By using outlined checks we can achieve a significant code size
improvement by moving the tag-based ASAN checks into separate
functions. Unlike the existing CONFIG_KASAN_OUTLINE mode these
functions have a custom calling convention that preserves most
registers and is specialized to the register containing the address
and the type of access, and as a result we can eliminate the code
size and performance overhead of a standard calling convention such
as AAPCS for these functions.
This change depends on a separate series of changes to Clang [1] to
support outlined checks in the kernel, although the change works fine
without them (we just don't get outlined checks). This is because the
flag -mllvm -hwasan-inline-all-checks=0 has no effect until the Clang
changes land. The flag was introduced in the Clang 9.0 timeframe as
part of the support for outlined checks in userspace and because our
minimum Clang version is 10.0 we can pass it unconditionally.
Outlined checks require a new runtime function with a custom calling
convention. Add this function to arch/arm64/lib.
I measured the code size of defconfig + tag-based KASAN, as well
as boot time (i.e. time to init launch) on a DragonBoard 845c with
an Android arm64 GKI kernel. The results are below:
code size boot time
CONFIG_KASAN_INLINE=y before 92824064 6.18s
CONFIG_KASAN_INLINE=y after 38822400 6.65s
CONFIG_KASAN_OUTLINE=y 39215616 11.48s
We can see straight away that specialized outlined checks beat the
existing CONFIG_KASAN_OUTLINE=y on both code size and boot time
for tag-based ASAN.
As for the comparison between CONFIG_KASAN_INLINE=y before and after
we saw similar performance numbers in userspace [2] and decided
that since the performance overhead is minimal compared to the
overhead of tag-based ASAN itself as well as compared to the code
size improvements we would just replace the inlined checks with the
specialized outlined checks without the option to select between them,
and that is what I have implemented in this patch.
Signed-off-by: Peter Collingbourne <pcc@google.com>
Acked-by: Andrey Konovalov <andreyknvl@gmail.com>
Reviewed-by: Mark Rutland <mark.rutland@arm.com>
Tested-by: Mark Rutland <mark.rutland@arm.com>
Link: https://linux-review.googlesource.com/id/I1a30036c70ab3c3ee78d75ed9b87ef7cdc3fdb76
Link: [1] https://reviews.llvm.org/D90426
Link: [2] https://reviews.llvm.org/D56954
Link: https://lore.kernel.org/r/20210526174927.2477847-3-pcc@google.com
Signed-off-by: Will Deacon <will@kernel.org>
Merge in stack unwinding work to cater for 8-byte aligned stack frames
which may be generated following optimisations to CONFIG_KASAN_OUTLINE.
* for-next/stacktrace:
arm64: stacktrace: Relax frame record alignment requirement to 8 bytes
arm64: Change the on_*stack functions to take a size argument
arm64: Implement stack trace termination record
Reliable stacktracing requires that we identify when a stacktrace is
terminated early. We can do this by ensuring all tasks have a final
frame record at a known location on their task stack, and checking
that this is the final frame record in the chain.
We'd like to use task_pt_regs(task)->stackframe as the final frame
record, as this is already setup upon exception entry from EL0. For
kernel tasks we need to consistently reserve the pt_regs and point x29
at this, which we can do with small changes to __primary_switched,
__secondary_switched, and copy_process().
Since the final frame record must be at a specific location, we must
create the final frame record in __primary_switched and
__secondary_switched rather than leaving this to start_kernel and
secondary_start_kernel. Thus, __primary_switched and
__secondary_switched will now show up in stacktraces for the idle tasks.
Since the final frame record is now identified by its location rather
than by its contents, we identify it at the start of unwind_frame(),
before we read any values from it.
External debuggers may terminate the stack trace when FP == 0. In the
pt_regs->stackframe, the PC is 0 as well. So, stack traces taken in the
debugger may print an extra record 0x0 at the end. While this is not
pretty, this does not do any harm. This is a small price to pay for
having reliable stack trace termination in the kernel. That said, gdb
does not show the extra record probably because it uses DWARF and not
frame pointers for stack traces.
Signed-off-by: Madhavan T. Venkataraman <madvenka@linux.microsoft.com>
Reviewed-by: Mark Brown <broonie@kernel.org>
[Mark: rebase, use ASM_BUG(), update comments, update commit message]
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Link: https://lore.kernel.org/r/20210510110026.18061-1-mark.rutland@arm.com
Signed-off-by: Will Deacon <will@kernel.org>
Pull perf fixes from Thomas Gleixner:
"Two perf fixes:
- Do not check the LBR_TOS MSR when setting up unrelated LBR MSRs as
this can cause malfunction when TOS is not supported
- Allocate the LBR XSAVE buffers along with the DS buffers upfront
because allocating them when adding an event can deadlock"
* tag 'perf-urgent-2021-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/x86/lbr: Remove cpuc->lbr_xsave allocation from atomic context
perf/x86: Avoid touching LBR_TOS MSR for Arch LBR
Pull locking fixes from Thomas Gleixner:
"Two locking fixes:
- Invoke the lockdep tracepoints in the correct place so the ordering
is correct again
- Don't leave the mutex WAITER bit stale when the last waiter is
dropping out early due to a signal as that forces all subsequent
lock operations needlessly into the slowpath until it's cleaned up
again"
* tag 'locking-urgent-2021-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
locking/mutex: clear MUTEX_FLAGS if wait_list is empty due to signal
locking/lockdep: Correct calling tracepoints
Pull irq fixes from Thomas Gleixner:
"A few fixes for irqchip drivers:
- Allocate interrupt descriptors correctly on Mainstone PXA when
SPARSE_IRQ is enabled; otherwise the interrupt association fails
- Make the APPLE AIC chip driver depend on APPLE
- Remove redundant error output on devm_ioremap_resource() failure"
* tag 'irq-urgent-2021-05-23' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
irqchip: Remove redundant error printing
irqchip/apple-aic: APPLE_AIC should depend on ARCH_APPLE
ARM: PXA: Fix cplds irqdesc allocation when using legacy mode
Pull x86 fixes from Borislav Petkov:
- Fix how SEV handles MMIO accesses by forwarding potential page faults
instead of killing the machine and by using the accessors with the
exact functionality needed when accessing memory.
- Fix a confusion with Clang LTO compiler switches passed to the it
- Handle the case gracefully when VMGEXIT has been executed in
userspace
* tag 'x86_urgent_for_v5.13_rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
x86/sev-es: Use __put_user()/__get_user() for data accesses
x86/sev-es: Forward page-faults which happen during emulation
x86/sev-es: Don't return NULL from sev_es_get_ghcb()
x86/build: Fix location of '-plugin-opt=' flags
x86/sev-es: Invalidate the GHCB after completing VMGEXIT
x86/sev-es: Move sev_es_put_ghcb() in prep for follow on patch
Pull powerpc fixes from Michael Ellerman:
- Fix breakage of strace (and other ptracers etc.) when using the new
scv ABI (Power9 or later with glibc >= 2.33).
- Fix early_ioremap() on 64-bit, which broke booting on some machines.
Thanks to Dmitry V. Levin, Nicholas Piggin, Alexey Kardashevskiy, and
Christophe Leroy.
* tag 'powerpc-5.13-4' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/64s/syscall: Fix ptrace syscall info with scv syscalls
powerpc/64s/syscall: Use pt_regs.trap to distinguish syscall ABI difference between sc and scv syscalls
powerpc: Fix early setup to make early_ioremap() work
Pull Kbuild fixes from Masahiro Yamada:
- Fix short log indentation for tools builds
- Fix dummy-tools to adjust to the latest stackprotector check
* tag 'kbuild-fixes-v5.13' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kbuild: dummy-tools: adjust to stricter stackprotector check
scripts/jobserver-exec: Fix a typo ("envirnoment")
tools build: Fix quiet cmd indentation
Merge misc fixes from Andrew Morton:
"10 patches.
Subsystems affected by this patch series: mm (pagealloc, gup, kasan,
and userfaultfd), ipc, selftests, watchdog, bitmap, procfs, and lib"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
userfaultfd: hugetlbfs: fix new flag usage in error path
lib: kunit: suppress a compilation warning of frame size
proc: remove Alexey from MAINTAINERS
linux/bits.h: fix compilation error with GENMASK
watchdog: reliable handling of timestamps
kasan: slab: always reset the tag in get_freepointer_safe()
tools/testing/selftests/exec: fix link error
ipc/mqueue, msg, sem: avoid relying on a stack reference past its expiry
Revert "mm/gup: check page posion status for coredump."
mm/shuffle: fix section mismatch warning
lib/bitfield_kunit.c: In function `test_bitfields_constants':
lib/bitfield_kunit.c:93:1: warning: the frame size of 7456 bytes is larger than 2048 bytes [-Wframe-larger-than=]
}
^
As the description of BITFIELD_KUNIT in lib/Kconfig.debug, it "Only useful
for kernel devs running the KUnit test harness, and not intended for
inclusion into a production build". Therefore, it is not worth modifying
variable 'test_bitfields_constants' to clear this warning. Just suppress
it.
Link: https://lkml.kernel.org/r/20210518094533.7652-1-thunder.leizhen@huawei.com
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Cc: Shuah Khan <skhan@linuxfoundation.org>
Cc: Vitor Massaru Iha <vitor@massaru.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit 9bf3bc949f ("watchdog: cleanup handling of false positives")
tried to handle a virtual host stopped by the host a more
straightforward and cleaner way.
But it introduced a risk of false softlockup reports. The virtual host
might be stopped at any time, for example between
kvm_check_and_clear_guest_paused() and is_softlockup(). As a result,
is_softlockup() might read the updated jiffies and detects a softlockup.
A solution might be to put back kvm_check_and_clear_guest_paused() after
is_softlockup() and detect it. But it would put back the cycle that
complicates the logic.
In fact, the handling of all the timestamps is not reliable. The code
does not guarantee when and how many times the timestamps are read. For
example, "period_ts" might be touched anytime also from NMI and re-read in
is_softlockup(). It works just by chance.
Fix all the problems by making the code even more explicit.
1. Make sure that "now" and "period_ts" timestamps are read only once.
They might be changed at anytime by NMI or when the virtual guest is
stopped by the host. Note that "now" timestamp does this implicitly
because "jiffies" is marked volatile.
2. "now" time must be read first. The state of "period_ts" will
decide whether it will be used or the period will get restarted.
3. kvm_check_and_clear_guest_paused() must be called before reading
"period_ts". It touches the variable when the guest was stopped.
As a result, "now" timestamp is used only when the watchdog was not
touched and the guest not stopped in the meantime. "period_ts" is
restarted in all other situations.
Link: https://lkml.kernel.org/r/YKT55gw+RZfyoFf7@alley
Fixes: 9bf3bc949f ("watchdog: cleanup handling of false positives")
Signed-off-by: Petr Mladek <pmladek@suse.com>
Reported-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Reviewed-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Fix the link error by adding '-static':
gcc -Wall -Wl,-z,max-page-size=0x1000 -pie load_address.c -o /home/yang/linux/tools/testing/selftests/exec/load_address_4096
/usr/bin/ld: /tmp/ccopEGun.o: relocation R_AARCH64_ADR_PREL_PG_HI21 against symbol `stderr@@GLIBC_2.17' which may bind externally can not be used when making a shared object; recompile with -fPIC
/usr/bin/ld: /tmp/ccopEGun.o(.text+0x158): unresolvable R_AARCH64_ADR_PREL_PG_HI21 relocation against symbol `stderr@@GLIBC_2.17'
/usr/bin/ld: final link failed: bad value
collect2: error: ld returned 1 exit status
make: *** [Makefile:25: tools/testing/selftests/exec/load_address_4096] Error 1
Link: https://lkml.kernel.org/r/20210514092422.2367367-1-yangyingliang@huawei.com
Fixes: 206e22f019 ("tools/testing/selftests: add self-test for verifying load alignment")
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Cc: Chris Kennelly <ckennelly@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
do_mq_timedreceive calls wq_sleep with a stack local address. The
sender (do_mq_timedsend) uses this address to later call pipelined_send.
This leads to a very hard to trigger race where a do_mq_timedreceive
call might return and leave do_mq_timedsend to rely on an invalid
address, causing the following crash:
RIP: 0010:wake_q_add_safe+0x13/0x60
Call Trace:
__x64_sys_mq_timedsend+0x2a9/0x490
do_syscall_64+0x80/0x680
entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x7f5928e40343
The race occurs as:
1. do_mq_timedreceive calls wq_sleep with the address of `struct
ext_wait_queue` on function stack (aliased as `ewq_addr` here) - it
holds a valid `struct ext_wait_queue *` as long as the stack has not
been overwritten.
2. `ewq_addr` gets added to info->e_wait_q[RECV].list in wq_add, and
do_mq_timedsend receives it via wq_get_first_waiter(info, RECV) to call
__pipelined_op.
3. Sender calls __pipelined_op::smp_store_release(&this->state,
STATE_READY). Here is where the race window begins. (`this` is
`ewq_addr`.)
4. If the receiver wakes up now in do_mq_timedreceive::wq_sleep, it
will see `state == STATE_READY` and break.
5. do_mq_timedreceive returns, and `ewq_addr` is no longer guaranteed
to be a `struct ext_wait_queue *` since it was on do_mq_timedreceive's
stack. (Although the address may not get overwritten until another
function happens to touch it, which means it can persist around for an
indefinite time.)
6. do_mq_timedsend::__pipelined_op() still believes `ewq_addr` is a
`struct ext_wait_queue *`, and uses it to find a task_struct to pass to
the wake_q_add_safe call. In the lucky case where nothing has
overwritten `ewq_addr` yet, `ewq_addr->task` is the right task_struct.
In the unlucky case, __pipelined_op::wake_q_add_safe gets handed a
bogus address as the receiver's task_struct causing the crash.
do_mq_timedsend::__pipelined_op() should not dereference `this` after
setting STATE_READY, as the receiver counterpart is now free to return.
Change __pipelined_op to call wake_q_add_safe on the receiver's
task_struct returned by get_task_struct, instead of dereferencing `this`
which sits on the receiver's stack.
As Manfred pointed out, the race potentially also exists in
ipc/msg.c::expunge_all and ipc/sem.c::wake_up_sem_queue_prepare. Fix
those in the same way.
Link: https://lkml.kernel.org/r/20210510102950.12551-1-varad.gautam@suse.com
Fixes: c5b2cbdbda ("ipc/mqueue.c: update/document memory barriers")
Fixes: 8116b54e7e ("ipc/sem.c: document and update memory barriers")
Fixes: 0d97a82ba8 ("ipc/msg.c: update and document memory barriers")
Signed-off-by: Varad Gautam <varad.gautam@suse.com>
Reported-by: Matthias von Faber <matthias.vonfaber@aox-tech.de>
Acked-by: Davidlohr Bueso <dbueso@suse.de>
Acked-by: Manfred Spraul <manfred@colorfullife.com>
Cc: Christian Brauner <christian.brauner@ubuntu.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
clang sometimes decides not to inline shuffle_zone(), but it calls a
__meminit function. Without the extra __meminit annotation we get this
warning:
WARNING: modpost: vmlinux.o(.text+0x2a86d4): Section mismatch in reference from the function shuffle_zone() to the function .meminit.text:__shuffle_zone()
The function shuffle_zone() references
the function __meminit __shuffle_zone().
This is often because shuffle_zone lacks a __meminit
annotation or the annotation of __shuffle_zone is wrong.
shuffle_free_memory() did not show the same problem in my tests, but it
could happen in theory as well, so mark both as __meminit.
Link: https://lkml.kernel.org/r/20210514135952.2928094-1-arnd@kernel.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Nathan Chancellor <nathan@kernel.org>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Wei Yang <richard.weiyang@linux.alibaba.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull block fixes from Jens Axboe:
- Fix BLKRRPART and deletion race (Gulam, Christoph)
- NVMe pull request (Christoph):
- nvme-tcp corruption and timeout fixes (Sagi Grimberg, Keith
Busch)
- nvme-fc teardown fix (James Smart)
- nvmet/nvme-loop memory leak fixes (Wu Bo)"
* tag 'block-5.13-2021-05-22' of git://git.kernel.dk/linux-block:
block: fix a race between del_gendisk and BLKRRPART
block: prevent block device lookups at the beginning of del_gendisk
nvme-fc: clear q_live at beginning of association teardown
nvme-tcp: rerun io_work if req_list is not empty
nvme-tcp: fix possible use-after-completion
nvme-loop: fix memory leak in nvme_loop_create_ctrl()
nvmet: fix memory leak in nvmet_alloc_ctrl()
Pull io_uring fixes from Jens Axboe:
"One fix for a regression with poll in this merge window, and another
just hardens the io-wq exit path a bit"
* tag 'io_uring-5.13-2021-05-22' of git://git.kernel.dk/linux-block:
io_uring: fortify tctx/io_wq cleanup
io_uring: don't modify req->poll for rw
Peter Collingbourne
Will Deacon
Madhavan T. Venkataraman
Linus Torvalds
Mike Kravetz
Zhen Lei
Alexey Dobriyan
Rikard Falkeborn
Petr Mladek
Alexander Potapenko
Yang Yingliang
Varad Gautam
Michal Hocko
Arnd Bergmann