Usual set of small fixes/logging improvements.
One bigger user reported fix, for inode <-> dirent inconsistencies
reported in fsck, after moving a subvolume that had been snapshotted.
-----BEGIN PGP SIGNATURE-----
iQIzBAABCgAdFiEEKnAFLkS8Qha+jvQrE6szbY3KbnYFAmgBYaIACgkQE6szbY3K
bnYcGQ/+K+LsEvGAZ5wtTwUN4KqJIYWhcHYcuLS2mHKf8PMbgYhL7TjmCwb9VWyr
0+GFQcJgfLsl++kX4j7CjG4gHd22aLiwbhMDmSt3r6c4aF29rG+zCpe4W1+7o60k
UIKokfbLUV6b+0vF5bA/W3PmtXK7S8E0yAPMfWxv4/sACu8RUvrUJtrUCKEWwLzC
bcrRGsN91456qNhCrOp3e4t3yZjiGtZIz+SbPYIxdNrZYIMURlGUm+f9sLH3O+2R
NKsi41sggo/TmgUyspH3KCtMT88IDbN07F7O9/zcxgtdfzfC9l9FI6HnvRVSHDOV
boFaH/NdRaIbg+O5kqZXYul+/EPXsYp5B77TL6KQ3jhv3q16uwpv9EL4v6HIwvz9
BTDOfI2y/+YWHMfrtzXgh3C9dZDPS7qxFFWjSjCs/lXwKVz46RjBWVmtQoTJSEmb
Ee29kBGMpkwmH8fqr5KQheJUIeYewpyTVeB6orgtshnrr+aezS6zunIbk7fJ6+Ng
Tc08H/Aqc2KGcyBS3KTLhbReQ1clQKGOqWJymeb1p2V3SMXfABMbh61B1VU1XulC
Al5B7/w/WPwb+T2XZIM2qbmeoRJ8OBara5RWkx4HN8pcYuWV8H6GWJtRJQD/eKSO
pOT5bz8z9N2n/otwrfLT5lfO2fNW1mULCAamn6iSzR+EDHyuaMU=
=pi0D
-----END PGP SIGNATURE-----
Merge tag 'bcachefs-2025-04-17' of git://evilpiepirate.org/bcachefs
Pull bcachefs fixes from Kent Overstreet:
"Usual set of small fixes/logging improvements.
One bigger user reported fix, for inode <-> dirent inconsistencies
reported in fsck, after moving a subvolume that had been snapshotted"
* tag 'bcachefs-2025-04-17' of git://evilpiepirate.org/bcachefs:
bcachefs: Fix snapshotting a subvolume, then renaming it
bcachefs: Add missing READ_ONCE() for metadata replicas
bcachefs: snapshot_node_missing is now autofix
bcachefs: Log message when incompat version requested but not enabled
bcachefs: Print version_incompat_allowed on startup
bcachefs: Silence extent_poisoned error messages
bcachefs: btree_root_unreadable_and_scan_found_nothing now AUTOFIX
bcachefs: fix bch2_dev_usage_full_read_fast()
bcachefs: Don't print data read retry success on non-errors
bcachefs: Add missing error handling
bcachefs: Prevent granting write refs when filesystem is read-only
- Include devices where the platform indicates PCI INTx is not routed
by setting pdev->irq to zero in the expanded virtualization of the
PCI pin register. This provides consistency in the INFO and
SET_IRQS ioctls. (Alex Williamson)
-----BEGIN PGP SIGNATURE-----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=Ce9h
-----END PGP SIGNATURE-----
Merge tag 'vfio-v6.15-rc3' of https://github.com/awilliam/linux-vfio
Pull vfio fix from Alex Williamson:
- Include devices where the platform indicates PCI INTx is not routed
by setting pdev->irq to zero in the expanded virtualization of the
PCI pin register. This provides consistency in the INFO and SET_IRQS
ioctls (Alex Williamson)
* tag 'vfio-v6.15-rc3' of https://github.com/awilliam/linux-vfio:
vfio/pci: Virtualize zero INTx PIN if no pdev->irq
'spi-cpha' was removed from this file. So, replace it in the description
with an existing example. Reformat the paragraph to adhere to max line
length.
Fixes: 233363aba7 ("spi/panel: dt-bindings: drop CPHA and CPOL from common properties")
Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Link: https://patch.msgid.link/20250417111630.53084-2-wsa+renesas@sang-engineering.com
Signed-off-by: Mark Brown <broonie@kernel.org>
A few more device specific fixes plus one trivial quirk. There's a
couple of patches for Tegra which avoid some fairly spectacular log spam
if the hardware breaks in ways which were actually seen in production,
plus a fix for the i.MX driver to propagate errors properly when setting
up the hardware. We also have a trivial patch marking the sun4i driver
as being compatible with GPIO chip selects.
-----BEGIN PGP SIGNATURE-----
iQEzBAABCgAdFiEEreZoqmdXGLWf4p/qJNaLcl1Uh9AFAmgBUaYACgkQJNaLcl1U
h9A3bQf+JlT85HXcrLiUmoQV2L4H09bzSdMTDR5xb5LweOOB4vp81zpxBerY3mPC
xqzHU66M+PXWN0NP8TcxMlLgJ1fZhx/0iET1U07gALGQjJUGHhYUGJxTjuhsnEfP
B/XU0R4FTTCbJ4FnlOTxEfA4sdfpMjByvn6Y0QJfdmcGIiI0dQo83RfZ3ETLgUaX
RFjYvibGkszqu0hDducx7I95gflZOazZ8y0xhli7Cf2ZfI4SDAWpQgVT238r1tlH
RiJbVf+GqIAMkO85Ucmx27mWVzIT8WrQCop6zZTuKW8xSJi/G7rKfynBgDEybxtM
kewqt/wHAyHFNT3uPJDmWt67CndzuA==
=7DAY
-----END PGP SIGNATURE-----
Merge tag 'spi-fix-v6.15-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi
Pull spi fixes from Mark Brown:
"A few more device specific fixes plus one trivial quirk.
There's a couple of patches for Tegra which avoid some fairly
spectacular log spam if the hardware breaks in ways which were
actually seen in production, plus a fix for the i.MX driver to
propagate errors properly when setting up the hardware.
We also have a trivial patch marking the sun4i driver as being
compatible with GPIO chip selects"
* tag 'spi-fix-v6.15-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
spi: spi-imx: Add check for spi_imx_setupxfer()
spi: tegra210-quad: add rate limiting and simplify timeout error message
spi: tegra210-quad: use WARN_ON_ONCE instead of WARN_ON for timeouts
spi: sun4i: add support for GPIO chip select lines
Loongson PCIe Root Ports don't advertise an ACS capability, but they do not
allow peer-to-peer transactions between Root Ports. Add an ACS quirk so
each Root Port can be in a separate IOMMU group.
Signed-off-by: Xianglai Li <lixianglai@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20250403040756.720409-1-chenhuacai@loongson.cn
Print the delay amount that pcie_wait_for_link_delay() is invoked with
instead of the hardcoded 1000ms value in the debug info print.
Fixes: 7b3ba09feb ("PCI/PM: Shorten pci_bridge_wait_for_secondary_bus() wait time for slow links")
Signed-off-by: Wilfred Mallawa <wilfred.mallawa@wdc.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Reviewed-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Link: https://patch.msgid.link/20250414001505.21243-2-wilfred.opensource@gmail.com
CE4100 PCI specific code has no 'pci' suffix in the filename,
intel_mid_pci.c is the only one that duplicates the folder name in its
filename, drop that redundancy.
While at it, group the respective modules in the Makefile.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Ingo Molnar <mingo@kernel.org>
Link: https://patch.msgid.link/20250407070321.3761063-1-andriy.shevchenko@linux.intel.com
We are enabling/disabling engine activity on per-GT basis, so any
errors should be also reported per GT, like:
[ ] xe 0000:00:02.0: [drm] GT0: PF: Failed to enable engine activity function stats (-ENOSPC)
[ ] xe 0000:00:02.0: [drm] GT1: PF: Failed to enable engine activity function stats (-ENOSPC)
Signed-off-by: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Umesh Nerlige Ramappa <umesh.nerlige.ramappa@intel.com>
Cc: Riana Tauro <riana.tauro@intel.com>
Reviewed-by: Riana Tauro <riana.tauro@intel.com>
Link: https://lore.kernel.org/r/20250414202347.1909-2-michal.wajdeczko@intel.com
In February 2003, historic commit
https://git.kernel.org/tglx/history/c/280c1c9a0ea4
("[PATCH] PCI Hotplug: Replace pcihpfs with sysfs.")
removed all invocations of __get_free_page() and free_page() from the PCI
hotplug core without also removing the #include <linux/pagemap.h>
directive.
It removed all invocations of kern_mount(), mntget() and mntput()
without also removing the #include <linux/mount.h> directive.
It removed all invocations of lookup_hash()
without also removing the #include <linux/namei.h> directive.
It removed all invocations of copy_to_user() and copy_from_user()
without also removing the #include <linux/uaccess.h> directive.
These #include directives are still unnecessary today, so drop them.
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/c19e25bf2cefecc14e0822c6a9bb3a7f546258bc.1744640331.git.lukas@wunner.de
ftrace_graph_ent.depth is int, but ftrace_graph_ent_entry.depth is
unsigned long. This confuses trace-cmd on 64-bit big-endian systems and
makes it print a huge amount of spaces. Fix this by using unsigned int,
which has a matching size, instead.
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Sven Schnelle <svens@linux.ibm.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Vasily Gorbik <gor@linux.ibm.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Link: https://lore.kernel.org/20250412221847.17310-2-iii@linux.ibm.com
Fixes: ff5c9c576e ("ftrace: Add support for function argument to graph tracer")
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
The maximum of the ftrace hash bits is made fls(32) in
register_ftrace_direct(), which seems illogical. So, we fix it by making
the max hash bits FTRACE_HASH_MAX_BITS instead.
Link: https://lore.kernel.org/20250413014444.36724-1-dongml2@chinatelecom.cn
Fixes: d05cb47066 ("ftrace: Fix modification of direct_function hash while in use")
Signed-off-by: Menglong Dong <dongml2@chinatelecom.cn>
Acked-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
The subops processing creates new hashes when adding and removing subops.
There were some places that the old hashes that were replaced were not
freed and this caused some memory leaks.
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Link: https://lore.kernel.org/20250417135939.245b128d@gandalf.local.home
Fixes: 0ae6b8ce20 ("ftrace: Fix accounting of subop hashes")
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
There's several locations that free a ftrace hash pointer but may be
referenced again. Reset them to EMPTY_HASH so that a u-a-f bug doesn't
happen.
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Link: https://lore.kernel.org/20250417110933.20ab718b@gandalf.local.home
Fixes: 0ae6b8ce20 ("ftrace: Fix accounting of subop hashes")
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
The reworking to fix and simplify the ftrace_startup_subops() and the
ftrace_shutdown_subops() made it possible for the filter_hash and
notrace_hash variables to be used uninitialized in a way that the compiler
did not catch it.
Initialize both filter_hash and notrace_hash to the EMPTY_HASH as that is
what they should be if they never are used.
Cc: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Link: https://lore.kernel.org/20250417104017.3aea66c2@gandalf.local.home
Reported-by: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Tested-by: Venkat Rao Bagalkote <venkat88@linux.ibm.com>
Fixes: 0ae6b8ce20 ("ftrace: Fix accounting of subop hashes")
Closes: https://lore.kernel.org/all/1db64a42-626d-4b3a-be08-c65e47333ce2@linux.ibm.com/
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Current release - regressions:
- 2 fixes for the netdev per-instance locking
- batman-adv: fix double-hold of meshif when getting enabled
Current release - new code bugs:
- Bluetooth: increment TX timestamping tskey always for stream sockets
- wifi: static analysis and build fixes for the new Intel sub-driver
Previous releases - regressions:
- net: fib_rules: fix iif / oif matching on L3 master (VRF) device
- ipv6: add exception routes to GC list in rt6_insert_exception()
- netfilter: conntrack: fix erroneous removal of offload bit
- Bluetooth:
- fix sending MGMT_EV_DEVICE_FOUND for invalid address
- l2cap: process valid commands in too long frame
- btnxpuart: Revert baudrate change in nxp_shutdown
Previous releases - always broken:
- ethtool: fix memory corruption during SFP FW flashing
- eth: hibmcge: fixes for link and MTU handling, pause frames etc.
- eth: igc: fixes for PTM (PCIe timestamping)
- dsa: b53: enable BPDU reception for management port
Misc:
- fixes for Netlink protocol schemas
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-----BEGIN PGP SIGNATURE-----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=EObk
-----END PGP SIGNATURE-----
Merge tag 'net-6.15-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Pull networking fixes from Jakub Kicinski:
"Including fixes from Bluetooth, CAN and Netfilter.
Current release - regressions:
- two fixes for the netdev per-instance locking
- batman-adv: fix double-hold of meshif when getting enabled
Current release - new code bugs:
- Bluetooth: increment TX timestamping tskey always for stream
sockets
- wifi: static analysis and build fixes for the new Intel sub-driver
Previous releases - regressions:
- net: fib_rules: fix iif / oif matching on L3 master (VRF) device
- ipv6: add exception routes to GC list in rt6_insert_exception()
- netfilter: conntrack: fix erroneous removal of offload bit
- Bluetooth:
- fix sending MGMT_EV_DEVICE_FOUND for invalid address
- l2cap: process valid commands in too long frame
- btnxpuart: Revert baudrate change in nxp_shutdown
Previous releases - always broken:
- ethtool: fix memory corruption during SFP FW flashing
- eth:
- hibmcge: fixes for link and MTU handling, pause frames etc
- igc: fixes for PTM (PCIe timestamping)
- dsa: b53: enable BPDU reception for management port
Misc:
- fixes for Netlink protocol schemas"
* tag 'net-6.15-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (81 commits)
net: ethernet: mtk_eth_soc: revise QDMA packet scheduler settings
net: ethernet: mtk_eth_soc: correct the max weight of the queue limit for 100Mbps
net: ethernet: mtk_eth_soc: reapply mdc divider on reset
net: ti: icss-iep: Fix possible NULL pointer dereference for perout request
net: ti: icssg-prueth: Fix possible NULL pointer dereference inside emac_xmit_xdp_frame()
net: ti: icssg-prueth: Fix kernel warning while bringing down network interface
netfilter: conntrack: fix erronous removal of offload bit
net: don't try to ops lock uninitialized devs
ptp: ocp: fix start time alignment in ptp_ocp_signal_set
net: dsa: avoid refcount warnings when ds->ops->tag_8021q_vlan_del() fails
net: dsa: free routing table on probe failure
net: dsa: clean up FDB, MDB, VLAN entries on unbind
net: dsa: mv88e6xxx: fix -ENOENT when deleting VLANs and MST is unsupported
net: dsa: mv88e6xxx: avoid unregistering devlink regions which were never registered
net: txgbe: fix memory leak in txgbe_probe() error path
net: bridge: switchdev: do not notify new brentries as changed
net: b53: enable BPDU reception for management port
netlink: specs: rt-neigh: prefix struct nfmsg members with ndm
netlink: specs: rt-link: adjust mctp attribute naming
netlink: specs: rtnetlink: attribute naming corrections
...
Subvolume roots and the dirents that point to them are special; they
don't obey the normal snapshot versioning rules because they cross
snapshot boundaries.
We don't keep around older versions of subvolume dirents on rename - we
don't need to, because subvolume dirents are only visible in the parent
subvolume, and we wouldn't be able to match up the different dirent and
inode versions due to crossing the snapshot ID boundary.
That means that when we rename a subvolume, that's been snapshotted, the
older version of the subvolume root will become dangling - it won't have
a dirent that points to it.
That's expected, we just need to tell fsck that this is ok.
Fixes: https://github.com/koverstreet/bcachefs/issues/856
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
kbuf imports have the front offset adjusted and segments removed, but
the tail segments are still included in the segment count that gets
passed in the iov_iter. As the segments aren't necessarily all the
same size, move importing to a separate helper and iterate the
mapped length to get an exact count.
Reviewed-by: Nitesh Shetty <nj.shetty@samsung.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Amery Hung says:
====================
bpf qdisc
Hi all,
This patchset aims to support implementing qdisc using bpf struct_ops.
This version takes a step back and only implements the minimum support
for bpf qdisc. 1) support of adding skb to bpf_list and bpf_rbtree
directly and 2) classful qdisc are deferred to future patchsets. In
addition, we only allow attaching bpf qdisc to root or mq for now.
This is to prevent accidentally breaking exisiting classful qdiscs
that rely on data in a child qdisc. This limit may be lifted in the
future after careful inspection.
* Overview *
This series supports implementing qdisc using bpf struct_ops. bpf qdisc
aims to be a flexible and easy-to-use infrastructure that allows users to
quickly experiment with different scheduling algorithms/policies. It only
requires users to implement core qdisc logic using bpf and implements the
mundane part for them. In addition, the ability to easily communicate
between qdisc and other components will also bring new opportunities for
new applications and optimizations.
* Performance of bpf qdisc *
This patchset includes two qdisc examples, bpf_fifo and bpf_fq, for
__testing__ purposes. For performance test, we compare selftests and their
kernel counterparts to give you a sense of the performance of qdisc
implemented in bpf.
The implementation of bpf_fq is fairly complex and slightly different from
fq so later we only compare the two fifo qdiscs. bpf_fq implements a
scheduling algorithm similar to fq before commit 29f834aa32 ("net_sched:
sch_fq: add 3 bands and WRR scheduling") was introduced. bpf_fifo uses a
single bpf_list as a queue instead of three queues for different
priorities in pfifo_fast. The time complexity of fifo however should be
similar since the queue selection time is negligible.
Test setup:
client -> qdisc -------------> server
~~~~~~~~~~~~~~~ ~~~~~~
nested VM1 @ DC1 VM2 @ DC2
Throghput: iperf3 -t 600, 5 times
Qdisc Average (GBits/sec)
---------- -------------------
pfifo_fast 12.52 ± 0.26
bpf_fifo 11.72 ± 0.32
fq 10.24 ± 0.13
bpf_fq 11.92 ± 0.64
Latency: sockperf pp --tcp -t 600, 5 times
Qdisc Average (usec)
---------- --------------
pfifo_fast 244.58 ± 7.93
bpf_fifo 244.92 ± 15.22
fq 234.30 ± 19.25
bpf_fq 221.34 ± 10.76
Looking at the two fifo qdiscs, the 6.4% drop in throughput in the bpf
implementatioin is consistent with previous observation (v8 throughput
test on a loopback device). This should be able to be mitigated by
supporting adding skb to bpf_list or bpf_rbtree directly in the future.
* Clean up skb in bpf qdisc during reset *
The current implementation relies on bpf qdisc implementors to correctly
release skbs in queues (bpf graphs or maps) in .reset, which might not be
a safe thing to do. The solution as Martin has suggested would be
supporting private data in struct_ops. This can also help simplifying
implementation of qdisc that works with mq. For examples, qdiscs in the
selftest mostly use global data. Therefore, even if user add multiple
qdisc instances under mq, they would still share the same queue.
====================
Link: https://patch.msgid.link/20250409214606.2000194-1-ameryhung@gmail.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Until we are certain that existing classful qdiscs work with bpf qdisc,
make sure we don't allow attaching a bpf qdisc to non root. Meanwhile,
attaching to mq is allowed.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://patch.msgid.link/20250409214606.2000194-11-ameryhung@gmail.com
This test implements a more sophisticated qdisc using bpf. The bpf fair-
queueing (fq) qdisc gives each flow an equal chance to transmit data. It
also respects the timestamp of skb for rate limiting.
Signed-off-by: Amery Hung <amery.hung@bytedance.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://patch.msgid.link/20250409214606.2000194-10-ameryhung@gmail.com
This selftest includes a bare minimum fifo qdisc, which simply enqueues
sk_buffs into the back of a bpf list and dequeues from the front of the
list.
Signed-off-by: Amery Hung <amery.hung@bytedance.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://patch.msgid.link/20250409214606.2000194-9-ameryhung@gmail.com
Extend struct bpf_tc_hook with handle, qdisc name and a new attach type,
BPF_TC_QDISC, to allow users to add or remove any qdisc specified in
addition to clsact.
Signed-off-by: Amery Hung <amery.hung@bytedance.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://patch.msgid.link/20250409214606.2000194-8-ameryhung@gmail.com
Do not allow users to attach bpf qdiscs to classful qdiscs. This is to
prevent accidentally breaking existings classful qdiscs if they rely on
some data in the child qdisc. This restriction can potentially be lifted
in the future. Note that, we still allow bpf qdisc to be attached to mq.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://patch.msgid.link/20250409214606.2000194-7-ameryhung@gmail.com
Add a kfunc to update Qdisc bstats when an skb is dequeued. The kfunc is
only available in .dequeue programs.
Signed-off-by: Amery Hung <amery.hung@bytedance.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://patch.msgid.link/20250409214606.2000194-6-ameryhung@gmail.com
Add a watchdog timer to bpf qdisc. The watchdog can be used to schedule
the execution of qdisc through kfunc, bpf_qdisc_schedule(). It can be
useful for building traffic shaping scheduling algorithm, where the time
the next packet will be dequeued is known.
The implementation relies on struct_ops gen_prologue/epilogue to patch bpf
programs provided by users. Operator specific prologue/epilogue kfuncs
are introduced instead of watchdog kfuncs so that it is easier to extend
prologue/epilogue in the future (writing C vs BPF bytecode).
Signed-off-by: Amery Hung <amery.hung@bytedance.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://patch.msgid.link/20250409214606.2000194-5-ameryhung@gmail.com
Add basic kfuncs for working on skb in qdisc.
Both bpf_qdisc_skb_drop() and bpf_kfree_skb() can be used to release
a reference to an skb. However, bpf_qdisc_skb_drop() can only be called
in .enqueue where a to_free skb list is available from kernel to defer
the release. bpf_kfree_skb() should be used elsewhere. It is also used
in bpf_obj_free_fields() when cleaning up skb in maps and collections.
bpf_skb_get_hash() returns the flow hash of an skb, which can be used
to build flow-based queueing algorithms.
Finally, allow users to create read-only dynptr via bpf_dynptr_from_skb().
Signed-off-by: Amery Hung <amery.hung@bytedance.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://patch.msgid.link/20250409214606.2000194-4-ameryhung@gmail.com
The recent advancement in bpf such as allocated objects, bpf list and bpf
rbtree has provided powerful and flexible building blocks to realize
sophisticated packet scheduling algorithms. As struct_ops now supports
core operators in Qdisc_ops, start allowing qdisc to be implemented using
bpf struct_ops with this patch. Users can implement Qdisc_ops.{enqueue,
dequeue, init, reset, destroy} in bpf and register the qdisc dynamically
into the kernel.
Co-developed-by: Cong Wang <cong.wang@bytedance.com>
Signed-off-by: Cong Wang <cong.wang@bytedance.com>
Signed-off-by: Amery Hung <amery.hung@bytedance.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://patch.msgid.link/20250409214606.2000194-3-ameryhung@gmail.com
Rename get_ctx_arg_idx to bpf_ctx_arg_idx, and allow others to call it.
No functional change.
Signed-off-by: Amery Hung <ameryhung@gmail.com>
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://patch.msgid.link/20250409214606.2000194-2-ameryhung@gmail.com
Compared to previous generations, Granite Rapids defines the RRL control
bits {en_patspr, noover, en} in different positions, adds an extra RRL set
for the new mode of the first patrol-scrub read error, and extends the
number of CORRERRCNT registers from 4 to 8, encoding one counter per
CORRERRCNT register.
Add a Granite Rapids reg_rrl configuration table and adjust the code to
accommodate the differences mentioned above for RRL support.
Signed-off-by: Qiuxu Zhuo <qiuxu.zhuo@intel.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Tested-by: Feng Xu <feng.f.xu@intel.com>
Link: https://lore.kernel.org/r/20250417150724.1170168-8-qiuxu.zhuo@intel.com
The "paravirt environment" is no longer in the tree. Axe that part of the
comment. Also add a blurb to remind readers that "USER_PMDS" refer to
the PTI user *copy* of the page tables, not the user *portion*.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: https://lore.kernel.org/all/20250414173240.5B1AB322%40davehans-spike.ostc.intel.com
There are a few too many levels of abstraction here.
First, just expand the PREALLOCATED_PMDS macro in place to make it
clear that it is only conditional on PTI.
Second, MAX_PREALLOCATED_PMDS is only used in one spot for an
on-stack allocation. It has a *maximum* value of 4. Do not bother
with the macro MAX() magic. Just set it to 4.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: https://lore.kernel.org/all/20250414173238.6E3CDA56%40davehans-spike.ostc.intel.com
Each mm_struct has its own copy of the page tables. When core mm code
makes changes to a copy of the page tables those changes sometimes
need to be synchronized with other mms' copies of the page tables. But
when this synchronization actually needs to happen is highly
architecture and configuration specific.
In cases where kernel PMDs are shared across processes
(SHARED_KERNEL_PMD) the core mm does not itself need to do that
synchronization for kernel PMD changes. The x86 code communicates
this by clearing the PGTBL_PMD_MODIFIED bit cleared in those
configs to avoid expensive synchronization.
The kernel is moving toward never sharing kernel PMDs on 32-bit.
Prepare for that and make 32-bit PAE always set PGTBL_PMD_MODIFIED,
even if there is no modification to synchronize. This obviously adds
some synchronization overhead in cases where the kernel page tables
are being changed.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: https://lore.kernel.org/all/20250414173237.EC790E95%40davehans-spike.ostc.intel.com
Kernel PMDs can either be shared across processes or private to a
process. On 64-bit, they are always shared. 32-bit non-PAE hardware
does not have PMDs, but the kernel logically squishes them into the
PGD and treats them as private. Here are the four cases:
64-bit: Shared
32-bit: non-PAE: Private
32-bit: PAE+ PTI: Private
32-bit: PAE+noPTI: Shared
Note that 32-bit is all "Private" except for PAE+noPTI being an
oddball. The 32-bit+PAE+noPTI case will be made like the rest of
32-bit shortly.
But until that can be done, temporarily treat the 32-bit+PAE+noPTI
case as Private. This will do unnecessary walks across pgd_list and
unnecessary PTE setting but should be otherwise harmless.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: https://lore.kernel.org/all/20250414173235.F63F50D1%40davehans-spike.ostc.intel.com
A hardware PAE PGD is only 32 bytes. A PGD is PAGE_SIZE in the other
paging modes. But for reasons*, the kernel _sometimes_ allocates a
whole page even though it only ever uses 32 bytes.
Make PAE less weird. Just allocate a page like the other paging modes.
This was already being done for PTI (and Xen in the past) and nobody
screamed that loudly about it so it can't be that bad.
* The original reason for PAGE_SIZE allocations for the PAE PGDs was
Xen's need to detect page table writes. But 32-bit PTI forced it too
for reasons I'm unclear about.
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: https://lore.kernel.org/all/20250414173234.D34F0C3E%40davehans-spike.ostc.intel.com
Make the {valid bit, overwritten status, number} of RRL registers and the
{number, offsets, widths} of per-channel CORRERRCNT registers configurable.
Refactor show_retry_rd_err_log() to use the configurable fields of struct
reg_rrl, making the code more scalable and simpler.
No functional changes intended.
Signed-off-by: Qiuxu Zhuo <qiuxu.zhuo@intel.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Tested-by: Feng Xu <feng.f.xu@intel.com>
Link: https://lore.kernel.org/r/20250417150724.1170168-7-qiuxu.zhuo@intel.com
Android has mounted the v1 cpuset controller using filesystem type
"cpuset" (not "cgroup") since 2015 [1], and depends on the resulting
behavior where the controller name is not added as a prefix for cgroupfs
files. [2]
Later, a problem was discovered where cpu hotplug onlining did not
affect the cpuset/cpus files, which Android carried an out-of-tree patch
to address for a while. An attempt was made to upstream this patch, but
the recommendation was to use the "cpuset_v2_mode" mount option
instead. [3]
An effort was made to do so, but this fails with "cgroup: Unknown
parameter 'cpuset_v2_mode'" because commit e1cba4b85d ("cgroup: Add
mount flag to enable cpuset to use v2 behavior in v1 cgroup") did not
update the special cased cpuset_mount(), and only the cgroup (v1)
filesystem type was updated.
Add parameter parsing to the cpuset filesystem type so that
cpuset_v2_mode works like the cgroup filesystem type:
$ mkdir /dev/cpuset
$ mount -t cpuset -ocpuset_v2_mode none /dev/cpuset
$ mount|grep cpuset
none on /dev/cpuset type cgroup (rw,relatime,cpuset,noprefix,cpuset_v2_mode,release_agent=/sbin/cpuset_release_agent)
[1] https://cs.android.com/android/_/android/platform/system/core/+/b769c8d24fd7be96f8968aa4c80b669525b930d3
[2] https://cs.android.com/android/platform/superproject/main/+/main:system/core/libprocessgroup/setup/cgroup_map_write.cpp;drc=2dac5d89a0f024a2d0cc46a80ba4ee13472f1681;l=192
[3] https://lore.kernel.org/lkml/f795f8be-a184-408a-0b5a-553d26061385@redhat.com/T/
Fixes: e1cba4b85d ("cgroup: Add mount flag to enable cpuset to use v2 behavior in v1 cgroup")
Signed-off-by: T.J. Mercier <tjmercier@google.com>
Acked-by: Waiman Long <longman@redhat.com>
Reviewed-by: Kamalesh Babulal <kamalesh.babulal@oracle.com>
Acked-by: Michal Koutný <mkoutny@suse.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Refactor enable_retry_rd_err_log() using helper functions for both
DDR and HBM, making the RRL control bits configurable instead of
hard-coded. Additionally, explicitly define the four RRL modes for
better readability.
No functional changes intended.
Signed-off-by: Qiuxu Zhuo <qiuxu.zhuo@intel.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Tested-by: Feng Xu <feng.f.xu@intel.com>
Link: https://lore.kernel.org/r/20250417150724.1170168-6-qiuxu.zhuo@intel.com
As the number of RRL (retry_rd_err_log) registers per memory channel
increases, the positions of the RRL control bits and the widths of the
RRL registers vary across different CPU generations. Adding RRL support
for a new CPU requires handling these differences throughout the
RRL-related code.
Structure the offsets, widths, control bit positions, set numbers, modes,
etc., of the per-channel RRL registers and make them configurable to
facilitate easier RRL support for new CPUs.
No functional changes are intended.
Signed-off-by: Qiuxu Zhuo <qiuxu.zhuo@intel.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Tested-by: Feng Xu <feng.f.xu@intel.com>
Link: https://lore.kernel.org/r/20250417150724.1170168-5-qiuxu.zhuo@intel.com
The i10nm_edac driver uses the default modes (either patrol scrub read
or on-demand read) of the RRL register sets configured by the BIOS.
Explicitly set the modes during the loading of the i10nm_edac driver with
the module parameter retry_rd_err_log=2.
Signed-off-by: Qiuxu Zhuo <qiuxu.zhuo@intel.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Tested-by: Feng Xu <feng.f.xu@intel.com>
Link: https://lore.kernel.org/r/20250417150724.1170168-4-qiuxu.zhuo@intel.com
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRTLbB6QfY48x44uB6AXGG7T9hjvgUCaAEWhAAKCRCAXGG7T9hj
vne3AQCXbD8VtBJZEZ6h3sxRYLYpDHaJa5B0NTNUwgbVDPG/pgD/ad8c8iOomlWT
EllZmgwobkMwz0XZHDjsBfHIYjA4AgE=
=aDuc
-----END PGP SIGNATURE-----
Merge tag 'for-linus-6.15a-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip
Pull xen fix from Juergen Gross:
"Just a single fix for the Xen multicall driver avoiding a percpu
variable referencing initdata by its initializer"
* tag 'for-linus-6.15a-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/xen/tip:
xen: fix multicall debug feature
When enabling the retry_rd_err_log (RRL) feature during the loading of the
i10nm_edac driver with the module parameter retry_rd_err_log=2 (Linux RRL
control mode), the default values of the control bits of RRL are saved so
that they can be restored during the unloading of the driver.
In the current code, the RRL of pseudo channel 1 of HBM overwrites pseudo
channel 0 during the loading of the driver, resulting in the loss of saved
RRL for pseudo channel 0. This causes the RRL of pseudo channel 0 of HBM to
be wrongly restored with the values from pseudo channel 1 when unloading
the driver.
Fix this issue by creating two separate groups of RRL control registers
per channel to save default RRL settings of two {sub-,pseudo-}channels.
Fixes: acd4cf68fe ("EDAC/i10nm: Retrieve and print retry_rd_err_log registers for HBM")
Signed-off-by: Qiuxu Zhuo <qiuxu.zhuo@intel.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
Tested-by: Feng Xu <feng.f.xu@intel.com>
Link: https://lore.kernel.org/r/20250417150724.1170168-3-qiuxu.zhuo@intel.com
Three small changes from further build testing
- Don't rely on the userspace uuid.h for the uapi header
- Fix sparse warnings in pds
- Typo in log message
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRRRCHOFoQz/8F5bUaFwuHvBreFYQUCaADznQAKCRCFwuHvBreF
YaqUAP9JEF8mbmT1WZhYfxq/IQAEIdIZ2uCEoZF2JoBa0Llh8wD/Zuyw3rdQDd/Y
j5Yhvj/rsbAA10ALpFNl8e/L38FN0AI=
=TJeD
-----END PGP SIGNATURE-----
Merge tag 'for-linus-fwctl' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma
Pull fwctl fixes from Jason Gunthorpe:
"Three small changes from further build testing:
- Don't rely on the userspace uuid.h for the uapi header
- Fix sparse warnings in pds
- Typo in log message"
* tag 'for-linus-fwctl' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
fwctl: Fix repeated device word in log message
pds_fwctl: Fix type and endian complaints
fwctl/cxl: Fix uuid_t usage in uapi
