Exchange between user and kernel memory is coded in assembly language.
Which means that such accesses won't be spotted by KASAN as a compiler
instruments only C code.
Add explicit KASAN checks to user memory access API to ensure that
userspace writes to (or reads from) a valid kernel memory.
Note: Unlike others strncpy_from_user() is written mostly in C and KASAN
sees memory accesses in it. However, it makes sense to add explicit
check for all @count bytes that *potentially* could be written to the
kernel.
[aryabinin@virtuozzo.com: move kasan check under the condition]
Link: http://lkml.kernel.org/r/1462869209-21096-1-git-send-email-aryabinin@virtuozzo.com
Link: http://lkml.kernel.org/r/1462538722-1574-4-git-send-email-aryabinin@virtuozzo.com
Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull x86 asm updates from Ingo Molnar:
"The main changes in this cycle were:
- MSR access API fixes and enhancements (Andy Lutomirski)
- early exception handling improvements (Andy Lutomirski)
- user-space FS/GS prctl usage fixes and improvements (Andy
Lutomirski)
- Remove the cpu_has_*() APIs and replace them with equivalents
(Borislav Petkov)
- task switch micro-optimization (Brian Gerst)
- 32-bit entry code simplification (Denys Vlasenko)
- enhance PAT handling in enumated CPUs (Toshi Kani)
... and lots of other cleanups/fixlets"
* 'x86-asm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (70 commits)
x86/arch_prctl/64: Restore accidentally removed put_cpu() in ARCH_SET_GS
x86/entry/32: Remove asmlinkage_protect()
x86/entry/32: Remove GET_THREAD_INFO() from entry code
x86/entry, sched/x86: Don't save/restore EFLAGS on task switch
x86/asm/entry/32: Simplify pushes of zeroed pt_regs->REGs
selftests/x86/ldt_gdt: Test set_thread_area() deletion of an active segment
x86/tls: Synchronize segment registers in set_thread_area()
x86/asm/64: Rename thread_struct's fs and gs to fsbase and gsbase
x86/arch_prctl/64: Remove FSBASE/GSBASE < 4G optimization
x86/segments/64: When load_gs_index fails, clear the base
x86/segments/64: When loadsegment(fs, ...) fails, clear the base
x86/asm: Make asm/alternative.h safe from assembly
x86/asm: Stop depending on ptrace.h in alternative.h
x86/entry: Rename is_{ia32,x32}_task() to in_{ia32,x32}_syscall()
x86/asm: Make sure verify_cpu() has a good stack
x86/extable: Add a comment about early exception handlers
x86/msr: Set the return value to zero when native_rdmsr_safe() fails
x86/paravirt: Make "unsafe" MSR accesses unsafe even if PARAVIRT=y
x86/paravirt: Add paravirt_{read,write}_msr()
x86/msr: Carry on after a non-"safe" MSR access fails
...
The x86 exception table sorting was changed in commit 29934b0fb8
("x86/extable: use generic search and sort routines") to use the arch
independent code in lib/extable.c. However, the patch was mangled
somehow on its way into the kernel from the last version posted at [1].
The committed version kind of attempted to incorporate the changes of
commit 548acf1923 ("x86/mm: Expand the exception table logic to allow
new handling options") as in _completely_ _ignoring_ the x86 specific
'handler' member of struct exception_table_entry. This effectively
broke the sorting as entries will only partly be swapped now.
Fortunately, the x86 Kconfig selects BUILDTIME_EXTABLE_SORT, so the
exception table doesn't need to be sorted at runtime. However, in case
that ever changes, we better not break the exception table sorting just
because of that.
[ Ard Biesheuvel points out that BUILDTIME_EXTABLE_SORT applies to the
core image only, but we still rely on the sorting routines for modules
in that case - Linus ]
Fix this by providing a swap_ex_entry_fixup() macro that takes care of
the 'handler' member.
[1] https://lkml.org/lkml/2016/1/27/232
Signed-off-by: Mathias Krause <minipli@googlemail.com>
Fixes: 29934b0fb8 ("x86/extable: use generic search and sort routines")
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Borislav Petkov <bp@suse.de>
Cc: H. Peter Anvin <hpa@linux.intel.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tony Luck <tony.luck@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This removes a bunch of assembly and adds some C code instead. It
changes the actual printouts on both 32-bit and 64-bit kernels, but
they still seem okay.
Tested-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: KVM list <kvm@vger.kernel.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: xen-devel <Xen-devel@lists.xen.org>
Link: http://lkml.kernel.org/r/4085070316fc3ab29538d3fcfe282648d1d4ee2e.1459605520.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
early_fixup_exception() is limited by the fact that it doesn't have a
real struct pt_regs. Change both the 32-bit and 64-bit asm and the
C code to pass and accept a real pt_regs.
Tested-by: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: KVM list <kvm@vger.kernel.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: xen-devel <Xen-devel@lists.xen.org>
Link: http://lkml.kernel.org/r/e3fb680fcfd5e23e38237e8328b64a25cc121d37.1459605520.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Replace the arch specific versions of search_extable() and
sort_extable() with calls to the generic ones, which now support
relative exception tables as well.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Acked-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Numerous 'call without frame pointer save/setup' warnings are introduced
by stacktool because of functions using the get_user() macro. Bad stack
traces could occur due to lack of or misplacement of stack frame setup
code.
This patch forces a stack frame to be created before the inline asm code
if CONFIG_FRAME_POINTER is enabled by listing the stack pointer as an
output operand for the get_user() inline assembly statement.
Signed-off-by: Chris J Arges <chris.j.arges@canonical.com>
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Reviewed-by: Borislav Petkov <bp@suse.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Bernd Petrovitsch <bernd@petrovitsch.priv.at>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Jiri Slaby <jslaby@suse.cz>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Michal Marek <mmarek@suse.cz>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Pedro Alves <palves@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: live-patching@vger.kernel.org
Link: http://lkml.kernel.org/r/bc85501f221ee512670797c7f110022e64b12c81.1453405861.git.jpoimboe@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Huge amounts of help from Andy Lutomirski and Borislav Petkov to
produce this. Andy provided the inspiration to add classes to the
exception table with a clever bit-squeezing trick, Boris pointed
out how much cleaner it would all be if we just had a new field.
Linus Torvalds blessed the expansion with:
' I'd rather not be clever in order to save just a tiny amount of space
in the exception table, which isn't really criticial for anybody. '
The third field is another relative function pointer, this one to a
handler that executes the actions.
We start out with three handlers:
1: Legacy - just jumps the to fixup IP
2: Fault - provide the trap number in %ax to the fixup code
3: Cleaned up legacy for the uaccess error hack
Signed-off-by: Tony Luck <tony.luck@intel.com>
Reviewed-by: Borislav Petkov <bp@suse.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/f6af78fcbd348cf4939875cfda9c19689b5e50b8.1455732970.git.tony.luck@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Expose an interface to allow users to mark several accesses together as
being user space accesses, allowing batching of the surrounding user
space access markers (SMAP on x86, PAN on arm64, domain register
switching on arm).
This is currently only used for the user string lenth and copying
functions, where the SMAP overhead on x86 drowned the actual user
accesses (only noticeable on newer microarchitectures that support SMAP
in the first place, of course).
* user access batching branch:
Use the new batched user accesses in generic user string handling
Add 'unsafe' user access functions for batched accesses
x86: reorganize SMAP handling in user space accesses
The naming is meant to discourage random use: the helper functions are
not really any more "unsafe" than the traditional double-underscore
functions (which need the address range checking), but they do need even
more infrastructure around them, and should not be used willy-nilly.
In addition to checking the access range, these user access functions
require that you wrap the user access with a "user_acess_{begin,end}()"
around it.
That allows architectures that implement kernel user access control
(x86: SMAP, arm64: PAN) to do the user access control in the wrapping
user_access_begin/end part, and then batch up the actual user space
accesses using the new interfaces.
The main (and hopefully only) use for these are for core generic access
helpers, initially just the generic user string functions
(strnlen_user() and strncpy_from_user()).
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This reorganizes how we do the stac/clac instructions in the user access
code. Instead of adding the instructions directly to the same inline
asm that does the actual user level access and exception handling, add
them at a higher level.
This is mainly preparation for the next step, where we will expose an
interface to allow users to mark several accesses together as being user
space accesses, but it does already clean up some code:
- the inlined trivial cases of copy_in_user() now do stac/clac just
once over the accesses: they used to do one pair around the user
space read, and another pair around the write-back.
- the {get,put}_user_ex() macros that are used with the catch/try
handling don't do any stac/clac at all, because that happens in the
try/catch surrounding them.
Other than those two cleanups that happened naturally from the
re-organization, this should not make any difference. Yet.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add a inlined __ variant of copy_from_user_nmi. The inlined variant allows
the user to:
- batch the access_ok() check for multiple accesses
- avoid having a pagefault_disable/enable() on every access if the
caller already ensures disabled page faults due to its context.
- get all the optimizations in copy_*_user() for small constant sized
transfers
It is just a define to __copy_from_user_inatomic().
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1445551641-13379-1-git-send-email-andi@firstfloor.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This should improve code quality a bit. It also shrinks the kernel text:
Before:
text data bss dec filename
21828379 5194760 1277952 28301091 vmlinux
After:
text data bss dec filename
21827997 5194760 1277952 28300709 vmlinux
... by 382 bytes.
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.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: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Link: http://lkml.kernel.org/r/f427b8002d932e5deab9055e0074bb4e7e80ee39.1444091584.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
GCC doesn't realize that get_user(), put_user(), and their __
variants are unlikely to fail. Tell it.
I noticed this while playing with the C entry code.
Before:
text data bss dec filename
21828763 5194760 1277952 28301475 vmlinux.baseline
After:
text data bss dec filename
21828379 5194760 1277952 28301091 vmlinux.new
The generated code shrunk by 384 bytes.
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.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: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Link: http://lkml.kernel.org/r/dc37bed7024319c3004d950d57151fca6aeacf97.1444091584.git.luto@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
virtio wants to read bitwise types from userspace using get_user. At the
moment this triggers sparse errors, since the value is passed through an
integer.
Fix that up using __force.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Pull x86 cpufeature and mpx updates from Peter Anvin:
"This includes the basic infrastructure for MPX (Memory Protection
Extensions) support, but does not include MPX support itself. It is,
however, a prerequisite for KVM support for MPX, which I believe will
be pushed later this merge window by the KVM team.
This includes moving the functionality in
futex_atomic_cmpxchg_inatomic() into a new function in uaccess.h so it
can be reused - this will be used by the final MPX patches.
The actual MPX functionality (map management and so on) will be pushed
in a future merge window, when ready"
* 'x86/mpx' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/intel/mpx: Remove unused LWP structure
x86, mpx: Add MPX related opcodes to the x86 opcode map
x86: replace futex_atomic_cmpxchg_inatomic() with user_atomic_cmpxchg_inatomic
x86: add user_atomic_cmpxchg_inatomic at uaccess.h
x86, xsave: Support eager-only xsave features, add MPX support
x86, cpufeature: Define the Intel MPX feature flag
gcc can under very specific circumstances realize that the code
sequence:
foo += bar;
if (foo < bar) ...
... is equivalent to a carry out from the addition. Tweak the
implementation of access_ok() (specifically __chk_range_not_ok()) to
make it more likely that gcc will make that connection. It isn't
fool-proof (sometimes gcc seems to think it can make better code with
lea, and ends up with a second comparison), still, but it seems to be
able to connect the two more frequently this way.
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/CA%2B55aFzPBdbfKovMT8Edr4SmE2_=%2BOKJFac9XW2awegogTkVTA@mail.gmail.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
It turns out that the assembly variant doesn't actually produce that
good code, presumably partly because it creates a long dependency
chain with no scheduling, and partly because we cannot get a flags
result out of gcc (which could be fixed with asm goto, but it turns
out not to be worth it.)
The C code allows gcc to schedule and generate multiple (easily
predictable) branches, and as a side benefit we can really optimize
the case where the size is constant.
Link: http://lkml.kernel.org/r/CA%2B55aFzPBdbfKovMT8Edr4SmE2_=%2BOKJFac9XW2awegogTkVTA@mail.gmail.com
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
This patch adds user_atomic_cmpxchg_inatomic() to use CMPXCHG
instruction against a user space address.
This generalizes the already existing futex_atomic_cmpxchg_inatomic()
so it can be used in other contexts. This will be used in the
upcoming support for Intel MPX (Memory Protection Extensions.)
[ hpa: replaced #ifdef inside a macro with IS_ENABLED() ]
Signed-off-by: Qiaowei Ren <qiaowei.ren@intel.com>
Link: http://lkml.kernel.org/r/1387002303-6620-1-git-send-email-qiaowei.ren@intel.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Similarly to copy_from_user(), where the range check is to
protect against kernel memory corruption, copy_to_user() can
benefit from such checking too: Here it protects against kernel
information leaks.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Cc: <arjan@linux.intel.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/5265059502000078000FC4F6@nat28.tlf.novell.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Commits 4a31276930 ("x86: Turn the
copy_from_user check into an (optional) compile time warning")
and 63312b6a6f ("x86: Add a
Kconfig option to turn the copy_from_user warnings into errors")
touched only the 32-bit variant of copy_from_user(), whereas the
original commit 9f0cf4adb6 ("x86:
Use __builtin_object_size() to validate the buffer size for
copy_from_user()") also added the same code to the 64-bit one.
Further the earlier conversion from an inline WARN() to the call
to copy_from_user_overflow() went a little too far: When the
number of bytes to be copied is not a constant (e.g. [looking at
3.11] in drivers/net/tun.c:__tun_chr_ioctl() or
drivers/pci/pcie/aer/aer_inject.c:aer_inject_write()), the
compiler will always have to keep the funtion call, and hence
there will always be a warning. By using __builtin_constant_p()
we can avoid this.
And then this slightly extends the effect of
CONFIG_DEBUG_STRICT_USER_COPY_CHECKS in that apart from
converting warnings to errors in the constant size case, it
retains the (possibly wrong) warnings in the non-constant size
case, such that if someone is prepared to get a few false
positives, (s)he'll be able to recover the current behavior
(except that these diagnostics now will never be converted to
errors).
Since the 32-bit variant (intentionally) didn't call
might_fault(), the unification results in this being called
twice now. Adding a suitable #ifdef would be the alternative if
that's a problem.
I'd like to point out though that with
__compiletime_object_size() being restricted to gcc before 4.6,
the whole construct is going to become more and more pointless
going forward. I would question however that commit
2fb0815c9e ("gcc4: disable
__compiletime_object_size for GCC 4.6+") was really necessary,
and instead this should have been dealt with as is done here
from the beginning.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/5265056D02000078000FC4F3@nat28.tlf.novell.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Update comment in uaccess.h to reflect the changes for clang support:
gcc only cares about the base register (most architectures don't
encode the size of the operation in the operands like x86 does, and so
it is treated effectively like a register number), whereas clang tries
to enforce the size -- but not for register pairs.
Link: http://lkml.kernel.org/r/1377803585-5913-3-git-send-email-dl9pf@gmx.de
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Cc: Jan-Simon Möller <dl9pf@gmx.de>
Clang does not support the "shortcut" we're taking here for gcc (see below).
The patch uses the macro _ASM_DX to do the job.
From arch/x86/include/asm/uaccess.h:
/*
* Careful: we have to cast the result to the type of the pointer
* for sign reasons.
*
* The use of %edx as the register specifier is a bit of a
* simplification, as gcc only cares about it as the starting point
* and not size: for a 64-bit value it will use %ecx:%edx on 32 bits
* (%ecx being the next register in gcc's x86 register sequence), and
* %rdx on 64 bits.
*/
[ hpa: I consider this a compatibility bug in clang as this reflects a
bit of a misunderstanding about how register strings are used by
gcc, but the workaround is straightforward and there is no
particular reason to not do it. ]
Signed-off-by: Jan-Simon Möller <dl9pf@gmx.de>
Link: http://lkml.kernel.org/r/1377803585-5913-3-git-send-email-dl9pf@gmx.de
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Put in a comment that explains that the use of asm("%edx") in
uaccess.h doesn't actually necessarily mean %edx alone.
Cc: Jamie Lokier <jamie@shareable.org>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: H. J. Lu <hjl.tools@gmail.com>
Link: http://lkml.kernel.org/r/511ACDFB.1050707@zytor.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Instead of using a bitfield, use an odd little trick using typeof,
__builtin_choose_expr, and sizeof. __builtin_choose_expr is
explicitly defined to not convert its type (its argument is required
to be a constant expression) so this should be well-defined.
The code is still not 100% preturbation-free versus the baseline
before 64-bit get_user(), but the differences seem to be very small,
mostly related to padding and to gcc deciding when to spill registers.
Cc: Jamie Lokier <jamie@shareable.org>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: H. J. Lu <hjl.tools@gmail.com>
Link: http://lkml.kernel.org/r/511A8922.6050908@zytor.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Even though it is never executed, gcc wants to warn for casting from
a large integer to a pointer. Furthermore, using a variable with
__typeof__() doesn't work because __typeof__ retains storage
specifiers (const, restrict, volatile).
However, we can declare a bitfield using sizeof(), which is legal
because sizeof() is a constant expression. This quiets the warning,
although the code generated isn't 100% identical from the baseline
before 96477b4 x86-32: Add support for 64bit get_user():
[x86-mb is baseline, x86-mm is this commit]
text data bss filename
113716147 15858380 35037184 tip.x86-mb/o.i386-allconfig/vmlinux
113716145 15858380 35037184 tip.x86-mm/o.i386-allconfig/vmlinux
12989837 3597944 12255232 tip.x86-mb/o.i386-modconfig/vmlinux
12989831 3597944 12255232 tip.x86-mm/o.i386-modconfig/vmlinux
1462784 237608 1401988 tip.x86-mb/o.i386-noconfig/vmlinux
1462837 237608 1401964 tip.x86-mm/o.i386-noconfig/vmlinux
7938994 553688 7639040 tip.x86-mb/o.i386-pae/vmlinux
7943136 557784 7639040 tip.x86-mm/o.i386-pae/vmlinux
7186126 510572 6574080 tip.x86-mb/o.i386/vmlinux
7186124 510572 6574080 tip.x86-mm/o.i386/vmlinux
103747269 33578856 65888256 tip.x86-mb/o.x86_64-allconfig/vmlinux
103746949 33578856 65888256 tip.x86-mm/o.x86_64-allconfig/vmlinux
12116695 11035832 20160512 tip.x86-mb/o.x86_64-modconfig/vmlinux
12116567 11035832 20160512 tip.x86-mm/o.x86_64-modconfig/vmlinux
1700790 380524 511808 tip.x86-mb/o.x86_64-noconfig/vmlinux
1700790 380524 511808 tip.x86-mm/o.x86_64-noconfig/vmlinux
12413612 1133376 1101824 tip.x86-mb/o.x86_64/vmlinux
12413484 1133376 1101824 tip.x86-mm/o.x86_64/vmlinux
Cc: Jamie Lokier <jamie@shareable.org>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20130209110031.GA17833@n2100.arm.linux.org.uk
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Implement __get_user_8() for x86-32. It will return the
64-bit result in edx:eax register pair, and ecx is used
to pass in the address and return the error value.
For consistency, change the register assignment for all
other __get_user_x() variants, so that address is passed in
ecx/rcx, the error value is returned in ecx/rcx, and eax/rax
contains the actual value.
[ hpa: I modified the patch so that it does NOT change the calling
conventions for the existing callsites, this also means that the code
is completely unchanged for 64 bits.
Instead, continue to use eax for address input/error output and use
the ecx:edx register pair for the output. ]
This is a partial refresh of a patch [1] by Jamie Lokier from
2004. Only the minimal changes to implement 64bit get_user()
were picked from the original patch.
[1] http://article.gmane.org/gmane.linux.kernel/198823
Originally-by: Jamie Lokier <jamie@shareable.org>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link:
http://lkml.kernel.org/r/1355312043-11467-1-git-send-email-ville.syrjala@linux.intel.com
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Convert #include "..." to #include <path/...> in kernel system headers.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Dave Jones <davej@redhat.com>
Signal handling contains a bunch of accesses to individual user space
items, which causes an excessive number of STAC and CLAC
instructions. Instead, let get/put_user_try ... get/put_user_catch()
contain the STAC and CLAC instructions.
This means that get/put_user_try no longer nests, and furthermore that
it is no longer legal to use user space access functions other than
__get/put_user_ex() inside those blocks. However, these macros are
x86-specific anyway and are only used in the signal-handling paths; a
simple reordering of moving the larger subroutine calls out of the
try...catch blocks resolves that problem.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Link: http://lkml.kernel.org/r/1348256595-29119-12-git-send-email-hpa@linux.intel.com
When Supervisor Mode Access Prevention (SMAP) is enabled, access to
userspace from the kernel is controlled by the AC flag. To make the
performance of manipulating that flag acceptable, there are two new
instructions, STAC and CLAC, to set and clear it.
This patch adds those instructions, via alternative(), when the SMAP
feature is enabled. It also adds X86_EFLAGS_AC unconditionally to the
SYSCALL entry mask; there is simply no reason to make that one
conditional.
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Link: http://lkml.kernel.org/r/1348256595-29119-9-git-send-email-hpa@linux.intel.com
This throws away the old x86-specific functions in favor of the generic
optimized version.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The generic strncpy_from_user() is not really optimal, since it is
designed to work on both little-endian and big-endian. And on
little-endian you can simplify much of the logic to find the first zero
byte, since little-endian arithmetic doesn't have to worry about the
carry bit propagating into earlier bytes (only later bytes, which we
don't care about).
But I have patches to make the generic routines use the architecture-
specific <asm/word-at-a-time.h> infrastructure, so that we can regain
the little-endian optimizations. But before we do that, switch over to
the generic routines to make the patches each do just one well-defined
thing.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Switch to using relative exception table entries on x86. On i386,
this has the advantage that the exception table entries don't need to
be relocated; on x86-64 this means the exception table entries take up
only half the space.
In either case, a 32-bit delta is sufficient, as the range of kernel
code addresses is limited.
Since part of the goal is to avoid needing to adjust the entries when
the kernel is relocated, the old trick of using addresses in the NULL
pointer range to indicate uaccess_err no longer works (and unlike RISC
architectures we can't use a flag bit); instead use an delta just
below +2G to indicate these special entries. The reach is still
limited to a single instruction.
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: David Daney <david.daney@cavium.com>
Link: http://lkml.kernel.org/r/CA%2B55aFyijf43qSu3N9nWHEBwaGbb7T2Oq9A=9EyR=Jtyqfq_cQ@mail.gmail.com
This merges the 32- and 64-bit versions of the x86 strncpy_from_user()
by just rewriting it in C rather than the ancient inline asm versions
that used lodsb/stosb and had been duplicated for (trivial) differences
between the 32-bit and 64-bit versions.
While doing that, it also speeds them up by doing the accesses a word at
a time. Finally, the new routines also properly handle the case of
hitting the end of the address space, which we have never done correctly
before (fs/namei.c has a hack around it for that reason).
Despite all these improvements, it actually removes more lines than it
adds, due to the de-duplication. Also, we no longer export (or define)
the legacy __strncpy_from_user() function (that was defined to not do
the user permission checks), since it's not actually used anywhere, and
the user address space checks are built in to the new code.
Other architecture maintainers have been notified that the old hack in
fs/namei.c will be going away in the 3.5 merge window, in case they
copied the x86 approach of being a bit cavalier about the end of the
address space.
Cc: linux-arch@vger.kernel.org
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Peter Anvin" <hpa@zytor.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
To make this work, we teach the page fault handler how to send
signals on failed uaccess. This only works for user addresses
(kernel addresses will never hit the page fault handler in the
first place), so we need to generate signals for those
separately.
This gets the tricky case right: if the user buffer spans
multiple pages and only the second page is invalid, we set
cr2 and si_addr correctly. UML relies on this behavior to
"fault in" pages as needed.
We steal a bit from thread_info.uaccess_err to enable this.
Before this change, uaccess_err was a 32-bit boolean value.
This fixes issues with UML when vsyscall=emulate.
Reported-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andy Lutomirski <luto@amacapital.net>
Cc: richard -rw- weinberger <richard.weinberger@gmail.com>
Cc: H. Peter Anvin <hpa@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/4c8f91de7ec5cd2ef0f59521a04e1015f11e42b4.1320712291.git.luto@amacapital.net
Signed-off-by: Ingo Molnar <mingo@elte.hu>
copy_from_user_nmi() is used in oprofile and perf. Moving it to other
library functions like copy_from_user(). As this is x86 code for 32
and 64 bits, create a new file usercopy.c for unified code.
Signed-off-by: Robert Richter <robert.richter@amd.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20110607172413.GJ20052@erda.amd.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Commit e66eed651f ("list: remove prefetching from regular list
iterators") removed the include of prefetch.h from list.h, which
uncovered several cases that had apparently relied on that rather
obscure header file dependency.
So this fixes things up a bit, using
grep -L linux/prefetch.h $(git grep -l '[^a-z_]prefetchw*(' -- '*.[ch]')
grep -L 'prefetchw*(' $(git grep -l 'linux/prefetch.h' -- '*.[ch]')
to guide us in finding files that either need <linux/prefetch.h>
inclusion, or have it despite not needing it.
There are more of them around (mostly network drivers), but this gets
many core ones.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
As reported in BZ #30352:
https://bugzilla.kernel.org/show_bug.cgi?id=30352
there's a kernel bug related to reading the last allowed page on x86_64.
The _copy_to_user() and _copy_from_user() functions use the following
check for address limit:
if (buf + size >= limit)
fail();
while it should be more permissive:
if (buf + size > limit)
fail();
That's because the size represents the number of bytes being
read/write from/to buf address AND including the buf address.
So the copy function will actually never touch the limit
address even if "buf + size == limit".
Following program fails to use the last page as buffer
due to the wrong limit check:
#include <sys/mman.h>
#include <sys/socket.h>
#include <assert.h>
#define PAGE_SIZE (4096)
#define LAST_PAGE ((void*)(0x7fffffffe000))
int main()
{
int fds[2], err;
void * ptr = mmap(LAST_PAGE, PAGE_SIZE, PROT_READ | PROT_WRITE,
MAP_ANONYMOUS | MAP_PRIVATE | MAP_FIXED, -1, 0);
assert(ptr == LAST_PAGE);
err = socketpair(AF_LOCAL, SOCK_STREAM, 0, fds);
assert(err == 0);
err = send(fds[0], ptr, PAGE_SIZE, 0);
perror("send");
assert(err == PAGE_SIZE);
err = recv(fds[1], ptr, PAGE_SIZE, MSG_WAITALL);
perror("recv");
assert(err == PAGE_SIZE);
return 0;
}
The other place checking the addr limit is the access_ok() function,
which is working properly. There's just a misleading comment
for the __range_not_ok() macro - which this patch fixes as well.
The last page of the user-space address range is a guard page and
Brian Gerst observed that the guard page itself due to an erratum on K8 cpus
(#121 Sequential Execution Across Non-Canonical Boundary Causes Processor
Hang).
However, the test code is using the last valid page before the guard page.
The bug is that the last byte before the guard page can't be read
because of the off-by-one error. The guard page is left in place.
This bug would normally not show up because the last page is
part of the process stack and never accessed via syscalls.
Signed-off-by: Jiri Olsa <jolsa@redhat.com>
Acked-by: Brian Gerst <brgerst@gmail.com>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: <stable@kernel.org>
Link: http://lkml.kernel.org/r/1305210630-7136-1-git-send-email-jolsa@redhat.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Move the handling of truncated %rip from an iret fault to the fault
entry path.
This allows x86-64 to use the standard search_extable() function.
Signed-off-by: Brian Gerst <brgerst@gmail.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Jan Beulich <jbeulich@novell.com>
LKML-Reference: <1255357103-5418-1-git-send-email-brgerst@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The movq instruction, generated by __put_user_asm() when used for
64-bit data, takes a sign-extended immediate ("e") not a zero-extended
immediate ("Z").
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Cc: Uros Bizjak <ubizjak@gmail.com>
Cc: stable@kernel.org
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>
Improve a few details in perfcounter call-chain recording that
makes use of fast-GUP:
- Use ACCESS_ONCE() to observe the pte value. ptes are fundamentally
racy and can be changed on another CPU, so we have to be careful
about how we access them. The PAE branch is already careful with
read-barriers - but the non-PAE and 64-bit side needs an
ACCESS_ONCE() to make sure the pte value is observed only once.
- make the checks a bit stricter so that we can feed it any kind of
cra^H^H^H user-space input ;-)
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: cleanup
In __put_user_size() macro errret is used for error value.
But if size is 8, errret isn't passed to__put_user_asm_u64().
This behavior is inconsistent.
Signed-off-by: Hiroshi Shimamoto <h-shimamoto@ct.jp.nec.com>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
In case of !CONFIG_X86_WP_WORKS_OK, __put_user_size_ex() is not defined.
Add macros for !CONFIG_X86_WP_WORKS_OK case.
Signed-off-by: Hiroshi Shimamoto <h-shimamoto@ct.jp.nec.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
