* make the reference from superblock to cgroup_root counting -
do cgroup_put() in cgroup_kill_sb() whether we'd done
percpu_ref_kill() or not; matching grab is done when we allocate
a new root. That gives the same refcounting rules for all callers
of cgroup_do_mount() - a reference to cgroup_root has been grabbed
by caller and it either is transferred to new superblock or dropped.
* have cgroup_kill_sb() treat an already killed refcount as "just
don't bother killing it, then".
* after successful cgroup_do_mount() have cgroup1_mount() recheck
if we'd raced with mount/umount from somebody else and cgroup_root
got killed. In that case we drop the superblock and bugger off
with -ERESTARTSYS, same as if we'd found it in the list already
dying.
* don't bother with delayed initialization of refcount - it's
unreliable and not needed. No need to prevent attempts to bump
the refcount if we find cgroup_root of another mount in progress -
sget will reuse an existing superblock just fine and if the
other sb manages to die before we get there, we'll catch
that immediately after cgroup_do_mount().
* don't bother with kernfs_pin_sb() - no need for doing that
either.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
same story as with last May fixes in sysfs (7b745a4e40
"unfuck sysfs_mount()"); new_sb is left uninitialized
in case of early errors in kernfs_mount_ns() and papering
over it by treating any error from kernfs_mount_ns() as
equivalent to !new_ns ends up conflating the cases when
objects had never been transferred to a superblock with
ones when that has happened and resulting new superblock
had been dropped. Easily fixed (same way as in sysfs
case). Additionally, there's a superblock leak on
kernfs_node_dentry() failure *and* a dentry leak inside
kernfs_node_dentry() itself - the latter on probably
impossible errors, but the former not impossible to trigger
(as the matter of fact, injecting allocation failures
at that point *does* trigger it).
Cc: stable@kernel.org
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
UNAME26 is a mechanism to report Linux's version as 2.6.x, for
compatibility with old/broken software. Due to the way it is
implemented, it would have to be updated after 5.0, to keep the
resulting versions unique. Linus Torvalds argued:
"Do we actually need this?
I'd rather let it bitrot, and just let it return random versions. It
will just start again at 2.4.60, won't it?
Anybody who uses UNAME26 for a 5.x kernel might as well think it's
still 4.x. The user space is so old that it can't possibly care about
differences between 4.x and 5.x, can it?
The only thing that matters is that it shows "2.4.<largeenough>",
which it will do regardless"
Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge misc fixes from Andrew Morton:
"14 fixes"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
mm, page_alloc: do not wake kswapd with zone lock held
hugetlbfs: revert "use i_mmap_rwsem for more pmd sharing synchronization"
hugetlbfs: revert "Use i_mmap_rwsem to fix page fault/truncate race"
mm: page_mapped: don't assume compound page is huge or THP
mm/memory.c: initialise mmu_notifier_range correctly
tools/vm/page_owner: use page_owner_sort in the use example
kasan: fix krealloc handling for tag-based mode
kasan: make tag based mode work with CONFIG_HARDENED_USERCOPY
kasan, arm64: use ARCH_SLAB_MINALIGN instead of manual aligning
mm, memcg: fix reclaim deadlock with writeback
mm/usercopy.c: no check page span for stack objects
slab: alien caches must not be initialized if the allocation of the alien cache failed
fork, memcg: fix cached_stacks case
zram: idle writeback fixes and cleanup
This changes the fork(2) syscall to record the process start_time after
initializing the basic task structure but still before making the new
process visible to user-space.
Technically, we could record the start_time anytime during fork(2). But
this might lead to scenarios where a start_time is recorded long before
a process becomes visible to user-space. For instance, with
userfaultfd(2) and TLS, user-space can delay the execution of fork(2)
for an indefinite amount of time (and will, if this causes network
access, or similar).
By recording the start_time late, it much closer reflects the point in
time where the process becomes live and can be observed by other
processes.
Lastly, this makes it much harder for user-space to predict and control
the start_time they get assigned. Previously, user-space could fork a
process and stall it in copy_thread_tls() before its pid is allocated,
but after its start_time is recorded. This can be misused to later-on
cycle through PIDs and resume the stalled fork(2) yielding a process
that has the same pid and start_time as a process that existed before.
This can be used to circumvent security systems that identify processes
by their pid+start_time combination.
Even though user-space was always aware that start_time recording is
flaky (but several projects are known to still rely on start_time-based
identification), changing the start_time to be recorded late will help
mitigate existing attacks and make it much harder for user-space to
control the start_time a process gets assigned.
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Tom Gundersen <teg@jklm.no>
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull more Kbuild updates from Masahiro Yamada:
- improve boolinit.cocci and use_after_iter.cocci semantic patches
- fix alignment for kallsyms
- move 'asm goto' compiler test to Kconfig and clean up jump_label
CONFIG option
- generate asm-generic wrappers automatically if arch does not
implement mandatory UAPI headers
- remove redundant generic-y defines
- misc cleanups
* tag 'kbuild-v4.21-3' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild:
kconfig: rename generated .*conf-cfg to *conf-cfg
kbuild: remove unnecessary stubs for archheader and archscripts
kbuild: use assignment instead of define ... endef for filechk_* rules
arch: remove redundant UAPI generic-y defines
kbuild: generate asm-generic wrappers if mandatory headers are missing
arch: remove stale comments "UAPI Header export list"
riscv: remove redundant kernel-space generic-y
kbuild: change filechk to surround the given command with { }
kbuild: remove redundant target cleaning on failure
kbuild: clean up rule_dtc_dt_yaml
kbuild: remove UIMAGE_IN and UIMAGE_OUT
jump_label: move 'asm goto' support test to Kconfig
kallsyms: lower alignment on ARM
scripts: coccinelle: boolinit: drop warnings on named constants
scripts: coccinelle: check for redeclaration
kconfig: remove unused "file" field of yylval union
nds32: remove redundant kernel-space generic-y
nios2: remove unneeded HAS_DMA define
Pull dma-mapping fixes from Christoph Hellwig:
"Fix various regressions introduced in this cycles:
- fix dma-debug tracking for the map_page / map_single
consolidatation
- properly stub out DMA mapping symbols for !HAS_DMA builds to avoid
link failures
- fix AMD Gart direct mappings
- setup the dma address for no kernel mappings using the remap
allocator"
* tag 'dma-mapping-4.21-1' of git://git.infradead.org/users/hch/dma-mapping:
dma-direct: fix DMA_ATTR_NO_KERNEL_MAPPING for remapped allocations
x86/amd_gart: fix unmapping of non-GART mappings
dma-mapping: remove a few unused exports
dma-mapping: properly stub out the DMA API for !CONFIG_HAS_DMA
dma-mapping: remove dmam_{declare,release}_coherent_memory
dma-mapping: implement dmam_alloc_coherent using dmam_alloc_attrs
dma-mapping: implement dma_map_single_attrs using dma_map_page_attrs
filechk_* rules often consist of multiple 'echo' lines. They must be
surrounded with { } or ( ) to work correctly. Otherwise, only the
string from the last 'echo' would be written into the target.
Let's take care of that in the 'filechk' in scripts/Kbuild.include
to clean up filechk_* rules.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Currently, CONFIG_JUMP_LABEL just means "I _want_ to use jump label".
The jump label is controlled by HAVE_JUMP_LABEL, which is defined
like this:
#if defined(CC_HAVE_ASM_GOTO) && defined(CONFIG_JUMP_LABEL)
# define HAVE_JUMP_LABEL
#endif
We can improve this by testing 'asm goto' support in Kconfig, then
make JUMP_LABEL depend on CC_HAS_ASM_GOTO.
Ugly #ifdef HAVE_JUMP_LABEL will go away, and CONFIG_JUMP_LABEL will
match to the real kernel capability.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Tested-by: Sedat Dilek <sedat.dilek@gmail.com>
Merge more updates from Andrew Morton:
- procfs updates
- various misc bits
- lib/ updates
- epoll updates
- autofs
- fatfs
- a few more MM bits
* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (58 commits)
mm/page_io.c: fix polled swap page in
checkpatch: add Co-developed-by to signature tags
docs: fix Co-Developed-by docs
drivers/base/platform.c: kmemleak ignore a known leak
fs: don't open code lru_to_page()
fs/: remove caller signal_pending branch predictions
mm/: remove caller signal_pending branch predictions
arch/arc/mm/fault.c: remove caller signal_pending_branch predictions
kernel/sched/: remove caller signal_pending branch predictions
kernel/locking/mutex.c: remove caller signal_pending branch predictions
mm: select HAVE_MOVE_PMD on x86 for faster mremap
mm: speed up mremap by 20x on large regions
mm: treewide: remove unused address argument from pte_alloc functions
initramfs: cleanup incomplete rootfs
scripts/gdb: fix lx-version string output
kernel/kcov.c: mark write_comp_data() as notrace
kernel/sysctl: add panic_print into sysctl
panic: add options to print system info when panic happens
bfs: extra sanity checking and static inode bitmap
exec: separate MM_ANONPAGES and RLIMIT_STACK accounting
...
We need to return a dma_addr_t even if we don't have a kernel mapping.
Do so by consolidating the phys_to_dma call in a single place and jump
to it from all the branches that return successfully.
Fixes: bfd56cd605 ("dma-mapping: support highmem in the generic remap allocator")
Reported-by: Liviu Dudau <liviu@dudau.co.uk
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Liviu Dudau <liviu@dudau.co.uk>
Since __sanitizer_cov_trace_const_cmp4 is marked as notrace, the
function called from __sanitizer_cov_trace_const_cmp4 shouldn't be
traceable either. ftrace_graph_caller() gets called every time func
write_comp_data() gets called if it isn't marked 'notrace'. This is the
backtrace from gdb:
#0 ftrace_graph_caller () at ../arch/arm64/kernel/entry-ftrace.S:179
#1 0xffffff8010201920 in ftrace_caller () at ../arch/arm64/kernel/entry-ftrace.S:151
#2 0xffffff8010439714 in write_comp_data (type=5, arg1=0, arg2=0, ip=18446743524224276596) at ../kernel/kcov.c:116
#3 0xffffff8010439894 in __sanitizer_cov_trace_const_cmp4 (arg1=<optimized out>, arg2=<optimized out>) at ../kernel/kcov.c:188
#4 0xffffff8010201874 in prepare_ftrace_return (self_addr=18446743524226602768, parent=0xffffff801014b918, frame_pointer=18446743524223531344) at ./include/generated/atomic-instrumented.h:27
#5 0xffffff801020194c in ftrace_graph_caller () at ../arch/arm64/kernel/entry-ftrace.S:182
Rework so that write_comp_data() that are called from
__sanitizer_cov_trace_*_cmp*() are marked as 'notrace'.
Commit 903e8ff867 ("kernel/kcov.c: mark funcs in __sanitizer_cov_trace_pc() as notrace")
missed to mark write_comp_data() as 'notrace'. When that patch was
created gcc-7 was used. In lib/Kconfig.debug
config KCOV_ENABLE_COMPARISONS
depends on $(cc-option,-fsanitize-coverage=trace-cmp)
That code path isn't hit with gcc-7. However, it were that with gcc-8.
Link: http://lkml.kernel.org/r/20181206143011.23719-1-anders.roxell@linaro.org
Signed-off-by: Anders Roxell <anders.roxell@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Co-developed-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Cc: Will Deacon <will.deacon@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Based on commit 401c636a0e ("kernel/hung_task.c: show all hung tasks
before panic"), we could get the call stack of hung task.
However, if the console loglevel is not high, we still can not see the
useful panic information in practice, and in most cases users don't set
console loglevel to high level.
This patch is to force console verbose before system panic, so that the
real useful information can be seen in the console, instead of being
like the following, which doesn't have hung task information.
INFO: task init:1 blocked for more than 120 seconds.
Tainted: G U W 4.19.0-quilt-2e5dc0ac-g51b6c21d76cc #1
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
Kernel panic - not syncing: hung_task: blocked tasks
CPU: 2 PID: 479 Comm: khungtaskd Tainted: G U W 4.19.0-quilt-2e5dc0ac-g51b6c21d76cc #1
Call Trace:
dump_stack+0x4f/0x65
panic+0xde/0x231
watchdog+0x290/0x410
kthread+0x12c/0x150
ret_from_fork+0x35/0x40
reboot: panic mode set: p,w
Kernel Offset: 0x34000000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff)
Link: http://lkml.kernel.org/r/27240C0AC20F114CBF8149A2696CBE4A6015B675@SHSMSX101.ccr.corp.intel.com
Signed-off-by: Chuansheng Liu <chuansheng.liu@intel.com>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Reviewed-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Tetsuo Handa <penguin-kernel@i-love.sakura.ne.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If the number of input parameters is less than the total parameters, an
EINVAL error will be returned.
For example, we use proc_doulongvec_minmax to pass up to two parameters
with kern_table:
{
.procname = "monitor_signals",
.data = &monitor_sigs,
.maxlen = 2*sizeof(unsigned long),
.mode = 0644,
.proc_handler = proc_doulongvec_minmax,
},
Reproduce:
When passing two parameters, it's work normal. But passing only one
parameter, an error "Invalid argument"(EINVAL) is returned.
[root@cl150 ~]# echo 1 2 > /proc/sys/kernel/monitor_signals
[root@cl150 ~]# cat /proc/sys/kernel/monitor_signals
1 2
[root@cl150 ~]# echo 3 > /proc/sys/kernel/monitor_signals
-bash: echo: write error: Invalid argument
[root@cl150 ~]# echo $?
1
[root@cl150 ~]# cat /proc/sys/kernel/monitor_signals
3 2
[root@cl150 ~]#
The following is the result after apply this patch. No error is
returned when the number of input parameters is less than the total
parameters.
[root@cl150 ~]# echo 1 2 > /proc/sys/kernel/monitor_signals
[root@cl150 ~]# cat /proc/sys/kernel/monitor_signals
1 2
[root@cl150 ~]# echo 3 > /proc/sys/kernel/monitor_signals
[root@cl150 ~]# echo $?
0
[root@cl150 ~]# cat /proc/sys/kernel/monitor_signals
3 2
[root@cl150 ~]#
There are three processing functions dealing with digital parameters,
__do_proc_dointvec/__do_proc_douintvec/__do_proc_doulongvec_minmax.
This patch deals with __do_proc_doulongvec_minmax, just as
__do_proc_dointvec does, adding a check for parameters 'left'. In
__do_proc_douintvec, its code implementation explicitly does not support
multiple inputs.
static int __do_proc_douintvec(...){
...
/*
* Arrays are not supported, keep this simple. *Do not* add
* support for them.
*/
if (vleft != 1) {
*lenp = 0;
return -EINVAL;
}
...
}
So, just __do_proc_doulongvec_minmax has the problem. And most use of
proc_doulongvec_minmax/proc_doulongvec_ms_jiffies_minmax just have one
parameter.
Link: http://lkml.kernel.org/r/1544081775-15720-1-git-send-email-cheng.lin130@zte.com.cn
Signed-off-by: Cheng Lin <cheng.lin130@zte.com.cn>
Acked-by: Luis Chamberlain <mcgrof@kernel.org>
Reviewed-by: Kees Cook <keescook@chromium.org>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Originally, the rule used to be that you'd have to do access_ok()
separately, and then user_access_begin() before actually doing the
direct (optimized) user access.
But experience has shown that people then decide not to do access_ok()
at all, and instead rely on it being implied by other operations or
similar. Which makes it very hard to verify that the access has
actually been range-checked.
If you use the unsafe direct user accesses, hardware features (either
SMAP - Supervisor Mode Access Protection - on x86, or PAN - Privileged
Access Never - on ARM) do force you to use user_access_begin(). But
nothing really forces the range check.
By putting the range check into user_access_begin(), we actually force
people to do the right thing (tm), and the range check vill be visible
near the actual accesses. We have way too long a history of people
trying to avoid them.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Now that the slow path DMA API calls are implemented out of line a few
helpers only used by them don't need to be exported anymore.
Signed-off-by: Christoph Hellwig <hch@lst.de>
dmam_alloc_coherent is just the default no-flags case of
dmam_alloc_attrs, so take advantage of this similar to the non-managed
version.
Signed-off-by: Christoph Hellwig <hch@lst.de>
And also switch the way we implement the unmap side around to stay
consistent. This ensures dma-debug works again because it records which
function we used for mapping to ensure it is also used for unmapping,
and also reduces further code duplication. Last but not least this
also officially allows calling dma_sync_single_* for mappings created
using dma_map_page, which is perfectly fine given that the sync calls
only take a dma_addr_t, but not a virtual address or struct page.
Fixes: 7f0fee242e ("dma-mapping: merge dma_unmap_page_attrs and dma_unmap_single_attrs")
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: LABBE Corentin <clabbe.montjoie@gmail.com>
Nobody has actually used the type (VERIFY_READ vs VERIFY_WRITE) argument
of the user address range verification function since we got rid of the
old racy i386-only code to walk page tables by hand.
It existed because the original 80386 would not honor the write protect
bit when in kernel mode, so you had to do COW by hand before doing any
user access. But we haven't supported that in a long time, and these
days the 'type' argument is a purely historical artifact.
A discussion about extending 'user_access_begin()' to do the range
checking resulted this patch, because there is no way we're going to
move the old VERIFY_xyz interface to that model. And it's best done at
the end of the merge window when I've done most of my merges, so let's
just get this done once and for all.
This patch was mostly done with a sed-script, with manual fix-ups for
the cases that weren't of the trivial 'access_ok(VERIFY_xyz' form.
There were a couple of notable cases:
- csky still had the old "verify_area()" name as an alias.
- the iter_iov code had magical hardcoded knowledge of the actual
values of VERIFY_{READ,WRITE} (not that they mattered, since nothing
really used it)
- microblaze used the type argument for a debug printout
but other than those oddities this should be a total no-op patch.
I tried to fix up all architectures, did fairly extensive grepping for
access_ok() uses, and the changes are trivial, but I may have missed
something. Any missed conversion should be trivially fixable, though.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pull networking fixes from David Miller:
"Several fixes here. Basically split down the line between newly
introduced regressions and long existing problems:
1) Double free in tipc_enable_bearer(), from Cong Wang.
2) Many fixes to nf_conncount, from Florian Westphal.
3) op->get_regs_len() can throw an error, check it, from Yunsheng
Lin.
4) Need to use GFP_ATOMIC in *_add_hash_mac_address() of fsl/fman
driver, from Scott Wood.
5) Inifnite loop in fib_empty_table(), from Yue Haibing.
6) Use after free in ax25_fillin_cb(), from Cong Wang.
7) Fix socket locking in nr_find_socket(), also from Cong Wang.
8) Fix WoL wakeup enable in r8169, from Heiner Kallweit.
9) On 32-bit sock->sk_stamp is not thread-safe, from Deepa Dinamani.
10) Fix ptr_ring wrap during queue swap, from Cong Wang.
11) Missing shutdown callback in hinic driver, from Xue Chaojing.
12) Need to return NULL on error from ip6_neigh_lookup(), from Stefano
Brivio.
13) BPF out of bounds speculation fixes from Daniel Borkmann"
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (57 commits)
ipv6: Consider sk_bound_dev_if when binding a socket to an address
ipv6: Fix dump of specific table with strict checking
bpf: add various test cases to selftests
bpf: prevent out of bounds speculation on pointer arithmetic
bpf: fix check_map_access smin_value test when pointer contains offset
bpf: restrict unknown scalars of mixed signed bounds for unprivileged
bpf: restrict stack pointer arithmetic for unprivileged
bpf: restrict map value pointer arithmetic for unprivileged
bpf: enable access to ax register also from verifier rewrite
bpf: move tmp variable into ax register in interpreter
bpf: move {prev_,}insn_idx into verifier env
isdn: fix kernel-infoleak in capi_unlocked_ioctl
ipv6: route: Fix return value of ip6_neigh_lookup() on neigh_create() error
net/hamradio/6pack: use mod_timer() to rearm timers
net-next/hinic:add shutdown callback
net: hns3: call hns3_nic_net_open() while doing HNAE3_UP_CLIENT
ip: validate header length on virtual device xmit
tap: call skb_probe_transport_header after setting skb->dev
ptr_ring: wrap back ->producer in __ptr_ring_swap_queue()
net: rds: remove unnecessary NULL check
...
Pull NFS client updates from Anna Schumaker:
"Stable bugfixes:
- xprtrdma: Yet another double DMA-unmap # v4.20
Features:
- Allow some /proc/sys/sunrpc entries without CONFIG_SUNRPC_DEBUG
- Per-xprt rdma receive workqueues
- Drop support for FMR memory registration
- Make port= mount option optional for RDMA mounts
Other bugfixes and cleanups:
- Remove unused nfs4_xdev_fs_type declaration
- Fix comments for behavior that has changed
- Remove generic RPC credentials by switching to 'struct cred'
- Fix crossing mountpoints with different auth flavors
- Various xprtrdma fixes from testing and auditing the close code
- Fixes for disconnect issues when using xprtrdma with krb5
- Clean up and improve xprtrdma trace points
- Fix NFS v4.2 async copy reboot recovery"
* tag 'nfs-for-4.21-1' of git://git.linux-nfs.org/projects/anna/linux-nfs: (63 commits)
sunrpc: convert to DEFINE_SHOW_ATTRIBUTE
sunrpc: Add xprt after nfs4_test_session_trunk()
sunrpc: convert unnecessary GFP_ATOMIC to GFP_NOFS
sunrpc: handle ENOMEM in rpcb_getport_async
NFS: remove unnecessary test for IS_ERR(cred)
xprtrdma: Prevent leak of rpcrdma_rep objects
NFSv4.2 fix async copy reboot recovery
xprtrdma: Don't leak freed MRs
xprtrdma: Add documenting comment for rpcrdma_buffer_destroy
xprtrdma: Replace outdated comment for rpcrdma_ep_post
xprtrdma: Update comments in frwr_op_send
SUNRPC: Fix some kernel doc complaints
SUNRPC: Simplify defining common RPC trace events
NFS: Fix NFSv4 symbolic trace point output
xprtrdma: Trace mapping, alloc, and dereg failures
xprtrdma: Add trace points for calls to transport switch methods
xprtrdma: Relocate the xprtrdma_mr_map trace points
xprtrdma: Clean up of xprtrdma chunk trace points
xprtrdma: Remove unused fields from rpcrdma_ia
xprtrdma: Cull dprintk() call sites
...
Jann reported that the original commit back in b2157399cc
("bpf: prevent out-of-bounds speculation") was not sufficient
to stop CPU from speculating out of bounds memory access:
While b2157399cc only focussed on masking array map access
for unprivileged users for tail calls and data access such
that the user provided index gets sanitized from BPF program
and syscall side, there is still a more generic form affected
from BPF programs that applies to most maps that hold user
data in relation to dynamic map access when dealing with
unknown scalars or "slow" known scalars as access offset, for
example:
- Load a map value pointer into R6
- Load an index into R7
- Do a slow computation (e.g. with a memory dependency) that
loads a limit into R8 (e.g. load the limit from a map for
high latency, then mask it to make the verifier happy)
- Exit if R7 >= R8 (mispredicted branch)
- Load R0 = R6[R7]
- Load R0 = R6[R0]
For unknown scalars there are two options in the BPF verifier
where we could derive knowledge from in order to guarantee
safe access to the memory: i) While </>/<=/>= variants won't
allow to derive any lower or upper bounds from the unknown
scalar where it would be safe to add it to the map value
pointer, it is possible through ==/!= test however. ii) another
option is to transform the unknown scalar into a known scalar,
for example, through ALU ops combination such as R &= <imm>
followed by R |= <imm> or any similar combination where the
original information from the unknown scalar would be destroyed
entirely leaving R with a constant. The initial slow load still
precedes the latter ALU ops on that register, so the CPU
executes speculatively from that point. Once we have the known
scalar, any compare operation would work then. A third option
only involving registers with known scalars could be crafted
as described in [0] where a CPU port (e.g. Slow Int unit)
would be filled with many dependent computations such that
the subsequent condition depending on its outcome has to wait
for evaluation on its execution port and thereby executing
speculatively if the speculated code can be scheduled on a
different execution port, or any other form of mistraining
as described in [1], for example. Given this is not limited
to only unknown scalars, not only map but also stack access
is affected since both is accessible for unprivileged users
and could potentially be used for out of bounds access under
speculation.
In order to prevent any of these cases, the verifier is now
sanitizing pointer arithmetic on the offset such that any
out of bounds speculation would be masked in a way where the
pointer arithmetic result in the destination register will
stay unchanged, meaning offset masked into zero similar as
in array_index_nospec() case. With regards to implementation,
there are three options that were considered: i) new insn
for sanitation, ii) push/pop insn and sanitation as inlined
BPF, iii) reuse of ax register and sanitation as inlined BPF.
Option i) has the downside that we end up using from reserved
bits in the opcode space, but also that we would require
each JIT to emit masking as native arch opcodes meaning
mitigation would have slow adoption till everyone implements
it eventually which is counter-productive. Option ii) and iii)
have both in common that a temporary register is needed in
order to implement the sanitation as inlined BPF since we
are not allowed to modify the source register. While a push /
pop insn in ii) would be useful to have in any case, it
requires once again that every JIT needs to implement it
first. While possible, amount of changes needed would also
be unsuitable for a -stable patch. Therefore, the path which
has fewer changes, less BPF instructions for the mitigation
and does not require anything to be changed in the JITs is
option iii) which this work is pursuing. The ax register is
already mapped to a register in all JITs (modulo arm32 where
it's mapped to stack as various other BPF registers there)
and used in constant blinding for JITs-only so far. It can
be reused for verifier rewrites under certain constraints.
The interpreter's tmp "register" has therefore been remapped
into extending the register set with hidden ax register and
reusing that for a number of instructions that needed the
prior temporary variable internally (e.g. div, mod). This
allows for zero increase in stack space usage in the interpreter,
and enables (restricted) generic use in rewrites otherwise as
long as such a patchlet does not make use of these instructions.
The sanitation mask is dynamic and relative to the offset the
map value or stack pointer currently holds.
There are various cases that need to be taken under consideration
for the masking, e.g. such operation could look as follows:
ptr += val or val += ptr or ptr -= val. Thus, the value to be
sanitized could reside either in source or in destination
register, and the limit is different depending on whether
the ALU op is addition or subtraction and depending on the
current known and bounded offset. The limit is derived as
follows: limit := max_value_size - (smin_value + off). For
subtraction: limit := umax_value + off. This holds because
we do not allow any pointer arithmetic that would
temporarily go out of bounds or would have an unknown
value with mixed signed bounds where it is unclear at
verification time whether the actual runtime value would
be either negative or positive. For example, we have a
derived map pointer value with constant offset and bounded
one, so limit based on smin_value works because the verifier
requires that statically analyzed arithmetic on the pointer
must be in bounds, and thus it checks if resulting
smin_value + off and umax_value + off is still within map
value bounds at time of arithmetic in addition to time of
access. Similarly, for the case of stack access we derive
the limit as follows: MAX_BPF_STACK + off for subtraction
and -off for the case of addition where off := ptr_reg->off +
ptr_reg->var_off.value. Subtraction is a special case for
the masking which can be in form of ptr += -val, ptr -= -val,
or ptr -= val. In the first two cases where we know that
the value is negative, we need to temporarily negate the
value in order to do the sanitation on a positive value
where we later swap the ALU op, and restore original source
register if the value was in source.
The sanitation of pointer arithmetic alone is still not fully
sufficient as is, since a scenario like the following could
happen ...
PTR += 0x1000 (e.g. K-based imm)
PTR -= BIG_NUMBER_WITH_SLOW_COMPARISON
PTR += 0x1000
PTR -= BIG_NUMBER_WITH_SLOW_COMPARISON
[...]
... which under speculation could end up as ...
PTR += 0x1000
PTR -= 0 [ truncated by mitigation ]
PTR += 0x1000
PTR -= 0 [ truncated by mitigation ]
[...]
... and therefore still access out of bounds. To prevent such
case, the verifier is also analyzing safety for potential out
of bounds access under speculative execution. Meaning, it is
also simulating pointer access under truncation. We therefore
"branch off" and push the current verification state after the
ALU operation with known 0 to the verification stack for later
analysis. Given the current path analysis succeeded it is
likely that the one under speculation can be pruned. In any
case, it is also subject to existing complexity limits and
therefore anything beyond this point will be rejected. In
terms of pruning, it needs to be ensured that the verification
state from speculative execution simulation must never prune
a non-speculative execution path, therefore, we mark verifier
state accordingly at the time of push_stack(). If verifier
detects out of bounds access under speculative execution from
one of the possible paths that includes a truncation, it will
reject such program.
Given we mask every reg-based pointer arithmetic for
unprivileged programs, we've been looking into how it could
affect real-world programs in terms of size increase. As the
majority of programs are targeted for privileged-only use
case, we've unconditionally enabled masking (with its alu
restrictions on top of it) for privileged programs for the
sake of testing in order to check i) whether they get rejected
in its current form, and ii) by how much the number of
instructions and size will increase. We've tested this by
using Katran, Cilium and test_l4lb from the kernel selftests.
For Katran we've evaluated balancer_kern.o, Cilium bpf_lxc.o
and an older test object bpf_lxc_opt_-DUNKNOWN.o and l4lb
we've used test_l4lb.o as well as test_l4lb_noinline.o. We
found that none of the programs got rejected by the verifier
with this change, and that impact is rather minimal to none.
balancer_kern.o had 13,904 bytes (1,738 insns) xlated and
7,797 bytes JITed before and after the change. Most complex
program in bpf_lxc.o had 30,544 bytes (3,817 insns) xlated
and 18,538 bytes JITed before and after and none of the other
tail call programs in bpf_lxc.o had any changes either. For
the older bpf_lxc_opt_-DUNKNOWN.o object we found a small
increase from 20,616 bytes (2,576 insns) and 12,536 bytes JITed
before to 20,664 bytes (2,582 insns) and 12,558 bytes JITed
after the change. Other programs from that object file had
similar small increase. Both test_l4lb.o had no change and
remained at 6,544 bytes (817 insns) xlated and 3,401 bytes
JITed and for test_l4lb_noinline.o constant at 5,080 bytes
(634 insns) xlated and 3,313 bytes JITed. This can be explained
in that LLVM typically optimizes stack based pointer arithmetic
by using K-based operations and that use of dynamic map access
is not overly frequent. However, in future we may decide to
optimize the algorithm further under known guarantees from
branch and value speculation. Latter seems also unclear in
terms of prediction heuristics that today's CPUs apply as well
as whether there could be collisions in e.g. the predictor's
Value History/Pattern Table for triggering out of bounds access,
thus masking is performed unconditionally at this point but could
be subject to relaxation later on. We were generally also
brainstorming various other approaches for mitigation, but the
blocker was always lack of available registers at runtime and/or
overhead for runtime tracking of limits belonging to a specific
pointer. Thus, we found this to be minimally intrusive under
given constraints.
With that in place, a simple example with sanitized access on
unprivileged load at post-verification time looks as follows:
# bpftool prog dump xlated id 282
[...]
28: (79) r1 = *(u64 *)(r7 +0)
29: (79) r2 = *(u64 *)(r7 +8)
30: (57) r1 &= 15
31: (79) r3 = *(u64 *)(r0 +4608)
32: (57) r3 &= 1
33: (47) r3 |= 1
34: (2d) if r2 > r3 goto pc+19
35: (b4) (u32) r11 = (u32) 20479 |
36: (1f) r11 -= r2 | Dynamic sanitation for pointer
37: (4f) r11 |= r2 | arithmetic with registers
38: (87) r11 = -r11 | containing bounded or known
39: (c7) r11 s>>= 63 | scalars in order to prevent
40: (5f) r11 &= r2 | out of bounds speculation.
41: (0f) r4 += r11 |
42: (71) r4 = *(u8 *)(r4 +0)
43: (6f) r4 <<= r1
[...]
For the case where the scalar sits in the destination register
as opposed to the source register, the following code is emitted
for the above example:
[...]
16: (b4) (u32) r11 = (u32) 20479
17: (1f) r11 -= r2
18: (4f) r11 |= r2
19: (87) r11 = -r11
20: (c7) r11 s>>= 63
21: (5f) r2 &= r11
22: (0f) r2 += r0
23: (61) r0 = *(u32 *)(r2 +0)
[...]
JIT blinding example with non-conflicting use of r10:
[...]
d5: je 0x0000000000000106 _
d7: mov 0x0(%rax),%edi |
da: mov $0xf153246,%r10d | Index load from map value and
e0: xor $0xf153259,%r10 | (const blinded) mask with 0x1f.
e7: and %r10,%rdi |_
ea: mov $0x2f,%r10d |
f0: sub %rdi,%r10 | Sanitized addition. Both use r10
f3: or %rdi,%r10 | but do not interfere with each
f6: neg %r10 | other. (Neither do these instructions
f9: sar $0x3f,%r10 | interfere with the use of ax as temp
fd: and %r10,%rdi | in interpreter.)
100: add %rax,%rdi |_
103: mov 0x0(%rdi),%eax
[...]
Tested that it fixes Jann's reproducer, and also checked that test_verifier
and test_progs suite with interpreter, JIT and JIT with hardening enabled
on x86-64 and arm64 runs successfully.
[0] Speculose: Analyzing the Security Implications of Speculative
Execution in CPUs, Giorgi Maisuradze and Christian Rossow,
https://arxiv.org/pdf/1801.04084.pdf
[1] A Systematic Evaluation of Transient Execution Attacks and
Defenses, Claudio Canella, Jo Van Bulck, Michael Schwarz,
Moritz Lipp, Benjamin von Berg, Philipp Ortner, Frank Piessens,
Dmitry Evtyushkin, Daniel Gruss,
https://arxiv.org/pdf/1811.05441.pdf
Fixes: b2157399cc ("bpf: prevent out-of-bounds speculation")
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
In check_map_access() we probe actual bounds through __check_map_access()
with offset of reg->smin_value + off for lower bound and offset of
reg->umax_value + off for the upper bound. However, even though the
reg->smin_value could have a negative value, the final result of the
sum with off could be positive when pointer arithmetic with known and
unknown scalars is combined. In this case we reject the program with
an error such as "R<x> min value is negative, either use unsigned index
or do a if (index >=0) check." even though the access itself would be
fine. Therefore extend the check to probe whether the actual resulting
reg->smin_value + off is less than zero.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
For unknown scalars of mixed signed bounds, meaning their smin_value is
negative and their smax_value is positive, we need to reject arithmetic
with pointer to map value. For unprivileged the goal is to mask every
map pointer arithmetic and this cannot reliably be done when it is
unknown at verification time whether the scalar value is negative or
positive. Given this is a corner case, the likelihood of breaking should
be very small.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Restrict stack pointer arithmetic for unprivileged users in that
arithmetic itself must not go out of bounds as opposed to the actual
access later on. Therefore after each adjust_ptr_min_max_vals() with
a stack pointer as a destination we simulate a check_stack_access()
of 1 byte on the destination and once that fails the program is
rejected for unprivileged program loads. This is analog to map
value pointer arithmetic and needed for masking later on.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Restrict map value pointer arithmetic for unprivileged users in that
arithmetic itself must not go out of bounds as opposed to the actual
access later on. Therefore after each adjust_ptr_min_max_vals() with a
map value pointer as a destination it will simulate a check_map_access()
of 1 byte on the destination and once that fails the program is rejected
for unprivileged program loads. We use this later on for masking any
pointer arithmetic with the remainder of the map value space. The
likelihood of breaking any existing real-world unprivileged eBPF
program is very small for this corner case.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Right now we are using BPF ax register in JIT for constant blinding as
well as in interpreter as temporary variable. Verifier will not be able
to use it simply because its use will get overridden from the former in
bpf_jit_blind_insn(). However, it can be made to work in that blinding
will be skipped if there is prior use in either source or destination
register on the instruction. Taking constraints of ax into account, the
verifier is then open to use it in rewrites under some constraints. Note,
ax register already has mappings in every eBPF JIT.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
This change moves the on-stack 64 bit tmp variable in ___bpf_prog_run()
into the hidden ax register. The latter is currently only used in JITs
for constant blinding as a temporary scratch register, meaning the BPF
interpreter will never see the use of ax. Therefore it is safe to use
it for the cases where tmp has been used earlier. This is needed to later
on allow restricted hidden use of ax in both interpreter and JITs.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Move prev_insn_idx and insn_idx from the do_check() function into
the verifier environment, so they can be read inside the various
helper functions for handling the instructions. It's easier to put
this into the environment rather than changing all call-sites only
to pass it along. insn_idx is useful in particular since this later
on allows to hold state in env->insn_aux_data[env->insn_idx].
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Pull seccomp updates from James Morris:
- Add SECCOMP_RET_USER_NOTIF
- seccomp fixes for sparse warnings and s390 build (Tycho)
* 'next-seccomp' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/linux-security:
seccomp, s390: fix build for syscall type change
seccomp: fix poor type promotion
samples: add an example of seccomp user trap
seccomp: add a return code to trap to userspace
seccomp: switch system call argument type to void *
seccomp: hoist struct seccomp_data recalculation higher
Pull kgdb updates from Daniel Thompson:
"Mostly clean ups although while Doug's was chasing down a odd lockdep
warning he also did some work to improved debugger resilience when
some CPUs fail to respond to the round up request.
The main changes are:
- Fixing a lockdep warning on architectures that cannot use an NMI
for the round up plus related changes to make CPU round up and all
CPU backtrace more resilient.
- Constify the arch ops tables
- A couple of other small clean ups
Two of the three patchsets here include changes that spill over into
arch/. Changes in the arch space are relatively narrow in scope (and
directly related to kgdb). Didn't get comprehensive acks but all
impacted maintainers were Cc:ed in good time"
* tag 'kgdb-4.21-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/danielt/linux:
kgdb/treewide: constify struct kgdb_arch arch_kgdb_ops
mips/kgdb: prepare arch_kgdb_ops for constness
kdb: use bool for binary state indicators
kdb: Don't back trace on a cpu that didn't round up
kgdb: Don't round up a CPU that failed rounding up before
kgdb: Fix kgdb_roundup_cpus() for arches who used smp_call_function()
kgdb: Remove irq flags from roundup
Pull tracing updates from Steven Rostedt:
- Rework of the kprobe/uprobe and synthetic events to consolidate all
the dynamic event code. This will make changes in the future easier.
- Partial rewrite of the function graph tracing infrastructure. This
will allow for multiple users of hooking onto functions to get the
callback (return) of the function. This is the ground work for having
kprobes and function graph tracer using one code base.
- Clean up of the histogram code that will facilitate adding more
features to the histograms in the future.
- Addition of str_has_prefix() and a few use cases. There currently is
a similar function strstart() that is used in a few places, but only
returns a bool and not a length. These instances will be removed in
the future to use str_has_prefix() instead.
- A few other various clean ups as well.
* tag 'trace-v4.21' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace: (57 commits)
tracing: Use the return of str_has_prefix() to remove open coded numbers
tracing: Have the historgram use the result of str_has_prefix() for len of prefix
tracing: Use str_has_prefix() instead of using fixed sizes
tracing: Use str_has_prefix() helper for histogram code
string.h: Add str_has_prefix() helper function
tracing: Make function ‘ftrace_exports’ static
tracing: Simplify printf'ing in seq_print_sym
tracing: Avoid -Wformat-nonliteral warning
tracing: Merge seq_print_sym_short() and seq_print_sym_offset()
tracing: Add hist trigger comments for variable-related fields
tracing: Remove hist trigger synth_var_refs
tracing: Use hist trigger's var_ref array to destroy var_refs
tracing: Remove open-coding of hist trigger var_ref management
tracing: Use var_refs[] for hist trigger reference checking
tracing: Change strlen to sizeof for hist trigger static strings
tracing: Remove unnecessary hist trigger struct field
tracing: Fix ftrace_graph_get_ret_stack() to use task and not current
seq_buf: Use size_t for len in seq_buf_puts()
seq_buf: Make seq_buf_puts() null-terminate the buffer
arm64: Use ftrace_graph_get_ret_stack() instead of curr_ret_stack
...
Pull scheduler fix from Ingo Molnar:
"This is a revert for a lockup in cgroups-intense workloads - the real
fixes will come later"
* 'sched-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
sched/fair: Fix infinite loop in update_blocked_averages() by reverting a9e7f6544b
Zhipeng Xie, Xie XiuQi and Sargun Dhillon reported lockups in the
scheduler under high loads, starting at around the v4.18 time frame,
and Zhipeng Xie tracked it down to bugs in the rq->leaf_cfs_rq_list
manipulation.
Do a (manual) revert of:
a9e7f6544b ("sched/fair: Fix O(nr_cgroups) in load balance path")
It turns out that the list_del_leaf_cfs_rq() introduced by this commit
is a surprising property that was not considered in followup commits
such as:
9c2791f936 ("sched/fair: Fix hierarchical order in rq->leaf_cfs_rq_list")
As Vincent Guittot explains:
"I think that there is a bigger problem with commit a9e7f6544b and
cfs_rq throttling:
Let take the example of the following topology TG2 --> TG1 --> root:
1) The 1st time a task is enqueued, we will add TG2 cfs_rq then TG1
cfs_rq to leaf_cfs_rq_list and we are sure to do the whole branch in
one path because it has never been used and can't be throttled so
tmp_alone_branch will point to leaf_cfs_rq_list at the end.
2) Then TG1 is throttled
3) and we add TG3 as a new child of TG1.
4) The 1st enqueue of a task on TG3 will add TG3 cfs_rq just before TG1
cfs_rq and tmp_alone_branch will stay on rq->leaf_cfs_rq_list.
With commit a9e7f6544b, we can del a cfs_rq from rq->leaf_cfs_rq_list.
So if the load of TG1 cfs_rq becomes NULL before step 2) above, TG1
cfs_rq is removed from the list.
Then at step 4), TG3 cfs_rq is added at the beginning of rq->leaf_cfs_rq_list
but tmp_alone_branch still points to TG3 cfs_rq because its throttled
parent can't be enqueued when the lock is released.
tmp_alone_branch doesn't point to rq->leaf_cfs_rq_list whereas it should.
So if TG3 cfs_rq is removed or destroyed before tmp_alone_branch
points on another TG cfs_rq, the next TG cfs_rq that will be added,
will be linked outside rq->leaf_cfs_rq_list - which is bad.
In addition, we can break the ordering of the cfs_rq in
rq->leaf_cfs_rq_list but this ordering is used to update and
propagate the update from leaf down to root."
Instead of trying to work through all these cases and trying to reproduce
the very high loads that produced the lockup to begin with, simplify
the code temporarily by reverting a9e7f6544b - which change was clearly
not thought through completely.
This (hopefully) gives us a kernel that doesn't lock up so people
can continue to enjoy their holidays without worrying about regressions. ;-)
[ mingo: Wrote changelog, fixed weird spelling in code comment while at it. ]
Analyzed-by: Xie XiuQi <xiexiuqi@huawei.com>
Analyzed-by: Vincent Guittot <vincent.guittot@linaro.org>
Reported-by: Zhipeng Xie <xiezhipeng1@huawei.com>
Reported-by: Sargun Dhillon <sargun@sargun.me>
Reported-by: Xie XiuQi <xiexiuqi@huawei.com>
Tested-by: Zhipeng Xie <xiezhipeng1@huawei.com>
Tested-by: Sargun Dhillon <sargun@sargun.me>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Acked-by: Vincent Guittot <vincent.guittot@linaro.org>
Cc: <stable@vger.kernel.org> # v4.13+
Cc: Bin Li <huawei.libin@huawei.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Fixes: a9e7f6544b ("sched/fair: Fix O(nr_cgroups) in load balance path")
Link: http://lkml.kernel.org/r/1545879866-27809-1-git-send-email-xiexiuqi@huawei.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
defcmd_in_progress is the state trace for command group processing
- within a command group or not - usable is an indicator if a command
set is valid (allocated/non-empty) - so use a bool for those binary
indication here.
Signed-off-by: Nicholas Mc Guire <hofrat@osadl.org>
Reviewed-by: Daniel Thompson <daniel.thompson@linaro.org>
Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org>
If you have a CPU that fails to round up and then run 'btc' you'll end
up crashing in kdb becaue we dereferenced NULL. Let's add a check.
It's wise to also set the task to NULL when leaving the debugger so
that if we fail to round up on a later entry into the debugger we
won't backtrace a stale task.
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Acked-by: Daniel Thompson <daniel.thompson@linaro.org>
Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org>
If we're using the default implementation of kgdb_roundup_cpus() that
uses smp_call_function_single_async() we can end up hanging
kgdb_roundup_cpus() if we try to round up a CPU that failed to round
up before.
Specifically smp_call_function_single_async() will try to wait on the
csd lock for the CPU that we're trying to round up. If the previous
round up never finished then that lock could still be held and we'll
just sit there hanging.
There's not a lot of use trying to round up a CPU that failed to round
up before. Let's keep a flag that indicates whether the CPU started
but didn't finish to round up before. If we see that flag set then
we'll skip the next round up.
In general we have a few goals here:
- We never want to end up calling smp_call_function_single_async()
when the csd is still locked. This is accomplished because
flush_smp_call_function_queue() unlocks the csd _before_ invoking
the callback. That means that when kgdb_nmicallback() runs we know
for sure the the csd is no longer locked. Thus when we set
"rounding_up = false" we know for sure that the csd is unlocked.
- If there are no timeouts rounding up we should never skip a round
up.
NOTE #1: In general trying to continue running after failing to round
up CPUs doesn't appear to be supported in the debugger. When I
simulate this I find that kdb reports "Catastrophic error detected"
when I try to continue. I can overrule and continue anyway, but it
should be noted that we may be entering the land of dragons here.
Possibly the "Catastrophic error detected" was added _because_ of the
future failure to round up, but even so this is an area of the code
that hasn't been strongly tested.
NOTE #2: I did a bit of testing before and after this change. I
introduced a 10 second hang in the kernel while holding a spinlock
that I could invoke on a certain CPU with 'taskset -c 3 cat /sys/...".
Before this change if I did:
- Invoke hang
- Enter debugger
- g (which warns about Catastrophic error, g again to go anyway)
- g
- Enter debugger
...I'd hang the rest of the 10 seconds without getting a debugger
prompt. After this change I end up in the debugger the 2nd time after
only 1 second with the standard warning about 'Timed out waiting for
secondary CPUs.'
I'll also note that once the CPU finished waiting I could actually
debug it (aka "btc" worked)
I won't promise that everything works perfectly if the errant CPU
comes back at just the wrong time (like as we're entering or exiting
the debugger) but it certainly seems like an improvement.
NOTE #3: setting 'kgdb_info[cpu].rounding_up = false' is in
kgdb_nmicallback() instead of kgdb_call_nmi_hook() because some
implementations override kgdb_call_nmi_hook(). It shouldn't hurt to
have it in kgdb_nmicallback() in any case.
NOTE #4: this logic is really only needed because there is no API call
like "smp_try_call_function_single_async()" or "smp_csd_is_locked()".
If such an API existed then we'd use it instead, but it seemed a bit
much to add an API like this just for kgdb.
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Acked-by: Daniel Thompson <daniel.thompson@linaro.org>
Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org>
When I had lockdep turned on and dropped into kgdb I got a nice splat
on my system. Specifically it hit:
DEBUG_LOCKS_WARN_ON(current->hardirq_context)
Specifically it looked like this:
sysrq: SysRq : DEBUG
------------[ cut here ]------------
DEBUG_LOCKS_WARN_ON(current->hardirq_context)
WARNING: CPU: 0 PID: 0 at .../kernel/locking/lockdep.c:2875 lockdep_hardirqs_on+0xf0/0x160
CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.19.0 #27
pstate: 604003c9 (nZCv DAIF +PAN -UAO)
pc : lockdep_hardirqs_on+0xf0/0x160
...
Call trace:
lockdep_hardirqs_on+0xf0/0x160
trace_hardirqs_on+0x188/0x1ac
kgdb_roundup_cpus+0x14/0x3c
kgdb_cpu_enter+0x53c/0x5cc
kgdb_handle_exception+0x180/0x1d4
kgdb_compiled_brk_fn+0x30/0x3c
brk_handler+0x134/0x178
do_debug_exception+0xfc/0x178
el1_dbg+0x18/0x78
kgdb_breakpoint+0x34/0x58
sysrq_handle_dbg+0x54/0x5c
__handle_sysrq+0x114/0x21c
handle_sysrq+0x30/0x3c
qcom_geni_serial_isr+0x2dc/0x30c
...
...
irq event stamp: ...45
hardirqs last enabled at (...44): [...] __do_softirq+0xd8/0x4e4
hardirqs last disabled at (...45): [...] el1_irq+0x74/0x130
softirqs last enabled at (...42): [...] _local_bh_enable+0x2c/0x34
softirqs last disabled at (...43): [...] irq_exit+0xa8/0x100
---[ end trace adf21f830c46e638 ]---
Looking closely at it, it seems like a really bad idea to be calling
local_irq_enable() in kgdb_roundup_cpus(). If nothing else that seems
like it could violate spinlock semantics and cause a deadlock.
Instead, let's use a private csd alongside
smp_call_function_single_async() to round up the other CPUs. Using
smp_call_function_single_async() doesn't require interrupts to be
enabled so we can remove the offending bit of code.
In order to avoid duplicating this across all the architectures that
use the default kgdb_roundup_cpus(), we'll add a "weak" implementation
to debug_core.c.
Looking at all the people who previously had copies of this code,
there were a few variants. I've attempted to keep the variants
working like they used to. Specifically:
* For arch/arc we passed NULL to kgdb_nmicallback() instead of
get_irq_regs().
* For arch/mips there was a bit of extra code around
kgdb_nmicallback()
NOTE: In this patch we will still get into trouble if we try to round
up a CPU that failed to round up before. We'll try to round it up
again and potentially hang when we try to grab the csd lock. That's
not new behavior but we'll still try to do better in a future patch.
Suggested-by: Daniel Thompson <daniel.thompson@linaro.org>
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Richard Kuo <rkuo@codeaurora.org>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Paul Burton <paul.burton@mips.com>
Cc: James Hogan <jhogan@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Rich Felker <dalias@libc.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Borislav Petkov <bp@alien8.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Acked-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Daniel Thompson <daniel.thompson@linaro.org>
Pull Kconfig updates from Masahiro Yamada:
- support -y option for merge_config.sh to avoid downgrading =y to =m
- remove S_OTHER symbol type, and touch include/config/*.h files correctly
- fix file name and line number in lexer warnings
- fix memory leak when EOF is encountered in quotation
- resolve all shift/reduce conflicts of the parser
- warn no new line at end of file
- make 'source' statement more strict to take only string literal
- rewrite the lexer and remove the keyword lookup table
- convert to SPDX License Identifier
- compile C files independently instead of including them from zconf.y
- fix various warnings of gconfig
- misc cleanups
* tag 'kconfig-v4.21' of git://git.kernel.org/pub/scm/linux/kernel/git/masahiroy/linux-kbuild: (39 commits)
kconfig: surround dbg_sym_flags with #ifdef DEBUG to fix gconf warning
kconfig: split images.c out of qconf.cc/gconf.c to fix gconf warnings
kconfig: add static qualifiers to fix gconf warnings
kconfig: split the lexer out of zconf.y
kconfig: split some C files out of zconf.y
kconfig: convert to SPDX License Identifier
kconfig: remove keyword lookup table entirely
kconfig: update current_pos in the second lexer
kconfig: switch to ASSIGN_VAL state in the second lexer
kconfig: stop associating kconf_id with yylval
kconfig: refactor end token rules
kconfig: stop supporting '.' and '/' in unquoted words
treewide: surround Kconfig file paths with double quotes
microblaze: surround string default in Kconfig with double quotes
kconfig: use T_WORD instead of T_VARIABLE for variables
kconfig: use specific tokens instead of T_ASSIGN for assignments
kconfig: refactor scanning and parsing "option" properties
kconfig: use distinct tokens for type and default properties
kconfig: remove redundant token defines
kconfig: rename depends_list to comment_option_list
...
Pull cgroup updates from Tejun Heo:
- Waiman's cgroup2 cpuset support has been finally merged closing one
of the last remaining feature gaps.
- cgroup.procs could show non-leader threads when cgroup2 threaded mode
was used in certain ways. I forgot to push the fix during the last
cycle.
- A patch to fix mount option parsing when all mount options have been
consumed by someone else (LSM).
- cgroup_no_v1 boot param can now block named cgroup1 hierarchies too.
* 'for-4.21' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/cgroup:
cgroup: Add named hierarchy disabling to cgroup_no_v1 boot param
cgroup: fix parsing empty mount option string
cpuset: Remove set but not used variable 'cs'
cgroup: fix CSS_TASK_ITER_PROCS
cgroup: Add .__DEBUG__. prefix to debug file names
cpuset: Minor cgroup2 interface updates
cpuset: Expose cpuset.cpus.subpartitions with cgroup_debug
cpuset: Add documentation about the new "cpuset.sched.partition" flag
cpuset: Use descriptive text when reading/writing cpuset.sched.partition
cpuset: Expose cpus.effective and mems.effective on cgroup v2 root
cpuset: Make generate_sched_domains() work with partition
cpuset: Make CPU hotplug work with partition
cpuset: Track cpusets that use parent's effective_cpus
cpuset: Add an error state to cpuset.sched.partition
cpuset: Add new v2 cpuset.sched.partition flag
cpuset: Simply allocation and freeing of cpumasks
cpuset: Define data structures to support scheduling partition
cpuset: Enable cpuset controller in default hierarchy
cgroup: remove unnecessary unlikely()
Pull driver core updates from Greg KH:
"Here is the "big" set of driver core patches for 4.21-rc1.
It's not really big, just a number of small changes for some reported
issues, some documentation updates to hopefully make it harder for
people to abuse the driver model, and some other minor cleanups.
All of these have been in linux-next for a while with no reported
issues"
* tag 'driver-core-4.21-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core:
mm, memory_hotplug: update a comment in unregister_memory()
component: convert to DEFINE_SHOW_ATTRIBUTE
sysfs: Disable lockdep for driver bind/unbind files
driver core: Add missing dev->bus->need_parent_lock checks
kobject: return error code if writing /sys/.../uevent fails
driver core: Move async_synchronize_full call
driver core: platform: Respect return code of platform_device_register_full()
kref/kobject: Improve documentation
drivers/base/memory.c: Use DEVICE_ATTR_RO and friends
driver core: Replace simple_strto{l,ul} by kstrtou{l,ul}
kernfs: Improve kernfs_notify() poll notification latency
kobject: Fix warnings in lib/kobject_uevent.c
kobject: drop unnecessary cast "%llu" for u64
driver core: fix comments for device_block_probing()
driver core: Replace simple_strtol by kstrtoint
Al Viro
Jonathan Neuschäfer
Linus Torvalds
Shakeel Butt
David Herrmann
Masahiro Yamada
Christoph Hellwig
Davidlohr Bueso
Anders Roxell
Feng Tang
Yi Wang
Tetsuo Handa
Liu, Chuansheng
Cheng Lin
Daniel Borkmann
Nicholas Mc Guire
Douglas Anderson