RISC-V Linux 内核及周边技术动态第 85 期

时间：20240331
编辑：晓怡
仓库：RISC-V Linux 内核技术调研活动
赞助：PLCT Lab, ISCAS

内核动态

RISC-V 架构支持

v1: bpf-next: Add 12-argument support for RV64 bpf trampoline

This patch adds 12 function arguments support for riscv64 bpf trampoline. The current bpf trampoline supports <= sizeof(u64) bytes scalar arguments [0] and <= 16 bytes struct arguments [1].

v1: iio: adc: add ADC driver for XuanTie TH1520 SoC

This patchset adds initial support for XuanTie TH1520 ADC driver. This is modified from XuanTie TH1520 Linux_SDK_V1.4.2(linux v5.10.113) The original author is Fugang Duan duanfugang.dfg@linux.alibaba.com

v3: clk: starfive: jh7100: Use clk_hw for external input clocks

The Starfive JH7100 clock driver does not use the DT “clocks” property to find the external main input clock, but instead relies on the name of the actual clock provider (“osc_sys”). This is fragile, and caused breakage when sanitizing clock node names in DTS.

v4: Unified cross-architecture kernel-mode FPU API

This series unifies the kernel-mode FPU API across several architectures by wrapping the existing functions (where needed) in consistently-named functions placed in a consistent header location, with mostly the same semantics: they can be called from preemptible or non-preemptible task context,

v13: riscv: sophgo: add clock support for sg2042

This series adds clock controller support for sophgo sg2042.

v2: riscv control-flow integrity for usermode

I had sent RFC patchset early this year (January) [7] to enable CPU assisted control-flow integrity for usermode on riscv. Since then I’ve been able to do more testing of the changes. As part of testing effort, compiled a rootfs with shadow stack and landing pad enabled (libraries and binaries) and booted to shell.

v6: riscv: sophgo: add dmamux support for Sophgo CV1800/SG2000 SoCs

Add dma multiplexer support for the Sophgo CV1800/SG2000 SoCs.

As the syscon device of CV1800 have a usb phy subdevices. The binding of the syscon can not be complete without the usb phy is finished. As a result, the binding of syscon is removed and will be evolved in its original series after the usb phy binding is fully explored.

v1: riscv: ftrace: make stack walk more robust.

The current stack walker in riscv implemented in walk_stackframe() provides the PC to a callback function when it unwinds the stacks. This doesn’t allow implementing stack walkers that need access to more information like the frame pointer, etc.

v1: riscv: Kconfig.socs: Deprecate SOC_CANAAN and use SOC_CANAAN_K210 for K210

Since SOC_FOO should be deprecated from patch [1], and cleanup for other SoCs is already in the mailing list [2,3,4], so we deprecate the use of SOC_CANAAN and use ARCH_CANAAN for SoCs vendored by Canaan instead from now on.

v1: ftrace: riscv: move from REGS to ARGS

This commit replaces riscv’s support for FTRACE_WITH_REGS with support for FTRACE_WITH_ARGS. This is required for the ongoing effort to stop relying on stop_machine() for RISCV’s implementation of ftrace.

v16: Refactoring Microchip PCIe driver and add StarFive PCIe

This patchset final purpose is add PCIe driver for StarFive JH7110 SoC. JH7110 using PLDA XpressRICH PCIe IP. Microchip PolarFire Using the same IP and have commit their codes, which are mixed with PLDA controller codes and Microchip platform codes.

v2: riscv: Call secondary mmu notifier when flushing the tlb

This is required to allow the IOMMU driver to correctly flush its own TLB.

回复: v9: Add timer driver for StarFive JH7110 RISC-V SoC

This patch serises are to add timer driver for the StarFive JH7110 RISC-V SoC. The first patch adds documentation to describe device tree bindings. The subsequent patch adds timer driver and support

v2: riscv: Various text patching improvements

Here are a few changes to minimize calls to stop_machine() and flush_icache_*() in the various text patching functions, as well as to simplify the code.

v1: Convert Tasklets to BH Workqueues

This patch series represents a significant shift in how asynchronous execution in the bottom half (BH) context is handled within the kernel. Traditionally, tasklets have been the go-to mechanism for such operations.

This series is a simple improvement for timer-clint and timer-riscv:

Add set_state_shutdown for timer-clint, this hook is used when switching clockevent from timer-clint to another timer.

v2: riscv: access_ok() optimization

This series optimizes access_ok() by defining TASK_SIZE_MAX. At Alex’s suggestion, I also tried making TASK_SIZE constant (specifically by making PGDIR_SHIFT a variable instead of a ternary expression, then replacing the load with an immediate using ALTERNATIVE).

v1: BeagleV Fire support

Wee series adding support for the BeagleV Fire. I’ve had a dts sitting locally for this for over a year for testing Auto Update and I meant to submit something to mainline once the board got announced publicly, but only got around to that now.

FAILED: Patch “clocksource/drivers/timer-riscv: Clear timer interrupt on timer initialization” failed to apply to 6.1-stable tree

The patch below does not apply to the 6.1-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to stable@vger.kernel.org.

v1: RISC-V: Test th.mxstatus.MAEE bit before enabling MAEE

Currently, the Linux kernel suffers from a boot regression when running on the c906 QEMU emulation. Details have been reported here by Björn Töpel:https://lists.gnu.org/archive/html/qemu-devel/2024-01/msg04766.html

v12: riscv: sophgo: add clock support for sg2042

This series adds clock controller support for sophgo sg2042.

v1: cache: sifive_ccache: Partially convert to a platform driver

Commit 8ec99b033147 (“irqchip/sifive-plic: Convert PLIC driver into a platform driver”) broke ccache initialization because the PLIC IRQ

进程调度

v1: sched/fair: Reset vlag in dequeue when PLAGE_LAG is disabled

The vlag is calculated in dequeue when PLAGE_LAG is disabled. If we enable the PLACE_LAG at some point, the old vlag of process will affect itself and other process. These are not in line with our original intention, where we expect the vlag of all processes to be calculated from 0 after the enable PLAGE_LAG.

v1: perf sched: Rename switches to count and add usage description, options for latency

Rename ‘Switches’ to ‘Count’ and document metrics shown for perf sched latency output. Also add options possible with perf sched latency.

v2: RESEND: sched/fair: simplify __calc_delta()

Commit 5e963f2bd4654a202a8a05aa3a86cb0300b10e6c (“sched/fair: Commit to EEVDF”) removed __calc_delta()’s use case where the input weight is not equal to NICE_0_LOAD. Now that weight is always NICE_0_LOAD, it is not required to have it as an input parameter. NICE_0_LOAD could be incorporated in __calc_delta() directly.

v1: sched/fair: allow disabling newidle_balance with sched_relax_domain_level

During the upgrade from Linux 5.4 we found a small (around 3%) performance regression which was tracked to commit c5b0a7eefc70150caf23e37bc9d639c68c87a097

v1: trace/sched: add tgid for sched_wakeup_template

By doing this, we are able to filter tasks by tgid while we are tracing wakeup events by ebpf or other methods.

v1: sched/fair: Combine EAS check with overutilized access

So modify the helper function to return this pattern. This is more readable code as it would say, do something when root domain is not overutilized. This function always return true when EAS is disabled.

v1: sched/fair: Simplify continue_balancing for newidle

newidle(CPU_NEWLY_IDLE) balancing doesn’t stop the load balancing if the continue_balancing flag is reset. Other two balancing (IDLE, BUSY) do that. newidle balance stops the load balancing if rq has a task or there is wakeup pending. The same checks are present in should_we_balance for newidle. Hence use the return value and simplify continue_balancing mechanism for newidle. Update the comment surrounding it as well.

v1: sched/eevdf: Curb wakeup preemption further

Bisection showed that the regression started at commit 147f3efaa241 (“sched/fair: Implement an EEVDF-like scheduling policy”) and this was reported [2]. Further narrowing down than the commit is hard due to the extent of changes in the commit. EEVDF has seen extensive development since, but the regression persists.

[GIT pull] sched/urgent for v6.9-rc1

A single update for the documentation of the base_slice_ns tunable to clarify that any value which is less than the tick slice has no effect because the scheduler tick is not guaranteed to happen within the set time slice.

内存管理

v1: SLUB: improve filling cpu partial a bit in get_partial_node()

This series is to remove the unnecessary check for filling cpu partial and improve the readability.

Introduce slub_get_cpu_partial() and dummy function to prevent compiler warning with CONFIG_SLUB_CPU_PARTIAL disabled. This is done in patch 2. Use the helper in patch 3 and 4.

Patch “mm/migrate: set swap entry values of THP tail pages properly.” has been added to the 6.1-stable tree

The tail pages in a THP can have swap entry information stored in their private field. When migrating to a new page, all tail pages of the new page need to update ->private to avoid future data corruption.

Patch “mm/migrate: set swap entry values of THP tail pages properly.” has been added to the 5.15-stable tree

This is a note to let you know that I’ve just added the patch titled

mm/migrate: set swap entry values of THP tail pages properly.

to the 5.15-stable tree which can be found at:http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

v2: mm/slub: Reduce memory consumption in extreme scenarios

When kmalloc_node() is called without __GFP_THISNODE and the target node lacks sufficient memory, SLUB allocates a folio from a different node other than the requested node, instead of taking a partial slab from it.

v12: Add AMD Secure Nested Paging (SEV-SNP) Hypervisor Support

This patchset is also available at:

https://github.com/amdese/linux/commits/snp-host-v12

and is based on top of the following series:

v2: selftests/mm: include strings.h for ffsl

Got a compilation error on Android for ffsl after 91b80cc5b39f (“selftests: mm: fix map_hugetlb failure on 64K page size systems”) included vm_util.h.

v1: mm: alloc_anon_folio: avoid doing vma_thp_gfp_mask in fallback cases

Fallback rates surpassing 90% have been observed on phones utilizing 64KiB CONT-PTE mTHP. In these scenarios, when one out of every 16 PTEs fails to allocate large folios, the remaining 15 PTEs fallback. Consequently, invoking vma_thp_gfp_mask seems redundant in such cases. Furthermore, abstaining from its use can also contribute to improved code readability.

v1: mm: huge_memory: add the missing folio_test_pmd_mappable() for THP split statistics

Now the mTHP can also be split or added into the deferred list, so add folio_test_pmd_mappable() validation for PMD mapped THP, to avoid confusion with PMD mapped THP related statistics.

v2: support multi-size THP numa balancing

This patchset tries to support mTHP numa balancing, as a simple solution to start, the NUMA balancing algorithm for mTHP will follow the THP strategy as the basic support. Please find details in each patch.

v9: Improved Memory Tier Creation for CPUless NUMA Nodes

When a memory device, such as CXL1.1 type3 memory, is emulated as normal memory (E820_TYPE_RAM), the memory device is indistinguishable from normal DRAM in terms of memory tiering with the current implementation. The current memory tiering assigns all detected normal memory nodes to the same DRAM tier. This results in normal memory devices with

v1: mempool: Modify mismatched function name

No functional modification involved.

mm/mempool.c:245: warning: expecting prototype for mempool_init(). Prototype was for mempool_init_noprof() instead. mm/mempool.c:271: warning: expecting prototype for mempool_create_node(). Prototype was for mempool_create_node_noprof() instead.

v6: netfs, cifs: Delegate high-level I/O to netfslib

Here are patches to convert cifs to use the netfslib library. I’ve tested them with and without a cache. Unfortunately, if “-o fsc” is specified a leak of a tcon object shows up, particularly with the generic/013 xfstest that prevents further testing. I’ve investigated this and found that the tcon leak is actually present upstream, but just goes unnoticed unless it also pins an fscache volume cookie.

v1: netfs, afs, 9p, cifs: Rework netfs to use ->writepages() to copy to cache

The primary purpose of these patches is to rework the netfslib writeback implementation such that pages read from the cache are written to the cache through ->writepages(), thereby allowing the fscache page flag to be retired.

v2: mm: add per-order mTHP alloc_success and alloc_fail counters

Profiling a system blindly with mTHP has become challenging due to the lack of visibility into its operations. Presenting the success rate of mTHP allocations appears to be pressing need.

v11: Support page table check PowerPC

Support page table check on all PowerPC platforms. This works by serialising assignments, reassignments and clears of page table entries at each level in order to ensure that anonymous mappings have at most one writable consumer, and likewise that file-backed mappings are not simultaneously also anonymous mappings.

v1: mTHP-friendly compression in zsmalloc and zram based on multi-pages

mTHP is generally considered to potentially waste memory due to fragmentation, but it may also serve as a source of memory savings. When large folios are compressed at a larger granularity, we observe a remarkable decrease in CPU utilization and a significant improvement in compression ratios.

v3: mm: workingset reporting

This patch series provides workingset reporting of user pages in lruvecs, of which coldness can be tracked by accessed bits and fd references. However, the concept of workingset applies generically to all types of memory, which could be kernel slab caches, discardable userspace caches (databases), or CXL.mem. Therefore, data sources might come from slab shrinkers, device drivers, or the userspace.

v1: mm: Use rwsem assertion macros for mmap_lock

This slightly strengthens our write assertion when lockdep is disabled. It also downgrades us from BUG_ON to WARN_ON, but I think that’s an improvement. I don’t think dumping the mm_struct was all that valuable; the call chain is what’s important.

v2: mm, netfs: Provide a means of invalidation without using launder_folio

mm, netfs: Provide a means of invalidation without using launder_folio

Implement a replacement for launder_folio. The key feature of invalidate_inode_pages2() is that it locks each folio individually, unmaps it to prevent mmap’d accesses interfering and calls the ->launder_folio() address_space op to flush it. This has problems: firstly, each folio is written individually as one or more small writes; secondly, adjacent folios cannot be added so easily into the laundry; thirdly, it’s yet another op to implement.

文件系统

v1: ext4: support adding multi-delalloc blocks

This patch series is the part 2 prepartory changes of the buffered IO iomap conversion, I picked them out from my buffered IO iomap conversion RFC series v3[1], and add bigalloc feature support.

v4: Fuse-BPF and plans on merging with Fuse Passthrough

I’ve recently gotten some time to re-focus on fuse-bpf efforts, and had some questions on how to best integrate with recent changes that have landed in the last year. I’ve included a rebased version (ontop of bpf-next e63985ecd226 (“bpf, riscv64/cfi: Support kCFI + BPF on riscv64”) of the old patchset for reference here.

v1: fuse: Add initial support for fs-verity

Note this doesn’t include support for FS_IOC_READ_VERITY_METADATA. I don’t know of an existing use of it and I figured it would be better not to include code that wouldn’t be used and tested. However if you feel like it should be added let me know.

网络设备

v1: net-next: ipvlan: handle NETDEV_DOWN event

In case of stacked devices, to help propagate the down link state from the parent/root device (to this leaf device), handle NETDEV_DOWN event like it is done now for NETDEV_UP.

v1: net-next: tcp/dccp: complete lockless accesses to sk->sk_max_ack_backlog

Since commit 099ecf59f05b (“net: annotate lockless accesses to sk->sk_max_ack_backlog”) decided to handle the sk_max_ack_backlog locklessly, there is one more function mostly called in TCP/DCCP cases. So this patch completes it:)

v2: net-next: Avoid explicit cpumask var allocation on stack

This is v2 of previous series[1] about cpumask var on stack for net subsystem.

v2: net-next: net: Add generic support for netdev LEDs

For some devices, the MAC controls the LEDs in the RJ45 connector, not the PHY. This patchset provides generic support for such LEDs, and adds the first user, mv88e6xxx.

v1: net-next: batman-adv: bypass empty buckets in batadv_purge_orig_ref()

Many syzbot reports are pointing to soft lockups in batadv_purge_orig_ref() [1]

Root cause is unknown, but we can avoid spending too much time there and perhaps get more interesting reports.

v2: net-next: caif: Use UTILITY_NAME_LENGTH instead of hard-coding 16

UTILITY_NAME_LENGTH is 16. So better use the former when defining the ‘utility_name’ array. This makes the intent clearer when it is used around line 260.

v1: bpf-next: selftests/bpf: Add sockaddr tests for kernel networking

In a follow up to these patches,

• commit 0bdf399342c5(“net: Avoid address overwrite in kernel_connect”)
• commit 86a7e0b69bd5(“net: prevent rewrite of msg_name in sock_sendmsg()”)
• commit c889a99a21bf(“net: prevent address rewrite in kernel_bind()”)
• commit 01b2885d9415(“net: Save and restore msg_namelen in sock_sendmsg”)

v1: net: mptcp: prevent BPF accessing lowat from a subflow socket.

The root cause of the issue is that bpf allows accessing mptcp-level proto_ops from a tcp subflow scope.

Fix the issue detecting the problematic call and preventing any action.

Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/482

v1: net-next: tools: ynl: add ynl_dump_empty() helper

Checking if dump is empty requires a couple of casts. Add a convenient wrapper.

Add an example use in the netdev sample, loopback is always present so an empty dump is an error.

v3: net: virtio_net: Do not send RSS key if it is not supported

There is a bug when setting the RSS options in virtio_net that can break the whole machine, getting the kernel into an infinite loop.

v1: net-next: page_pool: allow direct bulk recycling

Previously, there was no reliable way to check whether it’s safe to use direct PP cache. The drivers were passing @allow_direct to the PP recycling functions and that was it. Bulk recycling is used by xdp_return_frame_bulk() on .ndo_xdp_xmit() frames completion where the page origin is unknown, thus the direct recycling has never been tried.

v1: rhashtable: Improve grammar

Change “a” to “an” according to the usual rules, fix an “if” that was mistyped as “in”, improve grammar in “considerable slow” -> “considerably slower”.

v1: net: selftests: reuseaddr_conflict: add missing new line at the end of the output

The netdev CI runs in a VM and captures serial, so stdout and stderr get combined. Because there’s a missing new line in stderr the test ends up corrupting KTAP:

v1: net-next: tcp/dccp: do not care about families in inet_twsk_purge()

We lost ability to unload ipv6 module a long time ago.

Instead of calling expensive inet_twsk_purge() twice, we can handle all families in one round.

v1: net-next: inet: preserve const qualifier in inet_csk()

We can change inet_csk() to propagate its argument const qualifier, thanks to container_of_const().

v2: Documentation: networking: document ISO 15765-2:2016

While the in-kernel ISO 15765-2:2016 (ISO-TP) stack is fully functional and easy to use, no documentation exists for it.

v1: 6.1: octeontx2-af: Add validation of lmac

With the addition of new MAC blocks like CN10K RPM and CN10KB RPM_USX, LMACs are noncontiguous. Though in most of the functions, lmac validation checks exist but in few functions they are missing. The problem has been fixed by the following patch which can be cleanly applied to the 6.1.y branch.

v3: iwl-next: Introduce ETH56G PHY model for E825C products

E825C products have a different PHY model than E822, E823 and E810 products. This PHY is ETH56G and its support is necessary to have functional PTP stack for E825C products.

v2: 6.1.y: net: tls: handle backlogging of crypto requests

commit 8590541473188741055d27b955db0777569438e3 upstream

Since we’re setting the CRYPTO_TFM_REQ_MAY_BACKLOG flag on our requests to the crypto API, crypto_aead_{encrypt,decrypt} can return-EBUSY instead of -EINPROGRESS in valid situations.

v2: Documentation: Add reconnect process for VDUSE

Add a document explaining the reconnect process, including what the Userspace App needs to do and how it works with the kernel.

v1: net-next: ethtool: Max power support

Some ethernet modules use nonstandard power levels [1]. Extend ethtool module implementation to support new attributes that will allow user to change maximum power. Rename structures and functions to be more generic. Introduce an example of the new API in ice driver.

v1: wifi: mac80211: correctly document struct mesh_table

Currently kernel-doc -Wall reports:

net/mac80211/ieee80211_i.h:687: warning: missing initial short description on line:

• struct mesh_table

安全增强

v2: drm/radeon/radeon_display: Decrease the size of allocated memory

This is an effort to get rid of all multiplications from allocation functions in order to prevent integer overflows [1] [2].

v3: scsi: csiostor: Use kcalloc() instead of kzalloc()

Use 2-factor multiplication argument form kcalloc() instead of kzalloc().

v2: perf/x86/amd/uncore: Use kcalloc() instead of kzalloc()

As noted in the “Deprecated Interfaces, Language Features, Attributes, and Conventions” documentation [1], size calculations (especially multiplication) should not be performed in memory allocator (or similar) function arguments due to the risk of them overflowing. This could lead to values wrapping around and a smaller allocation being made than the caller was expecting. Using those allocations could lead to linear overflows of heap memory and other misbehaviors.

v2: dmaengine: pl08x: Use kcalloc() instead of kzalloc()

This is an effort to get rid of all multiplications from allocation functions in order to prevent integer overflows [1].

v1: perf/x86/intel/uncore: Prefer struct_size over open coded arithmetic

This is an effort to get rid of all multiplications from allocation functions in order to prevent integer overflows [1][2].

v1: nouveau/gsp: Avoid addressing beyond end of rpc->entries

Using the end of rpc->entries[] for addressing runs into both compile-time and run-time detection of accessing beyond the end of the array. Use the base pointer instead, since was allocated with the additional bytes for storing the strings. Avoids the following warning in future GCC releases with support for __counted_by:

v1: Squashfs: replace deprecated strncpy with strscpy

strncpy() is deprecated for use on NUL-terminated destination strings [1] and as such we should prefer more robust and less ambiguous string interfaces.

v1: Add sy7802 flash led driver

This series introduces a driver for the Silergy SY7802 charge pump used in the BQ Aquaris M5 and X5 smartphones.

v1: vmcore: replace strncpy with strtomem

strncpy() is in the process of being replaced as it is deprecated in some situations [1]. While the specific use of strncpy that this patch targets is not exactly deprecated, the real mission is to rid the kernel of all its uses.

v2: net-next: compiler_types: add Endianness-dependent __counted_by_{le,be}

Some structures contain flexible arrays at the end and the counter for them, but the counter has explicit Endianness and thus __counted_by() can’t be used directly.

v1: next: wifi: wil6210: Annotate struct wmi_set_link_monitor_cmd with __counted_by()

Prepare for the coming implementation by GCC and Clang of the __counted_by attribute. Flexible array members annotated with __counted_by can have their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family functions).

v4: Handle faults in KUnit tests

This patch series teaches KUnit to handle kthread faults as errors, and it brings a few related fixes and improvements.

v1: next: firewire: Annotate struct fw_iso_packet with __counted_by()

Prepare for the coming implementation by GCC and Clang of the __counted_by attribute. Flexible array members annotated with __counted_by can have their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family functions).

v1: next: fs: Annotate struct file_handle with __counted_by() and use struct_size()

Prepare for the coming implementation by GCC and Clang of the __counted_by attribute. Flexible array members annotated with __counted_by can have their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family functions).

v4: Add support for QoS configuration

This series adds QoS support for QNOC type device which can be found on SC7280 platform. It adds support for programming priority, priority forward disable and urgency forwarding. This helps in priortizing the traffic originating from different interconnect masters at NOC(Network On Chip).

异步 IO

v1: liburing: io_uring.h: Sync kernel header to fetch enum names

After a report by Ritesh (YoSTEALTH) on github, we named the enums in the io_uring uapi header. Sync the change into liburing.

v1: io_uring: return void from io_put_kbuf_comp()

The two callers don’t handle the return value of io_put_kbuf_comp(), so change its return type into void.

v1: io_uring: kill dead code in io_req_complete_post

Since commit 8f6c829491fe (“io_uring: remove struct io_tw_state::locked”), io_req_complete_post() is only called from io-wq submit work, where the request reference is guaranteed to be grabbed and won’t drop to zero in io_req_complete_post().

v2: fs: claw back a few FMODE_* bits

There’s a bunch of flags that are purely based on what the file operations support while also never being conditionally set or unset. IOW, they’re not subject to change for individual files. Imho, such flags don’t need to live in f_mode they might as well live in the fops structs itself.

v1: liburing: io_uring.h: Avoid anonymous enums

anonymous enums, while valid, confuses Cython (Python to C translator), as reported by Ritesh (YoSTEALTH) . Since people are using this, just name the existing enums.

v1: : fs: claw back a few FMODE_* bits

There’s a bunch of flags that are purely based on what the file operations support while also never being conditionally set or unset. IOW, they’re not subject to change for individual file opens. Imho, such flags don’t need to live in f_mode they might as well live in the fops structs itself.

v1: io_uring: refill request cache in memory order

The allocator will generally return memory in order, but __io_alloc_req_refill() then adds them to a stack and we’ll extract them in the opposite order. This obviously isn’t a huge deal

BPF

v4: perf/x86/amd: add LBR capture support outside of hardware events

Add AMD-specific implementation of perf_snapshot_branch_stack static call that allows LBR capture from arbitrary points in the kernel. This is utilized by BPF programs. See patch #3 for all the details.

v5: bpf-next: bpf: Add a generic bits iterator

Three new kfuncs, namely bpf_iter_bits_{new,next,destroy}, have been added for the new bpf_iter_bits functionality. These kfuncs enable the iteration of the bits from a given address and a given number of bits.

v2: bpf-next: selftests/bpf: make multi-uprobe tests work in RELEASE=1 mode

When BPF selftests are built in RELEASE=1 mode with -O2 optimization level, uprobe_multi binary, called from multi-uprobe tests is optimized to the point that all the thousands of target uprobe_multi_func_XXX functions are eliminated, breaking tests.

v1: bpf-next: Add internal-only BPF per-CPU instructions

Add two new BPF instructions for dealing with per-CPU memory.

One, BPF_LDX | BPF_ADDR_PERCPU | BPF_DW (where BPF_ADD_PERCPU is unused 0xe0 opcode), resolved provided per-CPU address (offset) to an absolute address where per-CPU data resides for “this” CPU. This is the most universal, and, strictly speaking, the only per-CPU BPF instruction necessary.

v1: bpf-next: bpf: Avoid kfree_rcu() under lock in bpf_lpm_trie.

bpf_lpm lock can be the same. timer_base lock can also be the same due to timer migration. but rcu krcp lock is always per-cpu, so it cannot be the same lock. Hence it’s a false positive. To avoid lockdep complain move kfree_rcu() after spin_unlock.

v1: kbuild: Avoid weak external linkage where possible

Weak external linkage is intended for cases where a symbol reference can remain unsatisfied in the final link. Taking the address of such a symbol should yield NULL if the reference was not satisfied.

v1: bpf-next: bpf: add a verbose message if map limit is reached

When more than 64 maps are used by a program the verifier return -E2BIG. Add a verbose message which highlights the error and also prints the actual limit.

v1: bpf-next: net: netfilter: Make ct zone id configurable for bpf ct helper functions

Add ct zone id to bpf_ct_opts so that arbitrary ct zone can be set for xdp/tc bpf ct helper functions bpf_{xdp,skb}ct_alloc and bpf{xdp,skb}_ct_lookup.

v1: bpf-next: bpf: Mark bpf prog stack with kmsan_unposion_memory in interpreter mode

syzbot reported uninit memory usages during map_{lookup,delete}_elem.

This should address different syzbot reports on the uninit “void *key” argument during map_{lookup,delete}_elem.

v1: First try to replace page_frag with page_frag_cache

This patchset tries to unfiy the page frag implementation by replacing page_frag with page_frag_cache for sk_page_frag() first. And will try to replace the rest of page_frag in the follow patchset.

v2: perf/x86/amd: support capturing LBR from software events

[0] added ability to capture LBR (Last Branch Records) on Intel CPUs from inside BPF program at pretty much any arbitrary point. This is extremely useful capability that allows to figure out otherwise hard-to-debug problems, because LBR is now available based on some application-defined conditions, not just hardware-supported events.

v1: bpf-next: selftests/bpf: Handle EAGAIN in bpf_tcp_ca

bpf_tcp_ca tests may emit EAGAIN sometimes. In that case, tests fail with “bytes != total_bytes” errors. Sending should continue, not break when errno is EAGAIN. This patch can make bpf_tcp_ca tests stable.

v5: net-next: Add minimal XDP support to TI AM65 CPSW Ethernet driver

This patch adds XDP support to TI AM65 CPSW Ethernet driver.

v1: bpf-next: bpf: freeze a task cgroup from bpf

This patch series adds support to freeze the task cgroup hierarchy that is on a default cgroup v2 without going through kernfs interface.

For some cases we want to freeze the cgroup of a task based on some signals, doing so from bpf is better than user space which could be too late.

v1: bpf-next: uprobe: uretprobe speed up

hi, as part of the effort on speeding up the uprobes [0] coming with return uprobe optimization by using syscall instead of the trap on the uretprobe trampoline.

v1: bpf,arena: Use helper sizeof_field in struct accessors

Use the well defined helper sizeof_field() to calculate the size of a struct member, instead of doing custom calculations.

v7: net-next: Device Memory TCP

This revision largely rebases on top of net-next and addresses the feedback RFCv6 received from folks, namely Jakub, Yunsheng, Arnd, David, & Pavel.

v1: bpf-next: bpf: support deferring bpf_link dealloc to after RCU grace period

BPF link for some program types is passed as a “context” which can be used by those BPF programs to look up additional information. E.g., for BPF raw tracepoints, link is used to fetch BPF cookie value, similarly for BPF multi-kprobes and multi-uprobes.

v2: bpf-next: bench: fast in-kernel triggering benchmarks

Remove “legacy” triggering benchmarks which rely on syscalls (and thus syscall overhead is a noticeable part of benchmark, unfortunately). Replace them with faster versions that rely on triggering BPF programs in-kernel through another simple “driver” BPF program. See patch #2 with comparison results.

v2: bpf-next: selftests/bpf: Enable cross platform testing for local vmtest

The variable $ARCH in the current script is platform semantics, not kernel semantics. Rename it to $PLATFORM so that we can easily use $ARCH in cross-compilation. For now, Using PLATFORM= and CROSS_COMPILE= options will enable cross platform testing:

v2: bpf-next: BPF: support mark in bpf_fib_lookup

This patch series adds policy routing support in bpf_fib_lookup. This is a useful functionality which was missing for a long time, as without it some networking setups can’t be implemented in BPF. One example can be found here [1].

v1: bpf-next: bpf: Mitigate latency spikes associated with freeing non-preallocated htab

Following the recent upgrade of one of our BPF programs, we encountered significant latency spikes affecting other applications running on the same host. After thorough investigation, we identified that these spikes were primarily caused by the prolonged duration required to free a non-preallocated htab with approximately 2 million keys.

v2: leds: trigger: legtrig-bpf: Add ledtrig-bpf trigger

This patch set adds a led trigger that interfaces with the bpf subsystem. It allows for BPF programs to control LED activity using bpf kfuncs. This functionality is useful in giving users a physical indication that a BPF program has performed an operation such as handling a packet or probe point.

周边技术动态

Qemu

v2: riscv: thead: Add th.sxstatus CSR emulation

The th.sxstatus CSR can be used to identify available custom extension on T-Head CPUs. The CSR is documented here:https://github.com/T-head-Semi/thead-extension-spec/pull/46

v3: target/riscv: Support Zve32x and Zve64x extensions

This patch series adds the support for Zve32x and Zvx64x and makes vector registers visible in GDB if any of the V/Zve/Zvk extensions is enabled.

v3: target/riscv/kvm/kvm-cpu.c: kvm_riscv_handle_sbi() fail with vendor-specific SBI

kvm_riscv_handle_sbi() may return not supported return code to not trigger qemu abort with vendor-specific sbi.

v1: riscv: thead: Add th.mxstatus CSR emulation

The th.mxstatus CSR can be used to identify available custom extension on T-Head CPUs. The CSR is documented here:https://github.com/T-head-Semi/thead-extension-spec/pull/45

Buildroot

v1: package/uclibc: bump to 1.0.47

Fixes riscv port. NPTL/TLS fixed. C++ applications now working. Added explicit_bzero and reallocarray.

U-Boot

v2: riscv: Move virtio scan to board_late_init()

When virtio_init() gets called from board_init() PCI isn’t ready. Thus, virtio-over-PCI (e.g. network interfaces) devices can’t be detected and used without additional virtio scan scan in the shell or a script.

v1: riscv: adds T-Head C9xx basic and GMAC support.

This patchset adds T-Head C9xx basic support in arch/riscv/, updates TH1520 Soc/Lichee Pi4A dts files for GMAC support. Also enable designware ethernet & realtek phy in default configs, and some boot env option for booting linux from Ethernet.

mx6cuboxi: failes to detect model

I am seeing model detection problems with the current git master.

U-Boot 2024.04-rc5 (Mar 26 2024 - 15:59:22 +0100)

GIT PULL: u-boot-riscv/master

The following changes since commit dde373bde392c38649c8c4420e0c98ef8d38d9dc:

Prepare v2024.04-rc5 (2024-03-25 21:56:50 -0400)

are available in the Git repository at:

https://source.denx.de/u-boot/custodians/u-boot-riscv.git

回复: v2: riscv: add support for Milk-V Mars board

With this patch series the VisionFive 2 U-Boot SPL will detect that it is running on a Milk-V board and patch the device-tree accordingly. This is the same approach that has been taken to handle the differences between the Visionfive 2 1.2B and 1.3A revisions.