linux.git
15 months agoRevert "kernfs: convert kernfs_idr_lock to an irq safe raw spinlock"
Tejun Heo [Tue, 9 Jan 2024 21:48:02 +0000 (11:48 -1000)]
Revert "kernfs: convert kernfs_idr_lock to an irq safe raw spinlock"

This reverts commit dad3fb67ca1cbef87ce700e83a55835e5921ce8a.

The commit converted kernfs_idr_lock to an IRQ-safe raw_spinlock because it
could be acquired while holding an rq lock through bpf_cgroup_from_id().
However, kernfs_idr_lock is held while doing GPF_NOWAIT allocations which
involves acquiring an non-IRQ-safe and non-raw lock leading to the following
lockdep warning:

  =============================
  [ BUG: Invalid wait context ]
  6.7.0-rc5-kzm9g-00251-g655022a45b1c #578 Not tainted
  -----------------------------
  swapper/0/0 is trying to lock:
  dfbcd488 (&c->lock){....}-{3:3}, at: local_lock_acquire+0x0/0xa4
  other info that might help us debug this:
  context-{5:5}
  2 locks held by swapper/0/0:
   #0: dfbc9c60 (lock){+.+.}-{3:3}, at: local_lock_acquire+0x0/0xa4
   #1: c0c012a8 (kernfs_idr_lock){....}-{2:2}, at: __kernfs_new_node.constprop.0+0x68/0x258
  stack backtrace:
  CPU: 0 PID: 0 Comm: swapper/0 Not tainted 6.7.0-rc5-kzm9g-00251-g655022a45b1c #578
  Hardware name: Generic SH73A0 (Flattened Device Tree)
   unwind_backtrace from show_stack+0x10/0x14
   show_stack from dump_stack_lvl+0x68/0x90
   dump_stack_lvl from __lock_acquire+0x3cc/0x168c
   __lock_acquire from lock_acquire+0x274/0x30c
   lock_acquire from local_lock_acquire+0x28/0xa4
   local_lock_acquire from ___slab_alloc+0x234/0x8a8
   ___slab_alloc from __slab_alloc.constprop.0+0x30/0x44
   __slab_alloc.constprop.0 from kmem_cache_alloc+0x7c/0x148
   kmem_cache_alloc from radix_tree_node_alloc.constprop.0+0x44/0xdc
   radix_tree_node_alloc.constprop.0 from idr_get_free+0x110/0x2b8
   idr_get_free from idr_alloc_u32+0x9c/0x108
   idr_alloc_u32 from idr_alloc_cyclic+0x50/0xb8
   idr_alloc_cyclic from __kernfs_new_node.constprop.0+0x88/0x258
   __kernfs_new_node.constprop.0 from kernfs_create_root+0xbc/0x154
   kernfs_create_root from sysfs_init+0x18/0x5c
   sysfs_init from mnt_init+0xc4/0x220
   mnt_init from vfs_caches_init+0x6c/0x88
   vfs_caches_init from start_kernel+0x474/0x528
   start_kernel from 0x0

Let's rever the commit. It's undesirable to spread out raw spinlock usage
anyway and the problem can be solved by protecting the lookup path with RCU
instead.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Andrea Righi <andrea.righi@canonical.com>
Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Link: http://lkml.kernel.org/r/CAMuHMdV=AKt+mwY7svEq5gFPx41LoSQZ_USME5_MEdWQze13ww@mail.gmail.com
Link: https://lore.kernel.org/r/20240109214828.252092-2-tj@kernel.org
Tested-by: Andrea Righi <andrea.righi@canonical.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agokernfs: convert kernfs_idr_lock to an irq safe raw spinlock
Andrea Righi [Fri, 29 Dec 2023 07:49:16 +0000 (08:49 +0100)]
kernfs: convert kernfs_idr_lock to an irq safe raw spinlock

bpf_cgroup_from_id() is basically a wrapper to cgroup_get_from_id(),
that is relying on kernfs to determine the right cgroup associated to
the target id.

As a kfunc, it has the potential to be attached to any function through
BPF, particularly in contexts where certain locks are held.

However, kernfs is not using an irq safe spinlock for kernfs_idr_lock,
that means any kernfs function that is acquiring this lock can be
interrupted and potentially hit bpf_cgroup_from_id() in the process,
triggering a deadlock.

For example, it is really easy to trigger a lockdep splat between
kernfs_idr_lock and rq->_lock, attaching a small BPF program to
__set_cpus_allowed_ptr_locked() that just calls bpf_cgroup_from_id():

 =====================================================
 WARNING: HARDIRQ-safe -> HARDIRQ-unsafe lock order detected
 6.7.0-rc7-virtme #5 Not tainted
 -----------------------------------------------------
 repro/131 [HC0[0]:SC0[0]:HE0:SE1] is trying to acquire:
 ffffffffb2dc4578 (kernfs_idr_lock){+.+.}-{2:2}, at: kernfs_find_and_get_node_by_id+0x1d/0x80

 and this task is already holding:
 ffff911cbecaf218 (&rq->__lock){-.-.}-{2:2}, at: task_rq_lock+0x50/0xc0
 which would create a new lock dependency:
  (&rq->__lock){-.-.}-{2:2} -> (kernfs_idr_lock){+.+.}-{2:2}

 but this new dependency connects a HARDIRQ-irq-safe lock:
  (&rq->__lock){-.-.}-{2:2}

 ... which became HARDIRQ-irq-safe at:
   lock_acquire+0xbf/0x2b0
   _raw_spin_lock_nested+0x2e/0x40
   scheduler_tick+0x5d/0x170
   update_process_times+0x9c/0xb0
   tick_periodic+0x27/0xe0
   tick_handle_periodic+0x24/0x70
   __sysvec_apic_timer_interrupt+0x64/0x1a0
   sysvec_apic_timer_interrupt+0x6f/0x80
   asm_sysvec_apic_timer_interrupt+0x1a/0x20
   memcpy+0xc/0x20
   arch_dup_task_struct+0x15/0x30
   copy_process+0x1ce/0x1eb0
   kernel_clone+0xac/0x390
   kernel_thread+0x6f/0xa0
   kthreadd+0x199/0x230
   ret_from_fork+0x31/0x50
   ret_from_fork_asm+0x1b/0x30

 to a HARDIRQ-irq-unsafe lock:
  (kernfs_idr_lock){+.+.}-{2:2}

 ... which became HARDIRQ-irq-unsafe at:
 ...
   lock_acquire+0xbf/0x2b0
   _raw_spin_lock+0x30/0x40
   __kernfs_new_node.isra.0+0x83/0x280
   kernfs_create_root+0xf6/0x1d0
   sysfs_init+0x1b/0x70
   mnt_init+0xd9/0x2a0
   vfs_caches_init+0xcf/0xe0
   start_kernel+0x58a/0x6a0
   x86_64_start_reservations+0x18/0x30
   x86_64_start_kernel+0xc5/0xe0
   secondary_startup_64_no_verify+0x178/0x17b

 other info that might help us debug this:

  Possible interrupt unsafe locking scenario:

        CPU0                    CPU1
        ----                    ----
   lock(kernfs_idr_lock);
                                local_irq_disable();
                                lock(&rq->__lock);
                                lock(kernfs_idr_lock);
   <Interrupt>
     lock(&rq->__lock);

  *** DEADLOCK ***

Prevent this deadlock condition converting kernfs_idr_lock to a raw irq
safe spinlock.

The performance impact of this change should be negligible and it also
helps to prevent similar deadlock conditions with any other subsystems
that may depend on kernfs.

Fixes: 332ea1f697be ("bpf: Add bpf_cgroup_from_id() kfunc")
Cc: stable <stable@kernel.org>
Signed-off-by: Andrea Righi <andrea.righi@canonical.com>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20231229074916.53547-1-andrea.righi@canonical.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agoclass: fix use-after-free in class_register()
Jing Xia [Wed, 20 Dec 2023 02:46:03 +0000 (10:46 +0800)]
class: fix use-after-free in class_register()

The lock_class_key is still registered and can be found in
lock_keys_hash hlist after subsys_private is freed in error
handler path.A task who iterate over the lock_keys_hash
later may cause use-after-free.So fix that up and unregister
the lock_class_key before kfree(cp).

On our platform, a driver fails to kset_register because of
creating duplicate filename '/class/xxx'.With Kasan enabled,
it prints a invalid-access bug report.

KASAN bug report:

BUG: KASAN: invalid-access in lockdep_register_key+0x19c/0x1bc
Write of size 8 at addr 15ffff808b8c0368 by task modprobe/252
Pointer tag: [15], memory tag: [fe]

CPU: 7 PID: 252 Comm: modprobe Tainted: G        W
 6.6.0-mainline-maybe-dirty #1

Call trace:
dump_backtrace+0x1b0/0x1e4
show_stack+0x2c/0x40
dump_stack_lvl+0xac/0xe0
print_report+0x18c/0x4d8
kasan_report+0xe8/0x148
__hwasan_store8_noabort+0x88/0x98
lockdep_register_key+0x19c/0x1bc
class_register+0x94/0x1ec
init_module+0xbc/0xf48 [rfkill]
do_one_initcall+0x17c/0x72c
do_init_module+0x19c/0x3f8
...
Memory state around the buggy address:
ffffff808b8c0100: 8a 8a 8a 8a 8a 8a 8a 8a 8a 8a 8a 8a 8a 8a 8a 8a
ffffff808b8c0200: 8a 8a 8a 8a 8a 8a 8a 8a fe fe fe fe fe fe fe fe
>ffffff808b8c0300: fe fe fe fe fe fe fe fe fe fe fe fe fe fe fe fe
                                     ^
ffffff808b8c0400: 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03 03

As CONFIG_KASAN_GENERIC is not set, Kasan reports invalid-access
not use-after-free here.In this case, modprobe is manipulating
the corrupted lock_keys_hash hlish where lock_class_key is already
freed before.

It's worth noting that this only can happen if lockdep is enabled,
which is not true for normal system.

Fixes: dcfbb67e48a2 ("driver core: class: use lock_class_key already present in struct subsys_private")
Cc: stable <stable@kernel.org>
Signed-off-by: Jing Xia <jing.xia@unisoc.com>
Signed-off-by: Xuewen Yan <xuewen.yan@unisoc.com>
Link: https://lore.kernel.org/r/20231220024603.186078-1-jing.xia@unisoc.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agoPM: clk: make pm_clk_add_notifier() take a const pointer
Greg Kroah-Hartman [Tue, 19 Dec 2023 14:07:23 +0000 (15:07 +0100)]
PM: clk: make pm_clk_add_notifier() take a const pointer

The driver core wants to work with const struct bus_type, so there's no
reason that pm_clk_add_notifier() should not also do the same thing,
considering that it just passes the pointer off to the driver core which
is expecting a const *.

Cc: Rafael J. Wysocki <rafael@kernel.org>
Link: https://lore.kernel.org/r/2023121922-triumph-exploit-f545@gregkh
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agoEDAC: constantify the struct bus_type usage
Greg Kroah-Hartman [Tue, 19 Dec 2023 13:13:10 +0000 (14:13 +0100)]
EDAC: constantify the struct bus_type usage

In many places in the edac code, struct bus_type pointers are passed
around and then eventually sent to the driver core, which can handle a
constant pointer.  So constantify all of the edac usage of these as well
because the data in them is never modified by the edac code either.

Cc: Borislav Petkov <bp@alien8.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: James Morse <james.morse@arm.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: Robert Richter <rric@kernel.org>
Cc: <linux-edac@vger.kernel.org>
Link: https://lore.kernel.org/r/2023121909-tribute-punctuate-4b22@gregkh
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agokernfs: fix reference to renamed function
Ahelenia Ziemiańska [Thu, 21 Dec 2023 14:15:42 +0000 (15:15 +0100)]
kernfs: fix reference to renamed function

commit c637b8acbe079edb477d887041755b489036f146 ("kernfs:
 s/sysfs/kernfs/ in internal functions and whatever is left")
renamed kernfs_file_open to kernfs_fop_open, but didn't update the
comment referencing it.

Signed-off-by: Ahelenia Ziemiańska <nabijaczleweli@nabijaczleweli.xyz>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/4f2wybrepigxjpuxj4bdkh3qmksetfioedit2bdrswf6b75ebb@tarta.nabijaczleweli.xyz
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agodriver core: device.h: fix Excess kernel-doc description warning
Randy Dunlap [Sat, 23 Dec 2023 05:05:32 +0000 (21:05 -0800)]
driver core: device.h: fix Excess kernel-doc description warning

Remove the @knode_class: line to prevent the kernel-doc warning:

include/linux/device.h:807: warning: Excess struct member 'knode_class' description in 'device'

Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Link: https://lore.kernel.org/r/20231223050532.13881-1-rdunlap@infradead.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agodriver core: class: fix Excess kernel-doc description warning
Randy Dunlap [Sat, 23 Dec 2023 05:05:22 +0000 (21:05 -0800)]
driver core: class: fix Excess kernel-doc description warning

Remove the @p: lines to prevent the kernel-doc warning:

include/linux/device/class.h:72: warning: Excess struct member 'p' description in 'class'

Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Link: https://lore.kernel.org/r/20231223050522.13867-1-rdunlap@infradead.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agodriver core: mark remaining local bus_type variables as const
Greg Kroah-Hartman [Tue, 19 Dec 2023 15:35:09 +0000 (16:35 +0100)]
driver core: mark remaining local bus_type variables as const

Now that the driver core can properly handle constant struct bus_type,
change the local driver core bus_type variables to be a constant
structure as well, placing them into read-only memory which can not be
modified at runtime.

Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Rafael J. Wysocki <rafael@kernel.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Kevin Hilman <khilman@kernel.org>
Cc: Ulf Hansson <ulf.hansson@linaro.org>
Cc: Len Brown <len.brown@intel.com>
Acked-by: William Breathitt Gray <william.gray@linaro.org>
Acked-by: Dave Ertman <david.m.ertman@intel.com>
Link: https://lore.kernel.org/r/2023121908-paver-follow-cc21@gregkh
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agodriver core: container: make container_subsys const
Greg Kroah-Hartman [Tue, 19 Dec 2023 15:03:20 +0000 (16:03 +0100)]
driver core: container: make container_subsys const

Now that the driver core can properly handle constant struct bus_type,
move the container_subsys variable to be a constant structure as well,
placing it into read-only memory which can not be modified at runtime.

Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Link: https://lore.kernel.org/r/2023121919-chatter-grumbling-9ef3@gregkh
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agodriver core: bus: constantify subsys_register() calls
Greg Kroah-Hartman [Tue, 19 Dec 2023 13:15:09 +0000 (14:15 +0100)]
driver core: bus: constantify subsys_register() calls

The functions subsys_register() and subsys_virtual_register() should be
taking a constant pointer to a struct bus_type, as they do not actually
modify anything in it, so fix up the function definitions to do so
properly.

This also changes the pointer type in struct subsys_interface to be
constant as well, as again, that's the proper signature of it.

Cc: Rafael J. Wysocki <rafael@kernel.org>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/2023121908-grove-genetics-f8af@gregkh
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agodriver core: bus: make bus_sort_breadthfirst() take a const pointer
Greg Kroah-Hartman [Tue, 19 Dec 2023 13:52:36 +0000 (14:52 +0100)]
driver core: bus: make bus_sort_breadthfirst() take a const pointer

For some reason, during the big "clean up the driver core for a const
struct bus_type" work, the bus_sort_breadthfirst() call was missed.  Fix
this up by changing the type to be a const * as it should be.

Cc: Rafael J. Wysocki <rafael@kernel.org>
Link: https://lore.kernel.org/r/2023121935-stinking-ditzy-fd5d@gregkh
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
15 months agokernfs: d_obtain_alias(NULL) will do the right thing...
Al Viro [Wed, 20 Dec 2023 05:22:29 +0000 (05:22 +0000)]
kernfs: d_obtain_alias(NULL) will do the right thing...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Link: https://lore.kernel.org/r/20231220052229.GH1674809@ZenIV
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodriver core: Better advertise dev_err_probe()
Uwe Kleine-König [Fri, 15 Dec 2023 17:45:41 +0000 (18:45 +0100)]
driver core: Better advertise dev_err_probe()

Describing the usage of dev_err_probe() as being (only?) "deemed
acceptable" has a bad connotation. In fact dev_err_probe() fulfills
three tasks:

 - handling of EPROBE_DEFER (even more than degrading to dev_dbg())
 - symbolic output of the error code
 - return err for compact error code paths

Advertise these better and claim the usage to be "fine" to get rid of
the bad connotation.

Acked-by: Rafael J. Wysocki <rafael@kernel.org>
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Link: https://lore.kernel.org/r/20231215174540.2438601-2-u.kleine-koenig@pengutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agokernfs: Convert kernfs_path_from_node_locked() from strlcpy() to strscpy()
Kees Cook [Tue, 12 Dec 2023 21:17:40 +0000 (13:17 -0800)]
kernfs: Convert kernfs_path_from_node_locked() from strlcpy() to strscpy()

One of the last remaining users of strlcpy() in the kernel is
kernfs_path_from_node_locked(), which passes back the problematic "length
we _would_ have copied" return value to indicate truncation.  Convert the
chain of all callers to use the negative return value (some of which
already doing this explicitly). All callers were already also checking
for negative return values, so the risk to missed checks looks very low.

In this analysis, it was found that cgroup1_release_agent() actually
didn't handle the "too large" condition, so this is technically also a
bug fix. :)

Here's the chain of callers, and resolution identifying each one as now
handling the correct return value:

kernfs_path_from_node_locked()
        kernfs_path_from_node()
                pr_cont_kernfs_path()
                        returns void
                kernfs_path()
                        sysfs_warn_dup()
                                return value ignored
                        cgroup_path()
                                blkg_path()
                                        bfq_bic_update_cgroup()
                                                return value ignored
                                TRACE_IOCG_PATH()
                                        return value ignored
                                TRACE_CGROUP_PATH()
                                        return value ignored
                                perf_event_cgroup()
                                        return value ignored
                                task_group_path()
                                        return value ignored
                                damon_sysfs_memcg_path_eq()
                                        return value ignored
                                get_mm_memcg_path()
                                        return value ignored
                                lru_gen_seq_show()
                                        return value ignored
                        cgroup_path_from_kernfs_id()
                                return value ignored
                cgroup_show_path()
                        already converted "too large" error to negative value
                cgroup_path_ns_locked()
                        cgroup_path_ns()
                                bpf_iter_cgroup_show_fdinfo()
                                        return value ignored
                                cgroup1_release_agent()
                                        wasn't checking "too large" error
                        proc_cgroup_show()
                                already converted "too large" to negative value

Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: Zefan Li <lizefan.x@bytedance.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Waiman Long <longman@redhat.com>
Cc: <cgroups@vger.kernel.org>
Co-developed-by: Azeem Shaikh <azeemshaikh38@gmail.com>
Signed-off-by: Azeem Shaikh <azeemshaikh38@gmail.com>
Link: https://lore.kernel.org/r/20231116192127.1558276-3-keescook@chromium.org
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231212211741.164376-3-keescook@chromium.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agokernfs: Convert kernfs_name_locked() from strlcpy() to strscpy()
Kees Cook [Tue, 12 Dec 2023 21:17:39 +0000 (13:17 -0800)]
kernfs: Convert kernfs_name_locked() from strlcpy() to strscpy()

strlcpy() reads the entire source buffer first. This read may exceed
the destination size limit. This is both inefficient and can lead
to linear read overflows if a source string is not NUL-terminated[1].
Additionally, it returns the size of the source string, not the
resulting size of the destination string. In an effort to remove strlcpy()
completely[2], replace strlcpy() here with strscpy().

Nothing actually checks the return value coming from kernfs_name_locked(),
so this has no impact on error paths. The caller hierarchy is:

kernfs_name_locked()
        kernfs_name()
                pr_cont_kernfs_name()
                        return value ignored
                cgroup_name()
                        current_css_set_cg_links_read()
                                return value ignored
                        print_page_owner_memcg()
                                return value ignored

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strlcpy
Link: https://github.com/KSPP/linux/issues/89
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: Azeem Shaikh <azeemshaikh38@gmail.com>
Link: https://lore.kernel.org/r/20231116192127.1558276-2-keescook@chromium.org
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231212211741.164376-2-keescook@chromium.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agokernfs: Convert kernfs_walk_ns() from strlcpy() to strscpy()
Kees Cook [Tue, 12 Dec 2023 21:17:38 +0000 (13:17 -0800)]
kernfs: Convert kernfs_walk_ns() from strlcpy() to strscpy()

strlcpy() reads the entire source buffer first. This read may exceed
the destination size limit. This is both inefficient and can lead
to linear read overflows if a source string is not NUL-terminated[1].
Additionally, it returns the size of the source string, not the
resulting size of the destination string. In an effort to remove strlcpy()
completely[2], replace strlcpy() here with strscpy().

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strlcpy
Link: https://github.com/KSPP/linux/issues/89
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Tejun Heo <tj@kernel.org>
Cc: Azeem Shaikh <azeemshaikh38@gmail.com>
Link: https://lore.kernel.org/r/20231116192127.1558276-1-keescook@chromium.org
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231212211741.164376-1-keescook@chromium.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoinitramfs: Expose retained initrd as sysfs file
Alexander Graf [Thu, 7 Dec 2023 23:56:54 +0000 (23:56 +0000)]
initramfs: Expose retained initrd as sysfs file

When the kernel command line option "retain_initrd" is set, we do not
free the initrd memory. However, we also don't expose it to anyone for
consumption. That leaves us in a weird situation where the only user of
this feature is ppc64 and arm64 specific kexec tooling.

To make it more generally useful, this patch adds a kobject to the
firmware object that contains the initrd context when "retain_initrd"
is set. That way, we can access the initrd any time after boot from
user space and for example hand it into kexec as --initrd parameter
if we want to reboot the same initrd. Or inspect it directly locally.

With this patch applied, there is a new /sys/firmware/initrd file when
the kernel was booted with an initrd and "retain_initrd" command line
option is set.

Signed-off-by: Alexander Graf <graf@amazon.com>
Tested-by: Bagas Sanjaya <bagasdotme@gmail.com>
Link: https://lore.kernel.org/r/20231207235654.16622-1-graf@amazon.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agofs/kernfs/dir: obey S_ISGID
Max Kellermann [Fri, 8 Dec 2023 09:33:10 +0000 (10:33 +0100)]
fs/kernfs/dir: obey S_ISGID

Handling of S_ISGID is usually done by inode_init_owner() in all other
filesystems, but kernfs doesn't use that function.  In kernfs, struct
kernfs_node is the primary data structure, and struct inode is only
created from it on demand.  Therefore, inode_init_owner() can't be
used and we need to imitate its behavior.

S_ISGID support is useful for the cgroup filesystem; it allows
subtrees managed by an unprivileged process to retain a certain owner
gid, which then enables sharing access to the subtree with another
unprivileged process.

--
v1 -> v2: minor coding style fix (comment)

Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20231208093310.297233-2-max.kellermann@ionos.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agokernel/cgroup: use kernfs_create_dir_ns()
Max Kellermann [Fri, 8 Dec 2023 09:33:09 +0000 (10:33 +0100)]
kernel/cgroup: use kernfs_create_dir_ns()

By passing the fsugid to kernfs_create_dir_ns(), we don't need
cgroup_kn_set_ugid() any longer.  That function was added for exactly
this purpose by commit 49957f8e2a43 ("cgroup: newly created dirs and
files should be owned by the creator").

Eliminating this piece of duplicate code means we benefit from future
improvements to kernfs_create_dir_ns(); for example, both are lacking
S_ISGID support currently, which my next patch will add to
kernfs_create_dir_ns().  It cannot (easily) be added to
cgroup_kn_set_ugid() because we can't dereference struct kernfs_iattrs
from there.

--
v1 -> v2: 12-digit commit id

Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Acked-by: Tejun Heo <tj@kernel.org>
Link: https://lore.kernel.org/r/20231208093310.297233-1-max.kellermann@ionos.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoMerge tag 'device_is_big_endian-6.8-rc1' into driver-core-next
Greg Kroah-Hartman [Fri, 15 Dec 2023 14:49:51 +0000 (15:49 +0100)]
Merge tag 'device_is_big_endian-6.8-rc1' into driver-core-next

Tag for the device_is_big_endian() addition to property.h

For others to be able to pull from in a stable way.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodevice property: Implement device_is_big_endian()
Andy Shevchenko [Wed, 25 Oct 2023 18:42:57 +0000 (21:42 +0300)]
device property: Implement device_is_big_endian()

Some users want to use the struct device pointer to see if the
device is big endian in terms of Open Firmware specifications,
i.e. if it has a "big-endian" property, or if the kernel was
compiled for BE *and* the device has a "native-endian" property.

Provide inline helper for the users.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20231025184259.250588-2-andriy.shevchenko@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodriver core: Emit reason for pending deferred probe
Uwe Kleine-König [Wed, 22 Nov 2023 09:33:33 +0000 (10:33 +0100)]
driver core: Emit reason for pending deferred probe

Ending a boot log with

platform 3f202000.mmc: deferred probe pending

is already a nice hint about the problem. Sometimes there is a more
detailed error indicator available, add that to the output.

Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Link: https://lore.kernel.org/r/20231122093332.274145-2-u.kleine-koenig@pengutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodriver core: Enable fw_devlink=rpm by default
Saravana Kannan [Mon, 13 Nov 2023 22:09:47 +0000 (14:09 -0800)]
driver core: Enable fw_devlink=rpm by default

fw_devlink=on has stabilized and is working correctly. Let's start using
device links created by fw_devlink to also enforce runtime PM ordering.

Signed-off-by: Saravana Kannan <saravanak@google.com>
Link: https://lore.kernel.org/r/20231113220948.80089-1-saravanak@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodevice property: fwnode_property_get_reference_args allows NULL args now
Sakari Ailus [Thu, 9 Nov 2023 10:10:10 +0000 (12:10 +0200)]
device property: fwnode_property_get_reference_args allows NULL args now

All three fwnode_property_get_reference_args() implemantations now allow
args argument to be NULL. Document this.

Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Link: https://lore.kernel.org/r/20231109101010.1329587-4-sakari.ailus@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agosoftware node: Let args be NULL in software_node_get_reference_args
Sakari Ailus [Thu, 9 Nov 2023 10:10:09 +0000 (12:10 +0200)]
software node: Let args be NULL in software_node_get_reference_args

fwnode_get_property_reference_args() may not be called with args argument
NULL and while OF already supports this. Add the missing NULL check.

The purpose is to be able to count the references.

Fixes: b06184acf751 ("software node: Add software_node_get_reference_args()")
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Link: https://lore.kernel.org/r/20231109101010.1329587-3-sakari.ailus@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoacpi: property: Let args be NULL in __acpi_node_get_property_reference
Sakari Ailus [Thu, 9 Nov 2023 10:10:08 +0000 (12:10 +0200)]
acpi: property: Let args be NULL in __acpi_node_get_property_reference

fwnode_get_property_reference_args() may not be called with args argument
NULL on ACPI, OF already supports this. Add the missing NULL checks and
document this.

The purpose is to be able to count the references.

Fixes: 977d5ad39f3e ("ACPI: Convert ACPI reference args to generic fwnode reference args")
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Link: https://lore.kernel.org/r/20231109101010.1329587-2-sakari.ailus@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agosoftware node: Remove usage of the deprecated ida_simple_xx() API
Christophe JAILLET [Wed, 1 Nov 2023 09:36:04 +0000 (10:36 +0100)]
software node: Remove usage of the deprecated ida_simple_xx() API

ida_alloc() and ida_free() should be preferred to the deprecated
ida_simple_get() and ida_simple_remove().

This is less verbose.

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/c7cdc3566c783d106138698b1e1923351fabace8.1698831275.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agofs/sysfs/dir.c : Fix typo in comment
Mukesh Ojha [Mon, 30 Oct 2023 10:21:14 +0000 (15:51 +0530)]
fs/sysfs/dir.c : Fix typo in comment

Typo correction
kboject => kobject

Signed-off-by: Mukesh Ojha <quic_mojha@quicinc.com>
Link: https://lore.kernel.org/r/1698661274-32540-1-git-send-email-quic_mojha@quicinc.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agobase/node.c: initialize the accessor list before registering
Gregory Price [Mon, 30 Oct 2023 04:42:39 +0000 (00:42 -0400)]
base/node.c: initialize the accessor list before registering

The current code registers the node as available in the node array
before initializing the accessor list.  This makes it so that
anything which might access the accessor list as a result of
allocations will cause an undefined memory access.

In one example, an extension to access hmat data during interleave
caused this undefined access as a result of a bulk allocation
that occurs during node initialization but before the accessor
list is initialized.

Initialize the accessor list before making the node generally
available to the global system.

Fixes: 08d9dbe72b1f ("node: Link memory nodes to their compute nodes")
Signed-off-by: Gregory Price <gregory.price@memverge.com>
Link: https://lore.kernel.org/r/20231030044239.971756-1-gregory.price@memverge.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agobase: soc: Remove usage of the deprecated ida_simple_xx() API
Christophe JAILLET [Wed, 1 Nov 2023 09:33:33 +0000 (10:33 +0100)]
base: soc: Remove usage of the deprecated ida_simple_xx() API

ida_alloc() and ida_free() should be preferred to the deprecated
ida_simple_get() and ida_simple_remove().

This is less verbose.

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/f0ef849446c9b3df7d6b16b1a58d089b4c17276c.1698831191.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoriscv: convert to use arch_cpu_is_hotpluggable()
Russell King (Oracle) [Tue, 21 Nov 2023 13:45:37 +0000 (13:45 +0000)]
riscv: convert to use arch_cpu_is_hotpluggable()

Convert riscv to use the arch_cpu_is_hotpluggable() helper rather than
arch_register_cpu().

Acked-by: Palmer Dabbelt <palmer@rivosinc.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Samuel Holland <samuel.holland@sifive.com>
Tested-by: Samuel Holland <samuel.holland@sifive.com> # On HiFive Unmatched
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R4L-00Ct0d-To@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoriscv: Switch over to GENERIC_CPU_DEVICES
James Morse [Tue, 21 Nov 2023 13:45:32 +0000 (13:45 +0000)]
riscv: Switch over to GENERIC_CPU_DEVICES

Now that GENERIC_CPU_DEVICES calls arch_register_cpu(), which can be
overridden by the arch code, switch over to this to allow common code
to choose when the register_cpu() call is made.

This allows topology_init() to be removed.

This is an intermediate step to the logic being moved to drivers/acpi,
where GENERIC_CPU_DEVICES will do the work when booting with acpi=off.

This patch also has the effect of moving the registration of CPUs from
subsys to driver core initialisation, prior to any initcalls running.

Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Acked-by: Palmer Dabbelt <palmer@rivosinc.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Samuel Holland <samuel.holland@sifive.com>
Tested-by: Samuel Holland <samuel.holland@sifive.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R4G-00Ct0M-PS@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoLoongArch: convert to use arch_cpu_is_hotpluggable()
Russell King (Oracle) [Tue, 21 Nov 2023 13:45:27 +0000 (13:45 +0000)]
LoongArch: convert to use arch_cpu_is_hotpluggable()

Convert loongarch to use the arch_cpu_is_hotpluggable() helper rather
than arch_register_cpu(). Also remove the export as nothing should be
using arch_register_cpu() outside of the core kernel/acpi code.

Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R4B-00Ct0G-Kk@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoLoongArch: Use the __weak version of arch_unregister_cpu()
James Morse [Tue, 21 Nov 2023 13:45:22 +0000 (13:45 +0000)]
LoongArch: Use the __weak version of arch_unregister_cpu()

LoongArch provides its own arch_unregister_cpu(). This clears the
hotpluggable flag, then unregisters the CPU.

It isn't necessary to clear the hotpluggable flag when unregistering
a cpu. unregister_cpu() writes NULL to the percpu cpu_sys_devices
pointer, meaning cpu_is_hotpluggable() will return false, as
get_cpu_device() has returned NULL.

Remove arch_unregister_cpu() and use the __weak version.

Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R46-00Ct0A-GJ@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoLoongArch: Switch over to GENERIC_CPU_DEVICES
James Morse [Tue, 21 Nov 2023 13:45:17 +0000 (13:45 +0000)]
LoongArch: Switch over to GENERIC_CPU_DEVICES

Now that GENERIC_CPU_DEVICES calls arch_register_cpu(), which can be
overridden by the arch code, switch over to this to allow common code
to choose when the register_cpu() call is made.

This allows topology_init() to be removed.

This is an intermediate step to the logic being moved to drivers/acpi,
where GENERIC_CPU_DEVICES will do the work when booting with acpi=off.

This is a subtle change. Originally:
- on boot, topology_init() would have marked present CPUs that
  io_master() is true for as hotplug-incapable.
- if a CPU is hotplugged that is an io_master(), it can later be
  hot-unplugged.

The new behaviour is that any CPU that io_master() is true for will
now always be marked as hotplug-incapable, thus even if it was
hotplugged, it can no longer be hot-unplugged.

This patch also has the effect of moving the registration of CPUs from
subsys to driver core initialisation, prior to any initcalls running.

Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R41-00Ct04-Bg@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agox86/topology: convert to use arch_cpu_is_hotpluggable()
Russell King (Oracle) [Tue, 21 Nov 2023 13:45:12 +0000 (13:45 +0000)]
x86/topology: convert to use arch_cpu_is_hotpluggable()

Convert x86 to use the arch_cpu_is_hotpluggable() helper rather than
arch_register_cpu().

Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R3w-00Cszy-6k@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agox86/topology: use weak version of arch_unregister_cpu()
Russell King (Oracle) [Tue, 21 Nov 2023 13:45:07 +0000 (13:45 +0000)]
x86/topology: use weak version of arch_unregister_cpu()

Since the x86 version of arch_unregister_cpu() is the same as the weak
version, drop the x86 specific version.

Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R3r-00Cszs-2R@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agox86/topology: Switch over to GENERIC_CPU_DEVICES
James Morse [Tue, 21 Nov 2023 13:45:01 +0000 (13:45 +0000)]
x86/topology: Switch over to GENERIC_CPU_DEVICES

Now that GENERIC_CPU_DEVICES calls arch_register_cpu(), which can be
overridden by the arch code, switch over to this to allow common code
to choose when the register_cpu() call is made.

x86's struct cpus come from struct x86_cpu, which has no other members
or users. Remove this and use the version defined by common code.

This is an intermediate step to the logic being moved to drivers/acpi,
where GENERIC_CPU_DEVICES will do the work when booting with acpi=off.

This patch also has the effect of moving the registration of CPUs from
subsys to driver core initialisation, prior to any initcalls running.

----
Changes since RFC:
 * Fixed the second copy of arch_register_cpu() used for non-hotplug
Changes since RFC v2:
 * Remove duplicate of the weak generic arch_register_cpu(), spotted
   by Jonathan Cameron. Add note about initialisation order change.
Changes since RFC v3:
 * Adapt to removal of EXPORT_SYMBOL()s

Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R3l-00Cszm-UA@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoarm64: convert to arch_cpu_is_hotpluggable()
Russell King (Oracle) [Tue, 21 Nov 2023 13:44:56 +0000 (13:44 +0000)]
arm64: convert to arch_cpu_is_hotpluggable()

Convert arm64 to use the arch_cpu_is_hotpluggable() helper rather than
arch_register_cpu().

Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R3g-00Cszg-PP@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoarm64: setup: Switch over to GENERIC_CPU_DEVICES using arch_register_cpu()
James Morse [Tue, 21 Nov 2023 13:44:51 +0000 (13:44 +0000)]
arm64: setup: Switch over to GENERIC_CPU_DEVICES using arch_register_cpu()

To allow ACPI's _STA value to hide CPUs that are present, but not
available to online right now due to VMM or firmware policy, the
register_cpu() call needs to be made by the ACPI machinery when ACPI
is in use. This allows it to hide CPUs that are unavailable from sysfs.

Switching to GENERIC_CPU_DEVICES is an intermediate step to allow all
five ACPI architectures to be modified at once.

Switch over to GENERIC_CPU_DEVICES, and provide an arch_register_cpu()
that populates the hotpluggable flag. arch_register_cpu() is also the
interface the ACPI machinery expects.

The struct cpu in struct cpuinfo_arm64 is never used directly, remove
it to use the one GENERIC_CPU_DEVICES provides.

This changes the CPUs visible in sysfs from possible to present, but
on arm64 smp_prepare_cpus() ensures these are the same.

This patch also has the effect of moving the registration of CPUs from
subsys to driver core initialisation, prior to any initcalls running.

Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R3b-00Csza-Ku@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodrivers: base: Print a warning instead of panic() when register_cpu() fails
James Morse [Tue, 21 Nov 2023 13:44:46 +0000 (13:44 +0000)]
drivers: base: Print a warning instead of panic() when register_cpu() fails

loongarch, mips, parisc, riscv and sh all print a warning if
register_cpu() returns an error. Architectures that use
GENERIC_CPU_DEVICES call panic() instead.

Errors in this path indicate something is wrong with the firmware
description of the platform, but the kernel is able to keep running.

Downgrade this to a warning to make it easier to debug this issue.

This will allow architectures that switching over to GENERIC_CPU_DEVICES
to drop their warning, but keep the existing behaviour.

Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R3W-00CszU-GM@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodrivers: base: Move cpu_dev_init() after node_dev_init()
James Morse [Tue, 21 Nov 2023 13:44:41 +0000 (13:44 +0000)]
drivers: base: Move cpu_dev_init() after node_dev_init()

NUMA systems require the node descriptions to be ready before CPUs are
registered. This is so that the node symlinks can be created in sysfs.

Currently no NUMA platform uses GENERIC_CPU_DEVICES, meaning that CPUs
are registered by arch code, instead of cpu_dev_init().

Move cpu_dev_init() after node_dev_init() so that NUMA architectures
can use GENERIC_CPU_DEVICES.

Signed-off-by: James Morse <james.morse@arm.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R3R-00CszO-C0@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodrivers: base: add arch_cpu_is_hotpluggable()
Russell King (Oracle) [Tue, 21 Nov 2023 13:44:36 +0000 (13:44 +0000)]
drivers: base: add arch_cpu_is_hotpluggable()

The differences between architecture specific implementations of
arch_register_cpu() are down to whether the CPU is hotpluggable or not.
Rather than overriding the weak version of arch_register_cpu(), provide
a function that can be used to provide this detail instead.

Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R3M-00CszH-6r@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodrivers: base: Implement weak arch_unregister_cpu()
James Morse [Tue, 21 Nov 2023 13:44:31 +0000 (13:44 +0000)]
drivers: base: Implement weak arch_unregister_cpu()

Add arch_unregister_cpu() to allow the ACPI machinery to call
unregister_cpu(). This is enough for arm64, riscv and loongarch, but
needs to be overridden by x86 and ia64 who need to do more work.

CC: Jean-Philippe Brucker <jean-philippe@linaro.org>
Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R3H-00CszC-2n@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodrivers: base: Allow parts of GENERIC_CPU_DEVICES to be overridden
James Morse [Tue, 21 Nov 2023 13:44:25 +0000 (13:44 +0000)]
drivers: base: Allow parts of GENERIC_CPU_DEVICES to be overridden

Architectures often have extra per-cpu work that needs doing
before a CPU is registered, often to determine if a CPU is
hotpluggable.

To allow the ACPI architectures to use GENERIC_CPU_DEVICES, move
the cpu_register() call into arch_register_cpu(), which is made __weak
so architectures with extra work can override it.
This aligns with the way x86, ia64 and loongarch register hotplug CPUs
when they become present.

Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R3B-00Csz6-Uh@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodrivers: base: Use present CPUs in GENERIC_CPU_DEVICES
James Morse [Tue, 21 Nov 2023 13:44:20 +0000 (13:44 +0000)]
drivers: base: Use present CPUs in GENERIC_CPU_DEVICES

Three of the five ACPI architectures create sysfs entries using
register_cpu() for present CPUs, whereas arm64, riscv and all
GENERIC_CPU_DEVICES do this for possible CPUs.

Registering a CPU is what causes them to show up in sysfs.

It makes very little sense to register all possible CPUs. Registering
a CPU is what triggers the udev notifications allowing user-space to
react to newly added CPUs.

To allow all five ACPI architectures to use GENERIC_CPU_DEVICES, change
it to use for_each_present_cpu().

Making the ACPI architectures use GENERIC_CPU_DEVICES is a pre-requisite
step to centralise their register_cpu() logic, before moving it into the
ACPI processor driver. When we add support for register CPUs from ACPI
in a later patch, we will avoid registering CPUs in this path.

Of the ACPI architectures that register possible CPUs, arm64 and riscv
do not support making possible CPUs present as they use the weak 'always
fails' version of arch_register_cpu().

Only two of the eight architectures that use GENERIC_CPU_DEVICES have a
distinction between present and possible CPUs.

The following architectures use GENERIC_CPU_DEVICES but are not SMP,
so possible == present:
 * m68k
 * microblaze
 * nios2

The following architectures use GENERIC_CPU_DEVICES and consider
possible == present:
 * csky: setup_smp()
 * processor_probe() sets possible for all CPUs and present for all CPUs
   except the boot cpu, which will have been done by
   init/main.c::start_kernel().

um appears to be a subarchitecture of x86.

The remaining architecture using GENERIC_CPU_DEVICES are:
 * openrisc and hexagon:
   where smp_init_cpus() makes all CPUs < NR_CPUS possible,
   whereas smp_prepare_cpus() only makes CPUs < setup_max_cpus present.

After this change, openrisc and hexagon systems that use the max_cpus
command line argument would not see the other CPUs present in sysfs.
This should not be a problem as these CPUs can't be brought online as
_cpu_up() checks cpu_present().

After this change, only CPUs which are present appear in sysfs.

Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R36-00Csz0-Px@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoACPI: Move ACPI_HOTPLUG_CPU to be disabled on arm64 and riscv
James Morse [Tue, 21 Nov 2023 13:44:15 +0000 (13:44 +0000)]
ACPI: Move ACPI_HOTPLUG_CPU to be disabled on arm64 and riscv

Neither arm64 nor riscv support physical hotadd of CPUs that were not
present at boot. For arm64 much of the platform description is in static
tables which do not have update methods. arm64 does support HOTPLUG_CPU,
which is backed by a firmware interface to turn CPUs on and off.

acpi_processor_hotadd_init() and acpi_processor_remove() are for adding
and removing CPUs that were not present at boot. arm64 systems that do this
are not supported as there is currently insufficient information in the
platform description. (e.g. did the GICR get removed too?)

arm64 currently relies on the MADT enabled flag check in map_gicc_mpidr()
to prevent CPUs that were not described as present at boot from being
added to the system. Similarly, riscv relies on the same check in
map_rintc_hartid(). Both architectures also rely on the weak 'always fails'
definitions of acpi_map_cpu() and arch_register_cpu().

Subsequent changes will redefine ACPI_HOTPLUG_CPU as making possible
CPUs present. Neither arm64 nor riscv support this.

Disable ACPI_HOTPLUG_CPU for arm64 and riscv by removing 'default y' and
selecting it on the other three ACPI architectures. This allows the weak
definitions of some symbols to be removed.

Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Shaoqin Huang <shahuang@redhat.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R31-00Csyt-Jq@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoLoongarch: remove arch_*register_cpu() exports
Russell King (Oracle) [Tue, 21 Nov 2023 13:44:10 +0000 (13:44 +0000)]
Loongarch: remove arch_*register_cpu() exports

arch_register_cpu() and arch_unregister_cpu() are not used by anything
that can be a module - they are used by drivers/base/cpu.c and
drivers/acpi/acpi_processor.c, neither of which can be a module.

Remove the exports.

Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R2w-00Csyn-E2@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agox86/topology: remove arch_*register_cpu() exports
Russell King (Oracle) [Tue, 21 Nov 2023 13:44:05 +0000 (13:44 +0000)]
x86/topology: remove arch_*register_cpu() exports

arch_register_cpu() and arch_unregister_cpu() are not used by anything
that can be a module - they are used by drivers/base/cpu.c and
drivers/acpi/acpi_processor.c, neither of which can be a module.

Remove the exports.

Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R2r-00Csyh-7B@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agox86: intel_epb: Don't rely on link order
James Morse [Tue, 21 Nov 2023 13:44:00 +0000 (13:44 +0000)]
x86: intel_epb: Don't rely on link order

intel_epb_init() is called as a subsys_initcall() to register cpuhp
callbacks. The callbacks make use of get_cpu_device() which will return
NULL unless register_cpu() has been called. register_cpu() is called
from topology_init(), which is also a subsys_initcall().

This is fragile. Moving the register_cpu() to a different
subsys_initcall()  leads to a NULL dereference during boot.

Make intel_epb_init() a late_initcall(), user-space can't provide a
policy before this point anyway.

Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Acked-by: "Rafael J. Wysocki" <rafael.j.wysocki@intel.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R2m-00Csyb-2S@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoarch_topology: Make register_cpu_capacity_sysctl() tolerant to late CPUs
James Morse [Tue, 21 Nov 2023 13:43:54 +0000 (13:43 +0000)]
arch_topology: Make register_cpu_capacity_sysctl() tolerant to late CPUs

register_cpu_capacity_sysctl() adds a property to sysfs that describes
the CPUs capacity. This is done from a subsys_initcall() that assumes
all possible CPUs are registered.

With CPU hotplug, possible CPUs aren't registered until they become
present, (or for arm64 enabled). This leads to messages during boot:
| register_cpu_capacity_sysctl: too early to get CPU1 device!
and once these CPUs are added to the system, the file is missing.

Move this to a cpuhp callback, so that the file is created once
CPUs are brought online. This covers CPUs that are added late by
mechanisms like hotplug.
One observable difference is the file is now missing for offline CPUs.

Signed-off-by: James Morse <james.morse@arm.com>
Reviewed-by: Gavin Shan <gshan@redhat.com>
Signed-off-by: "Russell King (Oracle)" <rmk+kernel@armlinux.org.uk>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/r/E1r5R2g-00CsyV-Ss@rmk-PC.armlinux.org.uk
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agodriver core: make device_is_dependent() static
Greg Kroah-Hartman [Tue, 28 Nov 2023 10:28:15 +0000 (10:28 +0000)]
driver core: make device_is_dependent() static

The function device_is_dependent() is only called by the driver core
internally and should not, at this time, be called by anyone else
outside of it, so mark it as static so as not to give driver authors the
wrong idea.

Cc: Saravana Kannan <saravanak@google.com>
Acked-by: "Rafael J. Wysocki" <rafael@kernel.org>
Link: https://lore.kernel.org/r/2023112815-faculty-thud-add8@gregkh
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
16 months agoLinux 6.7-rc3
Linus Torvalds [Mon, 27 Nov 2023 03:59:33 +0000 (19:59 -0800)]
Linux 6.7-rc3

16 months agoMerge tag 'trace-v6.7-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace...
Linus Torvalds [Mon, 27 Nov 2023 03:48:20 +0000 (19:48 -0800)]
Merge tag 'trace-v6.7-rc2' of git://git./linux/kernel/git/trace/linux-trace

Pull tracing fixes from Steven Rostedt::
 "Eventfs fixes:

   - With the usage of simple_recursive_remove() recommended by Al Viro,
     the code should not be calling "d_invalidate()" itself. Doing so is
     causing crashes. The code was calling d_invalidate() on the race of
     trying to look up a file while the parent was being deleted. This
     was detected, and the added dentry was having d_invalidate() called
     on it, but the deletion of the directory was also calling
     d_invalidate() on that same dentry.

   - A fix to not free the eventfs_inode (ei) until the last dput() was
     called on its ei->dentry made the ei->dentry exist even after it
     was marked for free by setting the ei->is_freed. But code elsewhere
     still was checking if ei->dentry was NULL if ei->is_freed is set
     and would trigger WARN_ON if that was the case. That's no longer
     true and there should not be any warnings when it is true.

   - Use GFP_NOFS for allocations done under eventfs_mutex. The
     eventfs_mutex can be taken on file system reclaim, make sure that
     allocations done under that mutex do not trigger file system
     reclaim.

   - Clean up code by moving the taking of inode_lock out of the helper
     functions and into where they are needed, and not use the parameter
     to know to take it or not. It must always be held but some callers
     of the helper function have it taken when they were called.

   - Warn if the inode_lock is not held in the helper functions.

   - Warn if eventfs_start_creating() is called without a parent. As
     eventfs is underneath tracefs, all files created will have a parent
     (the top one will have a tracefs parent).

  Tracing update:

   - Add Mathieu Desnoyers as an official reviewer of the tracing subsystem"

* tag 'trace-v6.7-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
  MAINTAINERS: TRACING: Add Mathieu Desnoyers as Reviewer
  eventfs: Make sure that parent->d_inode is locked in creating files/dirs
  eventfs: Do not allow NULL parent to eventfs_start_creating()
  eventfs: Move taking of inode_lock into dcache_dir_open_wrapper()
  eventfs: Use GFP_NOFS for allocation when eventfs_mutex is held
  eventfs: Do not invalidate dentry in create_file/dir_dentry()
  eventfs: Remove expectation that ei->is_freed means ei->dentry == NULL

16 months agoMerge tag 'parisc-for-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/delle...
Linus Torvalds [Sun, 26 Nov 2023 17:59:39 +0000 (09:59 -0800)]
Merge tag 'parisc-for-6.7-rc3' of git://git./linux/kernel/git/deller/parisc-linux

Pull parisc architecture fixes from Helge Deller:
 "This patchset fixes and enforces correct section alignments for the
  ex_table, altinstructions, parisc_unwind, jump_table and bug_table
  which are created by inline assembly.

  Due to not being correctly aligned at link & load time they can
  trigger unnecessarily the kernel unaligned exception handler at
  runtime. While at it, I switched the bug table to use relative
  addresses which reduces the size of the table by half on 64-bit.

  We still had the ENOSYM and EREMOTERELEASE errno symbols as left-overs
  from HP-UX, which now trigger build-issues with glibc. We can simply
  remove them.

  Most of the patches are tagged for stable kernel series.

  Summary:

   - Drop HP-UX ENOSYM and EREMOTERELEASE return codes to avoid glibc
     build issues

   - Fix section alignments for ex_table, altinstructions, parisc unwind
     table, jump_table and bug_table

   - Reduce size of bug_table on 64-bit kernel by using relative
     pointers"

* tag 'parisc-for-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:
  parisc: Reduce size of the bug_table on 64-bit kernel by half
  parisc: Drop the HP-UX ENOSYM and EREMOTERELEASE error codes
  parisc: Use natural CPU alignment for bug_table
  parisc: Ensure 32-bit alignment on parisc unwind section
  parisc: Mark lock_aligned variables 16-byte aligned on SMP
  parisc: Mark jump_table naturally aligned
  parisc: Mark altinstructions read-only and 32-bit aligned
  parisc: Mark ex_table entries 32-bit aligned in uaccess.h
  parisc: Mark ex_table entries 32-bit aligned in assembly.h

16 months agoMerge tag 'x86-urgent-2023-11-26' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sun, 26 Nov 2023 16:42:42 +0000 (08:42 -0800)]
Merge tag 'x86-urgent-2023-11-26' of git://git./linux/kernel/git/tip/tip

Pull x86 microcode fixes from Ingo Molnar:
 "Fix/enhance x86 microcode version reporting: fix the bootup log spam,
  and remove the driver version announcement to avoid version confusion
  when distros backport fixes"

* tag 'x86-urgent-2023-11-26' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/microcode: Rework early revisions reporting
  x86/microcode: Remove the driver announcement and version

16 months agoMerge tag 'perf-urgent-2023-11-26' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sun, 26 Nov 2023 16:34:12 +0000 (08:34 -0800)]
Merge tag 'perf-urgent-2023-11-26' of git://git./linux/kernel/git/tip/tip

Pull x86 perf event fix from Ingo Molnar:
 "Fix a bug in the Intel hybrid CPUs hardware-capabilities enumeration
  code resulting in non-working events on those platforms"

* tag 'perf-urgent-2023-11-26' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/x86/intel: Correct incorrect 'or' operation for PMU capabilities

16 months agoMerge tag 'locking-urgent-2023-11-26' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 26 Nov 2023 16:30:11 +0000 (08:30 -0800)]
Merge tag 'locking-urgent-2023-11-26' of git://git./linux/kernel/git/tip/tip

Pull locking fix from Ingo Molnar:
 "Fix lockdep block chain corruption resulting in KASAN warnings"

* tag 'locking-urgent-2023-11-26' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  lockdep: Fix block chain corruption

16 months agoMerge tag '6.7-rc2-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6
Linus Torvalds [Sun, 26 Nov 2023 16:22:27 +0000 (08:22 -0800)]
Merge tag '6.7-rc2-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6

Pull smb client fixes from Steve French:

 - use after free fix in releasing multichannel interfaces

 - fixes for special file types (report char, block, FIFOs properly when
   created e.g. by NFS to Windows)

 - fixes for reporting various special file types and symlinks properly
   when using SMB1

* tag '6.7-rc2-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
  smb: client: introduce cifs_sfu_make_node()
  smb: client: set correct file type from NFS reparse points
  smb: client: introduce ->parse_reparse_point()
  smb: client: implement ->query_reparse_point() for SMB1
  cifs: fix use after free for iface while disabling secondary channels

16 months agoMerge tag 'usb-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Linus Torvalds [Sun, 26 Nov 2023 02:22:42 +0000 (18:22 -0800)]
Merge tag 'usb-6.7-rc3' of git://git./linux/kernel/git/gregkh/usb

Pull USB / PHY / Thunderbolt fixes from Greg KH:
 "Here are a number of reverts, fixes, and new device ids for 6.7-rc3
  for the USB, PHY, and Thunderbolt driver subsystems. Include in here
  are:

   - reverts of some PHY drivers that went into 6.7-rc1 that shouldn't
     have been merged yet, the author is reworking them based on review
     comments as they were using older apis that shouldn't be used
     anymore for newer drivers

   - small thunderbolt driver fixes for reported issues

   - USB driver fixes for a variety of small issues in dwc3, typec,
     xhci, and other smaller drivers.

   - new device ids for usb-serial and onboard_usb_hub drivers.

  All of these have been in linux-next with no reported issues"

* tag 'usb-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb: (33 commits)
  USB: serial: option: add Luat Air72*U series products
  USB: dwc3: qcom: fix ACPI platform device leak
  USB: dwc3: qcom: fix software node leak on probe errors
  USB: dwc3: qcom: fix resource leaks on probe deferral
  USB: dwc3: qcom: simplify wakeup interrupt setup
  USB: dwc3: qcom: fix wakeup after probe deferral
  dt-bindings: usb: qcom,dwc3: fix example wakeup interrupt types
  usb: misc: onboard-hub: add support for Microchip USB5744
  dt-bindings: usb: microchip,usb5744: Add second supply
  usb: misc: ljca: Fix enumeration error on Dell Latitude 9420
  USB: serial: option: add Fibocom L7xx modules
  USB: xhci-plat: fix legacy PHY double init
  usb: typec: tipd: Supply also I2C driver data
  usb: xhci-mtk: fix in-ep's start-split check failure
  usb: dwc3: set the dma max_seg_size
  usb: config: fix iteration issue in 'usb_get_bos_descriptor()'
  usb: dwc3: add missing of_node_put and platform_device_put
  USB: dwc2: write HCINT with INTMASK applied
  usb: misc: ljca: Drop _ADR support to get ljca children devices
  usb: cdnsp: Fix deadlock issue during using NCM gadget
  ...

16 months agoMerge tag 'xfs-6.7-fixes-3' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux
Linus Torvalds [Sat, 25 Nov 2023 16:57:09 +0000 (08:57 -0800)]
Merge tag 'xfs-6.7-fixes-3' of git://git./fs/xfs/xfs-linux

Pull xfs fix from Chandan Babu:

 - Validate quota records recovered from the log before writing them to
   the disk.

* tag 'xfs-6.7-fixes-3' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
  xfs: dquot recovery does not validate the recovered dquot
  xfs: clean up dqblk extraction

16 months agoMerge tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux
Linus Torvalds [Sat, 25 Nov 2023 16:43:46 +0000 (08:43 -0800)]
Merge tag 'arm64-fixes' of git://git./linux/kernel/git/arm64/linux

Pull arm64 fixes from Catalin Marinas:

 - Fix "rodata=on" not disabling "rodata=full" on arm64

 - Add arm64 make dependency between vmlinuz.efi and Image, leading to
   occasional build failures previously (with parallel building)

 - Add newline to the output formatting of the za-fork kselftest

* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
  arm64: add dependency between vmlinuz.efi and Image
  kselftest/arm64: Fix output formatting for za-fork
  arm64: mm: Fix "rodata=on" when CONFIG_RODATA_FULL_DEFAULT_ENABLED=y

16 months agoMerge tag 'for-linus-6.7a-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Sat, 25 Nov 2023 16:32:44 +0000 (08:32 -0800)]
Merge tag 'for-linus-6.7a-rc3-tag' of git://git./linux/kernel/git/xen/tip

Pull xen fixes from Juergen Gross:

 - A small cleanup patch for the Xen privcmd driver

 - A fix for the swiotlb-xen driver which was missing the advertising of
   the maximum mapping length

 - A fix for Xen on Arm for a longstanding bug, which happened to occur
   only recently: a structure in percpu memory crossed a page boundary,
   which was rejected by the hypervisor

* tag 'for-linus-6.7a-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
  arm/xen: fix xen_vcpu_info allocation alignment
  xen: privcmd: Replace zero-length array with flex-array member and use __counted_by
  swiotlb-xen: provide the "max_mapping_size" method

16 months agoparisc: Reduce size of the bug_table on 64-bit kernel by half
Helge Deller [Thu, 23 Nov 2023 20:57:19 +0000 (21:57 +0100)]
parisc: Reduce size of the bug_table on 64-bit kernel by half

Enable GENERIC_BUG_RELATIVE_POINTERS which will store 32-bit relative
offsets to the bug address and the source file name instead of 64-bit
absolute addresses. This effectively reduces the size of the
bug_table[] array by half on 64-bit kernels.

Signed-off-by: Helge Deller <deller@gmx.de>
16 months agoparisc: Drop the HP-UX ENOSYM and EREMOTERELEASE error codes
Helge Deller [Thu, 23 Nov 2023 19:28:27 +0000 (20:28 +0100)]
parisc: Drop the HP-UX ENOSYM and EREMOTERELEASE error codes

Those return codes are only defined for the parisc architecture and
are leftovers from when we wanted to be HP-UX compatible.

They are not returned by any Linux kernel syscall but do trigger
problems with the glibc strerrorname_np() and strerror() functions as
reported in glibc issue #31080.

There is no need to keep them, so simply remove them.

Signed-off-by: Helge Deller <deller@gmx.de>
Reported-by: Bruno Haible <bruno@clisp.org>
Closes: https://sourceware.org/bugzilla/show_bug.cgi?id=31080
Cc: stable@vger.kernel.org
16 months agoparisc: Use natural CPU alignment for bug_table
Helge Deller [Mon, 20 Nov 2023 22:30:49 +0000 (23:30 +0100)]
parisc: Use natural CPU alignment for bug_table

Make sure that the __bug_table section gets 32- or 64-bit aligned,
depending if a 32- or 64-bit kernel is being built.
Mark it non-writeable and use .blockz instead of the .org assembler
directive to pad the struct.

Signed-off-by: Helge Deller <deller@gmx.de>
Cc: stable@vger.kernel.org # v6.0+
16 months agoparisc: Ensure 32-bit alignment on parisc unwind section
Helge Deller [Sat, 25 Nov 2023 08:16:02 +0000 (09:16 +0100)]
parisc: Ensure 32-bit alignment on parisc unwind section

Make sure the .PARISC.unwind section will be 32-bit aligned.

Signed-off-by: Helge Deller <deller@gmx.de>
Cc: stable@vger.kernel.org # v6.0+
16 months agoparisc: Mark lock_aligned variables 16-byte aligned on SMP
Helge Deller [Sat, 25 Nov 2023 08:11:56 +0000 (09:11 +0100)]
parisc: Mark lock_aligned variables 16-byte aligned on SMP

On parisc we need 16-byte alignment for variables which are used for
locking. Mark the __lock_aligned attribute acordingly so that the
.data..lock_aligned section will get that alignment in the generated
object files.

Signed-off-by: Helge Deller <deller@gmx.de>
Cc: stable@vger.kernel.org # v6.0+
16 months agoparisc: Mark jump_table naturally aligned
Helge Deller [Mon, 20 Nov 2023 22:14:39 +0000 (23:14 +0100)]
parisc: Mark jump_table naturally aligned

The jump_table stores two 32-bit words and one 32- (on 32-bit kernel)
or one 64-bit word (on 64-bit kernel).
Ensure that the last word is always 64-bit aligned on a 64-bit kernel
by aligning the whole structure on sizeof(long).

Signed-off-by: Helge Deller <deller@gmx.de>
Cc: stable@vger.kernel.org # v6.0+
16 months agoparisc: Mark altinstructions read-only and 32-bit aligned
Helge Deller [Mon, 20 Nov 2023 22:10:20 +0000 (23:10 +0100)]
parisc: Mark altinstructions read-only and 32-bit aligned

Signed-off-by: Helge Deller <deller@gmx.de>
Cc: stable@vger.kernel.org # v6.0+
16 months agoparisc: Mark ex_table entries 32-bit aligned in uaccess.h
Helge Deller [Mon, 20 Nov 2023 14:39:03 +0000 (15:39 +0100)]
parisc: Mark ex_table entries 32-bit aligned in uaccess.h

Add an align statement to tell the linker that all ex_table entries and as
such the whole ex_table section should be 32-bit aligned in vmlinux and modules.

Signed-off-by: Helge Deller <deller@gmx.de>
Cc: stable@vger.kernel.org # v6.0+
16 months agoparisc: Mark ex_table entries 32-bit aligned in assembly.h
Helge Deller [Mon, 20 Nov 2023 14:37:50 +0000 (15:37 +0100)]
parisc: Mark ex_table entries 32-bit aligned in assembly.h

Add an align statement to tell the linker that all ex_table entries and as
such the whole ex_table section should be 32-bit aligned in vmlinux and modules.

Signed-off-by: Helge Deller <deller@gmx.de>
Cc: stable@vger.kernel.org # v6.0+
16 months agoMerge tag 's390-6.7-3' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux
Linus Torvalds [Fri, 24 Nov 2023 19:44:50 +0000 (11:44 -0800)]
Merge tag 's390-6.7-3' of git://git./linux/kernel/git/s390/linux

Pull s390 updates from Alexander Gordeev:

 - Remove unnecessary assignment of the performance event last_tag.

 - Create missing /sys/firmware/ipl/* attributes when kernel is booted
   in dump mode using List-directed ECKD IPL.

 - Remove odd comment.

 - Fix s390-specific part of scripts/checkstack.pl script that only
   matches three-digit numbers starting with 3 or any higher number and
   skips any stack sizes smaller than 304 bytes.

* tag 's390-6.7-3' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux:
  scripts/checkstack.pl: match all stack sizes for s390
  s390: remove odd comment
  s390/ipl: add missing IPL_TYPE_ECKD_DUMP case to ipl_init()
  s390/pai: cleanup event initialization

16 months agoMerge tag 'acpi-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael...
Linus Torvalds [Fri, 24 Nov 2023 19:30:35 +0000 (11:30 -0800)]
Merge tag 'acpi-6.7-rc3' of git://git./linux/kernel/git/rafael/linux-pm

Pull ACPI fixes from Rafael Wysocki:
 "These add an ACPI IRQ override quirk for ASUS ExpertBook B1402CVA and
  fix an ACPI processor idle issue leading to triple-faults in Xen HVM
  guests and an ACPI backlight driver issue that causes GPUs to
  misbehave while their children power is being fixed up.

  Specifics:

   - Avoid powering up GPUs while attempting to fix up power for their
     children (Hans de Goede)

   - Use raw_safe_halt() instead of safe_halt() in acpi_idle_play_dead()
     so as to avoid triple-falts during CPU online in Xen HVM guests due
     to the setting of the hardirqs_enabled flag in safe_halt() (David
     Woodhouse)

   - Add an ACPI IRQ override quirk for ASUS ExpertBook B1402CVA (Hans
     de Goede)"

* tag 'acpi-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  ACPI: resource: Skip IRQ override on ASUS ExpertBook B1402CVA
  ACPI: video: Use acpi_device_fix_up_power_children()
  ACPI: PM: Add acpi_device_fix_up_power_children() function
  ACPI: processor_idle: use raw_safe_halt() in acpi_idle_play_dead()

16 months agoMerge tag 'pm-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Linus Torvalds [Fri, 24 Nov 2023 19:26:00 +0000 (11:26 -0800)]
Merge tag 'pm-6.7-rc3' of git://git./linux/kernel/git/rafael/linux-pm

Pull power management fix from Rafael Wysocki:
 "Fix a syntax error in the sleepgraph utility which causes it to exit
  early on every invocation (David Woodhouse)"

* tag 'pm-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:
  PM: tools: Fix sleepgraph syntax error

16 months agoMerge tag 'afs-fixes-20231124' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowe...
Linus Torvalds [Fri, 24 Nov 2023 18:40:03 +0000 (10:40 -0800)]
Merge tag 'afs-fixes-20231124' of git://git./linux/kernel/git/dhowells/linux-fs

Pull AFS fixes from David Howells:

 - Fix the afs_server_list struct to be cleaned up with RCU

 - Fix afs to translate a no-data result from a DNS lookup into ENOENT,
   not EDESTADDRREQ for consistency with OpenAFS

 - Fix afs to translate a negative DNS lookup result into ENOENT rather
   than EDESTADDRREQ

 - Fix file locking on R/O volumes to operate in local mode as the
   server doesn't handle exclusive locks on such files

 - Set SB_RDONLY on superblocks for RO and Backup volumes so that the
   VFS can see that they're read only

* tag 'afs-fixes-20231124' of git://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs:
  afs: Mark a superblock for an R/O or Backup volume as SB_RDONLY
  afs: Fix file locking on R/O volumes to operate in local mode
  afs: Return ENOENT if no cell DNS record can be found
  afs: Make error on cell lookup failure consistent with OpenAFS
  afs: Fix afs_server_list to be cleaned up with RCU

16 months agoMerge branches 'acpi-video' and 'acpi-processor' into acpi
Rafael J. Wysocki [Fri, 24 Nov 2023 18:16:22 +0000 (19:16 +0100)]
Merge branches 'acpi-video' and 'acpi-processor' into acpi

Merge ACPI backlight driver fixes and an ACPI processor driver fix for
6.7-rc3:

 - Avoid powering up GPUs while attempting to fix up power for their
   children (Hans de Goede).

 - Use raw_safe_halt() instead of safe_halt() in acpi_idle_play_dead()
   so as to avoid triple-falts during CPU online in Xen HVM guests due
   to the setting of the hardirqs_enabled flag in safe_halt() (David
   Woodhouse).

* acpi-video:
  ACPI: video: Use acpi_device_fix_up_power_children()
  ACPI: PM: Add acpi_device_fix_up_power_children() function

* acpi-processor:
  ACPI: processor_idle: use raw_safe_halt() in acpi_idle_play_dead()

16 months agoMerge tag 'vfs-6.7-rc3.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Linus Torvalds [Fri, 24 Nov 2023 17:45:40 +0000 (09:45 -0800)]
Merge tag 'vfs-6.7-rc3.fixes' of git://git./linux/kernel/git/vfs/vfs

Pull vfs fixes from Christian Brauner:

 - Avoid calling back into LSMs from vfs_getattr_nosec() calls.

   IMA used to query inode properties accessing raw inode fields without
   dedicated helpers. That was finally fixed a few releases ago by
   forcing IMA to use vfs_getattr_nosec() helpers.

   The goal of the vfs_getattr_nosec() helper is to query for attributes
   without calling into the LSM layer which would be quite problematic
   because incredibly IMA is called from __fput()...

     __fput()
       -> ima_file_free()

   What it does is to call back into the filesystem to update the file's
   IMA xattr. Querying the inode without using vfs_getattr_nosec() meant
   that IMA didn't handle stacking filesystems such as overlayfs
   correctly. So the switch to vfs_getattr_nosec() is quite correct. But
   the switch to vfs_getattr_nosec() revealed another bug when used on
   stacking filesystems:

     __fput()
       -> ima_file_free()
          -> vfs_getattr_nosec()
             -> i_op->getattr::ovl_getattr()
                -> vfs_getattr()
                   -> i_op->getattr::$WHATEVER_UNDERLYING_FS_getattr()
                      -> security_inode_getattr() # calls back into LSMs

   Now, if that __fput() happens from task_work_run() of an exiting task
   current->fs and various other pointer could already be NULL. So
   anything in the LSM layer relying on that not being NULL would be
   quite surprised.

   Fix that by passing the information that this is a security request
   through to the stacking filesystem by adding a new internal
   ATT_GETATTR_NOSEC flag. Now the callchain becomes:

     __fput()
       -> ima_file_free()
          -> vfs_getattr_nosec()
             -> i_op->getattr::ovl_getattr()
                -> if (AT_GETATTR_NOSEC)
                          vfs_getattr_nosec()
                   else
                          vfs_getattr()
                   -> i_op->getattr::$WHATEVER_UNDERLYING_FS_getattr()

 - Fix a bug introduced with the iov_iter rework from last cycle.

   This broke /proc/kcore by copying too much and without the correct
   offset.

 - Add a missing NULL check when allocating the root inode in
   autofs_fill_super().

 - Fix stable writes for multi-device filesystems (xfs, btrfs etc) and
   the block device pseudo filesystem.

   Stable writes used to be a superblock flag only, making it a per
   filesystem property. Add an additional AS_STABLE_WRITES mapping flag
   to allow for fine-grained control.

 - Ensure that offset_iterate_dir() returns 0 after reaching the end of
   a directory so it adheres to getdents() convention.

* tag 'vfs-6.7-rc3.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  libfs: getdents() should return 0 after reaching EOD
  xfs: respect the stable writes flag on the RT device
  xfs: clean up FS_XFLAG_REALTIME handling in xfs_ioctl_setattr_xflags
  block: update the stable_writes flag in bdev_add
  filemap: add a per-mapping stable writes flag
  autofs: add: new_inode check in autofs_fill_super()
  iov_iter: fix copy_page_to_iter_nofault()
  fs: Pass AT_GETATTR_NOSEC flag to getattr interface function

16 months agoMerge tag 'drm-fixes-2023-11-24' of git://anongit.freedesktop.org/drm/drm
Linus Torvalds [Fri, 24 Nov 2023 17:36:33 +0000 (09:36 -0800)]
Merge tag 'drm-fixes-2023-11-24' of git://anongit.freedesktop.org/drm/drm

Pull drm fixes from Dave Airlie:
 "Back to regular scheduled fixes pull request, mainly a bunch of msm,
  some i915 and otherwise a few scattered, one memory crasher in the
  nouveau GSP paths is helping stabilise that work.

  msm:
   - Fix the VREG_CTRL_1 for 4nm CPHY to match downstream
   - Remove duplicate call to drm_kms_helper_poll_init() in
     msm_drm_init()
   - Fix the safe_lut_tbl[] for sc8280xp to match downstream
   - Don't attach the drm_dp_set_subconnector_property() for eDP
   - Fix to attach drm_dp_set_subconnector_property() for DP. Otherwise
     there is a bootup crash on multiple targets
   - Remove unnecessary NULL check left behind during cleanup

  i915:
   - Fix race between DP MST connectore registration and setup
   - Fix GT memory leak on probe error path

  panel:
   - Fixes for innolux and auo,b101uan08.3 panel.
   - Fix Himax83102-j02 timings.

  ivpu:
   - Fix ivpu MMIO reset.

  ast:
   - AST fix on connetor disconnection.

  nouveau:
   - gsp memory corruption fix

  rockchip:
   - color fix"

* tag 'drm-fixes-2023-11-24' of git://anongit.freedesktop.org/drm/drm:
  nouveau/gsp: allocate enough space for all channel ids.
  drm/panel: boe-tv101wum-nl6: Fine tune Himax83102-j02 panel HFP and HBP
  drm/ast: Disconnect BMC if physical connector is connected
  accel/ivpu/37xx: Fix hangs related to MMIO reset
  drm/rockchip: vop: Fix color for RGB888/BGR888 format on VOP full
  drm/i915: do not clean GT table on error path
  drm/i915/dp_mst: Fix race between connector registration and setup
  drm/panel: simple: Fix Innolux G101ICE-L01 timings
  drm/panel: simple: Fix Innolux G101ICE-L01 bus flags
  drm/msm: remove unnecessary NULL check
  drm/panel: auo,b101uan08.3: Fine tune the panel power sequence
  drm/msm/dp: attach the DP subconnector property
  drm/msm/dp: don't touch DP subconnector property in eDP case
  drm/msm/dpu: Add missing safe_lut_tbl in sc8280xp catalog
  drm/msm: remove exra drm_kms_helper_poll_init() call
  drm/msm/dsi: use the correct VREG_CTRL_1 value for 4nm cphy

16 months agoMerge tag 'usb-serial-6.7-rc3' of https://git.kernel.org/pub/scm/linux/kernel/git...
Greg Kroah-Hartman [Fri, 24 Nov 2023 16:30:38 +0000 (16:30 +0000)]
Merge tag 'usb-serial-6.7-rc3' of https://git./linux/kernel/git/johan/usb-serial into usb-linus

Johan writes:

USB-serial fixes for 6.7-rc3

Here are a couple of modem device entry fixes and some new modem device
ids.

All have been in linux-next with no reported issues.

* tag 'usb-serial-6.7-rc3' of https://git.kernel.org/pub/scm/linux/kernel/git/johan/usb-serial: (329 commits)
  USB: serial: option: add Luat Air72*U series products
  USB: serial: option: add Fibocom L7xx modules
  USB: serial: option: fix FM101R-GL defines
  USB: serial: option: don't claim interface 4 for ZTE MF290
  Linux 6.7-rc2
  prctl: Disable prctl(PR_SET_MDWE) on parisc
  parisc/power: Fix power soft-off when running on qemu
  parisc: Replace strlcpy() with strscpy()
  NFSD: Fix checksum mismatches in the duplicate reply cache
  NFSD: Fix "start of NFS reply" pointer passed to nfsd_cache_update()
  NFSD: Update nfsd_cache_append() to use xdr_stream
  nfsd: fix file memleak on client_opens_release
  dm-crypt: start allocating with MAX_ORDER
  dm-verity: don't use blocking calls from tasklets
  dm-bufio: fix no-sleep mode
  dm-delay: avoid duplicate logic
  dm-delay: fix bugs introduced by kthread mode
  dm-delay: fix a race between delay_presuspend and delay_bio
  drm/amdgpu/gmc9: disable AGP aperture
  drm/amdgpu/gmc10: disable AGP aperture
  ...

16 months agoafs: Mark a superblock for an R/O or Backup volume as SB_RDONLY
David Howells [Thu, 2 Nov 2023 16:24:00 +0000 (16:24 +0000)]
afs: Mark a superblock for an R/O or Backup volume as SB_RDONLY

Mark a superblock that is for for an R/O or Backup volume as SB_RDONLY when
mounting it.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org

16 months agoafs: Fix file locking on R/O volumes to operate in local mode
David Howells [Wed, 1 Nov 2023 22:03:28 +0000 (22:03 +0000)]
afs: Fix file locking on R/O volumes to operate in local mode

AFS doesn't really do locking on R/O volumes as fileservers don't maintain
state with each other and thus a lock on a R/O volume file on one
fileserver will not be be visible to someone looking at the same file on
another fileserver.

Further, the server may return an error if you try it.

Fix this by doing what other AFS clients do and handle filelocking on R/O
volume files entirely within the client and don't touch the server.

Fixes: 6c6c1d63c243 ("afs: Provide mount-time configurable byte-range file locking emulation")
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org

16 months agoafs: Return ENOENT if no cell DNS record can be found
David Howells [Thu, 26 Oct 2023 00:25:07 +0000 (01:25 +0100)]
afs: Return ENOENT if no cell DNS record can be found

Make AFS return error ENOENT if no cell SRV or AFSDB DNS record (or
cellservdb config file record) can be found rather than returning
EDESTADDRREQ.

Also add cell name lookup info to the cursor dump.

Fixes: d5c32c89b208 ("afs: Fix cell DNS lookup")
Reported-by: Markus Suvanto <markus.suvanto@gmail.com>
Link: https://bugzilla.kernel.org/show_bug.cgi?id=216637
Signed-off-by: David Howells <dhowells@redhat.com>
Reviewed-by: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org

16 months agolockdep: Fix block chain corruption
Peter Zijlstra [Tue, 21 Nov 2023 11:41:26 +0000 (12:41 +0100)]
lockdep: Fix block chain corruption

Kent reported an occasional KASAN splat in lockdep. Mark then noted:

> I suspect the dodgy access is to chain_block_buckets[-1], which hits the last 4
> bytes of the redzone and gets (incorrectly/misleadingly) attributed to
> nr_large_chain_blocks.

That would mean @size == 0, at which point size_to_bucket() returns -1
and the above happens.

alloc_chain_hlocks() has 'size - req', for the first with the
precondition 'size >= rq', which allows the 0.

This code is trying to split a block, del_chain_block() takes what we
need, and add_chain_block() puts back the remainder, except in the
above case the remainder is 0 sized and things go sideways.

Fixes: 810507fe6fd5 ("locking/lockdep: Reuse freed chain_hlocks entries")
Reported-by: Kent Overstreet <kent.overstreet@linux.dev>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Kent Overstreet <kent.overstreet@linux.dev>
Link: https://lkml.kernel.org/r/20231121114126.GH8262@noisy.programming.kicks-ass.net
16 months agoMerge tag 'ata-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal...
Linus Torvalds [Fri, 24 Nov 2023 01:45:49 +0000 (17:45 -0800)]
Merge tag 'ata-6.7-rc3' of git://git./linux/kernel/git/dlemoal/libata

Pull ata fix from Damien Le Moal:

 - Add a missing error check in the adapter initialization of the
   pata_isapnp driver (Chen)

* tag 'ata-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/dlemoal/libata:
  ata: pata_isapnp: Add missing error check for devm_ioport_map()

16 months agoMerge tag 'block-6.7-2023-11-23' of git://git.kernel.dk/linux
Linus Torvalds [Fri, 24 Nov 2023 01:40:15 +0000 (17:40 -0800)]
Merge tag 'block-6.7-2023-11-23' of git://git.kernel.dk/linux

Pull block fixes from Jens Axboe:
 "A bit bigger than usual at this time, but nothing really earth
  shattering:

   - NVMe pull request via Keith:
       - TCP TLS fixes (Hannes)
       - Authentifaction fixes (Mark, Hannes)
       - Properly terminate target names (Christoph)

   - MD pull request via Song, fixing a raid5 corruption issue

   - Disentanglement of the dependency mess in nvme introduced with the
     tls additions. Now it should actually build on all configs (Arnd)

   - Series of bcache fixes (Coly)

   - Removal of a dead helper (Damien)

   - s390 dasd fix (Muhammad, Jan)

   - lockdep blk-cgroup fixes (Ming)"

* tag 'block-6.7-2023-11-23' of git://git.kernel.dk/linux: (33 commits)
  nvme: tcp: fix compile-time checks for TLS mode
  nvme: target: fix Kconfig select statements
  nvme: target: fix nvme_keyring_id() references
  nvme: move nvme_stop_keep_alive() back to original position
  nbd: pass nbd_sock to nbd_read_reply() instead of index
  s390/dasd: protect device queue against concurrent access
  s390/dasd: resolve spelling mistake
  block/null_blk: Fix double blk_mq_start_request() warning
  nvmet-tcp: always initialize tls_handshake_tmo_work
  nvmet: nul-terminate the NQNs passed in the connect command
  nvme: blank out authentication fabrics options if not configured
  nvme: catch errors from nvme_configure_metadata()
  nvme-tcp: only evaluate 'tls' option if TLS is selected
  nvme-auth: set explanation code for failure2 msgs
  nvme-auth: unlock mutex in one place only
  block: Remove blk_set_runtime_active()
  nbd: fix null-ptr-dereference while accessing 'nbd->config'
  nbd: factor out a helper to get nbd_config without holding 'config_lock'
  nbd: fold nbd config initialization into nbd_alloc_config()
  bcache: avoid NULL checking to c->root in run_cache_set()
  ...

16 months agoMerge tag 'io_uring-6.7-2023-11-23' of git://git.kernel.dk/linux
Linus Torvalds [Fri, 24 Nov 2023 01:36:29 +0000 (17:36 -0800)]
Merge tag 'io_uring-6.7-2023-11-23' of git://git.kernel.dk/linux

Pull io_uring fixes from Jens Axboe:
 "A fix for ensuring that LINKAT always propagates flags correctly, and
  a fix for an off-by-one in segment skipping for registered buffers.

  Both heading to stable as well"

* tag 'io_uring-6.7-2023-11-23' of git://git.kernel.dk/linux:
  io_uring: fix off-by one bvec index
  io_uring/fs: consider link->flags when getting path for LINKAT

16 months agoMerge tag 'for-linus-2023112301' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 24 Nov 2023 01:31:53 +0000 (17:31 -0800)]
Merge tag 'for-linus-2023112301' of git://git./linux/kernel/git/hid/hid

Pull HID fixes from Jiri Kosina:

 - revert of commit that caused regression to many Logitech unifying
   receiver users (Jiri Kosina)

 - power management fix for hid-mcp2221 (Hamish Martin)

 - fix for race condition between HID core and HID debug (Charles Yi)

 - a couple of assorted device-ID-specific quirks

* tag 'for-linus-2023112301' of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid:
  HID: multitouch: Add quirk for HONOR GLO-GXXX touchpad
  HID: hid-asus: reset the backlight brightness level on resume
  HID: hid-asus: add const to read-only outgoing usb buffer
  Revert "HID: logitech-dj: Add support for a new lightspeed receiver iteration"
  HID: add ALWAYS_POLL quirk for Apple kb
  HID: glorious: fix Glorious Model I HID report
  HID: fix HID device resource race between HID core and debugging support
  HID: apple: add Jamesdonkey and A3R to non-apple keyboards list
  HID: mcp2221: Allow IO to start during probe
  HID: mcp2221: Set driver data before I2C adapter add

16 months agoMerge tag 'drm-intel-fixes-2023-11-23' of git://anongit.freedesktop.org/drm/drm-intel...
Dave Airlie [Fri, 24 Nov 2023 01:18:28 +0000 (11:18 +1000)]
Merge tag 'drm-intel-fixes-2023-11-23' of git://anongit.freedesktop.org/drm/drm-intel into drm-fixes

drm/i915 fixes for v6.7-rc3:
- Fix race between DP MST connectore registration and setup
- Fix GT memory leak on probe error path

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/87y1eol98h.fsf@intel.com
16 months agoMerge tag 'drm-misc-fixes-2023-11-23' of git://anongit.freedesktop.org/drm/drm-misc...
Dave Airlie [Fri, 24 Nov 2023 01:14:56 +0000 (11:14 +1000)]
Merge tag 'drm-misc-fixes-2023-11-23' of git://anongit.freedesktop.org/drm/drm-misc into drm-fixes

Fixes for v6.7-rc3:
- Panel fixes for innolux and auo,b101uan08.3 panel.
- Fix ivpu MMIO reset.
- AST fix on connetor disconnection.
- nouveau gsp fix.
- rockchip color fix.
- Fix Himax83102-j02 timings.

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/12322257-2e0c-43d3-8241-876aafc10e4a@linux.intel.com
16 months agoMerge tag 'drm-msm-fixes-2023-11-21' of https://gitlab.freedesktop.org/drm/msm into...
Dave Airlie [Fri, 24 Nov 2023 00:37:45 +0000 (10:37 +1000)]
Merge tag 'drm-msm-fixes-2023-11-21' of https://gitlab.freedesktop.org/drm/msm into drm-fixes

Fixes for v6.7-rc3:

- Fix the VREG_CTRL_1 for 4nm CPHY to match downstream
- Remove duplicate call to drm_kms_helper_poll_init() in msm_drm_init()
- Fix the safe_lut_tbl[] for sc8280xp to match downstream
- Don't attach the drm_dp_set_subconnector_property() for eDP
- Fix to attach drm_dp_set_subconnector_property() for DP. Otherwise
  there is a bootup crash on multiple targets
- Remove unnecessary NULL check left behind during cleanup

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Rob Clark <robdclark@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/CAF6AEGtkna3P3mvaF53n2ARJACaXQU+OFfShayTrsUVmqCOmNQ@mail.gmail.com
16 months agoarm64: add dependency between vmlinuz.efi and Image
Masahiro Yamada [Sun, 19 Nov 2023 05:32:34 +0000 (14:32 +0900)]
arm64: add dependency between vmlinuz.efi and Image

A common issue in Makefile is a race in parallel building.

You need to be careful to prevent multiple threads from writing to the
same file simultaneously.

Commit 3939f3345050 ("ARM: 8418/1: add boot image dependencies to not
generate invalid images") addressed such a bad scenario.

A similar symptom occurs with the following command:

  $ make -j$(nproc) ARCH=arm64 Image vmlinuz.efi
    [ snip ]
    SORTTAB vmlinux
    OBJCOPY arch/arm64/boot/Image
    OBJCOPY arch/arm64/boot/Image
    AS      arch/arm64/boot/zboot-header.o
    PAD     arch/arm64/boot/vmlinux.bin
    GZIP    arch/arm64/boot/vmlinuz
    OBJCOPY arch/arm64/boot/vmlinuz.o
    LD      arch/arm64/boot/vmlinuz.efi.elf
    OBJCOPY arch/arm64/boot/vmlinuz.efi

The log "OBJCOPY arch/arm64/boot/Image" is displayed twice.

It indicates that two threads simultaneously enter arch/arm64/boot/
and write to arch/arm64/boot/Image.

It occasionally leads to a build failure:

  $ make -j$(nproc) ARCH=arm64 Image vmlinuz.efi
    [ snip ]
    SORTTAB vmlinux
    OBJCOPY arch/arm64/boot/Image
    PAD     arch/arm64/boot/vmlinux.bin
  truncate: Invalid number: 'arch/arm64/boot/vmlinux.bin'
  make[2]: *** [drivers/firmware/efi/libstub/Makefile.zboot:13:
  arch/arm64/boot/vmlinux.bin] Error 1
  make[2]: *** Deleting file 'arch/arm64/boot/vmlinux.bin'
  make[1]: *** [arch/arm64/Makefile:163: vmlinuz.efi] Error 2
  make[1]: *** Waiting for unfinished jobs....
  make: *** [Makefile:234: __sub-make] Error 2

vmlinuz.efi depends on Image, but such a dependency is not specified
in arch/arm64/Makefile.

Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Reviewed-by: SImon Glass <sjg@chromium.org>
Link: https://lore.kernel.org/r/20231119053234.2367621-1-masahiroy@kernel.org
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
16 months agoMerge tag 'net-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Linus Torvalds [Thu, 23 Nov 2023 18:40:13 +0000 (10:40 -0800)]
Merge tag 'net-6.7-rc3' of git://git./linux/kernel/git/netdev/net

Pull networking fixes from Jakub Kicinski:
 "Including fixes from bpf.

  Current release - regressions:

   - Revert "net: r8169: Disable multicast filter for RTL8168H and
     RTL8107E"

   - kselftest: rtnetlink: fix ip route command typo

  Current release - new code bugs:

   - s390/ism: make sure ism driver implies smc protocol in kconfig

   - two build fixes for tools/net

  Previous releases - regressions:

   - rxrpc: couple of ACK/PING/RTT handling fixes

  Previous releases - always broken:

   - bpf: verify bpf_loop() callbacks as if they are called unknown
     number of times

   - improve stability of auto-bonding with Hyper-V

   - account BPF-neigh-redirected traffic in interface statistics

  Misc:

   - net: fill in some more MODULE_DESCRIPTION()s"

* tag 'net-6.7-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (58 commits)
  tools: ynl: fix duplicate op name in devlink
  tools: ynl: fix header path for nfsd
  net: ipa: fix one GSI register field width
  tls: fix NULL deref on tls_sw_splice_eof() with empty record
  net: axienet: Fix check for partial TX checksum
  vsock/test: fix SEQPACKET message bounds test
  i40e: Fix adding unsupported cloud filters
  ice: restore timestamp configuration after device reset
  ice: unify logic for programming PFINT_TSYN_MSK
  ice: remove ptp_tx ring parameter flag
  amd-xgbe: propagate the correct speed and duplex status
  amd-xgbe: handle the corner-case during tx completion
  amd-xgbe: handle corner-case during sfp hotplug
  net: veth: fix ethtool stats reporting
  octeontx2-pf: Fix ntuple rule creation to direct packet to VF with higher Rx queue than its PF
  net: usb: qmi_wwan: claim interface 4 for ZTE MF290
  Revert "net: r8169: Disable multicast filter for RTL8168H and RTL8107E"
  net/smc: avoid data corruption caused by decline
  nfc: virtual_ncidev: Add variable to check if ndev is running
  dpll: Fix potential msg memleak when genlmsg_put_reply failed
  ...

16 months agosmb: client: introduce cifs_sfu_make_node()
Paulo Alcantara [Tue, 21 Nov 2023 23:12:55 +0000 (20:12 -0300)]
smb: client: introduce cifs_sfu_make_node()

Remove duplicate code and add new helper for creating special files in
SFU (Services for UNIX) format that can be shared by SMB1+ code.

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
16 months agosmb: client: set correct file type from NFS reparse points
Paulo Alcantara [Tue, 21 Nov 2023 23:12:54 +0000 (20:12 -0300)]
smb: client: set correct file type from NFS reparse points

Handle all file types in NFS reparse points as specified in MS-FSCC
2.1.2.6 Network File System (NFS) Reparse Data Buffer.

The client is now able to set all file types based on the parsed NFS
reparse point, which used to support only symlinks.  This works for
SMB1+.

Before patch:

$ mount.cifs //srv/share /mnt -o ...
$ ls -l /mnt
ls: cannot access 'block': Operation not supported
ls: cannot access 'char': Operation not supported
ls: cannot access 'fifo': Operation not supported
ls: cannot access 'sock': Operation not supported
total 1
l????????? ? ?    ?    ?            ? block
l????????? ? ?    ?    ?            ? char
-rwxr-xr-x 1 root root 5 Nov 18 23:22 f0
l????????? ? ?    ?    ?            ? fifo
l--------- 1 root root 0 Nov 18 23:23 link -> f0
l????????? ? ?    ?    ?            ? sock

After patch:

$ mount.cifs //srv/share /mnt -o ...
$ ls -l /mnt
total 1
brwxr-xr-x 1 root root  123,  123 Nov 18 00:34 block
crwxr-xr-x 1 root root 1234, 1234 Nov 18 00:33 char
-rwxr-xr-x 1 root root          5 Nov 18 23:22 f0
prwxr-xr-x 1 root root          0 Nov 18 23:23 fifo
lrwxr-xr-x 1 root root          0 Nov 18 23:23 link -> f0
srwxr-xr-x 1 root root          0 Nov 19  2023 sock

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
16 months agosmb: client: introduce ->parse_reparse_point()
Paulo Alcantara [Tue, 21 Nov 2023 23:12:53 +0000 (20:12 -0300)]
smb: client: introduce ->parse_reparse_point()

Parse reparse point into cifs_open_info_data structure and feed it
through cifs_open_info_to_fattr().

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
16 months agosmb: client: implement ->query_reparse_point() for SMB1
Paulo Alcantara [Tue, 21 Nov 2023 23:12:52 +0000 (20:12 -0300)]
smb: client: implement ->query_reparse_point() for SMB1

Reparse points are not limited to symlinks, so implement
->query_reparse_point() in order to handle different file types.

Signed-off-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
16 months agocifs: fix use after free for iface while disabling secondary channels
Ritvik Budhiraja [Tue, 21 Nov 2023 13:43:47 +0000 (19:13 +0530)]
cifs: fix use after free for iface while disabling secondary channels

We were deferencing iface after it has been released. Fix is to
release after all dereference instances have been encountered.

Signed-off-by: Ritvik Budhiraja <rbudhiraja@microsoft.com>
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <error27@gmail.com>
Closes: https://lore.kernel.org/r/202311110815.UJaeU3Tt-lkp@intel.com/
Signed-off-by: Steve French <stfrench@microsoft.com>
16 months agoUSB: serial: option: add Luat Air72*U series products
Asuna Yang [Wed, 22 Nov 2023 14:18:03 +0000 (22:18 +0800)]
USB: serial: option: add Luat Air72*U series products

Update the USB serial option driver support for Luat Air72*U series
products.

ID 1782:4e00 Spreadtrum Communications Inc. UNISOC-8910

T: Bus=01 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#= 13 Spd=480 MxCh= 0
D: Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs= 1
P: Vendor=1782 ProdID=4e00 Rev=00.00
S: Manufacturer=UNISOC
S: Product=UNISOC-8910
C: #Ifs= 5 Cfg#= 1 Atr=e0 MxPwr=400mA
I: If#= 0 Alt= 0 #EPs= 1 Cls=e0(wlcon) Sub=01 Prot=03 Driver=rndis_host
E: Ad=82(I) Atr=03(Int.) MxPS= 8 Ivl=4096ms
I: If#= 1 Alt= 0 #EPs= 2 Cls=0a(data ) Sub=00 Prot=00 Driver=rndis_host
E: Ad=01(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E: Ad=81(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I: If#= 2 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=00 Prot=00 Driver=option
E: Ad=02(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E: Ad=83(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I: If#= 3 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=00 Prot=00 Driver=option
E: Ad=03(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E: Ad=84(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms
I: If#= 4 Alt= 0 #EPs= 2 Cls=ff(vend.) Sub=00 Prot=00 Driver=option
E: Ad=04(O) Atr=02(Bulk) MxPS= 512 Ivl=0ms
E: Ad=85(I) Atr=02(Bulk) MxPS= 512 Ivl=0ms

If#= 2: AT
If#= 3: PPP + AT
If#= 4: Debug

Co-developed-by: Yangyu Chen <cyy@cyyself.name>
Signed-off-by: Yangyu Chen <cyy@cyyself.name>
Signed-off-by: Asuna Yang <SpriteOvO@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Johan Hovold <johan@kernel.org>