Rewrite the comment explaining why swiotlb copies the original buffer to
the TLB buffer before initiating DMA *from* the device, i.e. before the
device DMAs into the TLB buffer. The existing comment's argument that
preserving the original data can prevent a kernel memory leak is bogus.
If the driver that triggered the mapping _knows_ that the device will
overwrite the entire mapping, or the driver will consume only the written
parts, then copying from the original memory is completely pointless.
If neither of the above holds true, then copying from the original adds
value only if preserving the data is necessary for functional
correctness, or the driver explicitly initialized the original memory.
If the driver didn't initialize the memory, then copying the original
buffer to the TLB buffer simply changes what kernel data is leaked to
user space.
Writing the entire TLB buffer _does_ prevent leaking stale TLB buffer
data from a previous bounce, but that can be achieved by simply zeroing
the TLB buffer when grabbing a slot.
The real reason swiotlb ended up initializing the TLB buffer with the
original buffer is that it's necessary to make swiotlb operate as
transparently as possible, i.e. to behave as closely as possible to
hardware, and to avoid corrupting the original buffer, e.g. if the driver
knows the device will do partial writes and is relying on the unwritten
data to be preserved.
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Link: https://lore.kernel.org/all/ZN5elYQ5szQndN8n@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
If xfs_bmapi_write finds a delalloc extent at the requested range, it
tries to convert the entire delalloc extent to a real allocation.
But if the allocator cannot find a single free extent large enough to
cover the start block of the requested range, xfs_bmapi_write will
return 0 but leave *nimaps set to 0.
In that case we simply need to keep looping with the same startoffset_fsb
so that one of the following allocations will eventually reach the
requested range.
Note that this could affect any caller of xfs_bmapi_write that covers
an existing delayed allocation. As far as I can tell we do not have
any other such caller, though - the regular writeback path uses
xfs_bmapi_convert_delalloc to convert delayed allocations to real ones,
and direct I/O invalidates the page cache first.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: "Darrick J. Wong" <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
When abnormal drop_nlink are detected on the inode,
return error, to avoid corruption propagation.
Signed-off-by: Cheng Lin <cheng.lin130@zte.com.cn>
Reviewed-by: "Darrick J. Wong" <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
This is version 2 of [Omar's] XFS realtime allocator opimization patch
series.
Changes since v1 [1]:
- Fixed potential overflow in patch 4.
- Changed deprecated typedefs to normal struct names
- Fixed broken indentation
- Used xfs_fileoff_t instead of xfs_fsblock_t where appropriate.
- Added calls to xfs_rtbuf_cache_relse anywhere that the cache is used
instead of relying on the buffers being dirtied and thus attached to
the transaction.
- Clarified comments and commit messages in a few places.
- Added Darrick's Reviewed-bys.
Cover letter from v1:
Our distributed storage system uses XFS's realtime device support as a
way to split an XFS filesystem between an SSD and an HDD -- we configure
the HDD as the realtime device so that metadata goes on the SSD and data
goes on the HDD.
We've been running this in production for a few years now, so we have
some fairly fragmented filesystems. This has exposed various CPU
inefficiencies in the realtime allocator. These became even worse when
we experimented with using XFS_XFLAG_EXTSIZE to force files to be
allocated contiguously.
This series adds several optimizations that don't change the realtime
allocator's decisions, but make them happen more efficiently, mainly by
avoiding redundant work. We've tested these in production and measured
~10%% lower CPU utilization. Furthermore, it made it possible to use
XFS_XFLAG_EXTSIZE to force contiguous allocations -- without these
patches, our most fragmented systems would become unresponsive due to
high CPU usage in the realtime allocator, but with them, CPU utilization
is actually ~4-6%% lower than before, and disk I/O utilization is 15-20%%
lower.
Patches 2 and 3 are preparations for later optimizations; the remaining
patches are the optimizations themselves.
1: https://lore.kernel.org/linux-xfs/cover.1687296675.git.osandov@osandov.com/
v2.1: djwong rebased everything atop his own cleanups, added dave's rtalloc_args
v2.2: rebase with new apis and clean them up too
v2.3: move struct definition around for lolz
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQQ2qTKExjcn+O1o2YRKO3ySh0YRpgUCZTFTEQAKCRBKO3ySh0YR
pkbpAQD52XiQ7GXAiZSfJDStw1cr3C664AhmX/KdGtzHPmqxIQD/e8I0u1GEPezK
6qnipiiFBq/px/hqakx5t8VgHfucwwY=
=XdBg
-----END PGP SIGNATURE-----
Merge tag 'rtalloc-speedups-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into xfs-6.7-mergeA
xfs: CPU usage optimizations for realtime allocator [v2.3]
This is version 2 of [Omar's] XFS realtime allocator opimization patch
series.
Changes since v1 [1]:
- Fixed potential overflow in patch 4.
- Changed deprecated typedefs to normal struct names
- Fixed broken indentation
- Used xfs_fileoff_t instead of xfs_fsblock_t where appropriate.
- Added calls to xfs_rtbuf_cache_relse anywhere that the cache is used
instead of relying on the buffers being dirtied and thus attached to
the transaction.
- Clarified comments and commit messages in a few places.
- Added Darrick's Reviewed-bys.
Cover letter from v1:
Our distributed storage system uses XFS's realtime device support as a
way to split an XFS filesystem between an SSD and an HDD -- we configure
the HDD as the realtime device so that metadata goes on the SSD and data
goes on the HDD.
We've been running this in production for a few years now, so we have
some fairly fragmented filesystems. This has exposed various CPU
inefficiencies in the realtime allocator. These became even worse when
we experimented with using XFS_XFLAG_EXTSIZE to force files to be
allocated contiguously.
This series adds several optimizations that don't change the realtime
allocator's decisions, but make them happen more efficiently, mainly by
avoiding redundant work. We've tested these in production and measured
~10%% lower CPU utilization. Furthermore, it made it possible to use
XFS_XFLAG_EXTSIZE to force contiguous allocations -- without these
patches, our most fragmented systems would become unresponsive due to
high CPU usage in the realtime allocator, but with them, CPU utilization
is actually ~4-6%% lower than before, and disk I/O utilization is 15-20%%
lower.
Patches 2 and 3 are preparations for later optimizations; the remaining
patches are the optimizations themselves.
1: https://lore.kernel.org/linux-xfs/cover.1687296675.git.osandov@osandov.com/
v2.1: djwong rebased everything atop his own cleanups, added dave's rtalloc_args
v2.2: rebase with new apis and clean them up too
v2.3: move struct definition around for lolz
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
* tag 'rtalloc-speedups-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux:
xfs: don't look for end of extent further than necessary in xfs_rtallocate_extent_near()
xfs: don't try redundant allocations in xfs_rtallocate_extent_near()
xfs: limit maxlen based on available space in xfs_rtallocate_extent_near()
xfs: return maximum free size from xfs_rtany_summary()
xfs: invert the realtime summary cache
xfs: simplify rt bitmap/summary block accessor functions
xfs: simplify xfs_rtbuf_get calling conventions
xfs: cache last bitmap block in realtime allocator
xfs: consolidate realtime allocation arguments
Since the rtbitmap and rtsummary accessor functions have proven more
controversial than the rest of the macro refactoring, split the patchset
into two to make review easier.
v1.1: various cleanups suggested by hch
v1.2: rework the accessor functions to reduce the amount of cursor
tracking required, and create explicit bitmap/summary logging
functions
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQQ2qTKExjcn+O1o2YRKO3ySh0YRpgUCZTFTEQAKCRBKO3ySh0YR
prDOAQD5497tQO1iiA8bdO/kHGHwUWZub45q6+AwZRjzzc5BdAD8CgvZ/8F14hGC
i80DF+GsmFL95mNHwlB1FS+lhJMF8wk=
=R0rf
-----END PGP SIGNATURE-----
Merge tag 'refactor-rtbitmap-accessors-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into xfs-6.7-mergeA
xfs: refactor rtbitmap/summary accessors [v1.2]
Since the rtbitmap and rtsummary accessor functions have proven more
controversial than the rest of the macro refactoring, split the patchset
into two to make review easier.
v1.1: various cleanups suggested by hch
v1.2: rework the accessor functions to reduce the amount of cursor
tracking required, and create explicit bitmap/summary logging
functions
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
* tag 'refactor-rtbitmap-accessors-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux:
xfs: use accessor functions for summary info words
xfs: create helpers for rtsummary block/wordcount computations
xfs: use accessor functions for bitmap words
xfs: create a helper to handle logging parts of rt bitmap/summary blocks
In preparation for adding block headers and enforcing endian order in
rtbitmap and rtsummary blocks, replace open-coded geometry computations
and fugly macros with proper helper functions that can be typechecked.
Soon we'll be needing to add more complex logic to the helpers.
v1.1: various cleanups suggested by hch
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQQ2qTKExjcn+O1o2YRKO3ySh0YRpgUCZTFTEQAKCRBKO3ySh0YR
pn5HAP436EI/oXPv2LN6YgYXgbVEzhFdS4CsVE75L9Vs1Wux0wEAsKLyTNC/Re4d
rhpraQBZty0/YbTSzvXS0IxUIeOq7Qk=
=FzAh
-----END PGP SIGNATURE-----
Merge tag 'refactor-rtbitmap-macros-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into xfs-6.7-mergeA
xfs: refactor rtbitmap/summary macros [v1.1]
In preparation for adding block headers and enforcing endian order in
rtbitmap and rtsummary blocks, replace open-coded geometry computations
and fugly macros with proper helper functions that can be typechecked.
Soon we'll be needing to add more complex logic to the helpers.
v1.1: various cleanups suggested by hch
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
* tag 'refactor-rtbitmap-macros-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux:
xfs: create helpers for rtbitmap block/wordcount computations
xfs: convert rt summary macros to helpers
xfs: convert open-coded xfs_rtword_t pointer accesses to helper
xfs: remove XFS_BLOCKWSIZE and XFS_BLOCKWMASK macros
xfs: convert the rtbitmap block and bit macros to static inline functions
This series replaces all the open-coded integer division and
multiplication conversions between rt blocks and rt extents with calls
to static inline helpers. Having cleaned all that up, the helpers are
augmented to skip the expensive operations in favor of bit shifts and
masking if the rt extent size is a power of two.
v1.1: various cleanups suggested by hch
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQQ2qTKExjcn+O1o2YRKO3ySh0YRpgUCZTFTEQAKCRBKO3ySh0YR
pn8VAPwIqtQs6t19YAuIPy+Iwo0xls+G/XWTD+l+rWurtJBGKwD/csGP7Bt0w71i
UzJ+VhyFUacV4MJdyGCrkGi0tuwUjQU=
=zyjM
-----END PGP SIGNATURE-----
Merge tag 'refactor-rt-unit-conversions-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into xfs-6.7-mergeA
xfs: refactor rt extent unit conversions [v1.1]
This series replaces all the open-coded integer division and
multiplication conversions between rt blocks and rt extents with calls
to static inline helpers. Having cleaned all that up, the helpers are
augmented to skip the expensive operations in favor of bit shifts and
masking if the rt extent size is a power of two.
v1.1: various cleanups suggested by hch
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
* tag 'refactor-rt-unit-conversions-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux:
xfs: use shifting and masking when converting rt extents, if possible
xfs: create rt extent rounding helpers for realtime extent blocks
xfs: convert do_div calls to xfs_rtb_to_rtx helper calls
xfs: create helpers to convert rt block numbers to rt extent numbers
xfs: create a helper to convert extlen to rtextlen
xfs: create a helper to compute leftovers of realtime extents
xfs: create a helper to convert rtextents to rtblocks
A recent patch changed xfrm6_udp_encap_rcv to not
free the skb itself anymore but fogot the case
where xfrm4_udp_encap_rcv is called subsequently.
Fix this by moving the call to xfrm4_udp_encap_rcv
from __xfrm6_udp_encap_rcv to xfrm6_udp_encap_rcv.
Fixes: 221ddb723d ("xfrm: Support GRO for IPv6 ESP in UDP encapsulation")
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
The realtime code uses xfs_rtblock_t and xfs_fsblock_t in a lot of
places, and it's very confusing. Clean up all the type usage so that an
xfs_rtblock_t is always a block within the realtime volume, an
xfs_fileoff_t is always a file offset within a realtime metadata file,
and an xfs_rtxnumber_t is always a rt extent within the realtime volume.
v1.1: various cleanups suggested by hch
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQQ2qTKExjcn+O1o2YRKO3ySh0YRpgUCZTFTEQAKCRBKO3ySh0YR
pkdNAP0XCeqWYWQK7QspUy2PU0fJRtVtXi44hyz3DMPrKJg4MgEAgbuvy9iUoqXX
xeQ9hYBQS5N8+lSmgwRgMqTvEFTNgQU=
=cExb
-----END PGP SIGNATURE-----
Merge tag 'clean-up-realtime-units-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into xfs-6.7-mergeA
xfs: clean up realtime type usage [v1.1]
The realtime code uses xfs_rtblock_t and xfs_fsblock_t in a lot of
places, and it's very confusing. Clean up all the type usage so that an
xfs_rtblock_t is always a block within the realtime volume, an
xfs_fileoff_t is always a file offset within a realtime metadata file,
and an xfs_rtxnumber_t is always a rt extent within the realtime volume.
v1.1: various cleanups suggested by hch
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
* tag 'clean-up-realtime-units-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux:
xfs: convert rt extent numbers to xfs_rtxnum_t
xfs: rename xfs_verify_rtext to xfs_verify_rtbext
xfs: convert rt bitmap extent lengths to xfs_rtbxlen_t
xfs: convert rt bitmap/summary block numbers to xfs_fileoff_t
xfs: convert xfs_extlen_t to xfs_rtxlen_t in the rt allocator
xfs: move the xfs_rtbitmap.c declarations to xfs_rtbitmap.h
xfs: make sure maxlen is still congruent with prod when rounding down
xfs: fix units conversion error in xfs_bmap_del_extent_delay
This is a preparatory patchset that fixes a few miscellaneous bugs
before we start in on larger cleanups of realtime units usage.
v1.1: various cleanups suggested by hch
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQQ2qTKExjcn+O1o2YRKO3ySh0YRpgUCZTFTDQAKCRBKO3ySh0YR
psyEAQD4X4wWlxmJdgzI4ggmrh8AJyRrcy7vJgF3kBPm91RLFwD7BuX8vRjKim3y
p/90GWxBbymV76uL7XMGMBlFO2tE6w4=
=kflG
-----END PGP SIGNATURE-----
Merge tag 'realtime-fixes-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux into xfs-6.7-mergeA
xfs: minor bugfixes for rt stuff [v1.1]
This is a preparatory patchset that fixes a few miscellaneous bugs
before we start in on larger cleanups of realtime units usage.
v1.1: various cleanups suggested by hch
With a bit of luck, this should all go splendidly.
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Chandan Babu R <chandanbabu@kernel.org>
* tag 'realtime-fixes-6.7_2023-10-19' of https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfs-linux:
xfs: rt stubs should return negative errnos when rt disabled
xfs: prevent rt growfs when quota is enabled
xfs: hoist freeing of rt data fork extent mappings
xfs: bump max fsgeom struct version
Fix a typo in the path of this reference.
Fixes: 094f391013 ("docs: usb: Add documentation for the UVC Gadget")
Cc: Daniel Scally <dan.scally@ideasonboard.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Vegard Nossum <vegard.nossum@oracle.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Message-ID: <20231022185311.919325-1-vegard.nossum@oracle.com>
Add this to the section on fixing warnings.
Cc: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Signed-off-by: Vegard Nossum <vegard.nossum@oracle.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Message-ID: <20231022184910.919201-1-vegard.nossum@oracle.com>
In our CI testing, we use some commands as below to only turn a specific
type of warnings into errors, but we notice that kernel-doc warnings
are also turned into errors unexpectedly.
$ make KCFLAGS="-Werror=return-type" W=1 kernel/fork.o
kernel/fork.c:1406: warning: Function parameter or member 'mm' not described in 'set_mm_exe_file'
kernel/fork.c:1406: warning: Function parameter or member 'new_exe_file' not described in 'set_mm_exe_file'
kernel/fork.c:1441: warning: Function parameter or member 'mm' not described in 'replace_mm_exe_file'
kernel/fork.c:1441: warning: Function parameter or member 'new_exe_file' not described in 'replace_mm_exe_file'
kernel/fork.c:1491: warning: Function parameter or member 'mm' not described in 'get_mm_exe_file'
kernel/fork.c:1510: warning: Function parameter or member 'task' not described in 'get_task_exe_file'
kernel/fork.c:1534: warning: Function parameter or member 'task' not described in 'get_task_mm'
kernel/fork.c:2109: warning: bad line:
kernel/fork.c:2130: warning: Function parameter or member 'ret' not described in '__pidfd_prepare'
kernel/fork.c:2130: warning: Excess function parameter 'pidfd' description in '__pidfd_prepare'
kernel/fork.c:2179: warning: Function parameter or member 'ret' not described in 'pidfd_prepare'
kernel/fork.c:2179: warning: Excess function parameter 'pidfd' description in 'pidfd_prepare'
kernel/fork.c:3195: warning: expecting prototype for clone3(). Prototype was for sys_clone3() instead
13 warnings as Errors
make[3]: *** [scripts/Makefile.build:243: kernel/fork.o] Error 13
make[3]: *** Deleting file 'kernel/fork.o'
make[2]: *** [scripts/Makefile.build:480: kernel] Error 2
make[1]: *** [/root/linux/Makefile:1913: .] Error 2
make: *** [Makefile:234: __sub-make] Error 2
>From the git history, commit 2c12c8103d ("scripts/kernel-doc:
optionally treat warnings as errors") introduces a new command-line
option to make kernel-doc warnings into errors. It can also read the
KCFLAGS environment variable to decide whether to turn this option on,
but the regex used for matching may not be accurate enough. It can match
both "-Werror" and "-Werror=<diagnostic-type>", so the option is turned
on by mistake in the latter case.
Fix this by strictly matching the flag "-Werror": there must be a space
or start of string in the front, and a space or end of string at the
end. This can handle all the following cases correctly:
KCFLAGS="-Werror" make W=1 [MATCH]
KCFLAGS="-Werror=return-type" make W=1 [NO MATCH]
KCFLAGS="-Wcomment -Werror -Wundef" make W=1 [MATCH]
KCFLAGS="-Wcomment -Werror=return-type -Wundef" make W=1 [NO MATCH]
Fixes: 2c12c8103d ("scripts/kernel-doc: optionally treat warnings as errors")
Signed-off-by: Yujie Liu <yujie.liu@intel.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Message-ID: <20231019095637.2471840-1-yujie.liu@intel.com>
Remove the repeated word "the" in comments.
Signed-off-by: Charles Han <hanchunchao@inspur.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Message-ID: <20231018023046.30022-1-hanchunchao@inspur.com>
Add SPDX-License-Identifier to fix the checkpatch warning:
WARNING:SPDX_LICENSE_TAG: Missing or malformed SPDX-License-Identifier tag in line 1
\#26: FILE: Documentation/translations/zh_TW/dev-tools/index.rst:1:
+.. include:: ../disclaimer-zh_TW.rst
Signed-off-by: Min-Hua Chen <minhuadotchen@gmail.com>
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202310110859.tumJoXFl-lkp@intel.com/
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Message-ID: <20231011233757.181652-1-minhuadotchen@gmail.com>
Translate subsystem-apis.rst into Chinese.
The existence of this document is crucial. Without it, other Chinese
documents included in (such as sched-design-CFS.rst) will not be
displayed correctly in the left side of the web page.
Signed-off-by: Tang Yizhou <yizhou.tang@shopee.com>
Reviewed-by: Yanteng Si <siyanteng@loongson.cn>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Message-ID: <20231011152520.31079-1-tangyeechou@gmail.com>
ksmbd is missing supporting to convert filename included surrogate pair
characters. It triggers a "file or folder does not exist" error in
Windows client.
[Steps to Reproduce for bug]
1. Create surrogate pair file
touch $(echo -e '\xf0\x9d\x9f\xa3')
touch $(echo -e '\xf0\x9d\x9f\xa4')
2. Try to open these files in ksmbd share through Windows client.
This patch update unicode functions not to consider about surrogate pair
(and IVS).
Reviewed-by: Marios Makassikis <mmakassikis@freebox.fr>
Tested-by: Marios Makassikis <mmakassikis@freebox.fr>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
Physical ib_device does not have an underlying net_device, thus its
association with IPoIB net_device cannot be retrieved via
ops.get_netdev() or ib_device_get_by_netdev(). ksmbd reads physical
ib_device port GUID from the lower 16 bytes of the hardware addresses on
IPoIB net_device and match its underlying ib_device using ib_find_gid()
Signed-off-by: Kangjing Huang <huangkangjing@gmail.com>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Reviewed-by: Tom Talpey <tom@talpey.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Running smb2.rename test from Samba smbtorture suite against a kernel built
with lockdep triggers a "possible recursive locking detected" warning.
This is because mnt_want_write() is called twice with no mnt_drop_write()
in between:
-> ksmbd_vfs_mkdir()
-> ksmbd_vfs_kern_path_create()
-> kern_path_create()
-> filename_create()
-> mnt_want_write()
-> mnt_want_write()
Fix this by removing the mnt_want_write/mnt_drop_write calls from vfs
helpers that call kern_path_create().
Full lockdep trace below:
============================================
WARNING: possible recursive locking detected
6.6.0-rc5 #775 Not tainted
--------------------------------------------
kworker/1:1/32 is trying to acquire lock:
ffff888005ac83f8 (sb_writers#5){.+.+}-{0:0}, at: ksmbd_vfs_mkdir+0xe1/0x410
but task is already holding lock:
ffff888005ac83f8 (sb_writers#5){.+.+}-{0:0}, at: filename_create+0xb6/0x260
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(sb_writers#5);
lock(sb_writers#5);
*** DEADLOCK ***
May be due to missing lock nesting notation
4 locks held by kworker/1:1/32:
#0: ffff8880064e4138 ((wq_completion)ksmbd-io){+.+.}-{0:0}, at: process_one_work+0x40e/0x980
#1: ffff888005b0fdd0 ((work_completion)(&work->work)){+.+.}-{0:0}, at: process_one_work+0x40e/0x980
#2: ffff888005ac83f8 (sb_writers#5){.+.+}-{0:0}, at: filename_create+0xb6/0x260
#3: ffff8880057ce760 (&type->i_mutex_dir_key#3/1){+.+.}-{3:3}, at: filename_create+0x123/0x260
Cc: stable@vger.kernel.org
Fixes: 40b268d384 ("ksmbd: add mnt_want_write to ksmbd vfs functions")
Signed-off-by: Marios Makassikis <mmakassikis@freebox.fr>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
Fix argument list that the kdoc format and script verified in
ksmbd_vfs_setxattr().
fs/smb/server/vfs.c:929: warning: Function parameter or member 'path'
not described in 'ksmbd_vfs_setxattr'
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
If ksmbd_iov_pin_rsp fail, io vertor should be rollback.
This patch moves memory allocations to before setting the io vector
to avoid rollbacks.
Fixes: e2b76ab8b5 ("ksmbd: add support for read compound")
Signed-off-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
fs/smb/server/mgmt/user_config.h:21: Remove the unused field 'failed_login_count' from the ksmbd_user struct.
Signed-off-by: Cheng-Han Wu <hank20010209@gmail.com>
Acked-by: Namjae Jeon <linkinjeon@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
The NTLM authenticate message currently sets the NTLMSSP_NEGOTIATE_VERSION
flag but does not populate the VERSION structure. This commit fixes this
bug by ensuring that the flag is set and the version details are included
in the message.
Signed-off-by: Meetakshi Setiya <msetiya@microsoft.com>
Reviewed-by: Bharath SM <bharathsm@microsoft.com>
Reviewed-by: Paulo Alcantara (SUSE) <pc@manguebit.com>
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
For example:
touch -h -t 02011200 testfile
where testfile is a symlink would not change the timestamp, but
touch -t 02011200 testfile
does work to change the timestamp of the target
Suggested-by: David Howells <dhowells@redhat.com>
Reported-by: Micah Veilleux <micah.veilleux@iba-group.com>
Closes: https://bugzilla.samba.org/show_bug.cgi?id=14476
Cc: stable@vger.kernel.org
Signed-off-by: Steve French <stfrench@microsoft.com>
SOFTCONN tasks need to periodically check if the transport is still
connected, so that they can time out if that is not the case.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Fix up xs_wake_error() to close the socket when a hard error is being
reported. Usually, that means an ECONNRESET was received on a connection
attempt.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If connect() is returning ECONNRESET, it usually means that nothing is
listening on that port. If so, a rebind might be required in order to
obtain the new port on which the RPC service is listening.
Fixes: fd01b25979 ("SUNRPC: ECONNREFUSED should cause a rebind.")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If we're using the 'softerr' mount option, we may want to allow
layoutget to return EAGAIN to allow knfsd server threads to return a
JUKEBOX/DELAY error to the client instead of busy waiting.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
When using a 'softerr' mount, the NFSv4 client can get stuck waiting
forever while the server just returns NFS4ERR_DELAY. Among other things,
this causes the knfsd server threads to busy wait.
Add a parameter that tells the NFSv4 client how many times to retry
before giving up.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Add a compatible for the Qualcomm Kryo 465 found in SM7125.
Signed-off-by: David Wronek <davidwronek@gmail.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20231021071619.187374-1-davidwronek@gmail.com
Signed-off-by: Rob Herring <robh@kernel.org>
When building with W=1:
CC arch/m68k/68000/timers.o
arch/m68k/68000/timers.c:120:5: warning: no previous prototype for ‘m68328_hwclk’ [-Wmissing-prototypes]
int m68328_hwclk(int set, struct rtc_time *t)
^~~~~~~~~~~~
Include m68328.h to get prototype for m68328_hwclk().
Signed-off-by: Greg Ungerer <gerg@linux-m68k.org>
When building with W=1:
CC arch/m68k/68000/ints.o
arch/m68k/68000/ints.c:77:6: warning: no previous prototype for ‘process_int’ [-Wmissing-prototypes]
void process_int(int vec, struct pt_regs *fp)
^~~~~~~~~~~
arch/m68k/68000/ints.c:153:13: warning: no previous prototype for ‘trap_init’ [-Wmissing-prototypes]
void __init trap_init(void)
^~~~~~~~~
Include linux/cpu.h to get the prototype for taps_init().
Create a local ints.h for prototype of process_int(). Also mark
process_int() as asmlinkage, since it is called from the first level
interrupt assembly handler.
Signed-off-by: Greg Ungerer <gerg@linux-m68k.org>
When building with W=1:
CC arch/m68k/mm/mcfmmu.o
arch/m68k/mm/mcfmmu.c: In function ‘paging_init’:
arch/m68k/mm/mcfmmu.c:41:30: warning: variable ‘bootmem_end’ set but not used [-Wunused-but-set-variable]
unsigned long next_pgtable, bootmem_end;
^~~~~~~~~~~
Remove variable bootmem_end and its unused setting.
Signed-off-by: Greg Ungerer <gerg@linux-m68k.org>
When building with W=1:
CC arch/m68k/kernel/uboot.o
arch/m68k/kernel/uboot.c: In function ‘parse_uboot_commandline’:
arch/m68k/kernel/uboot.c:68:36: warning: variable ‘uboot_initrd_end’ set but not used [-Wunused-but-set-variable]
unsigned long uboot_initrd_start, uboot_initrd_end;
^~~~~~~~~~~~~~~~
arch/m68k/kernel/uboot.c:68:16: warning: variable ‘uboot_initrd_start’ set but not used [-Wunused-but-set-variable]
unsigned long uboot_initrd_start, uboot_initrd_end;
^~~~~~~~~~~~~~~~~~
arch/m68k/kernel/uboot.c:66:16: warning: variable ‘uboot_kbd’ set but not used [-Wunused-but-set-variable]
unsigned long uboot_kbd;
^~~~~~~~~
arch/m68k/kernel/uboot.c: At top level:
arch/m68k/kernel/uboot.c:90:13: warning: no previous prototype for ‘process_uboot_commandline’ [-Wmissing-prototypes]
__init void process_uboot_commandline(char *commandp, int size)
^~~~~~~~~~~~~~~~~~~~~~~~~
A couple of issues here. Firstly we already have a bootinfo.h that has a
prototype for process_uboot_commandline(), we should include that.
Secondly uboot_kbd is not used at all and can be removed. Thirdly the
conditional code based on CONFIG_BLK_DEV_INITRD means that sometimes
uboot_initrd_start and uboot_initrd_end are not needed. Make their
declaration and asignment conditional on CONFIG_BLK_DEV_INITRD same as
the code that uses them.
Signed-off-by: Greg Ungerer <gerg@linux-m68k.org>
When building with W=1:
CC arch/m68k/coldfire/intc.o
arch/m68k/coldfire/intc.c:83:6: warning: no previous prototype for ‘mcf_maskimr’ [-Wmissing-prototypes]
void mcf_maskimr(unsigned int mask)
^~~~~~~~~~~
The mcf_maskimr() function is only used within this file, make it static
to reduce name space pollution and fix warning.
Signed-off-by: Greg Ungerer <gerg@linux-m68k.org>
When building with W=1:
CC arch/m68k/coldfire/gpio.o
arch/m68k/coldfire/gpio.c:19:5: warning: no previous prototype for ‘__mcfgpio_get_value’ [-Wmissing-prototypes]
int __mcfgpio_get_value(unsigned gpio)
^~~~~~~~~~~~~~~~~~~
arch/m68k/coldfire/gpio.c:25:6: warning: no previous prototype for ‘__mcfgpio_set_value’ [-Wmissing-prototypes]
void __mcfgpio_set_value(unsigned gpio, int value)
^~~~~~~~~~~~~~~~~~~
arch/m68k/coldfire/gpio.c:50:5: warning: no previous prototype for ‘__mcfgpio_direction_input’ [-Wmissing-prototypes]
int __mcfgpio_direction_input(unsigned gpio)
^~~~~~~~~~~~~~~~~~~~~~~~~
arch/m68k/coldfire/gpio.c:65:5: warning: no previous prototype for ‘__mcfgpio_direction_output’ [-Wmissing-prototypes]
int __mcfgpio_direction_output(unsigned gpio, int value)
^~~~~~~~~~~~~~~~~~~~~~~~~~
arch/m68k/coldfire/gpio.c:96:5: warning: no previous prototype for ‘__mcfgpio_request’ [-Wmissing-prototypes]
int __mcfgpio_request(unsigned gpio)
^~~~~~~~~~~~~~~~~
arch/m68k/coldfire/gpio.c:102:6: warning: no previous prototype for ‘__mcfgpio_free’ [-Wmissing-prototypes]
void __mcfgpio_free(unsigned gpio)
^~~~~~~~~~~~~~
The local m68k asm version of gpio.h has prototypes for all of these,
but they are not always visible depending on the config options enabled.
Move the prototypes so they are always visible.
Signed-off-by: Greg Ungerer <gerg@linux-m68k.org>
When building with W=1:
arch/m68k/coldfire/vectors.c:43:13: warning: no previous prototype for ‘trap_init’ [-Wmissing-prototypes]
void __init trap_init(void)
^~~~~~~~~
Fix this by introducing a new header file "vectors.h" for holding the
prototypes of functions implemented in arch/m68k/coldfire/vectors.c.
Signed-off-by: Greg Ungerer <gerg@linux-m68k.org>
The ColdFire based Cleopatra family of boards use mostly the same
external pin arrangements as the NETtel board family. The build uses the
NETtel specific code as needed, but not all the conditional defines
allow for this. If you have the CONFIG_NETtel config option set
everything compiles as expected, but if you only select the
CONFIG_CLEOPATRA board type then you will get compile failures:
arch/m68k/coldfire/nettel.c: In function ‘nettel_smc91x_init’:
arch/m68k/coldfire/nettel.c:126:2: error: implicit declaration of function ‘mcf_setppdata’; did you mean ‘xas_set_update’? [-Werror=implicit-function-declaration]
mcf_setppdata(0, 0x0080);
^~~~~~~~~~~~~
xas_set_update
Fix the nettel.h include conditional checks to cover all board types.
This also means some code paths need to check for the 5407 SoC - since
one of the Cleopatra board types is based on that. It is very similar
to the 5307 specific code, and it can use that "as-is".
Signed-off-by: Greg Ungerer <gerg@linux-m68k.org>
The ROM region configuration settings used on some nommu m68k systems
(historically mostly 68328 (Dragonball) CPUs) default to an address
of 0. That can easily clash with default RAM address settings which
also default to 0. Of course that is invalid and those ranges overlap,
but if you make no value selection that is what you end up with. Those
default values produce a valid configuration but will fail compilation
like this:
m68k-linux-ld: section .rodata VMA [0000000000001000,0000000000262227] overlaps section .text VMA [0000000000000400,0000000000455e7f]
Looking at the platforms that use the ROM region configuration settings
it is clear that we can choose much better defaults than 0. By far the
most common ROM region settings are these:
CONFIG_ROMVEC=0x10c10000
CONFIG_ROMSTART=0x10c10400
So lets make these the default values.
It is still possible to configure overlapping ROM and RAM regions, but
at least the default selections are now valid.
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202305301407.z33zOjcG-lkp@intel.com/
Signed-off-by: Greg Ungerer <gerg@linux-m68k.org>
The memcpy() in bch2_bkey_append_ptr() is operating on an embedded fake
flexible array which looks to the compiler like it has 0 size. This
causes W=1 builds to emit warnings due to -Wstringop-overflow:
In file included from include/linux/string.h:254,
from include/linux/bitmap.h:11,
from include/linux/cpumask.h:12,
from include/linux/smp.h:13,
from include/linux/lockdep.h:14,
from include/linux/radix-tree.h:14,
from include/linux/backing-dev-defs.h:6,
from fs/bcachefs/bcachefs.h:182:
fs/bcachefs/extents.c: In function 'bch2_bkey_append_ptr':
include/linux/fortify-string.h:57:33: warning: writing 8 bytes into a region of size 0 [-Wstringop-overflow=]
57 | #define __underlying_memcpy __builtin_memcpy
| ^
include/linux/fortify-string.h:648:9: note: in expansion of macro '__underlying_memcpy'
648 | __underlying_##op(p, q, __fortify_size); \
| ^~~~~~~~~~~~~
include/linux/fortify-string.h:693:26: note: in expansion of macro '__fortify_memcpy_chk'
693 | #define memcpy(p, q, s) __fortify_memcpy_chk(p, q, s, \
| ^~~~~~~~~~~~~~~~~~~~
fs/bcachefs/extents.c:235:17: note: in expansion of macro 'memcpy'
235 | memcpy((void *) &k->v + bkey_val_bytes(&k->k),
| ^~~~~~
fs/bcachefs/bcachefs_format.h:287:33: note: destination object 'v' of size 0
287 | struct bch_val v;
| ^
Avoid making any structure changes and just replace the u64 copy into a
direct assignment, side-stepping the entire problem.
Cc: Kent Overstreet <kent.overstreet@linux.dev>
Cc: Brian Foster <bfoster@redhat.com>
Cc: linux-bcachefs@vger.kernel.org
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202309192314.VBsjiIm5-lkp@intel.com/
Link: https://lore.kernel.org/r/20231010235609.work.594-kees@kernel.org
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
For consistency with the rest of the reconstruct_alloc option, we should
be skipping all alloc keys.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Add a new lock for snapshot creation - this addresses a few races with
logged operations and snapshot deletion.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
In snapshot deleion, we have to pick new skiplist nodes for entries that
point to nodes being deleted.
The function that finds a new skiplist node, skipping over entries being
deleted, was incorrect: if n = 0, but the parent node is being deleted,
we also need to skip over that node.
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
Instead of using token pasting to generate methods for each superblock
section, just make the type a parameter to bch2_sb_field_get().
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
