linux-toradex.git/arch, branch v3.14.25

x86/mm: In the PTE swapout page reclaim case clear the accessed bit instead of flushing the TLB

2014-11-21T17:23:07+00:00

commit b13b1d2d8692b437203de7a404c6b809d2cc4d99 upstream.

We use the accessed bit to age a page at page reclaim time,
and currently we also flush the TLB when doing so.

But in some workloads TLB flush overhead is very heavy. In my
simple multithreaded app with a lot of swap to several pcie
SSDs, removing the tlb flush gives about 20% ~ 30% swapout
speedup.

Fortunately just removing the TLB flush is a valid optimization:
on x86 CPUs, clearing the accessed bit without a TLB flush
doesn't cause data corruption.

It could cause incorrect page aging and the (mistaken) reclaim of
hot pages, but the chance of that should be relatively low.

So as a performance optimization don't flush the TLB when
clearing the accessed bit, it will eventually be flushed by
a context switch or a VM operation anyway. [ In the rare
event of it not getting flushed for a long time the delay
shouldn't really matter because there's no real memory
pressure for swapout to react to. ]

Suggested-by: Linus Torvalds 
Signed-off-by: Shaohua Li 
Acked-by: Rik van Riel 
Acked-by: Mel Gorman 
Acked-by: Hugh Dickins 
Acked-by: Johannes Weiner 
Cc: linux-mm@kvack.org
Cc: Peter Zijlstra 
Link: http://lkml.kernel.org/r/20140408075809.GA1764@kernel.org
[ Rewrote the changelog and the code comments. ]
Signed-off-by: Ingo Molnar 
Signed-off-by: Mel Gorman 
Signed-off-by: Greg Kroah-Hartman

KVM: x86: Don't report guest userspace emulation error to userspace

2014-11-21T17:23:05+00:00

commit a2b9e6c1a35afcc0973acb72e591c714e78885ff upstream.

Commit fc3a9157d314 ("KVM: X86: Don't report L2 emulation failures to
user-space") disabled the reporting of L2 (nested guest) emulation failures to
userspace due to race-condition between a vmexit and the instruction emulator.
The same rational applies also to userspace applications that are permitted by
the guest OS to access MMIO area or perform PIO.

This patch extends the current behavior - of injecting a #UD instead of
reporting it to userspace - also for guest userspace code.

Signed-off-by: Nadav Amit 
Signed-off-by: Paolo Bonzini 
Signed-off-by: Greg Kroah-Hartman

perf/x86/intel: Use proper dTLB-load-misses event on IvyBridge

2014-11-21T17:23:05+00:00

commit 1996388e9f4e3444db8273bc08d25164d2967c21 upstream.

This was discussed back in February:

	https://lkml.org/lkml/2014/2/18/956

But I never saw a patch come out of it.

On IvyBridge we share the SandyBridge cache event tables, but the
dTLB-load-miss event is not compatible.  Patch it up after
the fact to the proper DTLB_LOAD_MISSES.DEMAND_LD_MISS_CAUSES_A_WALK

Signed-off-by: Vince Weaver 
Signed-off-by: Peter Zijlstra 
Cc: Arnaldo Carvalho de Melo 
Cc: Linus Torvalds 
Link: http://lkml.kernel.org/r/alpine.DEB.2.11.1407141528200.17214@vincent-weaver-1.umelst.maine.edu
Signed-off-by: Ingo Molnar 
Cc: Hou Pengyang 
Signed-off-by: Greg Kroah-Hartman

ARM: probes: fix instruction fetch order with

2014-11-21T17:23:05+00:00

commit 888be25402021a425da3e85e2d5a954d7509286e upstream.

If we are running BE8, the data and instruction endianness do not
match, so use  to correctly translate memory accesses
into ARM instructions.

Acked-by: Jon Medhurst 
Signed-off-by: Ben Dooks 
[taras.kondratiuk@linaro.org: fixed Thumb instruction fetch order]
Signed-off-by: Taras Kondratiuk 
[wangnan: backport to 3.10 and 3.14:
 - adjust context
 - backport all changes on arch/arm/kernel/probes.c to
   arch/arm/kernel/kprobes-common.c since we don't have
   commit c18377c303787ded44b7decd7dee694db0f205e9.
 - After the above adjustments, becomes same to Taras Kondratiuk's
   original patch:
     http://lists.linaro.org/pipermail/linaro-kernel/2014-January/010346.html
]
Signed-off-by: Wang Nan 
Signed-off-by: Greg Kroah-Hartman

x86, kaslr: Prevent .bss from overlaping initrd

2014-11-21T17:23:04+00:00

commit e6023367d779060fddc9a52d1f474085b2b36298 upstream.

When choosing a random address, the current implementation does not take into
account the reversed space for .bss and .brk sections. Thus the relocated kernel
may overlap other components in memory. Here is an example of the overlap from a
x86_64 kernel in qemu (the ranges of physical addresses are presented):

 Physical Address

    0x0fe00000                  --+--------------------+  <-- randomized base
                               /  |  relocated kernel  |
                   vmlinux.bin    | (from vmlinux.bin) |
    0x1336d000    (an ELF file)   +--------------------+--
                               \  |                    |  \
    0x1376d870                  --+--------------------+   |
                                  |    relocs table    |   |
    0x13c1c2a8                    +--------------------+   .bss and .brk
                                  |                    |   |
    0x13ce6000                    +--------------------+   |
                                  |                    |  /
    0x13f77000                    |       initrd       |--
                                  |                    |
    0x13fef374                    +--------------------+

The initrd image will then be overwritten by the memset during early
initialization:

[    1.655204] Unpacking initramfs...
[    1.662831] Initramfs unpacking failed: junk in compressed archive

This patch prevents the above situation by requiring a larger space when looking
for a random kernel base, so that existing logic can effectively avoids the
overlap.

[kees: switched to perl to avoid hex translation pain in mawk vs gawk]
[kees: calculated overlap without relocs table]

Fixes: 82fa9637a2 ("x86, kaslr: Select random position from e820 maps")
Reported-by: Fengguang Wu 
Signed-off-by: Junjie Mao 
Signed-off-by: Kees Cook 
Cc: Josh Triplett 
Cc: Matt Fleming 
Cc: Ard Biesheuvel 
Cc: Vivek Goyal 
Cc: Andi Kleen 
Link: http://lkml.kernel.org/r/1414762838-13067-1-git-send-email-eternal.n08@gmail.com
Signed-off-by: Thomas Gleixner 
Signed-off-by: Greg Kroah-Hartman

x86, microcode, AMD: Fix ucode patch stashing on 32-bit

2014-11-21T17:23:03+00:00

commit c0a717f23dccdb6e3b03471bc846fdc636f2b353 upstream.

Save the patch while we're running on the BSP instead of later, before
the initrd has been jettisoned. More importantly, on 32-bit we need to
access the physical address instead of the virtual.

This way we actually do find it on the APs instead of having to go
through the initrd each time.

Tested-by: Richard Hendershot 
Fixes: 5335ba5cf475 ("x86, microcode, AMD: Fix early ucode loading")
Signed-off-by: Borislav Petkov 
Signed-off-by: Greg Kroah-Hartman

x86, microcode, AMD: Fix early ucode loading on 32-bit

2014-11-21T17:23:03+00:00

commit 4750a0d112cbfcc744929f1530ffe3193436766c upstream.

Konrad triggered the following splat below in a 32-bit guest on an AMD
box. As it turns out, in save_microcode_in_initrd_amd() we're using the
*physical* address of the container *after* we have enabled paging and
thus we #PF in load_microcode_amd() when trying to access the microcode
container in the ramdisk range.

Because the ramdisk is exactly there:

[    0.000000] RAMDISK: [mem 0x35e04000-0x36ef9fff]

and we fault at 0x35e04304.

And since this guest doesn't relocate the ramdisk, we don't do the
computation which will give us the correct virtual address and we end up
with the PA.

So, we should actually be using virtual addresses on 32-bit too by the
time we're freeing the initrd. Do that then!

Unpacking initramfs...
BUG: unable to handle kernel paging request at 35d4e304
IP: [] load_microcode_amd+0x25/0x4a0
*pde = 00000000
Oops: 0000 [#1] SMP
Modules linked in:
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 3.17.1-302.fc21.i686 #1
Hardware name: Xen HVM domU, BIOS 4.4.1 10/01/2014
task: f5098000 ti: f50d0000 task.ti: f50d0000
EIP: 0060:[] EFLAGS: 00010246 CPU: 0
EIP is at load_microcode_amd+0x25/0x4a0
EAX: 00000000 EBX: f6e9ec4c ECX: 00001ec4 EDX: 00000000
ESI: f5d4e000 EDI: 35d4e2fc EBP: f50d1ed0 ESP: f50d1e94
 DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
CR0: 8005003b CR2: 35d4e304 CR3: 00e33000 CR4: 000406d0
Stack:
 00000000 00000000 f50d1ebc f50d1ec4 f5d4e000 c0d7735a f50d1ed0 15a3d17f
 f50d1ec4 00600f20 00001ec4 bfb83203 f6e9ec4c f5d4e000 c0d7735a f50d1ed8
 c0d80861 f50d1ee0 c0d80429 f50d1ef0 c0d889a9 f5d4e000 c0000000 f50d1f04
Call Trace:
? unpack_to_rootfs
? unpack_to_rootfs
save_microcode_in_initrd_amd
save_microcode_in_initrd
free_initrd_mem
populate_rootfs
? unpack_to_rootfs
do_one_initcall
? unpack_to_rootfs
? repair_env_string
? proc_mkdir
kernel_init_freeable
kernel_init
ret_from_kernel_thread
? rest_init

Reported-and-tested-by: Konrad Rzeszutek Wilk 
References: https://bugzilla.redhat.com/show_bug.cgi?id=1158204
Fixes: 75a1ba5b2c52 ("x86, microcode, AMD: Unify valid container checks")
Signed-off-by: Borislav Petkov 
Link: http://lkml.kernel.org/r/20141101100100.GA4462@pd.tnic
Signed-off-by: Thomas Gleixner 
Signed-off-by: Greg Kroah-Hartman

parisc: Use compat layer for msgctl, shmat, shmctl and semtimedop syscalls

2014-11-21T17:23:03+00:00

commit 2fe749f50b0bec07650ef135b29b1f55bf543869 upstream.

Switch over the msgctl, shmat, shmctl and semtimedop syscalls to use the compat
layer. The problem was found with the debian procenv package, which called
	shmctl(0, SHM_INFO, &info);
in which the shmctl syscall then overwrote parts of the surrounding areas on
the stack on which the info variable was stored and thus lead to a segfault
later on.

Additionally fix the definition of struct shminfo64 to use unsigned longs like
the other architectures. This has no impact on userspace since we only have a
32bit userspace up to now.

Signed-off-by: Helge Deller 
Cc: John David Anglin 
Signed-off-by: Greg Kroah-Hartman

Correct the race condition in aarch64_insn_patch_text_sync()

2014-11-21T17:23:02+00:00

commit 899d5933b2dd2720f2b20b01eaa07871aa6ad096 upstream.

When experimenting with patches to provide kprobes support for aarch64
smp machines would hang when inserting breakpoints into kernel code.
The hangs were caused by a race condition in the code called by
aarch64_insn_patch_text_sync().  The first processor in the
aarch64_insn_patch_text_cb() function would patch the code while other
processors were still entering the function and incrementing the
cpu_count field.  This resulted in some processors never observing the
exit condition and exiting the function.  Thus, processors in the
system hung.

The first processor to enter the patching function performs the
patching and signals that the patching is complete with an increment
of the cpu_count field. When all the processors have incremented the
cpu_count field the cpu_count will be num_cpus_online()+1 and they
will return to normal execution.

Fixes: ae16480785de arm64: introduce interfaces to hotpatch kernel and module code
Signed-off-by: William Cohen 
Acked-by: Will Deacon 
Signed-off-by: Catalin Marinas 
Signed-off-by: Greg Kroah-Hartman

arm64: __clear_user: handle exceptions on strb

2014-11-21T17:23:02+00:00

commit 97fc15436b36ee3956efad83e22a557991f7d19d upstream.

ARM64 currently doesn't fix up faults on the single-byte (strb) case of
__clear_user... which means that we can cause a nasty kernel panic as an
ordinary user with any multiple PAGE_SIZE+1 read from /dev/zero.
i.e.: dd if=/dev/zero of=foo ibs=1 count=1 (or ibs=65537, etc.)

This is a pretty obscure bug in the general case since we'll only
__do_kernel_fault (since there's no extable entry for pc) if the
mmap_sem is contended. However, with CONFIG_DEBUG_VM enabled, we'll
always fault.

if (!down_read_trylock(&mm->mmap_sem)) {
	if (!user_mode(regs) && !search_exception_tables(regs->pc))
		goto no_context;
retry:
	down_read(&mm->mmap_sem);
} else {
	/*
	 * The above down_read_trylock() might have succeeded in
	 * which
	 * case, we'll have missed the might_sleep() from
	 * down_read().
	 */
	might_sleep();
	if (!user_mode(regs) && !search_exception_tables(regs->pc))
		goto no_context;
}

Fix that by adding an extable entry for the strb instruction, since it
touches user memory, similar to the other stores in __clear_user.

Signed-off-by: Kyle McMartin 
Reported-by: Miloš Prchlík 
Signed-off-by: Catalin Marinas 
Signed-off-by: Greg Kroah-Hartman