Thomas Gleixner [Wed, 1 Mar 2023 21:07:48 +0000 (22:07 +0100)]
genirq/msi, platform-msi: Ensure that MSI descriptors are unreferenced
Miquel reported a warning in the MSI core which is triggered when
interrupts are freed via platform_msi_device_domain_free().
This code got reworked to use core functions for freeing the MSI
descriptors, but nothing took care to clear the msi_desc->irq entry, which
then triggers the warning in msi_free_msi_desc() which uses desc->irq to
validate that the descriptor has been torn down. The same issue exists in
msi_domain_populate_irqs().
Up to the point that msi_free_msi_descs() grew a warning for this case,
this went un-noticed.
Provide the counterpart of msi_domain_populate_irqs() and invoke it in
platform_msi_device_domain_free() before freeing the interrupts and MSI
descriptors and also in the error path of msi_domain_populate_irqs().
Vasily Gorbik [Wed, 1 Mar 2023 16:58:06 +0000 (17:58 +0100)]
s390/kprobes: fix current_kprobe never cleared after kprobes reenter
Recent test_kprobe_missed kprobes kunit test uncovers the following
problem. Once kprobe is triggered from another kprobe (kprobe reenter),
all future kprobes on this cpu are considered as kprobe reenter, thus
pre_handler and post_handler are not being called and kprobes are counted
as "missed".
Commit b9599798f953 ("[S390] kprobes: activation and deactivation")
introduced a simpler scheme for kprobes (de)activation and status
tracking by using push_kprobe/pop_kprobe, which supposed to work for
both initial kprobe entry as well as kprobe reentry and helps to avoid
handling those two cases differently. The problem is that a sequence of
calls in case of kprobes reenter:
push_kprobe() <- NULL (current_kprobe)
push_kprobe() <- kprobe1 (current_kprobe)
pop_kprobe() -> kprobe1 (current_kprobe)
pop_kprobe() -> kprobe1 (current_kprobe)
leaves "kprobe1" as "current_kprobe" on this cpu, instead of setting it
to NULL. In fact push_kprobe/pop_kprobe can only store a single state
(there is just one prev_kprobe in kprobe_ctlblk). Which is a hack but
sufficient, there is no need to have another prev_kprobe just to store
NULL. To make a simple and backportable fix simply reset "prev_kprobe"
when kprobe is poped from this "stack". No need to worry about
"kprobe_status" in this case, because its value is only checked when
current_kprobe != NULL.
Vasily Gorbik [Wed, 1 Mar 2023 01:23:08 +0000 (02:23 +0100)]
s390/kprobes: fix irq mask clobbering on kprobe reenter from post_handler
Recent test_kprobe_missed kprobes kunit test uncovers the following error
(reported when CONFIG_DEBUG_ATOMIC_SLEEP is enabled):
BUG: sleeping function called from invalid context at kernel/locking/mutex.c:580
in_atomic(): 0, irqs_disabled(): 1, non_block: 0, pid: 662, name: kunit_try_catch
preempt_count: 0, expected: 0
RCU nest depth: 0, expected: 0
no locks held by kunit_try_catch/662.
irq event stamp: 280
hardirqs last enabled at (279): [<00000003e60a3d42>] __do_pgm_check+0x17a/0x1c0
hardirqs last disabled at (280): [<00000003e3bd774a>] kprobe_exceptions_notify+0x27a/0x318
softirqs last enabled at (0): [<00000003e3c5c890>] copy_process+0x14a8/0x4c80
softirqs last disabled at (0): [<0000000000000000>] 0x0
CPU: 46 PID: 662 Comm: kunit_try_catch Tainted: G N 6.2.0-173644-g44c18d77f0c0 #2
Hardware name: IBM 3931 A01 704 (LPAR)
Call Trace:
[<00000003e60a3a00>] dump_stack_lvl+0x120/0x198
[<00000003e3d02e82>] __might_resched+0x60a/0x668
[<00000003e60b9908>] __mutex_lock+0xc0/0x14e0
[<00000003e60bad5a>] mutex_lock_nested+0x32/0x40
[<00000003e3f7b460>] unregister_kprobe+0x30/0xd8
[<00000003e51b2602>] test_kprobe_missed+0xf2/0x268
[<00000003e51b5406>] kunit_try_run_case+0x10e/0x290
[<00000003e51b7dfa>] kunit_generic_run_threadfn_adapter+0x62/0xb8
[<00000003e3ce30f8>] kthread+0x2d0/0x398
[<00000003e3b96afa>] __ret_from_fork+0x8a/0xe8
[<00000003e60ccada>] ret_from_fork+0xa/0x40
The reason for this error report is that kprobes handling code failed
to restore irqs.
The problem is that when kprobe is triggered from another kprobe
post_handler current sequence of enable_singlestep / disable_singlestep
is the following:
enable_singlestep <- original kprobe (saves kprobe_saved_imask)
enable_singlestep <- kprobe triggered from post_handler (clobbers kprobe_saved_imask)
disable_singlestep <- kprobe triggered from post_handler (restores kprobe_saved_imask)
disable_singlestep <- original kprobe (restores wrong clobbered kprobe_saved_imask)
There is just one kprobe_ctlblk per cpu and both calls saves and
loads irq mask to kprobe_saved_imask. To fix the problem simply move
resume_execution (which calls disable_singlestep) before calling
post_handler. This also fixes the problem that post_handler is called
with pt_regs which were not yet adjusted after single-stepping.
Uwe Kleine-König [Fri, 17 Feb 2023 09:53:17 +0000 (10:53 +0100)]
watchdog: at91rm9200: Only warn once about problems in .remove()
The single difference between returning 0 and returning an error code in
a platform remove callback is that in the latter case the platform core
emits a warning about the error being ignored.
at91wdt_remove() already emits a warning in the error case, so suppress
the more generic (and less helpful) one by returning 0.
MT7621 SoC has a system controller node. Watchdog need to access to reset
status register. Ralink architecture and related driver are old and from
the beggining they are using some architecture dependent operations for
accessing this shared registers through 'asm/mach-ralink/ralink_regs.h'
header file. However this is not ideal from a driver perspective which can
just access to the system controller registers in an arch independent way
using regmap syscon APIs. Update Kconfig accordingly to select new added
dependencies and allow driver to be compile tested.
dt-bindings: watchdog: mt7621-wdt: add phandle to access system controller registers
MT7621 SoC provides a system controller node for accessing to some registers.
Add a phandle in this node to avoid using MIPS related arch operations and
includes in watchdog driver code.
bobzhou [Mon, 27 Feb 2023 07:30:54 +0000 (15:30 +0800)]
drm/amdgpu/vcn: fix compilation issue with legacy gcc
This patch is used to fix following compilation issue with legacy gcc
error: ‘for’ loop initial declarations are only allowed in C99 mode
for (int i = 0; i < adev->vcn.num_vcn_inst; ++i) {
Horatio Zhang [Fri, 24 Feb 2023 05:55:44 +0000 (13:55 +0800)]
drm/amdgpu: fix ttm_bo calltrace warning in psp_hw_fini
The call trace occurs when the amdgpu is removed after
the mode1 reset. During mode1 reset, from suspend to resume,
there is no need to reinitialize the ta firmware buffer
which caused the bo pin_count increase redundantly.
Tom Rix [Fri, 24 Feb 2023 16:45:19 +0000 (11:45 -0500)]
drm/amdgpu: remove unused variable ring
building with gcc and W=1 reports
drivers/gpu/drm/amd/amdgpu/vcn_v4_0.c:81:29: error: variable
‘ring’ set but not used [-Werror=unused-but-set-variable]
81 | struct amdgpu_ring *ring;
| ^~~~
tiancyin [Wed, 8 Feb 2023 06:10:04 +0000 (14:10 +0800)]
drm/amd/display: fix dm irq error message in gpu recover
[Why]
Variable adev->crtc_irq.num_types was initialized as the value of
adev->mode_info.num_crtc at early_init stage, later at hw_init stage,
the num_crtc changed due to the display pipe harvest on some SKUs,
but the num_types was not updated accordingly, that cause below error
in gpu recover.
*ERROR* amdgpu_dm_set_crtc_irq_state: crtc is NULL at id :3
*ERROR* amdgpu_dm_set_crtc_irq_state: crtc is NULL at id :3
*ERROR* amdgpu_dm_set_crtc_irq_state: crtc is NULL at id :3
*ERROR* amdgpu_dm_set_pflip_irq_state: crtc is NULL at id :3
*ERROR* amdgpu_dm_set_pflip_irq_state: crtc is NULL at id :3
*ERROR* amdgpu_dm_set_pflip_irq_state: crtc is NULL at id :3
*ERROR* amdgpu_dm_set_pflip_irq_state: crtc is NULL at id :3
*ERROR* amdgpu_dm_set_vupdate_irq_state: crtc is NULL at id :3
*ERROR* amdgpu_dm_set_vupdate_irq_state: crtc is NULL at id :3
*ERROR* amdgpu_dm_set_vupdate_irq_state: crtc is NULL at id :3
[How]
Defer the initialization of num_types to eliminate the error logs.
Harry Wentland [Mon, 12 Dec 2022 18:02:25 +0000 (13:02 -0500)]
drm/amd/display: Don't restrict bpc to 8 bpc
This will let us pass the kms_hdr.bpc_switch IGT
test.
The reason the bpc restriction was required is
historical. At one point in time we were not falling
back to a lower bpc when we didn't have enough
bandwidth for the maximum bpc reported by a display.
This meant that we couldn't enable some high refresh
modes unless we limitted the bpc.
Starting with this patch the issue is fixed:
commit cbd14ae7ea93 ("drm/amd/display: Fix
incorrectly pruned modes with deep color")
This patch implemented a fallback mechanism if mode
validation failed at the max bpc. This means users
now automatically get all modes that can be supported
by at least 6 bpc. The driver will enable the mode
with the highest possible bpc that is supported by
the display.
v2:
- explain why this is no longer needed (Michel)
- refer to commit that fixed bpc fallback (Michel)
Mark Hawrylak [Sun, 19 Feb 2023 05:02:00 +0000 (16:02 +1100)]
drm/radeon: Fix eDP for single-display iMac11,2
Apple iMac11,2 (mid 2010) also with Radeon HD-4670 that has the same
issue as iMac10,1 (late 2009) where the internal eDP panel stays dark on
driver load. This patch treats iMac11,2 the same as iMac10,1,
so the eDP panel stays active.
Additional steps:
Kernel boot parameter radeon.nomodeset=0 required to keep the eDP
panel active.
Alexandre Ghiti [Tue, 16 Mar 2021 19:34:20 +0000 (15:34 -0400)]
riscv: Bump COMMAND_LINE_SIZE value to 1024
Increase COMMAND_LINE_SIZE as the current default value is too low
for syzbot kernel command line.
There has been considerable discussion on this patch that has led to a
larger patch set removing COMMAND_LINE_SIZE from the uapi headers on all
ports. That's not quite done yet, but it's gotten far enough we're
confident this is not a uABI change so this is safe.
David Howells [Tue, 28 Feb 2023 22:38:38 +0000 (22:38 +0000)]
cifs: Fix memory leak in direct I/O
When __cifs_readv() and __cifs_writev() extract pages from a user-backed
iterator into a BVEC-type iterator, they set ->bv_need_unpin to note
whether they need to unpin the pages later. However, in both cases they
examine the BVEC-type iterator and not the source iterator - and so
bv_need_unpin doesn't get set and the pages are leaked.
I think this may be responsible for the generic/208 xfstest failing
occasionally with:
with the page refcount going negative. This is less unlikely than it seems
because the page is being pinned, not simply got, and so the refcount
increased by 1024 each time, and so only needs to be called around ~2097152
for the refcount to go negative.
Further, the test program (aio-dio-invalidate-failure) uses a 32MiB static
buffer and all the PTEs covering it refer to the same page because it's
never written to.
The warning in try_grab_page():
if (WARN_ON_ONCE(folio_ref_count(folio) <= 0))
return -ENOMEM;
then trips and prevents us ever using the page again for DIO at least.
Paulo Alcantara [Tue, 28 Feb 2023 22:01:55 +0000 (19:01 -0300)]
cifs: prevent data race in cifs_reconnect_tcon()
Make sure to get an up-to-date TCP_Server_Info::nr_targets value prior
to waiting the server to be reconnected in cifs_reconnect_tcon(). It
is set in cifs_tcp_ses_needs_reconnect() and protected by
TCP_Server_Info::srv_lock.
Create a new cifs_wait_for_server_reconnect() helper that can be used
by both SMB2+ and CIFS reconnect code.
Paulo Alcantara [Tue, 28 Feb 2023 22:01:54 +0000 (19:01 -0300)]
cifs: improve checking of DFS links over STATUS_OBJECT_NAME_INVALID
Do not map STATUS_OBJECT_NAME_INVALID to -EREMOTE under non-DFS
shares, or 'nodfs' mounts or CONFIG_CIFS_DFS_UPCALL=n builds.
Otherwise, in the slow path, get a referral to figure out whether it
is an actual DFS link.
This could be simply reproduced under a non-DFS share by running the
following
David Howells [Mon, 27 Feb 2023 13:04:54 +0000 (13:04 +0000)]
iov: Fix netfs_extract_user_to_sg()
Fix the loop check in netfs_extract_user_to_sg() for extraction from
user-backed iterators to do the body if npages > 0, not if npages < 0
(which it can never be).
This isn't currently used by cifs, which only ever extracts data from BVEC,
KVEC and XARRAY iterators at this level, user-backed iterators having being
decanted into BVEC iterators at a higher level to accommodate the work
being done in a kernel thread.
Found by smatch:
fs/netfs/iterator.c:139 netfs_extract_user_to_sg() warn: unsigned 'npages' is never less than zero.
David Howells [Mon, 27 Feb 2023 13:04:53 +0000 (13:04 +0000)]
cifs: Fix cifs_write_back_from_locked_folio()
cifs_write_back_from_locked_folio() should return the number of bytes read,
but returns the result of ->async_writev(), which will be 0 on success. As
it happens, this doesn't prevent cifs_writepages_region() from working as
it will then examine and ignore the pages that are no longer dirty rather
than just skipping over them.
Shyam Prasad N [Tue, 27 Dec 2022 14:04:29 +0000 (14:04 +0000)]
cifs: match even the scope id for ipv6 addresses
match_address function matches the scope id for ipv6 addresses,
but cifs_match_ipaddr (which is another function used for comparison)
does not use scope id. Doing so with this change.
Applications need to be able to program the SBI implementation specific
or custom firmware events in addition to the standard firmware events.
Remove a check in the driver that prohibits the programming of the custom
firmware events.
Linus Torvalds [Wed, 1 Mar 2023 19:03:44 +0000 (11:03 -0800)]
Merge tag 'nfsd-6.3-1' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux
Pull nfsd fix from Chuck Lever:
- Make new GSS Kerberos Kunit tests work on non-x86 platforms
* tag 'nfsd-6.3-1' of git://git.kernel.org/pub/scm/linux/kernel/git/cel/linux:
SUNRPC: Properly terminate test case arrays
SUNRPC: Let Kunit tests run with some enctypes compiled out
Gerald Schaefer [Mon, 27 Feb 2023 19:03:00 +0000 (20:03 +0100)]
s390/extmem: return correct segment type in __segment_load()
Commit f05f62d04271f ("s390/vmem: get rid of memory segment list")
reshuffled the call to vmem_add_mapping() in __segment_load(), which now
overwrites rc after it was set to contain the segment type code.
As result, __segment_load() will now always return 0 on success, which
corresponds to the segment type code SEG_TYPE_SW, i.e. a writeable
segment. This results in a kernel crash when loading a read-only segment
as dcssblk block device, and trying to write to it.
Instead of reshuffling code again, make sure to return the segment type
on success, and also describe this rather delicate and unexpected logic
in the function comment. Also initialize new segtype variable with
invalid value, to prevent possible future confusion.
Dan Carpenter [Mon, 27 Feb 2023 10:07:09 +0000 (13:07 +0300)]
cpufreq: apple-soc: Fix an IS_ERR() vs NULL check
The of_iomap() function returns NULL if it fails. It never returns
error pointers. Fix the check accordingly.
Fixes: 6286bbb40576 ("cpufreq: apple-soc: Add new driver to control Apple SoC CPU P-states") Signed-off-by: Dan Carpenter <[email protected]> Reviewed-by: Eric Curtin <[email protected]> Signed-off-by: Rafael J. Wysocki <[email protected]>
Randy Dunlap [Sun, 26 Feb 2023 05:39:52 +0000 (21:39 -0800)]
thermal: intel: BXT_PMIC: select REGMAP instead of depending on it
REGMAP is a hidden (not user visible) symbol. Users cannot set it
directly thru "make *config", so drivers should select it instead of
depending on it if they need it.
Consistently using "select" or "depends on" can also help reduce
Kconfig circular dependency issues.
Therefore, change the use of "depends on REGMAP" to "select REGMAP".
Fixes: b474303ffd57 ("thermal: add Intel BXT WhiskeyCove PMIC thermal driver") Signed-off-by: Randy Dunlap <[email protected]> Signed-off-by: Rafael J. Wysocki <[email protected]>
ACPI: x86: utils: Add Cezanne to the list for forcing StorageD3Enable
commit 018d6711c26e4 ("ACPI: x86: Add a quirk for Dell Inspiron 14 2-in-1
for StorageD3Enable") introduced a quirk to allow a system with ambiguous
use of _ADR 0 to force StorageD3Enable.
It was reported that several more Dell systems suffered the same symptoms.
As the list is continuing to grow but these are all Cezanne systems,
instead add Cezanne to the CPU list to apply the StorageD3Enable property
and remove the whole list.
It was also reported that an HP system only has StorageD3Enable on the ACPI
device for the first NVME disk, not the second.
Linus Torvalds [Tue, 28 Feb 2023 19:39:09 +0000 (11:39 -0800)]
capability: just use a 'u64' instead of a 'u32[2]' array
Back in 2008 we extended the capability bits from 32 to 64, and we did
it by extending the single 32-bit capability word from one word to an
array of two words. It was then obfuscated by hiding the "2" behind two
macro expansions, with the reasoning being that maybe it gets extended
further some day.
That reasoning may have been valid at the time, but the last thing we
want to do is to extend the capability set any more. And the array of
values not only causes source code oddities (with loops to deal with
it), but also results in worse code generation. It's a lose-lose
situation.
So just change the 'u32[2]' into a 'u64' and be done with it.
We still have to deal with the fact that the user space interface is
designed around an array of these 32-bit values, but that was the case
before too, since the array layouts were different (ie user space
doesn't use an array of 32-bit values for individual capability masks,
but an array of 32-bit slices of multiple masks).
So that marshalling of data is actually simplified too, even if it does
remain somewhat obscure and odd.
This was all triggered by my reaction to the new "cap_isidentical()"
introduced recently. By just using a saner data structure, it went from
Linus Torvalds [Wed, 1 Mar 2023 17:44:22 +0000 (09:44 -0800)]
Merge tag 'sh-for-v6.3-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/glaubitz/sh-linux
Pull sh updates from John Paul Adrian Glaubitz:
- regression fix in connection with the rtl8169 driver on SuperH boards
that was introduced when the driver was switched to use
devm_clk_get_optional_enabled() to simplify the code (Geert
Uytterhoeven)
- build warning fix to allow the kernel to be built with CONFIG_WERROR
enabled (Michael Karcher)
* tag 'sh-for-v6.3-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/glaubitz/sh-linux:
sh: clk: Fix clk_enable() to return 0 on NULL clk
sh: intc: Avoid spurious sizeof-pointer-div warning
Linus Torvalds [Wed, 1 Mar 2023 17:27:00 +0000 (09:27 -0800)]
Merge tag 'loongarch-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson
Pull LoongArch updates from Huacai Chen:
- Make -mstrict-align configurable
- Add kernel relocation and KASLR support
- Add single kernel image implementation for kdump
- Add hardware breakpoints/watchpoints support
- Add kprobes/kretprobes/kprobes_on_ftrace support
- Add LoongArch support for some selftests.
* tag 'loongarch-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson: (23 commits)
selftests/ftrace: Add LoongArch kprobe args string tests support
selftests/seccomp: Add LoongArch selftesting support
tools: Add LoongArch build infrastructure
samples/kprobes: Add LoongArch support
LoongArch: Mark some assembler symbols as non-kprobe-able
LoongArch: Add kprobes on ftrace support
LoongArch: Add kretprobes support
LoongArch: Add kprobes support
LoongArch: Simulate branch and PC* instructions
LoongArch: ptrace: Add hardware single step support
LoongArch: ptrace: Add function argument access API
LoongArch: ptrace: Expose hardware breakpoints to debuggers
LoongArch: Add hardware breakpoints/watchpoints support
LoongArch: kdump: Add crashkernel=YM handling
LoongArch: kdump: Add single kernel image implementation
LoongArch: Add support for kernel address space layout randomization (KASLR)
LoongArch: Add support for kernel relocation
LoongArch: Add la_abs macro implementation
LoongArch: Add JUMP_VIRT_ADDR macro implementation to avoid using la.abs
LoongArch: Use la.pcrel instead of la.abs when it's trivially possible
...
Linus Torvalds [Wed, 1 Mar 2023 17:13:00 +0000 (09:13 -0800)]
Merge tag 'uml-for-linus-6.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/uml/linux
Pull UML updates from Richard Weinberger:
- Add support for rust (yay!)
- Add support for LTO
- Add platform bus support to virtio-pci
- Various virtio fixes
- Coding style, spelling cleanups
* tag 'uml-for-linus-6.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/uml/linux: (27 commits)
Documentation: rust: Fix arch support table
uml: vector: Remove unused definitions VECTOR_{WRITE,HEADERS}
um: virt-pci: properly remove PCI device from bus
um: virtio_uml: move device breaking into workqueue
um: virtio_uml: mark device as unregistered when breaking it
um: virtio_uml: free command if adding to virtqueue failed
UML: define RUNTIME_DISCARD_EXIT
virt-pci: add platform bus support
um-virt-pci: Make max delay configurable
um: virt-pci: implement pcibios_get_phb_of_node()
um: Support LTO
um: put power options in a menu
um: Use CFLAGS_vmlinux
um: Prevent building modules incompatible with MODVERSIONS
um: Avoid pcap multiple definition errors
um: Make the definition of cpu_data more compatible
x86: um: vdso: Add '%rcx' and '%r11' to the syscall clobber list
rust: arch/um: Add support for CONFIG_RUST under x86_64 UML
rust: arch/um: Disable FP/SIMD instruction to match x86
rust: arch/um: Use 'pie' relocation mode under UML
...
Linus Torvalds [Wed, 1 Mar 2023 17:06:51 +0000 (09:06 -0800)]
Merge tag 'ubifs-for-linus-6.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rw/ubifs
Pull jffs2, ubi and ubifs updates from Richard Weinberger:
"JFFS2:
- Fix memory corruption in error path
- Spelling and coding style fixes
UBI:
- Switch to BLK_MQ_F_BLOCKING in ubiblock
- Wire up partent device (for sysfs)
- Multiple UAF bugfixes
- Fix for an infinite loop in WL error path
UBIFS:
- Fix for multiple memory leaks in error paths
- Fixes for wrong space accounting
- Minor cleanups
- Spelling and coding style fixes"
* tag 'ubifs-for-linus-6.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rw/ubifs: (36 commits)
ubi: block: Fix a possible use-after-free bug in ubiblock_create()
ubifs: make kobj_type structures constant
mtd: ubi: block: wire-up device parent
mtd: ubi: wire-up parent MTD device
ubi: use correct names in function kernel-doc comments
ubi: block: set BLK_MQ_F_BLOCKING
jffs2: Fix list_del corruption if compressors initialized failed
jffs2: Use function instead of macro when initialize compressors
jffs2: fix spelling mistake "neccecary"->"necessary"
ubifs: Fix kernel-doc
ubifs: Fix some kernel-doc comments
UBI: Fastmap: Fix kernel-doc
ubi: ubi_wl_put_peb: Fix infinite loop when wear-leveling work failed
ubi: Fix UAF wear-leveling entry in eraseblk_count_seq_show()
ubi: fastmap: Fix missed fm_anchor PEB in wear-leveling after disabling fastmap
ubifs: ubifs_releasepage: Remove ubifs_assert(0) to valid this process
ubifs: ubifs_writepage: Mark page dirty after writing inode failed
ubifs: dirty_cow_znode: Fix memleak in error handling path
ubifs: Re-statistic cleaned znode count if commit failed
ubi: Fix permission display of the debugfs files
...
Linus Torvalds [Wed, 1 Mar 2023 16:52:49 +0000 (08:52 -0800)]
Merge tag '9p-6.3-for-linus-part1' of git://git.kernel.org/pub/scm/linux/kernel/git/ericvh/v9fs
Pull 9p updates from Eric Van Hensbergen:
- some fixes and cleanup setting up for a larger set of performance
patches I've been working on
- a contributed fixes relating to 9p/rdma
- some contributed fixes relating to 9p/xen
* tag '9p-6.3-for-linus-part1' of git://git.kernel.org/pub/scm/linux/kernel/git/ericvh/v9fs:
fs/9p: fix error reporting in v9fs_dir_release
net/9p: fix bug in client create for .L
9p/rdma: unmap receive dma buffer in rdma_request()/post_recv()
9p/xen: fix connection sequence
9p/xen: fix version parsing
fs/9p: Expand setup of writeback cache to all levels
net/9p: Adjust maximum MSIZE to account for p9 header
Linus Torvalds [Wed, 1 Mar 2023 16:42:27 +0000 (08:42 -0800)]
Merge tag 'exfat-for-6.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/exfat
Pull exfat updates from Namjae Jeon:
- Handle vendor extension and allocation entries as unrecognized benign
secondary entries
- Fix wrong ->i_blocks on devices with non-512 byte sector
- Add the check to avoid returning -EIO from exfat_readdir() at current
position exceeding the directory size
- Fix a bug that reach the end of the directory stream at a position
not aligned with the dentry size
- Redefine DIR_DELETED as 0xFFFFFFF7, the bad cluster number
- Two cleanup fixes and fix cluster leakage in error handling
* tag 'exfat-for-6.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/linkinjeon/exfat:
exfat: fix the newly allocated clusters are not freed in error handling
exfat: don't print error log in normal case
exfat: remove unneeded code from exfat_alloc_cluster()
exfat: handle unreconized benign secondary entries
exfat: fix inode->i_blocks for non-512 byte sector size device
exfat: redefine DIR_DELETED as the bad cluster number
exfat: fix reporting fs error when reading dir beyond EOF
exfat: fix unexpected EOF while reading dir
Jens Axboe [Wed, 1 Mar 2023 14:49:59 +0000 (07:49 -0700)]
Merge tag 'nvme-6.3-2022-03-01' of git://git.infradead.org/nvme into for-6.3/block
Pull NVMe fixes from Christoph:
"nvme fixes for Linux 6.3
- don't access released socket during error recovery (Akinobu Mita)
- bring back auto-removal of deleted namespaces during sequential scan
(Christoph Hellwig)
- fix an error code in nvme_auth_process_dhchap_challenge
(Dan Carpenter)
- show well known discovery name (Daniel Wagner)
- add a missing endianess conversion in effects masking (Keith Busch)"
* tag 'nvme-6.3-2022-03-01' of git://git.infradead.org/nvme:
nvme-fabrics: show well known discovery name
nvme-tcp: don't access released socket during error recovery
nvme-auth: fix an error code in nvme_auth_process_dhchap_challenge()
nvme: bring back auto-removal of deleted namespaces during sequential scan
nvme: fix sparse warning on effects masking
In etdm dai driver, dai_etdm_parse_of() function is used to parse dts
properties to get parameters. There are two for-loops which are
sepearately for all etdm and etdm input only cases. In etdm in only
loop, dai_id is not initialized, so it keeps the value intiliazed in
another loop.
In the patch, add the missing initialization to fix the unexpected
parsing problem.
In etdm dai driver, dai_etdm_parse_of() function is used to parse dts
properties to get parameters. There are two for-loops which are
sepearately for all etdm and etdm input only cases. In etdm in only
loop, dai_id is not initialized, so it keeps the value intiliazed in
another loop.
In the patch, add the missing initialization to fix the unexpected
parsing problem.
Tom Lendacky [Wed, 22 Feb 2023 16:39:39 +0000 (10:39 -0600)]
virt/sev-guest: Return -EIO if certificate buffer is not large enough
Commit
47894e0fa6a5 ("virt/sev-guest: Prevent IV reuse in the SNP guest driver")
changed the behavior associated with the return value when the caller
does not supply a large enough certificate buffer. Prior to the commit a
value of -EIO was returned. Now, 0 is returned. This breaks the
established ABI with the user.
Change the code to detect the buffer size error and return -EIO.
Björn Töpel [Tue, 28 Feb 2023 18:42:10 +0000 (19:42 +0100)]
riscv, lib: Fix Zbb strncmp
The Zbb optimized strncmp has two parts; a fast path that does XLEN/8B
per iteration, and a slow that does one byte per iteration.
The idea is to compare aligned XLEN chunks for most of strings, and do
the remainder tail in the slow path.
The Zbb strncmp has two issues in the fast path:
Incorrect remainder handling (wrong compare): Assume that the string
length is 9. On 64b systems, the fast path should do one iteration,
and one iteration in the slow path. Instead, both were done in the
fast path, which lead to incorrect results. An example:
strncmp("/dev/vda", "/dev/", 5);
Correct by changing "bgt" to "bge".
Missing NULL checks in the second string: This could lead to incorrect
results for:
Linus Torvalds [Wed, 1 Mar 2023 00:08:30 +0000 (16:08 -0800)]
Merge tag 'xfs-6.3-merge-4' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux
Pull moar xfs updates from Darrick Wong:
"This contains a fix for a deadlock in the allocator. It continues the
slow march towards being able to offline AGs, and it refactors the
interface to the xfs allocator to be less indirection happy.
Summary:
- Fix a deadlock in the free space allocator due to the AG-walking
algorithm forgetting to follow AG-order locking rules
- Make the inode allocator prefer existing free inodes instead of
failing to allocate new inode chunks when free space is low
- Set minleft correctly when setting allocator parameters for bmap
changes
- Fix uninitialized variable access in the getfsmap code
- Make a distinction between active and passive per-AG structure
references. For now, active references are taken to perform some
work in an AG on behalf of a high level operation; passive
references are used by lower level code to finish operations
started by other threads. Eventually this will become part of
online shrink
- Split out all the different allocator strategies into separate
functions to move us away from design antipattern of filling out a
huge structure for various differentish things and issuing a single
function multiplexing call
- Various cleanups in the filestreams allocator code, which we might
very well want to deprecate instead of continuing
- Fix a bug with the agi rotor code that was introduced earlier in
this series"
* tag 'xfs-6.3-merge-4' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux: (44 commits)
xfs: restore old agirotor behavior
xfs: fix uninitialized variable access
xfs: refactor the filestreams allocator pick functions
xfs: return a referenced perag from filestreams allocator
xfs: pass perag to filestreams tracing
xfs: use for_each_perag_wrap in xfs_filestream_pick_ag
xfs: track an active perag reference in filestreams
xfs: factor out MRU hit case in xfs_filestream_select_ag
xfs: remove xfs_filestream_select_ag() longest extent check
xfs: merge new filestream AG selection into xfs_filestream_select_ag()
xfs: merge filestream AG lookup into xfs_filestream_select_ag()
xfs: move xfs_bmap_btalloc_filestreams() to xfs_filestreams.c
xfs: use xfs_bmap_longest_free_extent() in filestreams
xfs: get rid of notinit from xfs_bmap_longest_free_extent
xfs: factor out filestreams from xfs_bmap_btalloc_nullfb
xfs: convert trim to use for_each_perag_range
xfs: convert xfs_alloc_vextent_iterate_ags() to use perag walker
xfs: move the minimum agno checks into xfs_alloc_vextent_check_args
xfs: fold xfs_alloc_ag_vextent() into callers
xfs: move allocation accounting to xfs_alloc_vextent_set_fsbno()
...
* tag 'i3c/for-6.3' of git://git.kernel.org/pub/scm/linux/kernel/git/i3c/linux:
i3c: master: dw: stop hardcoding initial speed
i3c: transfer pid from boardinfo to device info
Linus Torvalds [Wed, 1 Mar 2023 00:01:08 +0000 (16:01 -0800)]
Merge tag 'docs-6.3-2' of git://git.lwn.net/linux
Pull Documentation stragglers from Jonathan Corbet:
"A handful of documentation patches that were ready before the merge
window, but which I didn't get merged for the first round:
- A recommendation from Thorsten (also akpm) on use of Link tags to
point out problem reports
- Some front-page formatting tweaks
- Another Spanish translation
- One typo(ish) fix"
* tag 'docs-6.3-2' of git://git.lwn.net/linux:
docs: recommend using Link: whenever using Reported-by:
Documentation: front page: use recommended heading adornments
docs/sp_SP: Add process programming-language translation
docs: locking: refer to the actual existing config names
Jack Chen [Thu, 16 Feb 2023 15:10:57 +0000 (10:10 -0500)]
i3c: master: dw: stop hardcoding initial speed
Bus-speed could be default(12.5MHz) or defined by users in dts.
Dw-i3c-master should not hard-code the initial speed to be
I3C_BUS_TYP_I3C_SCL_RATE (12.5MHz)
And because of Synopsys's I3C controller limit (hcnt/lcnt register
length) and core-clk provided, there is a limit to bus speed, too.
For example, when core-clk is 250 MHz, the bus speed cannot be
lowered below 1MHz.
Tested: tested with an i3c sensor and captured with a logic analyzer.
Jiaxun Yang [Tue, 28 Feb 2023 19:34:59 +0000 (19:34 +0000)]
MIPS: Workaround clang inline compat branch issue
Clang is unable to handle the situation that a chunk of inline
assembly ends with a compat branch instruction and then compiler
generates another control transfer instruction immediately after
this compat branch. The later instruction will end up in forbidden
slot and cause exception.
Workaround by add a option to control the use of compact branch.
Currently it's selected by CC_IS_CLANG and hopefully we can change
it to a version check in future if clang manages to fix it.
mips: dts: ralink: mt7621: add phandle to system controller node for watchdog
To allow to access system controller registers from watchdog driver code
add a phandle in the watchdog 'wdt' node. This avoid using arch dependent
operations in driver code.
Nick Alcock [Tue, 28 Feb 2023 13:02:01 +0000 (13:02 +0000)]
powercap: remove MODULE_LICENSE in non-modules
Since commit 8b41fc4454e ("kbuild: create modules.builtin without
Makefile.modbuiltin or tristate.conf"), MODULE_LICENSE declarations
are used to identify modules. As a consequence, uses of the macro
in non-modules will cause modprobe to misidentify their containing
object file as a module when it is not (false positives), and modprobe
might succeed rather than failing with a suitable error message.
So remove it in the files in this commit, none of which can be built as
modules.
Nick Alcock [Fri, 24 Feb 2023 15:07:49 +0000 (15:07 +0000)]
cpufreq: intel_pstate: remove MODULE_LICENSE in non-modules
Since commit 8b41fc4454e ("kbuild: create modules.builtin without
Makefile.modbuiltin or tristate.conf"), MODULE_LICENSE declarations
are used to identify modules. As a consequence, uses of the macro
in non-modules will cause modprobe to misidentify their containing
object file as a module when it is not (false positives), and modprobe
might succeed rather than failing with a suitable error message.
So remove it in the files in this commit, none of which can be built as
modules.
This is a collection of fixes I came up after glancing through an
initial test run with the snappily named Kukui Jacuzzi SKU16 Chromebook
on KernelCI. There are more issues flagged, this is just what I fixed
thus far.
To start capture on Microchip PDMC the enable bits for each supported
microphone need to be set. After this bit is set the PDMC starts to
receive data from microphones and it considers this data as valid data.
Thus if microphones are not ready the PDMC captures anyway data from its
lines. This data is interpreted by the human ear as poc noises.
To avoid this the following software workaround need to be applied when
starting capture:
1/ enable PDMC channel
2/ wait 150ms
3/ execute 16 dummy reads from RHR
4/ clear interrupts
5/ enable interrupts
6/ enable DMA channel
For this workaround to work step 6 need to be executed at the end.
For step 6 was added patch 1/3 from this series. With this, component
DAI driver sets its struct snd_soc_component_driver::start_dma_last = 1
and proper action is taken based on this flag when starting DAI trigger
vs DMA.
This is a collection of fixes I came up after glancing through an
initial test run with the snappily named Kukui Jacuzzi SKU16 Chromebook
on KernelCI. There are more issues flagged, this is just what I fixed
thus far.
This is a collection of fixes I came up after glancing through an
initial test run with the Spherion Chromebook on KernelCI. There are
more issues flagged, this is just what I fixed thus far - the volume
controls on the MT6359 have issues for example, and a lot of controls
aren't marked as Switches like they should be.
Linus Torvalds [Tue, 28 Feb 2023 17:12:47 +0000 (09:12 -0800)]
Merge tag 'pwm/for-6.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/thierry.reding/linux-pwm
Pull pwm updates from Thierry Reding:
"This rather small set of changes includes some minor fixes and
improvements.
The AB8500 driver gained support for reading the initial hardware
state and the Synopsys DesignWare driver received some work to prepare
for device tree and platform support"
* tag 'pwm/for-6.3-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/thierry.reding/linux-pwm:
pwm: dwc: Use devm_pwmchip_add()
pwm: dwc: Move memory allocation to own function
pwm: dwc: Change &pci->dev to dev in probe
dt-bindings: pwm: Document Synopsys DesignWare snps,pwm-dw-apb-timers-pwm2
pwm: iqs620a: Replace one remaining instance of regmap_update_bits()
pwm: ab8500: Implement .get_state()
pwm: ab8500: Fix calculation of duty and period
pwm: lp3943: Drop unused i2c include
dt-bindings: pwm: mediatek: Convert pwm-mediatek to DT schema
pwm: stm32-lp: fix the check on arr and cmp registers update
pwm: Move pwm_capture() dummy to restore order
pwm: sifive: Always let the first pwm_apply_state succeed
Linus Torvalds [Tue, 28 Feb 2023 17:05:47 +0000 (09:05 -0800)]
Merge tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4
Pull ext4 updates from Ted Ts'o:
"Improve performance for ext4 by allowing multiple process to perform
direct I/O writes to preallocated blocks by using a shared inode lock
instead of taking an exclusive lock.
In addition, multiple bug fixes and cleanups"
* tag 'ext4_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
ext4: fix incorrect options show of original mount_opt and extend mount_opt2
ext4: Fix possible corruption when moving a directory
ext4: init error handle resource before init group descriptors
ext4: fix task hung in ext4_xattr_delete_inode
jbd2: fix data missing when reusing bh which is ready to be checkpointed
ext4: update s_journal_inum if it changes after journal replay
ext4: fail ext4_iget if special inode unallocated
ext4: fix function prototype mismatch for ext4_feat_ktype
ext4: remove unnecessary variable initialization
ext4: fix inode tree inconsistency caused by ENOMEM
ext4: refuse to create ea block when umounted
ext4: optimize ea_inode block expansion
ext4: remove dead code in updating backup sb
ext4: dio take shared inode lock when overwriting preallocated blocks
ext4: don't show commit interval if it is zero
ext4: use ext4_fc_tl_mem in fast-commit replay path
ext4: improve xattr consistency checking and error reporting
Heiko Stuebner [Wed, 8 Feb 2023 22:53:28 +0000 (23:53 +0100)]
RISC-V: improve string-function assembly
Adapt the suggestions for the assembly string functions that Andrew
suggested but that I didn't manage to include into the series that
got applied.
This includes improvements to two comments, removal of unneeded labels
and moving one instruction slightly higher to contradict an
explanatory comment.
Fix warn: iterator used outside loop: 'xfer'. 'xfer' variable contain
invalid value in few conditions. Complete transfer within DATA phase
in successful case and at the end for failed transfer.
Arnd Bergmann [Mon, 27 Feb 2023 08:58:26 +0000 (09:58 +0100)]
ASoC: zl38060 add gpiolib dependency
Without gpiolib, this driver fails to link:
arm-linux-gnueabi-ld: sound/soc/codecs/zl38060.o: in function `chip_gpio_get':
zl38060.c:(.text+0x30): undefined reference to `gpiochip_get_data'
arm-linux-gnueabi-ld: sound/soc/codecs/zl38060.o: in function `zl38_spi_probe':
zl38060.c:(.text+0xa18): undefined reference to `devm_gpiochip_add_data_with_key'
This appears to have been in the driver since the start, but is hard to
hit in randconfig testing since gpiolib is almost always selected by something
else.
Mark Brown [Sat, 25 Feb 2023 21:48:13 +0000 (21:48 +0000)]
ASoC: sam9g20ek: Disable capture unless building with microphone input
Without modification the AT91SAM9G20-EK has no capture support, none of the
inputs of the CODEC are wired to anything to useful and there are no paths
supporting loopback. Since the audio is clocked from the CODEC and the DAPM
inputs are marked as unusable this means that capture will fail to transfer
any data as the ADC path can't be powered up.
Flag this in the device description so apps don't see unusable capture
support, guarded with the existing optional define for mic input.
Mark Brown [Fri, 24 Feb 2023 14:03:58 +0000 (14:03 +0000)]
ASoC: mt8192: Fix range for sidetone positive gain
The Sidetone_Positive_Gain_dB control reports a range of 0..100 as valid
but the put() function rejects anything larger than 24. Fix this.
There are numerous other problems with this control, the name is very non
idiomatic and it should be a TLV, but it's ABI so probably we should leave
those alone.
Mark Brown [Fri, 24 Feb 2023 14:03:55 +0000 (14:03 +0000)]
ASoC: mt8192: Remove spammy log messages
There are a lot of info level log messages in the mt8192 ADDA driver which
are trivially triggerable from userspace, many in normal operation. Remove
these to avoid spamming the console.
Claudiu Beznea [Tue, 28 Feb 2023 11:01:45 +0000 (13:01 +0200)]
ASoC: mchp-pdmc: fix poc noise at capture startup
Microchip PDMC IP doesn't filter microphone noises on startup. By default,
it captures data received from digital microphones after
the MCHP_PDMC_MR.EN bits are set. Thus when enable is set on PDMC side the
digital microphones might not be ready yet and PDMC captures data from then
in this time. This data captured is poc noise. To avoid this the software
workaround is to the following:
1/ enable PDMC channel
2/ wait 150ms (on SAMA7G5-EK setup)
3/ execute 16 dummy reads from RHR
4/ clear interrupts
5/ enable interrupts
6/ enable DMA channel
PDMC can work with different types of microphones, thus different boards
could have different microphones. Depending on microphone type the PDMC
would need to wait longer or shorter period (at startup) than the default
chosen one to filter unwanted noise. Thus add microchip,startup-delay-us
binding to let PDMC users to specify startup delay.
Mark Brown [Sun, 26 Feb 2023 12:49:57 +0000 (12:49 +0000)]
ASoC: mt8183: Fix event generation for I2S DAI operations
ALSA control put() operations should return 0 if the value changed so that
events can be generated appropriately for userspace but the custom control
in the MT8183 I2S DAI driver doesn't do that, fix it.
The HPx Mux enumerations define values 5, 6 and 7 but describe them as
"undefined" and map them to the value 0 on writing. Given the descriptions
and behaviour it seems that these values are invalid and should not be
present in the register, the current behaviour is detected as problematic
by mixer-test:
Mark Brown [Sun, 26 Feb 2023 12:47:57 +0000 (12:47 +0000)]
ASoC: mt6358: Validate Wake on Voice 2 writes
Currently the Wake on Voice 2 control accepts and stores any value written
but it reports that only 0 and 1 are valid values. Reject any out of range
values written by userspace.
Mark Brown [Sun, 26 Feb 2023 12:47:56 +0000 (12:47 +0000)]
ASoC: mt6358: Fix event generation for wake on voice stage 2 switch
ALSA control put() operations should return 0 if the value changed so that
events can be generated appropriately for userspace but the custom control
for wake on voice stage 2 doesn't do this, fix it.
Daniel Wagner [Tue, 21 Feb 2023 16:51:06 +0000 (17:51 +0100)]
nvme-fabrics: show well known discovery name
The kernel always logs the unique subsystem name for a discovery
controller, even in the case user space asked for the well known.
This has lead to confusion as the logs of nvme-cli and the kernel
logs didn't match.
First, nvme-cli connects to the well known discovery controller to
figure out if it supports TP8013. If so then nvme-cli disconnects and
connects to the unique discovery controller. Currently, the kernel show
that user space connected twice to the unique one.
To avoid further confusion, show the well known discovery controller if
user space asked for it:
Akinobu Mita [Sun, 26 Feb 2023 12:42:54 +0000 (21:42 +0900)]
nvme-tcp: don't access released socket during error recovery
While the error recovery work is temporarily failing reconnect attempts,
running the 'nvme list' command causes a kernel NULL pointer dereference
by calling getsockname() with a released socket.
During error recovery work, the nvme tcp socket is released and a new one
created, so it is not safe to access the socket without proper check.
Signed-off-by: Akinobu Mita <[email protected]> Fixes: 02c57a82c008 ("nvme-tcp: print actual source IP address through sysfs "address" attr") Reviewed-by: Martin Belanger <[email protected]> Reviewed-by: Hannes Reinecke <[email protected]> Signed-off-by: Christoph Hellwig <[email protected]>
Dan Carpenter [Thu, 16 Feb 2023 12:14:49 +0000 (15:14 +0300)]
nvme-auth: fix an error code in nvme_auth_process_dhchap_challenge()
This function was transitioned from returning NVMe status codes to
returning traditional kernel error codes. However, this particular
return now accidentally returns positive error codes like ENOMEM instead
of negative -ENOMEM.
Fixes: b0ef1b11d390 ("nvme-auth: don't use NVMe status codes") Signed-off-by: Dan Carpenter <[email protected]> Signed-off-by: Christoph Hellwig <[email protected]>
nvme: bring back auto-removal of deleted namespaces during sequential scan
Bring back the check of the Identify Namespace return value for the
legacy NVMe 1.0-style sequential scanning. While NVMe 1.0 does not
support namespace management, there are "modern" cloud solutions like
Google Cloud Platform that claim the obsolete 1.0 compliance for no
good reason while supporting proprietary sideband namespace management.
Joseph Qi [Tue, 28 Feb 2023 04:54:59 +0000 (12:54 +0800)]
io_uring: fix fget leak when fs don't support nowait buffered read
Heming reported a BUG when using io_uring doing link-cp on ocfs2. [1]
Do the following steps can reproduce this BUG:
mount -t ocfs2 /dev/vdc /mnt/ocfs2
cp testfile /mnt/ocfs2/
./link-cp /mnt/ocfs2/testfile /mnt/ocfs2/testfile.1
umount /mnt/ocfs2
Then umount will fail, and it outputs:
umount: /mnt/ocfs2: target is busy.
While tracing umount, it blames mnt_get_count() not return as expected.
Do a deep investigation for fget()/fput() on related code flow, I've
finally found that fget() leaks since ocfs2 doesn't support nowait
buffered read.
io_issue_sqe
|-io_assign_file // do fget() first
|-io_read
|-io_iter_do_read
|-ocfs2_file_read_iter // return -EOPNOTSUPP
|-kiocb_done
|-io_rw_done
|-__io_complete_rw_common // set REQ_F_REISSUE
|-io_resubmit_prep
|-io_req_prep_async // override req->file, leak happens
This was introduced by commit a196c78b5443 in v5.18. Fix it by don't
re-assign req->file if it has already been assigned.
This happens because ioc_refresh_params() is being called without
a properly initialized ioc->rqos, which is happening later in the callee
side.
ioc_refresh_params() -> ioc_autop_idx() tries to access
ioc->rqos.disk->queue but ioc->rqos.disk is NULL, causing the BUG above.
Create function, called ioc_refresh_params_disk(), that is similar to
ioc_refresh_params() but where the "struct gendisk" could be passed as
an explicit argument. This function will be called when ioc->rqos.disk
is not initialized.
Heiko Carstens [Mon, 27 Feb 2023 17:56:10 +0000 (18:56 +0100)]
s390/smp: perform cpu reset before delegating work to target cpu
Clear CPU state (e.g. all TLB entries, prefetched instructions, etc.)
of the target CPU, however without clearing register contents before
starting any work on it.
This puts the target CPU in a more defined state compared to the
current Stop + Restart sigp orders.
If a machine check interrupt hits while user process is
running __s390_handle_mcck() helper function is called
directly from the interrupt handler and terminates the
current process by calling make_task_dead() routine.
The make_task_dead() is not allowed to be called from
interrupt context which forces the machine check handler
switch to the kernel stack and enable local interrupts
first.
The __s390_handle_mcck() could also be called to service
pending work, but this time from the external interrupts
handler. It is the machine check handler that establishes
the work and schedules the external interrupt, therefore
the machine check interrupt itself should be disabled
while reading out the corresponding variable:
However, local_mcck_disable() does not have effect when
__s390_handle_mcck() is called directly form the machine
check handler, since the machine check interrupt is still
disabled. Therefore, it is not the opening bracket to the
following local_mcck_enable() function.
Simplify the user process termination flow by scheduling
the external interrupt and killing the affected process
from the interrupt context.
Assume a kernel-generated signal is always delivered and
ignore a value returned by do_send_sig_info() funciton.
Heiko Carstens [Fri, 24 Feb 2023 10:02:37 +0000 (11:02 +0100)]
s390/cpum_sf: use READ_ONCE_ALIGNED_128() instead of 128-bit cmpxchg
Use READ_ONCE_ALIGNED_128() to read the previous value in front of a
128-bit cmpxchg loop, instead of (mis-)using a 128-bit cmpxchg operation to
do the same.
Heiko Carstens [Fri, 24 Feb 2023 10:02:36 +0000 (11:02 +0100)]
s390/rwonce: add READ_ONCE_ALIGNED_128() macro
Add an s390 specific READ_ONCE_ALIGNED_128() helper, which can be used for
fast block concurrent (atomic) 128-bit accesses.
The used lpq instruction requires 128-bit alignment. This is also the
reason why the compiler doesn't emit this instruction if __READ_ONCE() is
used for 128-bit accesses.
Ard Biesheuvel [Thu, 23 Feb 2023 20:41:01 +0000 (21:41 +0100)]
arm64: kaslr: don't pretend KASLR is enabled if offset < MIN_KIMG_ALIGN
Our virtual KASLR displacement is a randomly chosen multiple of
2 MiB plus an offset that is equal to the physical placement modulo 2
MiB. This arrangement ensures that we can always use 2 MiB block
mappings (or contiguous PTE mappings for 16k or 64k pages) to map the
kernel.
This means that a KASLR offset of less than 2 MiB is simply the product
of this physical displacement, and no randomization has actually taken
place. Currently, we use 'kaslr_offset() > 0' to decide whether or not
randomization has occurred, and so we misidentify this case.
If the kernel image placement is not randomized, modules are allocated
from a dedicated region below the kernel mapping, which is only used for
modules and not for other vmalloc() or vmap() calls.
When randomization is enabled, the kernel image is vmap()'ed randomly
inside the vmalloc region, and modules are allocated in the vicinity of
this mapping to ensure that relative references are always in range.
However, unlike the dedicated module region below the vmalloc region,
this region is not reserved exclusively for modules, and so ordinary
vmalloc() calls may end up overlapping with it. This should rarely
happen, given that vmalloc allocates bottom up, although it cannot be
ruled out entirely.
The misidentified case results in a placement of the kernel image within
2 MiB of its default address. However, the logic that randomizes the
module region is still invoked, and this could result in the module
region overlapping with the start of the vmalloc region, instead of
using the dedicated region below it. If this happens, a single large
vmalloc() or vmap() call will use up the entire region, and leave no
space for loading modules after that.
Since commit 82046702e288 ("efi/libstub/arm64: Replace 'preferred'
offset with alignment check"), this is much more likely to occur on
systems that boot via EFI but lack an implementation of the EFI RNG
protocol, as in that case, the EFI stub will decide to leave the image
where it found it, and the EFI firmware uses 64k alignment only.
Fix this, by correctly identifying the case where the virtual
displacement is a result of the physical displacement only.
Yuezhang Mo [Fri, 10 Feb 2023 05:33:44 +0000 (13:33 +0800)]
exfat: fix the newly allocated clusters are not freed in error handling
In error handling 'free_cluster', before num_alloc clusters allocated,
p_chain->size will not updated and always 0, thus the newly allocated
clusters are not freed.
Yuezhang Mo [Thu, 9 Feb 2023 09:44:50 +0000 (17:44 +0800)]
exfat: don't print error log in normal case
When allocating a new cluster, exFAT first allocates from the
next cluster of the last cluster of the file. If the last cluster
of the file is the last cluster of the volume, allocate from the
first cluster. This is a normal case, but the following error log
will be printed. It makes users confused, so this commit removes
the error log.
[1960905.181545] exFAT-fs (sdb1): hint_cluster is invalid (262130)
projects / linux.git / log