linux.git/mm/vmalloc.c, branch v5.18

mm/vmalloc: huge vmalloc backing pages should be split rather than compound

2022-04-22T16:20:16+00:00

Huge vmalloc higher-order backing pages were allocated with __GFP_COMP
in order to allow the sub-pages to be refcounted by callers such as
"remap_vmalloc_page [sic]" (remap_vmalloc_range).

However a similar problem exists for other struct page fields callers
use, for example fb_deferred_io_fault() takes a vmalloc'ed page and
not only refcounts it but uses ->lru, ->mapping, ->index.

This is not compatible with compound sub-pages, and can cause bad page
state issues like

  BUG: Bad page state in process swapper/0  pfn:00743
  page:(____ptrval____) refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x743
  flags: 0x7ffff000000000(node=0|zone=0|lastcpupid=0x7ffff)
  raw: 007ffff000000000 c00c00000001d0c8 c00c00000001d0c8 0000000000000000
  raw: 0000000000000000 0000000000000000 00000000ffffffff 0000000000000000
  page dumped because: corrupted mapping in tail page
  Modules linked in:
  CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.18.0-rc3-00082-gfc6fff4a7ce1-dirty #2810
  Call Trace:
    dump_stack_lvl+0x74/0xa8 (unreliable)
    bad_page+0x12c/0x170
    free_tail_pages_check+0xe8/0x190
    free_pcp_prepare+0x31c/0x4e0
    free_unref_page+0x40/0x1b0
    __vunmap+0x1d8/0x420
    ...

The correct approach is to use split high-order pages for the huge
vmalloc backing. These allow callers to treat them in exactly the same
way as individually-allocated order-0 pages.

Link: https://lore.kernel.org/all/14444103-d51b-0fb3-ee63-c3f182f0b546@molgen.mpg.de/
Signed-off-by: Nicholas Piggin 
Cc: Paul Menzel 
Cc: Song Liu 
Cc: Rick Edgecombe 
Signed-off-by: Linus Torvalds

vmalloc: replace VM_NO_HUGE_VMAP with VM_ALLOW_HUGE_VMAP

2022-04-19T19:08:57+00:00

Huge page backed vmalloc memory could benefit performance in many cases.
However, some users of vmalloc may not be ready to handle huge pages for
various reasons: hardware constraints, potential pages split, etc.
VM_NO_HUGE_VMAP was introduced to allow vmalloc users to opt-out huge
pages.  However, it is not easy to track down all the users that require
the opt-out, as the allocation are passed different stacks and may cause
issues in different layers.

To address this issue, replace VM_NO_HUGE_VMAP with an opt-in flag,
VM_ALLOW_HUGE_VMAP, so that users that benefit from huge pages could ask
specificially.

Also, remove vmalloc_no_huge() and add opt-in helper vmalloc_huge().

Fixes: fac54e2bfb5b ("x86/Kconfig: Select HAVE_ARCH_HUGE_VMALLOC with HAVE_ARCH_HUGE_VMAP")
Link: https://lore.kernel.org/netdev/14444103-d51b-0fb3-ee63-c3f182f0b546@molgen.mpg.de/"
Reviewed-by: Christoph Hellwig 
Signed-off-by: Song Liu 
Reviewed-by: Rik van Riel 
Signed-off-by: Linus Torvalds

mm/vmalloc: fix spinning drain_vmap_work after reading from /proc/vmcore

2022-04-15T21:49:56+00:00

Commit 3ee48b6af49c ("mm, x86: Saving vmcore with non-lazy freeing of
vmas") introduced set_iounmap_nonlazy(), which sets vmap_lazy_nr to
lazy_max_pages() + 1, ensuring that any future vunmaps() immediately
purge the vmap areas instead of doing it lazily.

Commit 690467c81b1a ("mm/vmalloc: Move draining areas out of caller
context") moved the purging from the vunmap() caller to a worker thread.
Unfortunately, set_iounmap_nonlazy() can cause the worker thread to spin
(possibly forever).  For example, consider the following scenario:

 1. Thread reads from /proc/vmcore. This eventually calls
    __copy_oldmem_page() -> set_iounmap_nonlazy(), which sets
    vmap_lazy_nr to lazy_max_pages() + 1.

 2. Then it calls free_vmap_area_noflush() (via iounmap()), which adds 2
    pages (one page plus the guard page) to the purge list and
    vmap_lazy_nr. vmap_lazy_nr is now lazy_max_pages() + 3, so the
    drain_vmap_work is scheduled.

 3. Thread returns from the kernel and is scheduled out.

 4. Worker thread is scheduled in and calls drain_vmap_area_work(). It
    frees the 2 pages on the purge list. vmap_lazy_nr is now
    lazy_max_pages() + 1.

 5. This is still over the threshold, so it tries to purge areas again,
    but doesn't find anything.

 6. Repeat 5.

If the system is running with only one CPU (which is typicial for kdump)
and preemption is disabled, then this will never make forward progress:
there aren't any more pages to purge, so it hangs.  If there is more
than one CPU or preemption is enabled, then the worker thread will spin
forever in the background.  (Note that if there were already pages to be
purged at the time that set_iounmap_nonlazy() was called, this bug is
avoided.)

This can be reproduced with anything that reads from /proc/vmcore
multiple times.  E.g., vmcore-dmesg /proc/vmcore.

It turns out that improvements to vmap() over the years have obsoleted
the need for this "optimization".  I benchmarked `dd if=/proc/vmcore
of=/dev/null` with 4k and 1M read sizes on a system with a 32GB vmcore.
The test was run on 5.17, 5.18-rc1 with a fix that avoided the hang, and
5.18-rc1 with set_iounmap_nonlazy() removed entirely:

    |5.17  |5.18+fix|5.18+removal
  4k|40.86s|  40.09s|      26.73s
  1M|24.47s|  23.98s|      21.84s

The removal was the fastest (by a wide margin with 4k reads).  This
patch removes set_iounmap_nonlazy().

Link: https://lkml.kernel.org/r/52f819991051f9b865e9ce25605509bfdbacadcd.1649277321.git.osandov@fb.com
Fixes: 690467c81b1a  ("mm/vmalloc: Move draining areas out of caller context")
Signed-off-by: Omar Sandoval 
Acked-by: Chris Down 
Reviewed-by: Uladzislau Rezki (Sony) 
Reviewed-by: Christoph Hellwig 
Acked-by: Baoquan He 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

kasan, vmalloc: only tag normal vmalloc allocations

2022-03-25T02:06:48+00:00

The kernel can use to allocate executable memory.  The only supported
way to do that is via __vmalloc_node_range() with the executable bit set
in the prot argument.  (vmap() resets the bit via pgprot_nx()).

Once tag-based KASAN modes start tagging vmalloc allocations, executing
code from such allocations will lead to the PC register getting a tag,
which is not tolerated by the kernel.

Only tag the allocations for normal kernel pages.

[andreyknvl@google.com: pass KASAN_VMALLOC_PROT_NORMAL to kasan_unpoison_vmalloc()]
  Link: https://lkml.kernel.org/r/9230ca3d3e40ffca041c133a524191fd71969a8d.1646233925.git.andreyknvl@google.com
[andreyknvl@google.com: support tagged vmalloc mappings]
  Link: https://lkml.kernel.org/r/2f6605e3a358cf64d73a05710cb3da356886ad29.1646233925.git.andreyknvl@google.com
[andreyknvl@google.com: don't unintentionally disabled poisoning]
  Link: https://lkml.kernel.org/r/de4587d6a719232e83c760113e46ed2d4d8da61e.1646757322.git.andreyknvl@google.com

Link: https://lkml.kernel.org/r/fbfd9939a4dc375923c9a5c6b9e7ab05c26b8c6b.1643047180.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov 
Acked-by: Marco Elver 
Cc: Alexander Potapenko 
Cc: Andrey Ryabinin 
Cc: Catalin Marinas 
Cc: Dmitry Vyukov 
Cc: Evgenii Stepanov 
Cc: Mark Rutland 
Cc: Peter Collingbourne 
Cc: Vincenzo Frascino 
Cc: Will Deacon 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

kasan, vmalloc: add vmalloc tagging for HW_TAGS

2022-03-25T02:06:48+00:00

Add vmalloc tagging support to HW_TAGS KASAN.

The key difference between HW_TAGS and the other two KASAN modes when it
comes to vmalloc: HW_TAGS KASAN can only assign tags to physical memory.
The other two modes have shadow memory covering every mapped virtual
memory region.

Make __kasan_unpoison_vmalloc() for HW_TAGS KASAN:

 - Skip non-VM_ALLOC mappings as HW_TAGS KASAN can only tag a single
   mapping of normal physical memory; see the comment in the function.

 - Generate a random tag, tag the returned pointer and the allocation,
   and initialize the allocation at the same time.

 - Propagate the tag into the page stucts to allow accesses through
   page_address(vmalloc_to_page()).

The rest of vmalloc-related KASAN hooks are not needed:

 - The shadow-related ones are fully skipped.

 - __kasan_poison_vmalloc() is kept as a no-op with a comment.

Poisoning and zeroing of physical pages that are backing vmalloc()
allocations are skipped via __GFP_SKIP_KASAN_UNPOISON and
__GFP_SKIP_ZERO: __kasan_unpoison_vmalloc() does that instead.

Enabling CONFIG_KASAN_VMALLOC with HW_TAGS is not yet allowed.

Link: https://lkml.kernel.org/r/d19b2e9e59a9abc59d05b72dea8429dcaea739c6.1643047180.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov 
Co-developed-by: Vincenzo Frascino 
Signed-off-by: Vincenzo Frascino 
Acked-by: Marco Elver 
Cc: Alexander Potapenko 
Cc: Andrey Ryabinin 
Cc: Catalin Marinas 
Cc: Dmitry Vyukov 
Cc: Evgenii Stepanov 
Cc: Mark Rutland 
Cc: Peter Collingbourne 
Cc: Will Deacon 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

kasan, vmalloc: unpoison VM_ALLOC pages after mapping

2022-03-25T02:06:47+00:00

Make KASAN unpoison vmalloc mappings after they have been mapped in when
it's possible: for vmalloc() (indentified via VM_ALLOC) and vm_map_ram().

The reasons for this are:

 - For vmalloc() and vm_map_ram(): pages don't get unpoisoned in case
   mapping them fails.

 - For vmalloc(): HW_TAGS KASAN needs pages to be mapped to set tags via
   kasan_unpoison_vmalloc().

As a part of these changes, the return value of __vmalloc_node_range() is
changed to area->addr.  This is a non-functional change, as
__vmalloc_area_node() returns area->addr anyway.

Link: https://lkml.kernel.org/r/fcb98980e6fcd3c4be6acdcb5d6110898ef28548.1643047180.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov 
Reviewed-by: Alexander Potapenko 
Acked-by: Marco Elver 
Cc: Andrey Ryabinin 
Cc: Catalin Marinas 
Cc: Dmitry Vyukov 
Cc: Evgenii Stepanov 
Cc: Mark Rutland 
Cc: Peter Collingbourne 
Cc: Vincenzo Frascino 
Cc: Will Deacon 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

kasan, vmalloc, arm64: mark vmalloc mappings as pgprot_tagged

2022-03-25T02:06:47+00:00

HW_TAGS KASAN relies on ARM Memory Tagging Extension (MTE).  With MTE, a
memory region must be mapped as MT_NORMAL_TAGGED to allow setting memory
tags via MTE-specific instructions.

Add proper protection bits to vmalloc() allocations.  These allocations
are always backed by page_alloc pages, so the tags will actually be
getting set on the corresponding physical memory.

Link: https://lkml.kernel.org/r/983fc33542db2f6b1e77b34ca23448d4640bbb9e.1643047180.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov 
Co-developed-by: Vincenzo Frascino 
Signed-off-by: Vincenzo Frascino 
Acked-by: Marco Elver 
Cc: Alexander Potapenko 
Cc: Andrey Ryabinin 
Cc: Catalin Marinas 
Cc: Dmitry Vyukov 
Cc: Evgenii Stepanov 
Cc: Mark Rutland 
Cc: Peter Collingbourne 
Cc: Will Deacon 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

kasan, vmalloc: add vmalloc tagging for SW_TAGS

2022-03-25T02:06:47+00:00

Add vmalloc tagging support to SW_TAGS KASAN.

 - __kasan_unpoison_vmalloc() now assigns a random pointer tag, poisons
   the virtual mapping accordingly, and embeds the tag into the returned
   pointer.

 - __get_vm_area_node() (used by vmalloc() and vmap()) and
   pcpu_get_vm_areas() save the tagged pointer into vm_struct->addr
   (note: not into vmap_area->addr).

   This requires putting kasan_unpoison_vmalloc() after
   setup_vmalloc_vm[_locked](); otherwise the latter will overwrite the
   tagged pointer. The tagged pointer then is naturally propagateed to
   vmalloc() and vmap().

 - vm_map_ram() returns the tagged pointer directly.

As a result of this change, vm_struct->addr is now tagged.

Enabling KASAN_VMALLOC with SW_TAGS is not yet allowed.

Link: https://lkml.kernel.org/r/4a78f3c064ce905e9070c29733aca1dd254a74f1.1643047180.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov 
Acked-by: Marco Elver 
Cc: Alexander Potapenko 
Cc: Andrey Ryabinin 
Cc: Catalin Marinas 
Cc: Dmitry Vyukov 
Cc: Evgenii Stepanov 
Cc: Mark Rutland 
Cc: Peter Collingbourne 
Cc: Vincenzo Frascino 
Cc: Will Deacon 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

kasan, vmalloc: reset tags in vmalloc functions

2022-03-25T02:06:47+00:00

In preparation for adding vmalloc support to SW/HW_TAGS KASAN, reset
pointer tags in functions that use pointer values in range checks.

vread() is a special case here.  Despite the untagging of the addr pointer
in its prologue, the accesses performed by vread() are checked.

Instead of accessing the virtual mappings though addr directly, vread()
recovers the physical address via page_address(vmalloc_to_page()) and
acceses that.  And as page_address() recovers the pointer tag, the
accesses get checked.

Link: https://lkml.kernel.org/r/046003c5f683cacb0ba18e1079e9688bb3dca943.1643047180.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov 
Acked-by: Marco Elver 
Cc: Alexander Potapenko 
Cc: Andrey Ryabinin 
Cc: Catalin Marinas 
Cc: Dmitry Vyukov 
Cc: Evgenii Stepanov 
Cc: Mark Rutland 
Cc: Peter Collingbourne 
Cc: Vincenzo Frascino 
Cc: Will Deacon 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

kasan, x86, arm64, s390: rename functions for modules shadow

2022-03-25T02:06:47+00:00

Rename kasan_free_shadow to kasan_free_module_shadow and
kasan_module_alloc to kasan_alloc_module_shadow.

These functions are used to allocate/free shadow memory for kernel modules
when KASAN_VMALLOC is not enabled.  The new names better reflect their
purpose.

Also reword the comment next to their declaration to improve clarity.

Link: https://lkml.kernel.org/r/36db32bde765d5d0b856f77d2d806e838513fe84.1643047180.git.andreyknvl@google.com
Signed-off-by: Andrey Konovalov 
Acked-by: Catalin Marinas 
Acked-by: Marco Elver 
Cc: Alexander Potapenko 
Cc: Andrey Ryabinin 
Cc: Dmitry Vyukov 
Cc: Evgenii Stepanov 
Cc: Mark Rutland 
Cc: Peter Collingbourne 
Cc: Vincenzo Frascino 
Cc: Will Deacon 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds