult requiring special casing to distinguish between PROTNONE
and NUMA ptes based on context.
Fundamentally, we only need the _PAGE_NUMA bit to tell the difference
between an entry that is really unmapped and a page that is protected
for NUMA hinting faults as if the PTE is not present then a fault will
be trapped.
Swap PTEs on x86-64 use the bits after _PAGE_GLOBAL for the offset.
This patch shrinks the maximum possible swap size and uses the bit to
uniquely distinguish between NUMA hinting ptes and swap ptes.
Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: David Vrabel <david.vrabel@citrix.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Peter Anvin <hpa@zytor.com>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Steven Noonan <steven@uplinklabs.net>
Cc: Rik van Riel <riel@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Cyrill Gorcunov <gorcunov@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull x86 mm changes from Peter Anvin:
"This is a huge set of several partly interrelated (and concurrently
developed) changes, which is why the branch history is messier than
one would like.
The *really* big items are two humonguous patchsets mostly developed
by Yinghai Lu at my request, which completely revamps the way we
create initial page tables. In particular, rather than estimating how
much memory we will need for page tables and then build them into that
memory -- a calculation that has shown to be incredibly fragile -- we
now build them (on 64 bits) with the aid of a "pseudo-linear mode" --
a #PF handler which creates temporary page tables on demand.
This has several advantages:
1. It makes it much easier to support things that need access to data
very early (a followon patchset uses this to load microcode way
early in the kernel startup).
2. It allows the kernel and all the kernel data objects to be invoked
from above the 4 GB limit. This allows kdump to work on very large
systems.
3. It greatly reduces the difference between Xen and native (Xen's
equivalent of the #PF handler are the temporary page tables created
by the domain builder), eliminating a bunch of fragile hooks.
The patch series also gets us a bit closer to W^X.
Additional work in this pull is the 64-bit get_user() work which you
were also involved with, and a bunch of cleanups/speedups to
__phys_addr()/__pa()."
* 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (105 commits)
x86, mm: Move reserving low memory later in initialization
x86, doc: Clarify the use of asm("%edx") in uaccess.h
x86, mm: Redesign get_user with a __builtin_choose_expr hack
x86: Be consistent with data size in getuser.S
x86, mm: Use a bitfield to mask nuisance get_user() warnings
x86/kvm: Fix compile warning in kvm_register_steal_time()
x86-32: Add support for 64bit get_user()
x86-32, mm: Remove reference to alloc_remap()
x86-32, mm: Remove reference to resume_map_numa_kva()
x86-32, mm: Rip out x86_32 NUMA remapping code
x86/numa: Use __pa_nodebug() instead
x86: Don't panic if can not alloc buffer for swiotlb
mm: Add alloc_bootmem_low_pages_nopanic()
x86, 64bit, mm: hibernate use generic mapping_init
x86, 64bit, mm: Mark data/bss/brk to nx
x86: Merge early kernel reserve for 32bit and 64bit
x86: Add Crash kernel low reservation
x86, kdump: Remove crashkernel range find limit for 64bit
memblock: Add memblock_mem_size()
x86, boot: Not need to check setup_header version for setup_data
...
Converting macros to functions unhide type problems before
changes will be integrated and trigger problems on other
architectures.
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This patch is meant to clean-up the fact that we have several functions in
page_64_types.h which really don't belong there. I found this issue when I
had tried to replace __phys_addr with an inline function. It resulted in the
realmode bits generating compile warnings about types. In order to resolve
that I am relocating the address translation to page_64.h since this is in
keeping with where these functions are located in 32 bit.
In addtion I have relocated several functions defined in init_64.c to
pgtable_64.h as this seems to be where most of the functions related to
memory initialization were already located.
[ hpa: added missing #include <asm/pgtable.h> to apic_numachip.c,
as reported by Yinghai Lu. ]
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Link: http://lkml.kernel.org/r/20121116215244.8521.31505.stgit@ahduyck-cp1.jf.intel.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Daniel J Blueman <daniel@numascale-asia.com>
The transparent huge page code passes a PMD pointer in as the third
argument of update_mmu_cache(), which expects a PTE pointer.
This never got noticed because X86 implements update_mmu_cache() as a
macro and thus we don't get any type checking, and X86 is the only
architecture which supports transparent huge pages currently.
Before other architectures can support transparent huge pages properly we
need to add a new interface which will take a PMD pointer as the third
argument rather than a PTE pointer.
[akpm@linux-foundation.org: implement update_mm_cache_pmd() for s390]
Signed-off-by: David S. Miller <davem@davemloft.net>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Use a more current logging style:
- Bare printks should have a KERN_<LEVEL> for consistency's sake
- Add pr_fmt where appropriate
- Neaten some macro definitions
- Convert some Ok output to OK
- Use "%s: ", __func__ in pr_fmt for summit
- Convert some printks to pr_<level>
Message output is not identical in all cases.
Signed-off-by: Joe Perches <joe@perches.com>
Cc: levinsasha928@gmail.com
Link: http://lkml.kernel.org/r/1337655007.24226.10.camel@joe2Laptop
[ merged two similar patches, tidied up the changelog ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Add support for transparent hugepages to x86 32bit.
Share the same VM_ bitflag for VM_MAPPED_COPY. mm/nommu.c will never
support transparent hugepages.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Reviewed-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Lately I've been working to make KVM use hugepages transparently without
the usual restrictions of hugetlbfs. Some of the restrictions I'd like to
see removed:
1) hugepages have to be swappable or the guest physical memory remains
locked in RAM and can't be paged out to swap
2) if a hugepage allocation fails, regular pages should be allocated
instead and mixed in the same vma without any failure and without
userland noticing
3) if some task quits and more hugepages become available in the
buddy, guest physical memory backed by regular pages should be
relocated on hugepages automatically in regions under
madvise(MADV_HUGEPAGE) (ideally event driven by waking up the
kernel deamon if the order=HPAGE_PMD_SHIFT-PAGE_SHIFT list becomes
not null)
4) avoidance of reservation and maximization of use of hugepages whenever
possible. Reservation (needed to avoid runtime fatal faliures) may be ok for
1 machine with 1 database with 1 database cache with 1 database cache size
known at boot time. It's definitely not feasible with a virtualization
hypervisor usage like RHEV-H that runs an unknown number of virtual machines
with an unknown size of each virtual machine with an unknown amount of
pagecache that could be potentially useful in the host for guest not using
O_DIRECT (aka cache=off).
hugepages in the virtualization hypervisor (and also in the guest!) are
much more important than in a regular host not using virtualization,
becasue with NPT/EPT they decrease the tlb-miss cacheline accesses from 24
to 19 in case only the hypervisor uses transparent hugepages, and they
decrease the tlb-miss cacheline accesses from 19 to 15 in case both the
linux hypervisor and the linux guest both uses this patch (though the
guest will limit the addition speedup to anonymous regions only for
now...). Even more important is that the tlb miss handler is much slower
on a NPT/EPT guest than for a regular shadow paging or no-virtualization
scenario. So maximizing the amount of virtual memory cached by the TLB
pays off significantly more with NPT/EPT than without (even if there would
be no significant speedup in the tlb-miss runtime).
The first (and more tedious) part of this work requires allowing the VM to
handle anonymous hugepages mixed with regular pages transparently on
regular anonymous vmas. This is what this patch tries to achieve in the
least intrusive possible way. We want hugepages and hugetlb to be used in
a way so that all applications can benefit without changes (as usual we
leverage the KVM virtualization design: by improving the Linux VM at
large, KVM gets the performance boost too).
The most important design choice is: always fallback to 4k allocation if
the hugepage allocation fails! This is the _very_ opposite of some large
pagecache patches that failed with -EIO back then if a 64k (or similar)
allocation failed...
Second important decision (to reduce the impact of the feature on the
existing pagetable handling code) is that at any time we can split an
hugepage into 512 regular pages and it has to be done with an operation
that can't fail. This way the reliability of the swapping isn't decreased
(no need to allocate memory when we are short on memory to swap) and it's
trivial to plug a split_huge_page* one-liner where needed without
polluting the VM. Over time we can teach mprotect, mremap and friends to
handle pmd_trans_huge natively without calling split_huge_page*. The fact
it can't fail isn't just for swap: if split_huge_page would return -ENOMEM
(instead of the current void) we'd need to rollback the mprotect from the
middle of it (ideally including undoing the split_vma) which would be a
big change and in the very wrong direction (it'd likely be simpler not to
call split_huge_page at all and to teach mprotect and friends to handle
hugepages instead of rolling them back from the middle). In short the
very value of split_huge_page is that it can't fail.
The collapsing and madvise(MADV_HUGEPAGE) part will remain separated and
incremental and it'll just be an "harmless" addition later if this initial
part is agreed upon. It also should be noted that locking-wise replacing
regular pages with hugepages is going to be very easy if compared to what
I'm doing below in split_huge_page, as it will only happen when
page_count(page) matches page_mapcount(page) if we can take the PG_lock
and mmap_sem in write mode. collapse_huge_page will be a "best effort"
that (unlike split_huge_page) can fail at the minimal sign of trouble and
we can try again later. collapse_huge_page will be similar to how KSM
works and the madvise(MADV_HUGEPAGE) will work similar to
madvise(MADV_MERGEABLE).
The default I like is that transparent hugepages are used at page fault
time. This can be changed with
/sys/kernel/mm/transparent_hugepage/enabled. The control knob can be set
to three values "always", "madvise", "never" which mean respectively that
hugepages are always used, or only inside madvise(MADV_HUGEPAGE) regions,
or never used. /sys/kernel/mm/transparent_hugepage/defrag instead
controls if the hugepage allocation should defrag memory aggressively
"always", only inside "madvise" regions, or "never".
The pmd_trans_splitting/pmd_trans_huge locking is very solid. The
put_page (from get_user_page users that can't use mmu notifier like
O_DIRECT) that runs against a __split_huge_page_refcount instead was a
pain to serialize in a way that would result always in a coherent page
count for both tail and head. I think my locking solution with a
compound_lock taken only after the page_first is valid and is still a
PageHead should be safe but it surely needs review from SMP race point of
view. In short there is no current existing way to serialize the O_DIRECT
final put_page against split_huge_page_refcount so I had to invent a new
one (O_DIRECT loses knowledge on the mapping status by the time gup_fast
returns so...). And I didn't want to impact all gup/gup_fast users for
now, maybe if we change the gup interface substantially we can avoid this
locking, I admit I didn't think too much about it because changing the gup
unpinning interface would be invasive.
If we ignored O_DIRECT we could stick to the existing compound refcounting
code, by simply adding a get_user_pages_fast_flags(foll_flags) where KVM
(and any other mmu notifier user) would call it without FOLL_GET (and if
FOLL_GET isn't set we'd just BUG_ON if nobody registered itself in the
current task mmu notifier list yet). But O_DIRECT is fundamental for
decent performance of virtualized I/O on fast storage so we can't avoid it
to solve the race of put_page against split_huge_page_refcount to achieve
a complete hugepage feature for KVM.
Swap and oom works fine (well just like with regular pages ;). MMU
notifier is handled transparently too, with the exception of the young bit
on the pmd, that didn't have a range check but I think KVM will be fine
because the whole point of hugepages is that EPT/NPT will also use a huge
pmd when they notice gup returns pages with PageCompound set, so they
won't care of a range and there's just the pmd young bit to check in that
case.
NOTE: in some cases if the L2 cache is small, this may slowdown and waste
memory during COWs because 4M of memory are accessed in a single fault
instead of 8k (the payoff is that after COW the program can run faster).
So we might want to switch the copy_huge_page (and clear_huge_page too) to
not temporal stores. I also extensively researched ways to avoid this
cache trashing with a full prefault logic that would cow in 8k/16k/32k/64k
up to 1M (I can send those patches that fully implemented prefault) but I
concluded they're not worth it and they add an huge additional complexity
and they remove all tlb benefits until the full hugepage has been faulted
in, to save a little bit of memory and some cache during app startup, but
they still don't improve substantially the cache-trashing during startup
if the prefault happens in >4k chunks. One reason is that those 4k pte
entries copied are still mapped on a perfectly cache-colored hugepage, so
the trashing is the worst one can generate in those copies (cow of 4k page
copies aren't so well colored so they trashes less, but again this results
in software running faster after the page fault). Those prefault patches
allowed things like a pte where post-cow pages were local 4k regular anon
pages and the not-yet-cowed pte entries were pointing in the middle of
some hugepage mapped read-only. If it doesn't payoff substantially with
todays hardware it will payoff even less in the future with larger l2
caches, and the prefault logic would blot the VM a lot. If one is
emebdded transparent_hugepage can be disabled during boot with sysfs or
with the boot commandline parameter transparent_hugepage=0 (or
transparent_hugepage=2 to restrict hugepages inside madvise regions) that
will ensure not a single hugepage is allocated at boot time. It is simple
enough to just disable transparent hugepage globally and let transparent
hugepages be allocated selectively by applications in the MADV_HUGEPAGE
region (both at page fault time, and if enabled with the
collapse_huge_page too through the kernel daemon).
This patch supports only hugepages mapped in the pmd, archs that have
smaller hugepages will not fit in this patch alone. Also some archs like
power have certain tlb limits that prevents mixing different page size in
the same regions so they will not fit in this framework that requires
"graceful fallback" to basic PAGE_SIZE in case of physical memory
fragmentation. hugetlbfs remains a perfect fit for those because its
software limits happen to match the hardware limits. hugetlbfs also
remains a perfect fit for hugepage sizes like 1GByte that cannot be hoped
to be found not fragmented after a certain system uptime and that would be
very expensive to defragment with relocation, so requiring reservation.
hugetlbfs is the "reservation way", the point of transparent hugepages is
not to have any reservation at all and maximizing the use of cache and
hugepages at all times automatically.
Some performance result:
vmx andrea # LD_PRELOAD=/usr/lib64/libhugetlbfs.so HUGETLB_MORECORE=yes HUGETLB_PATH=/mnt/huge/ ./largep
ages3
memset page fault 1566023
memset tlb miss 453854
memset second tlb miss 453321
random access tlb miss 41635
random access second tlb miss 41658
vmx andrea # LD_PRELOAD=/usr/lib64/libhugetlbfs.so HUGETLB_MORECORE=yes HUGETLB_PATH=/mnt/huge/ ./largepages3
memset page fault 1566471
memset tlb miss 453375
memset second tlb miss 453320
random access tlb miss 41636
random access second tlb miss 41637
vmx andrea # ./largepages3
memset page fault 1566642
memset tlb miss 453417
memset second tlb miss 453313
random access tlb miss 41630
random access second tlb miss 41647
vmx andrea # ./largepages3
memset page fault 1566872
memset tlb miss 453418
memset second tlb miss 453315
random access tlb miss 41618
random access second tlb miss 41659
vmx andrea # echo 0 > /proc/sys/vm/transparent_hugepage
vmx andrea # ./largepages3
memset page fault 2182476
memset tlb miss 460305
memset second tlb miss 460179
random access tlb miss 44483
random access second tlb miss 44186
vmx andrea # ./largepages3
memset page fault 2182791
memset tlb miss 460742
memset second tlb miss 459962
random access tlb miss 43981
random access second tlb miss 43988
============
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/time.h>
#define SIZE (3UL*1024*1024*1024)
int main()
{
char *p = malloc(SIZE), *p2;
struct timeval before, after;
gettimeofday(&before, NULL);
memset(p, 0, SIZE);
gettimeofday(&after, NULL);
printf("memset page fault %Lu\n",
(after.tv_sec-before.tv_sec)*1000000UL +
after.tv_usec-before.tv_usec);
gettimeofday(&before, NULL);
memset(p, 0, SIZE);
gettimeofday(&after, NULL);
printf("memset tlb miss %Lu\n",
(after.tv_sec-before.tv_sec)*1000000UL +
after.tv_usec-before.tv_usec);
gettimeofday(&before, NULL);
memset(p, 0, SIZE);
gettimeofday(&after, NULL);
printf("memset second tlb miss %Lu\n",
(after.tv_sec-before.tv_sec)*1000000UL +
after.tv_usec-before.tv_usec);
gettimeofday(&before, NULL);
for (p2 = p; p2 < p+SIZE; p2 += 4096)
*p2 = 0;
gettimeofday(&after, NULL);
printf("random access tlb miss %Lu\n",
(after.tv_sec-before.tv_sec)*1000000UL +
after.tv_usec-before.tv_usec);
gettimeofday(&before, NULL);
for (p2 = p; p2 < p+SIZE; p2 += 4096)
*p2 = 0;
gettimeofday(&after, NULL);
printf("random access second tlb miss %Lu\n",
(after.tv_sec-before.tv_sec)*1000000UL +
after.tv_usec-before.tv_usec);
return 0;
}
============
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add needed pmd mangling functions with symmetry with their pte
counterparts. pmdp_splitting_flush() is the only new addition on the pmd_
methods and it's needed to serialize the VM against split_huge_page. It
simply atomically sets the splitting bit in a similar way
pmdp_clear_flush_young atomically clears the accessed bit.
pmdp_splitting_flush() also has to flush the tlb to make it effective
against gup_fast, but it wouldn't really require to flush the tlb too.
Just the tlb flush is the simplest operation we can invoke to serialize
pmdp_splitting_flush() against gup_fast.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
These returns 0 at compile time when the config option is disabled, to
allow gcc to eliminate the transparent hugepage function calls at compile
time without additional #ifdefs (only the export of those functions have
to be visible to gcc but they won't be required at link time and
huge_memory.o can be not built at all).
_PAGE_BIT_UNUSED1 is never used for pmd, only on pte.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Since we no longer need to provide KM_type, the whole pte_*map_nested()
API is now redundant, remove it.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Chris Metcalf <cmetcalf@tilera.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: David Miller <davem@davemloft.net>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
No behavior change.
Move some of vmalloc_sync_all() code into a new function
sync_global_pgds() that will be useful for memory hotplug.
Signed-off-by: Haicheng Li <haicheng.li@linux.intel.com>
LKML-Reference: <4C6E4ECD.1090607@linux.intel.com>
Reviewed-by: Wu Fengguang <fengguang.wu@intel.com>
Reviewed-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
No real bugs, just some dead code and some fixups.
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
On VIVT ARM, when we have multiple shared mappings of the same file
in the same MM, we need to ensure that we have coherency across all
copies. We do this via make_coherent() by making the pages
uncacheable.
This used to work fine, until we allowed highmem with highpte - we
now have a page table which is mapped as required, and is not available
for modification via update_mmu_cache().
Ralf Beache suggested getting rid of the PTE value passed to
update_mmu_cache():
On MIPS update_mmu_cache() calls __update_tlb() which walks pagetables
to construct a pointer to the pte again. Passing a pte_t * is much
more elegant. Maybe we might even replace the pte argument with the
pte_t?
Ben Herrenschmidt would also like the pte pointer for PowerPC:
Passing the ptep in there is exactly what I want. I want that
-instead- of the PTE value, because I have issue on some ppc cases,
for I$/D$ coherency, where set_pte_at() may decide to mask out the
_PAGE_EXEC.
So, pass in the mapped page table pointer into update_mmu_cache(), and
remove the PTE value, updating all implementations and call sites to
suit.
Includes a fix from Stephen Rothwell:
sparc: fix fallout from update_mmu_cache API change
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Acked-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
The discussion about using "access_ok()" in get_user_pages_fast() (see
commit 7f81890687: "x86: don't use
'access_ok()' as a range check in get_user_pages_fast()" for details and
end result), made us notice that x86-64 was really being very sloppy
about virtual address checking.
So be way more careful and straightforward about masking x86-64 virtual
addresses:
- All the VIRTUAL_MASK* variants now cover half of the address
space, it's not like we can use the full mask on a signed
integer, and the larger mask just invites mistakes when
applying it to either half of the 48-bit address space.
- /proc/kcore's kc_offset_to_vaddr() becomes a lot more
obvious when it transforms a file offset into a
(kernel-half) virtual address.
- Unify/simplify the 32-bit and 64-bit USER_DS definition to
be based on TASK_SIZE_MAX.
This cleanup and more careful/obvious user virtual address checking also
uncovered a buglet in the x86-64 implementation of strnlen_user(): it
would do an "access_ok()" check on the whole potential area, even if the
string itself was much shorter, and thus return an error even for valid
strings. Our sloppy checking had hidden this.
So this fixes 'strnlen_user()' to do this properly, the same way we
already handled user strings in 'strncpy_from_user()'. Namely by just
checking the first byte, and then relying on fault handling for the
rest. That always works, since we impose a guard page that cannot be
mapped at the end of the user space address space (and even if we
didn't, we'd have the address space hole).
Acked-by: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Nick Piggin <npiggin@suse.de>
Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Impact: cleanup, no code changed
- syscalls.h update declarations due to unifications
- irq.c declare smp_generic_interrupt() before it gets used
- process.c declare sys_fork() and sys_vfork() before they get used
- tsc.c rename tsc_khz shadowed variable
- apic/probe_32.c declare apic_default before it gets used
- apic/nmi.c prev_nmi_count should be unsigned
- apic/io_apic.c declare smp_irq_move_cleanup_interrupt() before it gets used
- mm/init.c declare direct_gbpages and free_initrd_mem before they get used
Signed-off-by: Jaswinder Singh Rajput <jaswinder@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* commit 'remotes/tip/x86/paravirt': (175 commits)
xen: use direct ops on 64-bit
xen: make direct versions of irq_enable/disable/save/restore to common code
xen: setup percpu data pointers
xen: fix 32-bit build resulting from mmu move
x86/paravirt: return full 64-bit result
x86, percpu: fix kexec with vmlinux
x86/vmi: fix interrupt enable/disable/save/restore calling convention.
x86/paravirt: don't restore second return reg
xen: setup percpu data pointers
x86: split loading percpu segments from loading gdt
x86: pass in cpu number to switch_to_new_gdt()
x86: UV fix uv_flush_send_and_wait()
x86/paravirt: fix missing callee-save call on pud_val
x86/paravirt: use callee-saved convention for pte_val/make_pte/etc
x86/paravirt: implement PVOP_CALL macros for callee-save functions
x86/paravirt: add register-saving thunks to reduce caller register pressure
x86/paravirt: selectively save/restore regs around pvops calls
x86: fix paravirt clobber in entry_64.S
x86/pvops: add a paravirt_ident functions to allow special patching
xen: move remaining mmu-related stuff into mmu.c
...
Conflicts:
arch/x86/mach-voyager/voyager_smp.c
arch/x86/mm/fault.c
Impact: cleanup
Unify pmd_pfn. Unfortunately it can't be demacroed because it has a
cyclic dependency on linux/mm.h:page_to_nid().
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
