aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2021-05-04powerpc/powernv/memtrace: Fix dcache flushingSandipan Das1-2/+2
Trace memory is cleared and the corresponding dcache lines are flushed after allocation. However, this should not be done using the PFN. This adds the missing conversion to virtual address. Fixes: 2ac02e5ecec0 ("powerpc/mm: Remove dcache flush from memory remove.") Signed-off-by: Sandipan Das <[email protected]> Reviewed-by: Aneesh Kumar K.V <[email protected]> Signed-off-by: Michael Ellerman <[email protected]> Link: https://lore.kernel.org/r/[email protected]
2021-05-04powerpc/kexec_file: Use current CPU info while setting up FDTSourabh Jain1-0/+92
kexec_file_load() uses initial_boot_params in setting up the device tree for the kernel to be loaded. Though initial_boot_params holds info about CPUs at the time of boot, it doesn't account for hot added CPUs. So, kexec'ing with kexec_file_load() syscall leaves the kexec'ed kernel with inaccurate CPU info. If kdump kernel is loaded with kexec_file_load() syscall and the system crashes on a hot added CPU, the capture kernel hangs failing to identify the boot CPU, with no output. To avoid this from happening, extract current CPU info from of_root device node and use it for setting up the fdt in kexec_file_load case. Fixes: 6ecd0163d360 ("powerpc/kexec_file: Add appropriate regions for memory reserve map") Cc: [email protected] # v5.9+ Signed-off-by: Sourabh Jain <[email protected]> Reviewed-by: Hari Bathini <[email protected]> Signed-off-by: Michael Ellerman <[email protected]> Link: https://lore.kernel.org/r/[email protected]
2021-05-04ALSA: hda/realtek: Add fixup for HP OMEN laptopTakashi Iwai1-0/+23
HP OMEN dc0019-ur with codec SSID 103c:84da requires the pin config overrides and the existing mic/mute LED setup. This patch implements those in the fixup table. BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=212733 Cc: <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Takashi Iwai <[email protected]>
2021-05-04dt-bindings: PCI: Add SiFive FU740 PCIe host controllerGreentime Hu1-0/+113
Add PCIe host controller DT bindings of SiFive FU740. Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greentime Hu <[email protected]> Signed-off-by: Lorenzo Pieralisi <[email protected]> Reviewed-by: Rob Herring <[email protected]>
2021-05-04MAINTAINERS: Add maintainers for SiFive FU740 PCIe driverGreentime Hu1-0/+8
Here add maintainer information for SiFive FU740 PCIe driver. Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greentime Hu <[email protected]> Signed-off-by: Lorenzo Pieralisi <[email protected]>
2021-05-04clk: sifive: Use reset-simple in prci driver for PCIe driverGreentime Hu4-0/+20
We use reset-simple in this patch so that pcie driver can use devm_reset_control_get() to get this reset data structure and use reset_control_deassert() to deassert pcie_power_up_rst_n. Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greentime Hu <[email protected]> Signed-off-by: Lorenzo Pieralisi <[email protected]> Reviewed-by: Philipp Zabel <[email protected]> Acked-by: Stephen Boyd <[email protected]>
2021-05-04clk: sifive: Add pcie_aux clock in prci driver for PCIe driverGreentime Hu5-1/+63
We add pcie_aux clock in this patch so that pcie driver can use clk_prepare_enable() and clk_disable_unprepare() to enable and disable pcie_aux clock. Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greentime Hu <[email protected]> Signed-off-by: Lorenzo Pieralisi <[email protected]> Acked-by: Stephen Boyd <[email protected]>
2021-05-04ALSA: hda/realtek: Fix speaker amp on HP Envy AiO 32Takashi Iwai1-0/+35
HP Envy AiO 32-a12xxx has an external amp that is controlled via GPIO bit 0x04. However, unlike other devices, this amp seems to shut down itself after the certain period, hence the OS needs to up/down the bit dynamically only during the actual playback. This patch adds the control of the GPIO bit via the existing pcm_hook mechanism. Ideally it should be triggered at the actual stream start, but we have only the state change at prepare/cleanup, so use those for switching the GPIO bit on/off. This should be good enough for the purpose, and was actually confirmed to work fine. BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=212873 Cc: <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Takashi Iwai <[email protected]>
2021-05-04ALSA: hda/realtek: Fix silent headphone output on ASUS UX430UATakashi Iwai1-0/+18
It was reported that the headphone output on ASUS UX430UA (SSID 1043:1740) with ALC295 codec is silent while the speaker works. After the investigation, it turned out that the DAC assignment has to be fixed on this machine; unlike others, it expects DAC 0x02 to be assigned to the speaker pin 0x07 while DAC 0x03 to headphone pin 0x21. This patch provides a fixup for the fixed DAC/pin mapping for this device. BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=212933 Cc: <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Takashi Iwai <[email protected]>
2021-05-04virtio_pci_modern: correct sparse tags for notifyMichael S. Tsirkin2-7/+6
When switching virtio_pci_modern to use a helper for mappings we lost an __iomem tag. Restore it. Reported-by: kernel test robot <[email protected]> Fixes: 9e3bb9b79a71 ("virtio_pci_modern: introduce helper to map vq notify area") Signed-off-by: Michael S. Tsirkin <[email protected]>
2021-05-04virtio_pci_modern: __force cast the notify mappingMichael S. Tsirkin1-1/+1
When switching virtio_pci_modern to use a helper for mappings we lost an __iomem tag. We should restore it. However, virtio_pci_modern is playing tricks by hiding an iomem pointer in a regular vq->priv pointer. Which is okay as long as it's all contained within a single file, but we need to __force cast the value otherwise we'll get sparse warnings. Reported-by: kernel test robot <[email protected]> Fixes: 7dca6c0ea96b ("virtio-pci library: switch to use vp_modern_map_vq_notify()") Signed-off-by: Michael S. Tsirkin <[email protected]>
2021-05-04nvmet: remove unsupported command noiseKeith Busch1-3/+3
Nothing can stop a host from submitting invalid commands. The target just needs to respond with an appropriate status, but that's not a target error. Demote invalid command messages to the debug level so these events don't spam the kernel logs. Reported-by: Yi Zhang <[email protected]> Signed-off-by: Keith Busch <[email protected]> Reviewed-by: Klaus Jensen <[email protected]> Reviewed-by: Chaitanya Kulkarni <[email protected]> Signed-off-by: Christoph Hellwig <[email protected]>
2021-05-04nvme-multipath: reset bdev to ns head when failoverDaniel Wagner1-0/+3
When a request finally completes in end_io() after it has failed over, the bdev pointer can be stale and thus the system can crash. Set the bdev back to ns head, so the request is map to an active path when resubmitted. Signed-off-by: Daniel Wagner <[email protected]> Reviewed-by: Hannes Reinecke <[email protected]> Signed-off-by: Christoph Hellwig <[email protected]>
2021-05-04nvme-pci: fix controller reset hang when racing with nvme_timeoutTao Chiu1-0/+3
reset_work() in nvme-pci may hang forever in the following scenario: 1) A reset caused by a command timeout occurs due to a controller being temporarily irresponsive. 2) nvme_reset_work() restarts admin queue at nvme_alloc_admin_tags(). At the same time, a user-submitted admin command is queued and waiting for completion. Then, reset_work() changes its state to CONNECTING, and submits an identify command. 3) However, the controller does still not respond to any command, causing a timeout being fired at the user-submitted command. Unfortunately, nvme_timeout() does not see the completion on cq, and any timeout that takes place under CONNECTING state causes a controller shutdown. 4) Normally, the identify command in reset_work() would be canceled with SC_HOST_ABORTED by nvme_dev_disable(), then reset_work can tear down the controller accordingly. But the controller happens to return online and respond the identify command before nvme_dev_disable() should have been reaped it off. 5) reset_work() continues to setup_io_queues() as it observes no error in init_identify(). However, the admin queue has already been quiesced in dev_disable(). Thus, any following commands would be blocked forever in blk_execute_rq(). This can be fixed by restricting usercmd commands when controller is not in a LIVE state in nvme_queue_rq(), as what has been done previously in fabrics. ``` nvme_reset_work(): | nvme_alloc_admin_tags() | | nvme_submit_user_cmd(): nvme_init_identify(): | ... __nvme_submit_sync_cmd(): | ... | ... ---------------------------------------> nvme_timeout(): (Controller starts reponding commands) | nvme_dev_disable(, true): nvme_setup_io_queues(): | __nvme_submit_sync_cmd(): | (hung in blk_execute_rq | since run_hw_queue sees | queue quiesced) | ``` Signed-off-by: Tao Chiu <[email protected]> Signed-off-by: Cody Wong <[email protected]> Reviewed-by: Leon Chien <[email protected]> Reviewed-by: Keith Busch <[email protected]> Signed-off-by: Christoph Hellwig <[email protected]>
2021-05-04nvme: move the fabrics queue ready check routines to coreTao Chiu8-78/+83
queue_rq() in pci only checks if the dispatched queue (nvmeq) is ready, e.g. not being suspended. Since nvme_alloc_admin_tags() in reset flow restarts the admin queue, users are able to submit admin commands to a controller before reset_work() completes. Commands submitted under this condition may interfere with commands that performs identify, IO queue setup in reset_work(), and may result in a hang described in the following patch. As seen in the fabrics, user commands are prevented from being executed under inproper controller states. We may reuse this logic to maintain a clear admin queue during reset_work(). Signed-off-by: Tao Chiu <[email protected]> Signed-off-by: Cody Wong <[email protected]> Reviewed-by: Leon Chien <[email protected]> Reviewed-by: Keith Busch <[email protected]> Signed-off-by: Christoph Hellwig <[email protected]>
2021-05-04nvme: avoid memset for passthrough requestsKanchan Joshi1-4/+3
nvme_clear_nvme_request() clears the nvme_command, which is unncessary for passthrough requests as nvme_command is overwritten immediately. Move clearing part from this helper to the caller, so that double memset for passthrough requests is avoided. Signed-off-by: Kanchan Joshi <[email protected]> Signed-off-by: Christoph Hellwig <[email protected]>
2021-05-04nvme: add nvme_get_ns helperKanchan Joshi1-2/+7
Add a helper to avoid opencoding ns->kref increment. Decrement is already done via nvme_put_ns helper. Signed-off-by: Kanchan Joshi <[email protected]> Reviewed-by: Chaitanya Kulkarni <[email protected]> Signed-off-by: Christoph Hellwig <[email protected]>
2021-05-04nvme: fix controller ioctl through ns_headMinwoo Im3-47/+41
In multipath case, we should consider namespace attachment with controllers in a subsystem when we find out the live controller for the namespace. This patch manually reverted the commit 3557a4409701 ("nvme: don't bother to look up a namespace for controller ioctls") with few more updates to nvme_ns_head_chr_ioctl which has been newly updated. Fixes: 3557a4409701 ("nvme: don't bother to look up a namespace for controller ioctls") Cc: Christoph Hellwig <[email protected]> Signed-off-by: Minwoo Im <[email protected]> Signed-off-by: Christoph Hellwig <[email protected]>
2021-05-03Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpfDavid S. Miller6-30/+52
Daniel Borkmann says: ==================== pull-request: bpf 2021-05-04 The following pull-request contains BPF updates for your *net* tree. We've added 5 non-merge commits during the last 4 day(s) which contain a total of 6 files changed, 52 insertions(+), 30 deletions(-). The main changes are: 1) Fix libbpf overflow when processing BPF ring buffer in case of extreme application behavior, from Brendan Jackman. 2) Fix potential data leakage of uninitialized BPF stack under speculative execution, from Daniel Borkmann. 3) Fix off-by-one when validating xsk pool chunks, from Xuan Zhuo. 4) Fix snprintf BPF selftest with a pid filter to avoid racing its output test buffer, from Florent Revest. ==================== Signed-off-by: David S. Miller <[email protected]>
2021-05-04powerpc/64s/radix: Enable huge vmalloc mappingsNicholas Piggin3-5/+16
This reduces TLB misses by nearly 30x on a `git diff` workload on a 2-node POWER9 (59,800 -> 2,100) and reduces CPU cycles by 0.54%, due to vfs hashes being allocated with 2MB pages. Signed-off-by: Nicholas Piggin <[email protected]> Reviewed-by: Christophe Leroy <[email protected]> Acked-by: Michael Ellerman <[email protected]> Signed-off-by: Michael Ellerman <[email protected]> Link: https://lore.kernel.org/r/[email protected]
2021-05-03dt-bindings: display: renesas,du: Add missing power-domains propertyGeert Uytterhoeven1-0/+3
"make dtbs_check" complains: arch/arm/boot/dts/r8a7779-marzen.dt.yaml: display@fff80000: 'power-domains' does not match any of the regexes: 'pinctrl-[0-9]+' arch/arm64/boot/dts/renesas/r8a77970-v3msk.dt.yaml: display@feb00000: 'power-domains' does not match any of the regexes: 'pinctrl-[0-9]+' arch/arm64/boot/dts/renesas/r8a77970-eagle.dt.yaml: display@feb00000: 'power-domains' does not match any of the regexes: 'pinctrl-[0-9]+' arch/arm64/boot/dts/renesas/r8a77980-condor.dt.yaml: display@feb00000: 'power-domains' does not match any of the regexes: 'pinctrl-[0-9]+' arch/arm64/boot/dts/renesas/r8a77980-v3hsk.dt.yaml: display@feb00000: 'power-domains' does not match any of the regexes: 'pinctrl-[0-9]+' Fix this by documenting the power-domains property. Fixes: 99d66127fad25ebb ("dt-bindings: display: renesas,du: Convert binding to YAML") Signed-off-by: Geert Uytterhoeven <[email protected]> Reviewed-by: Laurent Pinchart <[email protected]> Link: https://lore.kernel.org/r/600d42256515f180bc84b72e8bdb5c5d9126ab62.1619700459.git.geert+renesas@glider.be Signed-off-by: Rob Herring <[email protected]>
2021-05-03Enlisted oprofile version line removedBhaskar Chowdhury1-1/+0
Enlisted oprofile version line removed. Signed-off-by: Bhaskar Chowdhury <[email protected]> Link: https://lore.kernel.org/r/35c4436f0f1b3072d3016148ce1461905b6f782b.1619181632.git.unixbhaskar@gmail.com Signed-off-by: Jonathan Corbet <[email protected]>
2021-05-03oprofiled version output line removed from the listBhaskar Chowdhury1-1/+0
Oprofiled version output line removed from the list. Signed-off-by: Bhaskar Chowdhury <[email protected]> Link: https://lore.kernel.org/r/8d1928ff2fea29d67143d235839a5e845e4402c9.1619181632.git.unixbhaskar@gmail.com Signed-off-by: Jonathan Corbet <[email protected]>
2021-05-03Removed the oprofiled version optionBhaskar Chowdhury1-1/+0
Removed the oprofiled version option Signed-off-by: Bhaskar Chowdhury <[email protected]> Link: https://lore.kernel.org/r/c98fa38b74bdd8ab16d35862895dac5f5a535f94.1619181632.git.unixbhaskar@gmail.com Signed-off-by: Jonathan Corbet <[email protected]>
2021-05-03docs: reporting-issues.rst: CC subsystem and maintainers on regressionsThorsten Leemhuis1-19/+30
When reporting a regression, users ideally should CC the subsystem and its maintainers, as that will ensure they get aware of the regression quickly. And if the culprit is known, they should also CC everyone who signed if off; the text mentioned the latter in once place already, but forgot to do so in two other areas where it's relevant. While fixing this also remind readers to check the mailing list archives for issues that need to be reported to a bug tracker, as someone might have reported it by mail only. All of this got triggered by a recent report where these changes would have made a difference. Signed-off-by: Thorsten Leemhuis <[email protected]> Link: https://lore.kernel.org/lkml/[email protected]/ Link: https://lore.kernel.org/lkml/CAJZ5v0hX2StQVttAciHYH-urUH+Hi92z9z2ZbcNgQPt0E2Jpwg@mail.gmail.com/ Link: https://lore.kernel.org/r/dd13f10c30e79e550215e53a8103406daec4e593.1618482489.git.linux@leemhuis.info Signed-off-by: Jonathan Corbet <[email protected]>
2021-05-03docs: correct URL to bios and kernel developer's guideAnatoly Pugachev1-1/+1
correct URL to bios and kernel developer's guide on amd.com site Signed-off-by: Anatoly Pugachev <[email protected]> Link: https://lore.kernel.org/r/[email protected] [jc: fixed resulting sphinx warning] Signed-off-by: Jonathan Corbet <[email protected]>
2021-05-03ktest: Add KTEST section to MAINTAINERS fileSteven Rostedt (VMware)1-0/+6
As I wanted to add John Hawley as a co-maintainer for ktest, I found that there never was a KTEST section in the MAINTAINERS file. Add one! Signed-off-by: Steven Rostedt (VMware) <[email protected]>
2021-05-03ktest: Re-arrange the code blocks for better discoverabilityJohn 'Warthog9' Hawley (VMware)1-142/+154
Perl, as with most scripting languages, is fairly flexible in how / where you can define things, and it will (for the most part) do what you would expect it to do. This however can lead to situations, like with ktest, where things get muddled over time. This pushes the variable definitions back up to the top, followed by functions, with the main script executables down at the bottom, INSTEAD of being somewhat mish-mashed together in certain places. This mostly has the advantage of making it more obvious where things are initially defined, what functions are there, and ACTUALLY where the main script starts executing, and should make this a little more approachable. Signed-off-by: John 'Warthog9' Hawley (VMware) <[email protected]> Signed-off-by: Steven Rostedt (VMware) <[email protected]>
2021-05-03ktest: Further consistency cleanupsJohn 'Warthog9' Hawley (VMware)1-49/+36
This cleans up some additional whitespace pieces that to be more consistent, as well as moving a curly brace around, and some 'or' statements to match the rest of the file (usually or goes at the end of the line vs. at the beginning) Signed-off-by: John 'Warthog9' Hawley (VMware) <[email protected]> Signed-off-by: Steven Rostedt (VMware) <[email protected]>
2021-05-03ktest: Fixing indentation to match expected patternJohn 'Warthog9' Hawley (VMware)1-94/+92
This is a followup to "ktest: Adding editor hints to improve consistency" to actually adjust the existing indentation to match the, now, expected pattern (first column 4 spaces, 2nd tab, 3rd tab + 4 spaces, etc). This should, at least help, keep things consistent going forward now. Signed-off-by: John 'Warthog9' Hawley (VMware) <[email protected]> Signed-off-by: Steven Rostedt (VMware) <[email protected]>
2021-05-03ktest: Adding editor hints to improve consistencyJohn 'Warthog9' Hawley (VMware)1-0/+9
Emacs and Vi(m) have different styles of dealing with perl syntax which can lead to slightly inconsistent indentation, and makes the code slightly harder to read. Emacs assumes a more perl recommended standard of 4 spaces (1 column) or tab (two column) indentation. Vi(m) tends to favor just normal spaces or tabs depending on what was being used. This gives the basic hinting to Emacs and Vim to do what is expected to be basically consistent. Emacs: - Explicitly flip into perl mode, cperl would require more adjustments Vi(m): - Set softtabs=4 which will flip it over to doing indentation the way you would expect from Emacs Signed-off-by: John 'Warthog9' Hawley (VMware) <[email protected]> Signed-off-by: Steven Rostedt (VMware) <[email protected]>
2021-05-03ktest: Add example config for using VMware VMsJohn 'Warthog9' Hawley (VMware)1-0/+137
This duplicates the KVM/Qemu config with specific notes for how to use it with VMware VMs on Workstation, Player, or Fusion. The main thing to be aware of is how the serial port is exposed which is a unix pipe, and will need something like ncat to get into ktest's monitoring Signed-off-by: John 'Warthog9' Hawley (VMware) <[email protected]> Signed-off-by: Steven Rostedt (VMware) <[email protected]>
2021-05-03ktest: Minor cleanup with uninitialized variable $build_optionsJohn 'Warthog9' Hawley (VMware)1-0/+3
Signed-off-by: John 'Warthog9' Hawley (VMware) <[email protected]> Signed-off-by: Steven Rostedt (VMware) <[email protected]>
2021-05-03docs/core-api: Consistent code styleYanteng Si1-13/+13
all `example` in this file should be replaced with ``example``. Signed-off-by: Yanteng Si <[email protected]> Acked-by: Matthias Maennich <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Jonathan Corbet <[email protected]>
2021-05-03docs/zh_CN: Adjust order and content of zh_CN/index.rstWu XiangCheng1-10/+158
Adjust order and content of zh_CN/index.rst to make it clear, complete introductions and TODOLists. Signed-off-by: Wu XiangCheng <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Jonathan Corbet <[email protected]>
2021-05-03Documentation: input: joydev file correctionsRandy Dunlap2-20/+20
Fix typos, grammar, punctuation in Documentation/input/joydev/*.rst files. Signed-off-by: Randy Dunlap <[email protected]> Cc: Dmitry Torokhov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Jonathan Corbet <[email protected]>
2021-05-04xsk: Fix for xp_aligned_validate_desc() when len == chunk_sizeXuan Zhuo1-4/+3
When desc->len is equal to chunk_size, it is legal. But when the xp_aligned_validate_desc() got chunk_end from desc->addr + desc->len pointing to the next chunk during the check, it caused the check to fail. This problem was first introduced in bbff2f321a86 ("xsk: new descriptor addressing scheme"). Later in 2b43470add8c ("xsk: Introduce AF_XDP buffer allocation API") this piece of code was moved into the new function called xp_aligned_validate_desc(). This function was then moved into xsk_queue.h via 26062b185eee ("xsk: Explicitly inline functions and move definitions"). Fixes: bbff2f321a86 ("xsk: new descriptor addressing scheme") Signed-off-by: Xuan Zhuo <[email protected]> Signed-off-by: Daniel Borkmann <[email protected]> Acked-by: Magnus Karlsson <[email protected]> Link: https://lore.kernel.org/bpf/[email protected]
2021-05-03netfilter: arptables: use pernet ops struct during unregisterFlorian Westphal3-6/+4
Like with iptables and ebtables, hook unregistration has to use the pernet ops struct, not the template. This triggered following splat: hook not found, pf 3 num 0 WARNING: CPU: 0 PID: 224 at net/netfilter/core.c:480 __nf_unregister_net_hook+0x1eb/0x610 net/netfilter/core.c:480 [..] nf_unregister_net_hook net/netfilter/core.c:502 [inline] nf_unregister_net_hooks+0x117/0x160 net/netfilter/core.c:576 arpt_unregister_table_pre_exit+0x67/0x80 net/ipv4/netfilter/arp_tables.c:1565 Fixes: f9006acc8dfe5 ("netfilter: arp_tables: pass table pointer via nf_hook_ops") Reported-by: [email protected] Signed-off-by: Florian Westphal <[email protected]> Signed-off-by: Pablo Neira Ayuso <[email protected]>
2021-05-03netfilter: xt_SECMARK: add new revision to fix structure layoutPablo Neira Ayuso2-19/+75
This extension breaks when trying to delete rules, add a new revision to fix this. Fixes: 5e6874cdb8de ("[SECMARK]: Add xtables SECMARK target") Signed-off-by: Phil Sutter <[email protected]> Signed-off-by: Pablo Neira Ayuso <[email protected]>
2021-05-03Merge tag 'for-5.13/parisc' of ↵Linus Torvalds6-104/+17
git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux Pull parisc architecture updates from Helge Deller: - switch to generic syscall header scripts - minor typo fix in setup.c * tag 'for-5.13/parisc' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux: parisc: Fix typo in setup.c parisc: syscalls: switch to generic syscallhdr.sh parisc: syscalls: switch to generic syscalltbl.sh
2021-05-03Documentation: ABI: sysfs-class-net-qmi: document pass-through fileDaniele Palmas1-0/+16
Add documentation for /sys/class/net/<iface>/qmi/pass_through Signed-off-by: Daniele Palmas <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-05-03Revert "drivers/net/wan/hdlc_fr: Fix a double free in pvc_xmit"Xie He1-3/+2
This reverts commit 1b479fb80160 ("drivers/net/wan/hdlc_fr: Fix a double free in pvc_xmit"). 1. This commit is incorrect. "__skb_pad" will NOT free the skb on failure when its "free_on_error" parameter is "false". 2. This commit claims to fix my commit. But it didn't CC me?? Fixes: 1b479fb80160 ("drivers/net/wan/hdlc_fr: Fix a double free in pvc_xmit") Cc: Lv Yunlong <[email protected]> Signed-off-by: Xie He <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-05-03Merge branch 'sctp-race-fix'David S. Miller1-16/+22
Xin Long says: ==================== sctp: fix the race condition in sctp_destroy_sock in a proper way The original fix introduced a dead lock, and has to be removed in Patch 1/2, and we will get a proper way to fix it in Patch 2/2. ==================== Signed-off-by: David S. Miller <[email protected]>
2021-05-03sctp: delay auto_asconf init until binding the first addrXin Long1-14/+17
As Or Cohen described: If sctp_destroy_sock is called without sock_net(sk)->sctp.addr_wq_lock held and sp->do_auto_asconf is true, then an element is removed from the auto_asconf_splist without any proper locking. This can happen in the following functions: 1. In sctp_accept, if sctp_sock_migrate fails. 2. In inet_create or inet6_create, if there is a bpf program attached to BPF_CGROUP_INET_SOCK_CREATE which denies creation of the sctp socket. This patch is to fix it by moving the auto_asconf init out of sctp_init_sock(), by which inet_create()/inet6_create() won't need to operate it in sctp_destroy_sock() when calling sk_common_release(). It also makes more sense to do auto_asconf init while binding the first addr, as auto_asconf actually requires an ANY addr bind, see it in sctp_addr_wq_timeout_handler(). This addresses CVE-2021-23133. Fixes: 610236587600 ("bpf: Add new cgroup attach type to enable sock modifications") Reported-by: Or Cohen <[email protected]> Signed-off-by: Xin Long <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-05-03Revert "net/sctp: fix race condition in sctp_destroy_sock"Xin Long1-5/+8
This reverts commit b166a20b07382b8bc1dcee2a448715c9c2c81b5b. This one has to be reverted as it introduced a dead lock, as syzbot reported: CPU0 CPU1 ---- ---- lock(&net->sctp.addr_wq_lock); lock(slock-AF_INET6); lock(&net->sctp.addr_wq_lock); lock(slock-AF_INET6); CPU0 is the thread of sctp_addr_wq_timeout_handler(), and CPU1 is that of sctp_close(). The original issue this commit fixed will be fixed in the next patch. Reported-by: [email protected] Signed-off-by: Xin Long <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-05-03net: hsr: check skb can contain struct hsr_ethhdr in fill_frame_infoPhillip Potter1-0/+4
Check at start of fill_frame_info that the MAC header in the supplied skb is large enough to fit a struct hsr_ethhdr, as otherwise this is not a valid HSR frame. If it is too small, return an error which will then cause the callers to clean up the skb. Fixes a KMSAN-found uninit-value bug reported by syzbot at: https://syzkaller.appspot.com/bug?id=f7e9b601f1414f814f7602a82b6619a8d80bce3f Reported-by: [email protected] Signed-off-by: Phillip Potter <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-05-03sctp: fix a SCTP_MIB_CURRESTAB leak in sctp_sf_do_dupcook_bXin Long1-1/+2
Normally SCTP_MIB_CURRESTAB is always incremented once asoc enter into ESTABLISHED from the state < ESTABLISHED and decremented when the asoc is being deleted. However, in sctp_sf_do_dupcook_b(), the asoc's state can be changed to ESTABLISHED from the state >= ESTABLISHED where it shouldn't increment SCTP_MIB_CURRESTAB. Otherwise, one asoc may increment MIB_CURRESTAB multiple times but only decrement once at the end. I was able to reproduce it by using scapy to do the 4-way shakehands, after that I replayed the COOKIE-ECHO chunk with 'peer_vtag' field changed to different values, and SCTP_MIB_CURRESTAB was incremented multiple times and never went back to 0 even when the asoc was freed. This patch is to fix it by only incrementing SCTP_MIB_CURRESTAB when the state < ESTABLISHED in sctp_sf_do_dupcook_b(). Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Reported-by: Marcelo Ricardo Leitner <[email protected]> Signed-off-by: Xin Long <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-05-03Merge branch 'sctp-bad-revert'David S. Miller2-8/+4
Xin Long says: ==================== sctp: fix the incorrect revert commit 35b4f24415c8 ("sctp: do asoc update earlier in sctp_sf_do_dupcook_a") only keeps the SHUTDOWN and COOKIE-ACK with the same asoc, not transport. So instead of revert commit 145cb2f7177d ("sctp: Fix bundling of SHUTDOWN with COOKIE-ACK"), we should revert 12dfd78e3a74 ("sctp: Fix SHUTDOWN CTSN Ack in the peer restart case"). ==================== Signed-off-by: David S. Miller <[email protected]>
2021-05-03Revert "sctp: Fix SHUTDOWN CTSN Ack in the peer restart case"Xin Long1-5/+1
This reverts commit 12dfd78e3a74825e6f0bc8df7ef9f938fbc6bfe3. This can be reverted as shutdown and cookie_ack chunk are using the same asoc since commit 35b4f24415c8 ("sctp: do asoc update earlier in sctp_sf_do_dupcook_a"). Reported-by: Jere Leppänen <[email protected]> Signed-off-by: Xin Long <[email protected]> Signed-off-by: David S. Miller <[email protected]>
2021-05-03Revert "Revert "sctp: Fix bundling of SHUTDOWN with COOKIE-ACK""Xin Long1-3/+3
This reverts commit 7e9269a5acec6d841d22e12770a0b02db4f5d8f2. As Jere notice, commit 35b4f24415c8 ("sctp: do asoc update earlier in sctp_sf_do_dupcook_a") only keeps the SHUTDOWN and COOKIE-ACK with the same asoc, not transport. So we have to bring this patch back. Reported-by: Jere Leppänen <[email protected]> Signed-off-by: Xin Long <[email protected]> Signed-off-by: David S. Miller <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.