platform/kernel/linux-starfive.git
17 months agoapparmor: fix profile verification and enable it
John Johansen [Mon, 17 Apr 2023 09:57:55 +0000 (02:57 -0700)]
apparmor: fix profile verification and enable it

The transition table size was not being set by compat mappings
resulting in the profile verification code not being run. Unfortunately
the checks were also buggy not being correctly updated from the old
accept perms, to the new layout.

Also indicate to userspace that the kernel has the permstable verification
fixes.

BugLink: http://bugs.launchpad.net/bugs/2017903
Fixes: 670f31774ab6 ("apparmor: verify permission table indexes")
Signed-off-by: John Johansen <john.johansen@canonical.com>
Reviewed-by: Jon Tourville <jontourville@me.com>
17 months agoapparmor: fix policy_compat permission remap with extended permissions
John Johansen [Fri, 10 Mar 2023 23:59:45 +0000 (15:59 -0800)]
apparmor: fix policy_compat permission remap with extended permissions

If the extended permission table is present we should not be attempting
to do a compat_permission remap as the compat_permissions are not
stored in the dfa accept states.

Fixes: fd1b2b95a211 ("apparmor: add the ability for policy to specify a permission table")
Signed-off-by: John Johansen <john.johansen@canonical.com>
Reviewed-by: Jon Tourville <jontourville@me.com>
17 months agoapparmor: aa_buffer: Convert 1-element array to flexible array
Kees Cook [Thu, 11 May 2023 21:34:45 +0000 (14:34 -0700)]
apparmor: aa_buffer: Convert 1-element array to flexible array

In the ongoing effort to convert all fake flexible arrays to proper
flexible arrays, replace aa_buffer's 1-element "buffer" member with a
flexible array.

Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: John Johansen <john.johansen@canonical.com>
17 months agoapparmor: add missing failure check in compute_xmatch_perms
John Johansen [Fri, 14 Apr 2023 07:24:47 +0000 (00:24 -0700)]
apparmor: add missing failure check in compute_xmatch_perms

Add check for failure to allocate the permission table.

Fixes: caa9f579ca72 ("apparmor: isolate policy backwards compatibility to its own file")
Signed-off-by: John Johansen <john.johansen@canonical.com>
17 months agoapparmor: fix missing error check for rhashtable_insert_fast
Danila Chernetsov [Tue, 4 Apr 2023 19:05:49 +0000 (19:05 +0000)]
apparmor: fix missing error check for rhashtable_insert_fast

 rhashtable_insert_fast() could return err value when memory allocation is
 failed. but unpack_profile() do not check values and this always returns
 success value. This patch just adds error check code.

Found by Linux Verification Center (linuxtesting.org) with SVACE.

Fixes: e025be0f26d5 ("apparmor: support querying extended trusted helper extra data")
Signed-off-by: Danila Chernetsov <listdansp@mail.ru>
Signed-off-by: John Johansen <john.johansen@canonical.com>
17 months agoapparmor: Return directly after a failed kzalloc() in two functions
Markus Elfring [Wed, 29 Mar 2023 09:50:44 +0000 (11:50 +0200)]
apparmor: Return directly after a failed kzalloc() in two functions

1. Return directly after a call of the function “kzalloc” failed
   at the beginning in these function implementations.

2. Omit extra initialisations (for a few local variables)
   which became unnecessary with this refactoring.

This issue was detected by using the Coccinelle software.

Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: John Johansen <john.johansen@canonical.com>
17 months agoAppArmor: Fix some kernel-doc comments
Yang Li [Fri, 3 Mar 2023 01:28:33 +0000 (09:28 +0800)]
AppArmor: Fix some kernel-doc comments

Make the description of @table to @strs in function unpack_trans_table()
to silence the warnings:

security/apparmor/policy_unpack.c:456: warning: Function parameter or member 'strs' not described in 'unpack_trans_table'
security/apparmor/policy_unpack.c:456: warning: Excess function parameter 'table' description in 'unpack_trans_table'

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Link: https://bugzilla.openanolis.cn/show_bug.cgi?id=4332
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>
17 months agoapparmor: fix use of strcpy in policy_unpack_test
Rae Moar [Fri, 27 Jan 2023 20:12:19 +0000 (20:12 +0000)]
apparmor: fix use of strcpy in policy_unpack_test

Replace the use of strcpy() in build_aa_ext_struct() in
policy_unpack_test.c with strscpy().

strscpy() is the safer method to use to ensure the buffer does not
overflow. This was found by kernel test robot:
https://lore.kernel.org/all/202301040348.NbfVsXO0-lkp@intel.com/.

Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Rae Moar <rmoar@google.com>
Signed-off-by: John Johansen <john.johansen@canonical.com>
17 months agoMerge patch series "riscv: some CMO alternative related clean up"
Palmer Dabbelt [Thu, 6 Jul 2023 17:32:08 +0000 (10:32 -0700)]
Merge patch series "riscv: some CMO alternative related clean up"

These cleanups came up as part of the discussion on the "riscv: Reduce
ARCH_KMALLOC_MINALIGN to 8" patch set, but that needs additional work
and thus will be delayed at least a cycle.

* b4-shazam-merge:
  riscv: mm: mark noncoherent_supported as __ro_after_init
  riscv: mm: mark CBO relate initialization funcs as __init
  riscv: errata: thead: only set cbom size & noncoherent during boot

Link: https://lore.kernel.org/linux-riscv/20230526165958.908-1-jszhang@kernel.org/
Link: https://lore.kernel.org/r/20230614165504.532-1-jszhang@kernel.org
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
17 months agoriscv: mm: mark noncoherent_supported as __ro_after_init
Jisheng Zhang [Wed, 14 Jun 2023 16:55:04 +0000 (00:55 +0800)]
riscv: mm: mark noncoherent_supported as __ro_after_init

The noncoherent_supported indicates whether the HW is coherent or not,
it won't change after booting, mark it as __ro_after_init.

Signed-off-by: Jisheng Zhang <jszhang@kernel.org>
Reviewed-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20230614165504.532-4-jszhang@kernel.org
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
17 months agoriscv: mm: mark CBO relate initialization funcs as __init
Jisheng Zhang [Wed, 14 Jun 2023 16:55:03 +0000 (00:55 +0800)]
riscv: mm: mark CBO relate initialization funcs as __init

The two functions cbo_get_block_size() and riscv_init_cbo_blocksizes()
are only called during booting, mark them as __init.

Signed-off-by: Jisheng Zhang <jszhang@kernel.org>
Reviewed-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20230614165504.532-3-jszhang@kernel.org
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
17 months agoriscv: errata: thead: only set cbom size & noncoherent during boot
Jisheng Zhang [Wed, 14 Jun 2023 16:55:02 +0000 (00:55 +0800)]
riscv: errata: thead: only set cbom size & noncoherent during boot

The CBOM size and whether the HW is noncoherent is known and
determined during booting and won't change after that.

Signed-off-by: Jisheng Zhang <jszhang@kernel.org>
Reviewed-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20230614165504.532-2-jszhang@kernel.org
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
17 months agoriscv: Select HAVE_ARCH_USERFAULTFD_MINOR
Samuel Holland [Sat, 24 Jun 2023 06:03:20 +0000 (23:03 -0700)]
riscv: Select HAVE_ARCH_USERFAULTFD_MINOR

This allocates the VM flag needed to support the userfaultfd minor fault
functionality. Because the flag bit is >= bit 32, it can only be enabled
for 64-bit kernels. See commit 7677f7fd8be7 ("userfaultfd: add minor
fault registration mode") for more information.

Signed-off-by: Samuel Holland <samuel.holland@sifive.com>
Link: https://lore.kernel.org/r/20230624060321.3401504-1-samuel.holland@sifive.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
17 months agoRISC-V: Document the ISA string parsing rules for ACPI
Palmer Dabbelt [Thu, 29 Jun 2023 03:17:05 +0000 (20:17 -0700)]
RISC-V: Document the ISA string parsing rules for ACPI

We've had a ton of issues around the ISA string parsing rules elsewhere
in RISC-V, so let's at least be clear about what the rules are so we can
try and avoid more issues.

Link: https://lore.kernel.org/r/CAK9=C2Vy-4V9kgnga98tiC3TeHkR2LFPakyBbS8s_h3_Z=ieyQ@mail.gmail.com/
Link: https://lore.kernel.org/r/20230629031705.15575-1-palmer@rivosinc.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
17 months agoMerge branches 'acpi-bus' and 'acpi-scan'
Rafael J. Wysocki [Thu, 6 Jul 2023 17:24:06 +0000 (19:24 +0200)]
Merge branches 'acpi-bus' and 'acpi-scan'

Merge additional ACPI device enumeration code changes for 6.5-rc1.

 - Make acpi_companion_match() return a const pointer and update its
   callers accordingly (Andy Shevchenko).

 - Move the extern declaration of the acpi_root variable to a header
   file so as to address a compiler warning (Andy Shevchenko).

 - Address compiler warnings in the ACPI device enumeration code by
   adding a missing header file include to it (Ben Dooks).

 - Refine the SMB0001 quirk in the ACPI device enumeration code so as to
   address an i2c-scmi driver regression (Andy Shevchenko).

 - Clean up two pieces of the ACPI device enumeration code (Andy
   Shevchenko).

* acpi-bus:
  ACPI: bus: Constify acpi_companion_match() returned value

* acpi-scan:
  ACPI: scan: Use the acpi_match_acpi_device() helper
  ACPI: platform: Move SMB0001 HID to the header and reuse
  ACPI: platform: Ignore SMB0001 only when it has resources
  ACPI: bus: Introduce acpi_match_acpi_device() helper
  ACPI: scan: fix undeclared variable warnings by including sleep.h
  ACPI: scan: Move acpi_root to internal header

17 months agoMerge tag 'asm-generic-6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/arnd...
Linus Torvalds [Thu, 6 Jul 2023 17:06:04 +0000 (10:06 -0700)]
Merge tag 'asm-generic-6.5' of git://git./linux/kernel/git/arnd/asm-generic

Pull asm-generic updates from Arnd Bergmann:
 "These are cleanups for architecture specific header files:

   - the comments in include/linux/syscalls.h have gone out of sync and
     are really pointless, so these get removed

   - The asm/bitsperlong.h header no longer needs to be architecture
     specific on modern compilers, so use a generic version for newer
     architectures that use new enough userspace compilers

   - A cleanup for virt_to_pfn/virt_to_bus to have proper type checking,
     forcing the use of pointers"

* tag 'asm-generic-6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/asm-generic:
  syscalls: Remove file path comments from headers
  tools arch: Remove uapi bitsperlong.h of hexagon and microblaze
  asm-generic: Unify uapi bitsperlong.h for arm64, riscv and loongarch
  m68k/mm: Make pfn accessors static inlines
  arm64: memory: Make virt_to_pfn() a static inline
  ARM: mm: Make virt_to_pfn() a static inline
  asm-generic/page.h: Make pfn accessors static inlines
  xen/netback: Pass (void *) to virt_to_page()
  netfs: Pass a pointer to virt_to_page()
  cifs: Pass a pointer to virt_to_page() in cifsglob
  cifs: Pass a pointer to virt_to_page()
  riscv: mm: init: Pass a pointer to virt_to_page()
  ARC: init: Pass a pointer to virt_to_pfn() in init
  m68k: Pass a pointer to virt_to_pfn() virt_to_page()
  fs/proc/kcore.c: Pass a pointer to virt_addr_valid()

17 months agoMerge tag 'soc-fixes-6.5-1' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc
Linus Torvalds [Thu, 6 Jul 2023 16:56:53 +0000 (09:56 -0700)]
Merge tag 'soc-fixes-6.5-1' of git://git./linux/kernel/git/soc/soc

Pull SoC fixes from Arnd Bergmann:
 "There are three small fixes that came up sincie the past week:

   - an incorrect bit offset in ixp4xx bus driver

   - a riscv randconfig regression in the thead platform I merged

   - whitespace fixes for some dts files"

* tag 'soc-fixes-6.5-1' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc:
  bus: ixp4xx: fix IXP4XX_EXP_T1_MASK
  ARM: dts: st: add missing space before {
  RISC-V: make ARCH_THEAD preclude XIP_KERNEL

17 months agoMerge tag 'dmaengine-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul...
Linus Torvalds [Thu, 6 Jul 2023 16:48:31 +0000 (09:48 -0700)]
Merge tag 'dmaengine-6.5-rc1' of git://git./linux/kernel/git/vkoul/dmaengine

Pull dmaengine updates from Vinod Koul:
 "New support:
   - TI J721S2 CSI BCDMA support

  Updates:
   - Native HDMI support for dw edma driver
   - ste dma40 updates for supporting proper SRAM handle in DT
   - removal of dma device chancnt setting in drivers"

* tag 'dmaengine-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine: (28 commits)
  dmaengine: sprd: Don't set chancnt
  dmaengine: hidma: Don't set chancnt
  dmaengine: plx_dma: Don't set chancnt
  dmaengine: axi-dmac: Don't set chancnt
  dmaengine: dw-axi-dmac: Don't set chancnt
  dmaengine: qcom: bam_dma: allow omitting num-{channels,ees}
  dmaengine: dw-edma: Add HDMA DebugFS support
  dmaengine: dw-edma: Add support for native HDMA
  dmaengine: dw-edma: Create a new dw_edma_core_ops structure to abstract controller operation
  dmaengine: dw-edma: Rename dw_edma_core_ops structure to dw_edma_plat_ops
  dmaengine: ste_dma40: use proper format string for resource_size_t
  dmaengine: make QCOM_HIDMA depend on HAS_IOMEM
  dmaengine: ste_dma40: fix typo in enum documentation
  dmaengine: ste_dma40: use correct print specfier for resource_size_t
  MAINTAINERS: Add myself as the DW eDMA driver reviewer
  MAINTAINERS: Add Manivannan to DW eDMA driver maintainers list
  MAINTAINERS: Demote Gustavo Pimentel to DW EDMA driver reviewer
  dmaengine: ti: k3-udma: Add support for J721S2 CSI BCDMA instance
  dt-bindings: dma: ti: Add J721S2 BCDMA
  dmaengine: ti: k3-psil-j721s2: Add PSI-L thread map for main CPSW2G
  ...

17 months agoMerge tag 'linux-watchdog-6.5-rc1' of git://www.linux-watchdog.org/linux-watchdog
Linus Torvalds [Thu, 6 Jul 2023 16:44:33 +0000 (09:44 -0700)]
Merge tag 'linux-watchdog-6.5-rc1' of git://linux-watchdog.org/linux-watchdog

Pull watchdog updates from Wim Van Sebroeck:

 - add Xilinx Versal watchdog

 - support Hygon FCH/SCH (Server Controller Hub)

 - convert GPL notices to SPDX identifiers

 - other improvements

* tag 'linux-watchdog-6.5-rc1' of git://www.linux-watchdog.org/linux-watchdog:
  watchdog: sp5100_tco: support Hygon FCH/SCH (Server Controller Hub)
  dt-bindings: watchdog: restrict node name suffixes
  MAINTAINERS: Add support for Xilinx versal watchdog
  watchdog: xilinx_wwdt: Add Versal window watchdog support
  dt-bindings: watchdog: xlnx,versal-wwdt: Add versal watchdog
  watchdog: ziirave_wdt: Switch i2c driver back to use .probe()
  watchdog: ibmasr: Replace GPL license notice with SPDX identifier
  watchdog: Convert GPL 2.0 notice to SPDX identifier
  watchdog: loongson1_wdt: Add DT support

17 months agoALSA: jack: Fix mutex call in snd_jack_report()
Takashi Iwai [Thu, 6 Jul 2023 15:53:57 +0000 (17:53 +0200)]
ALSA: jack: Fix mutex call in snd_jack_report()

snd_jack_report() is supposed to be callable from an IRQ context, too,
and it's indeed used in that way from virtsnd driver.  The fix for
input_dev race in commit 1b6a6fc5280e ("ALSA: jack: Access input_dev
under mutex"), however, introduced a mutex lock in snd_jack_report(),
and this resulted in a potential sleep-in-atomic.

For addressing that problem, this patch changes the relevant code to
use the object get/put and removes the mutex usage.  That is,
snd_jack_report(), it takes input_get_device() and leaves with
input_put_device() for assuring the input_dev being assigned.

Although the whole mutex could be reduced, we keep it because it can
be still a protection for potential races between creation and
deletion.

Fixes: 1b6a6fc5280e ("ALSA: jack: Access input_dev under mutex")
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Closes: https://lore.kernel.org/r/cf95f7fe-a748-4990-8378-000491b40329@moroto.mountain
Tested-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20230706155357.3470-1-tiwai@suse.de
Signed-off-by: Takashi Iwai <tiwai@suse.de>
17 months agoMIPS: dts: add missing space before {
Krzysztof Kozlowski [Wed, 5 Jul 2023 14:57:24 +0000 (16:57 +0200)]
MIPS: dts: add missing space before {

Add missing whitespace between node name/label and opening {.

Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
17 months agodt-bindings: watchdog: Add Loongson-1 watchdog
Keguang Zhang [Thu, 11 May 2023 12:11:58 +0000 (20:11 +0800)]
dt-bindings: watchdog: Add Loongson-1 watchdog

Add devicetree binding document for Loongson-1 watchdog.

Signed-off-by: Keguang Zhang <keguang.zhang@gmail.com>
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lkml.kernel.org/r/20230511121159.463645-2-keguang.zhang@gmail.com
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Wim Van Sebroeck <wim@linux-watchdog.org>
17 months agoMerge tag 'nf-23-07-06' of git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf
Paolo Abeni [Thu, 6 Jul 2023 09:02:57 +0000 (11:02 +0200)]
Merge tag 'nf-23-07-06' of git://git./linux/kernel/git/netfilter/nf

Pablo Neira Ayuso says:

====================
Netfilter fixes for net

The following patchset contains Netfilter fixes for net:

1) Fix missing overflow use refcount checks in nf_tables.

2) Do not set IPS_ASSURED for IPS_NAT_CLASH entries in GRE tracker,
   from Florian Westphal.

3) Bail out if nf_ct_helper_hash is NULL before registering helper,
   from Florent Revest.

4) Use siphash() instead siphash_4u64() to fix performance regression,
   also from Florian.

5) Do not allow to add rules to removed chains via ID,
   from Thadeu Lima de Souza Cascardo.

6) Fix oob read access in byteorder expression, also from Thadeu.

netfilter pull request 23-07-06

====================

Link: https://lore.kernel.org/r/20230705230406.52201-1-pablo@netfilter.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
17 months agonet: mvneta: fix txq_map in case of txq_number==1
Klaus Kudielka [Wed, 5 Jul 2023 05:37:12 +0000 (07:37 +0200)]
net: mvneta: fix txq_map in case of txq_number==1

If we boot with mvneta.txq_number=1, the txq_map is set incorrectly:
MVNETA_CPU_TXQ_ACCESS(1) refers to TX queue 1, but only TX queue 0 is
initialized. Fix this.

Fixes: 50bf8cb6fc9c ("net: mvneta: Configure XPS support")
Signed-off-by: Klaus Kudielka <klaus.kudielka@gmail.com>
Reviewed-by: Michal Kubiak <michal.kubiak@intel.com>
Link: https://lore.kernel.org/r/20230705053712.3914-1-klaus.kudielka@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
17 months agoMerge tag 'drm-misc-next-fixes-2023-06-29' of git://anongit.freedesktop.org/drm/drm...
Dave Airlie [Thu, 6 Jul 2023 08:01:47 +0000 (18:01 +1000)]
Merge tag 'drm-misc-next-fixes-2023-06-29' of git://anongit.freedesktop.org/drm/drm-misc into drm-next

Short summary of fixes pull:

 * fbdev: Fix module infos on sparc

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20230629113814.GA10448@linux-uq9g
17 months agoMerge tag 'sh-for-v6.5-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/glaubit...
Linus Torvalds [Thu, 6 Jul 2023 04:45:32 +0000 (21:45 -0700)]
Merge tag 'sh-for-v6.5-tag1' of git://git./linux/kernel/git/glaubitz/sh-linux

Pull sh updates from Adrian Glaubitz:
 "Fix a compiler warning in the J2 probing code and a fix by Sergey
  Shtylyov to avoid using IRQ0 on SH3 and SH4 targets. Masahiro Yamada
  made some clean-up in the build system to address reports by the 0day
  bot.

  The most notable changes come from Artur Rojek who addressed a number
  of issues in the DMA code, in particular a fix for the DMA channel
  offset calculation that was introduced in in 7f47c7189b3e ("sh: dma:
  More legacy cpu dma chainsawing.") in 2012!

  Together with another change to correct the number of DMA channels for
  each SuperH SoC according to specification, Artur's series unbreaks
  the kernel on the SH7709 SoC allowing Linux to boot on the HP Jornada
  680 handheld again.

  Summary:

   - Provide unxlate_dev_mem_ptr() in asm/io.h

   - dma: Correct the number of DMA channels for SH7709

   - dma: Drop incorrect SH_DMAC_BASE1 definition for SH4

   - dma: Fix DMA channel offset calculation

   - Remove compiler flag duplication

   - Refactor header include path addition

   - Move build rule for cchips/hd6446x/ to arch/sh/Kbuild

   - Fix -Wmissing-include-dirs warnings for various platforms

   - Avoid using IRQ0 on SH3 and SH4

   - j2: Use ioremap() to translate device tree address into kernel
     memory"

* tag 'sh-for-v6.5-tag1' of git://git.kernel.org/pub/scm/linux/kernel/git/glaubitz/sh-linux:
  sh: Provide unxlate_dev_mem_ptr() in asm/io.h
  sh: dma: Correct the number of DMA channels for SH7709
  sh: dma: Drop incorrect SH_DMAC_BASE1 definition for SH4
  sh: dma: Fix DMA channel offset calculation
  sh: Remove compiler flag duplication
  sh: Refactor header include path addition
  sh: Move build rule for cchips/hd6446x/ to arch/sh/Kbuild
  sh: Fix -Wmissing-include-dirs warnings for various platforms
  sh: Avoid using IRQ0 on SH3 and SH4
  sh: j2: Use ioremap() to translate device tree address into kernel memory

17 months agoMerge tag 'phy-for-6.5_v2' of git://git.kernel.org/pub/scm/linux/kernel/git/phy/linux-phy
Linus Torvalds [Thu, 6 Jul 2023 04:38:13 +0000 (21:38 -0700)]
Merge tag 'phy-for-6.5_v2' of git://git./linux/kernel/git/phy/linux-phy

Pull phy updates from Vinod Koul:
 "New Support:
   - Debugfs support for phy core and mediatek driver
   - Hisilicon inno-usb2-phy driver supporting Hi3798MV100
   - Qualcomm SGMII SerDes PHY driver, SM6115 & QCM2290 QMP-USB support,
     SA8775P USB PHY & USB3 UNI support, QUSB2 support for IPQ9574,
     IPQ9574 USB3 PHY

  UpdatesL
   - Sparx5 serdes phy power optimzation
   - cadence salvo usb properties and updates and torrent DP with PCIe &
     USB support
   - Yaml conversion for Broadcom kona USB bindings and MXS USB binding"

* tag 'phy-for-6.5_v2' of git://git.kernel.org/pub/scm/linux/kernel/git/phy/linux-phy: (67 commits)
  dt-bindings: phy: brcm,brcmstb-usb-phy: Fix error in "compatible" conditional schema
  dt-bindings: phy: mixel,mipi-dsi-phy: Remove assigned-clock* properties
  dt-bindings: phy: intel,combo-phy: restrict node name suffixes
  dt-bindings: phy: qcom,usb-hs-phy: Add compatible
  phy: tegra: xusb: check return value of devm_kzalloc()
  phy: qcom: qmp-combo: fix Display Port PHY configuration for SM8550
  phy: qcom: add the SGMII SerDes PHY driver
  dt-bindings: phy: describe the Qualcomm SGMII PHY
  phy: qualcomm: fix indentation in Makefile
  phy: usb: suppress OC condition for 7439b2
  phy: usb: Turn off phy when port is in suspend
  phy: tegra: xusb: Clear the driver reference in usb-phy dev
  dt-bindings: phy: mxs-usb-phy: add imx8ulp and imx8qm compatible
  dt-bindings: phy: mxs-usb-phy: convert to DT schema format
  dt-bindings: phy: qcom,qmp-usb: fix bindings error
  dt-bindings: phy: qcom,qmp-ufs: fix the sc8180x regs
  dt-bindings: phy: qcom,qmp-pcie: fix the sc8180x regs
  phy: mediatek: tphy: add debugfs files
  phy: core: add debugfs files
  phy: fsl-imx8mp-usb: add support for phy tuning
  ...

17 months agocifs: Add a laundromat thread for cached directories
Ronnie Sahlberg [Thu, 6 Jul 2023 02:32:24 +0000 (12:32 +1000)]
cifs: Add a laundromat thread for cached directories

and drop cached directories after 30 seconds

Signed-off-by: Ronnie Sahlberg <lsahlber@redhat.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
17 months agoMerge branch 'Fix for check_max_stack_depth'
Alexei Starovoitov [Thu, 6 Jul 2023 02:14:54 +0000 (19:14 -0700)]
Merge branch 'Fix for check_max_stack_depth'

Kumar Kartikeya Dwivedi says:

====================
Fix for a bug in check_max_stack_depth which allows bypassing the
512-byte stack limit.
====================

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
17 months agoselftests/bpf: Add selftest for check_stack_max_depth bug
Kumar Kartikeya Dwivedi [Wed, 5 Jul 2023 14:47:30 +0000 (20:17 +0530)]
selftests/bpf: Add selftest for check_stack_max_depth bug

Use the bpf_timer_set_callback helper to mark timer_cb as an async
callback, and put a direct call to timer_cb in the main subprog.

As the check_stack_max_depth happens after the do_check pass, the order
does not matter. Without the previous fix, the test passes successfully.

Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://lore.kernel.org/r/20230705144730.235802-3-memxor@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
17 months agobpf: Fix max stack depth check for async callbacks
Kumar Kartikeya Dwivedi [Wed, 5 Jul 2023 14:47:29 +0000 (20:17 +0530)]
bpf: Fix max stack depth check for async callbacks

The check_max_stack_depth pass happens after the verifier's symbolic
execution, and attempts to walk the call graph of the BPF program,
ensuring that the stack usage stays within bounds for all possible call
chains. There are two cases to consider: bpf_pseudo_func and
bpf_pseudo_call. In the former case, the callback pointer is loaded into
a register, and is assumed that it is passed to some helper later which
calls it (however there is no way to be sure), but the check remains
conservative and accounts the stack usage anyway. For this particular
case, asynchronous callbacks are skipped as they execute asynchronously
when their corresponding event fires.

The case of bpf_pseudo_call is simpler and we know that the call is
definitely made, hence the stack depth of the subprog is accounted for.

However, the current check still skips an asynchronous callback even if
a bpf_pseudo_call was made for it. This is erroneous, as it will miss
accounting for the stack usage of the asynchronous callback, which can
be used to breach the maximum stack depth limit.

Fix this by only skipping asynchronous callbacks when the instruction is
not a pseudo call to the subprog.

Fixes: 7ddc80a476c2 ("bpf: Teach stack depth check about async callbacks.")
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://lore.kernel.org/r/20230705144730.235802-2-memxor@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
17 months agonetfilter: nf_tables: prevent OOB access in nft_byteorder_eval
Thadeu Lima de Souza Cascardo [Wed, 5 Jul 2023 21:05:35 +0000 (18:05 -0300)]
netfilter: nf_tables: prevent OOB access in nft_byteorder_eval

When evaluating byteorder expressions with size 2, a union with 32-bit and
16-bit members is used. Since the 16-bit members are aligned to 32-bit,
the array accesses will be out-of-bounds.

It may lead to a stack-out-of-bounds access like the one below:

[   23.095215] ==================================================================
[   23.095625] BUG: KASAN: stack-out-of-bounds in nft_byteorder_eval+0x13c/0x320
[   23.096020] Read of size 2 at addr ffffc90000007948 by task ping/115
[   23.096358]
[   23.096456] CPU: 0 PID: 115 Comm: ping Not tainted 6.4.0+ #413
[   23.096770] Call Trace:
[   23.096910]  <IRQ>
[   23.097030]  dump_stack_lvl+0x60/0xc0
[   23.097218]  print_report+0xcf/0x630
[   23.097388]  ? nft_byteorder_eval+0x13c/0x320
[   23.097577]  ? kasan_addr_to_slab+0xd/0xc0
[   23.097760]  ? nft_byteorder_eval+0x13c/0x320
[   23.097949]  kasan_report+0xc9/0x110
[   23.098106]  ? nft_byteorder_eval+0x13c/0x320
[   23.098298]  __asan_load2+0x83/0xd0
[   23.098453]  nft_byteorder_eval+0x13c/0x320
[   23.098659]  nft_do_chain+0x1c8/0xc50
[   23.098852]  ? __pfx_nft_do_chain+0x10/0x10
[   23.099078]  ? __kasan_check_read+0x11/0x20
[   23.099295]  ? __pfx___lock_acquire+0x10/0x10
[   23.099535]  ? __pfx___lock_acquire+0x10/0x10
[   23.099745]  ? __kasan_check_read+0x11/0x20
[   23.099929]  nft_do_chain_ipv4+0xfe/0x140
[   23.100105]  ? __pfx_nft_do_chain_ipv4+0x10/0x10
[   23.100327]  ? lock_release+0x204/0x400
[   23.100515]  ? nf_hook.constprop.0+0x340/0x550
[   23.100779]  nf_hook_slow+0x6c/0x100
[   23.100977]  ? __pfx_nft_do_chain_ipv4+0x10/0x10
[   23.101223]  nf_hook.constprop.0+0x334/0x550
[   23.101443]  ? __pfx_ip_local_deliver_finish+0x10/0x10
[   23.101677]  ? __pfx_nf_hook.constprop.0+0x10/0x10
[   23.101882]  ? __pfx_ip_rcv_finish+0x10/0x10
[   23.102071]  ? __pfx_ip_local_deliver_finish+0x10/0x10
[   23.102291]  ? rcu_read_lock_held+0x4b/0x70
[   23.102481]  ip_local_deliver+0xbb/0x110
[   23.102665]  ? __pfx_ip_rcv+0x10/0x10
[   23.102839]  ip_rcv+0x199/0x2a0
[   23.102980]  ? __pfx_ip_rcv+0x10/0x10
[   23.103140]  __netif_receive_skb_one_core+0x13e/0x150
[   23.103362]  ? __pfx___netif_receive_skb_one_core+0x10/0x10
[   23.103647]  ? mark_held_locks+0x48/0xa0
[   23.103819]  ? process_backlog+0x36c/0x380
[   23.103999]  __netif_receive_skb+0x23/0xc0
[   23.104179]  process_backlog+0x91/0x380
[   23.104350]  __napi_poll.constprop.0+0x66/0x360
[   23.104589]  ? net_rx_action+0x1cb/0x610
[   23.104811]  net_rx_action+0x33e/0x610
[   23.105024]  ? _raw_spin_unlock+0x23/0x50
[   23.105257]  ? __pfx_net_rx_action+0x10/0x10
[   23.105485]  ? mark_held_locks+0x48/0xa0
[   23.105741]  __do_softirq+0xfa/0x5ab
[   23.105956]  ? __dev_queue_xmit+0x765/0x1c00
[   23.106193]  do_softirq.part.0+0x49/0xc0
[   23.106423]  </IRQ>
[   23.106547]  <TASK>
[   23.106670]  __local_bh_enable_ip+0xf5/0x120
[   23.106903]  __dev_queue_xmit+0x789/0x1c00
[   23.107131]  ? __pfx___dev_queue_xmit+0x10/0x10
[   23.107381]  ? find_held_lock+0x8e/0xb0
[   23.107585]  ? lock_release+0x204/0x400
[   23.107798]  ? neigh_resolve_output+0x185/0x350
[   23.108049]  ? mark_held_locks+0x48/0xa0
[   23.108265]  ? neigh_resolve_output+0x185/0x350
[   23.108514]  neigh_resolve_output+0x246/0x350
[   23.108753]  ? neigh_resolve_output+0x246/0x350
[   23.109003]  ip_finish_output2+0x3c3/0x10b0
[   23.109250]  ? __pfx_ip_finish_output2+0x10/0x10
[   23.109510]  ? __pfx_nf_hook+0x10/0x10
[   23.109732]  __ip_finish_output+0x217/0x390
[   23.109978]  ip_finish_output+0x2f/0x130
[   23.110207]  ip_output+0xc9/0x170
[   23.110404]  ip_push_pending_frames+0x1a0/0x240
[   23.110652]  raw_sendmsg+0x102e/0x19e0
[   23.110871]  ? __pfx_raw_sendmsg+0x10/0x10
[   23.111093]  ? lock_release+0x204/0x400
[   23.111304]  ? __mod_lruvec_page_state+0x148/0x330
[   23.111567]  ? find_held_lock+0x8e/0xb0
[   23.111777]  ? find_held_lock+0x8e/0xb0
[   23.111993]  ? __rcu_read_unlock+0x7c/0x2f0
[   23.112225]  ? aa_sk_perm+0x18a/0x550
[   23.112431]  ? filemap_map_pages+0x4f1/0x900
[   23.112665]  ? __pfx_aa_sk_perm+0x10/0x10
[   23.112880]  ? find_held_lock+0x8e/0xb0
[   23.113098]  inet_sendmsg+0xa0/0xb0
[   23.113297]  ? inet_sendmsg+0xa0/0xb0
[   23.113500]  ? __pfx_inet_sendmsg+0x10/0x10
[   23.113727]  sock_sendmsg+0xf4/0x100
[   23.113924]  ? move_addr_to_kernel.part.0+0x4f/0xa0
[   23.114190]  __sys_sendto+0x1d4/0x290
[   23.114391]  ? __pfx___sys_sendto+0x10/0x10
[   23.114621]  ? __pfx_mark_lock.part.0+0x10/0x10
[   23.114869]  ? lock_release+0x204/0x400
[   23.115076]  ? find_held_lock+0x8e/0xb0
[   23.115287]  ? rcu_is_watching+0x23/0x60
[   23.115503]  ? __rseq_handle_notify_resume+0x6e2/0x860
[   23.115778]  ? __kasan_check_write+0x14/0x30
[   23.116008]  ? blkcg_maybe_throttle_current+0x8d/0x770
[   23.116285]  ? mark_held_locks+0x28/0xa0
[   23.116503]  ? do_syscall_64+0x37/0x90
[   23.116713]  __x64_sys_sendto+0x7f/0xb0
[   23.116924]  do_syscall_64+0x59/0x90
[   23.117123]  ? irqentry_exit_to_user_mode+0x25/0x30
[   23.117387]  ? irqentry_exit+0x77/0xb0
[   23.117593]  ? exc_page_fault+0x92/0x140
[   23.117806]  entry_SYSCALL_64_after_hwframe+0x6e/0xd8
[   23.118081] RIP: 0033:0x7f744aee2bba
[   23.118282] Code: d8 64 89 02 48 c7 c0 ff ff ff ff eb b8 0f 1f 00 f3 0f 1e fa 41 89 ca 64 8b 04 25 18 00 00 00 85 c0 75 15 b8 2c 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 7e c3 0f 1f 44 00 00 41 54 48 83 ec 30 44 89
[   23.119237] RSP: 002b:00007ffd04a7c9f8 EFLAGS: 00000246 ORIG_RAX: 000000000000002c
[   23.119644] RAX: ffffffffffffffda RBX: 00007ffd04a7e0a0 RCX: 00007f744aee2bba
[   23.120023] RDX: 0000000000000040 RSI: 000056488e9e6300 RDI: 0000000000000003
[   23.120413] RBP: 000056488e9e6300 R08: 00007ffd04a80320 R09: 0000000000000010
[   23.120809] R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000040
[   23.121219] R13: 00007ffd04a7dc38 R14: 00007ffd04a7ca00 R15: 00007ffd04a7e0a0
[   23.121617]  </TASK>
[   23.121749]
[   23.121845] The buggy address belongs to the virtual mapping at
[   23.121845]  [ffffc90000000000ffffc90000009000) created by:
[   23.121845]  irq_init_percpu_irqstack+0x1cf/0x270
[   23.122707]
[   23.122803] The buggy address belongs to the physical page:
[   23.123104] page:0000000072ac19f0 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x24a09
[   23.123609] flags: 0xfffffc0001000(reserved|node=0|zone=1|lastcpupid=0x1fffff)
[   23.123998] page_type: 0xffffffff()
[   23.124194] raw: 000fffffc0001000 ffffea0000928248 ffffea0000928248 0000000000000000
[   23.124610] raw: 0000000000000000 0000000000000000 00000001ffffffff 0000000000000000
[   23.125023] page dumped because: kasan: bad access detected
[   23.125326]
[   23.125421] Memory state around the buggy address:
[   23.125682]  ffffc90000007800: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
[   23.126072]  ffffc90000007880: 00 00 00 00 00 f1 f1 f1 f1 f1 f1 00 00 f2 f2 00
[   23.126455] >ffffc90000007900: 00 00 00 00 00 00 00 00 00 f2 f2 f2 f2 00 00 00
[   23.126840]                                               ^
[   23.127138]  ffffc90000007980: 00 00 00 00 00 00 00 00 00 00 00 00 00 f3 f3 f3
[   23.127522]  ffffc90000007a00: f3 00 00 00 00 00 00 00 00 00 00 00 f1 f1 f1 f1
[   23.127906] ==================================================================
[   23.128324] Disabling lock debugging due to kernel taint

Using simple s16 pointers for the 16-bit accesses fixes the problem. For
the 32-bit accesses, src and dst can be used directly.

Fixes: 96518518cc41 ("netfilter: add nftables")
Cc: stable@vger.kernel.org
Reported-by: Tanguy DUBROCA (@SidewayRE) from @Synacktiv working with ZDI
Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@canonical.com>
Reviewed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
17 months agoMerge tag 'net-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Linus Torvalds [Wed, 5 Jul 2023 22:44:45 +0000 (15:44 -0700)]
Merge tag 'net-6.5-rc1' of git://git./linux/kernel/git/netdev/net

Pull networking fixes from Jakub Kicinski:
 "Including fixes from bluetooth, bpf and wireguard.

  Current release - regressions:

   - nvme-tcp: fix comma-related oops after sendpage changes

  Current release - new code bugs:

   - ptp: make max_phase_adjustment sysfs device attribute invisible
     when not supported

  Previous releases - regressions:

   - sctp: fix potential deadlock on &net->sctp.addr_wq_lock

   - mptcp:
      - ensure subflow is unhashed before cleaning the backlog
      - do not rely on implicit state check in mptcp_listen()

  Previous releases - always broken:

   - net: fix net_dev_start_xmit trace event vs skb_transport_offset()

   - Bluetooth:
      - fix use-bdaddr-property quirk
      - L2CAP: fix multiple UaFs
      - ISO: use hci_sync for setting CIG parameters
      - hci_event: fix Set CIG Parameters error status handling
      - hci_event: fix parsing of CIS Established Event
      - MGMT: fix marking SCAN_RSP as not connectable

   - wireguard: queuing: use saner cpu selection wrapping

   - sched: act_ipt: various bug fixes for iptables <> TC interactions

   - sched: act_pedit: add size check for TCA_PEDIT_PARMS_EX

   - dsa: fixes for receiving PTP packets with 8021q and sja1105 tagging

   - eth: sfc: fix null-deref in devlink port without MAE access

   - eth: ibmvnic: do not reset dql stats on NON_FATAL err

  Misc:

   - xsk: honor SO_BINDTODEVICE on bind"

* tag 'net-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (70 commits)
  nfp: clean mc addresses in application firmware when closing port
  selftests: mptcp: pm_nl_ctl: fix 32-bit support
  selftests: mptcp: depend on SYN_COOKIES
  selftests: mptcp: userspace_pm: report errors with 'remove' tests
  selftests: mptcp: userspace_pm: use correct server port
  selftests: mptcp: sockopt: return error if wrong mark
  selftests: mptcp: sockopt: use 'iptables-legacy' if available
  selftests: mptcp: connect: fail if nft supposed to work
  mptcp: do not rely on implicit state check in mptcp_listen()
  mptcp: ensure subflow is unhashed before cleaning the backlog
  s390/qeth: Fix vipa deletion
  octeontx-af: fix hardware timestamp configuration
  net: dsa: sja1105: always enable the send_meta options
  net: dsa: tag_sja1105: fix MAC DA patching from meta frames
  net: Replace strlcpy with strscpy
  pptp: Fix fib lookup calls.
  mlxsw: spectrum_router: Fix an IS_ERR() vs NULL check
  net/sched: act_pedit: Add size check for TCA_PEDIT_PARMS_EX
  xsk: Honor SO_BINDTODEVICE on bind
  ptp: Make max_phase_adjustment sysfs device attribute invisible when not supported
  ...

17 months agoMerge tag 'f2fs-for-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk...
Linus Torvalds [Wed, 5 Jul 2023 21:14:37 +0000 (14:14 -0700)]
Merge tag 'f2fs-for-6.5-rc1' of git://git./linux/kernel/git/jaegeuk/f2fs

Pull f2fs updates from Jaegeuk Kim:
 "In this cycle, we've mainly investigated the zoned block device
  support along with patches such as correcting write pointers between
  f2fs and storage, adding asynchronous zone reset flow, and managing
  the number of open zones.

  Other than them, f2fs adds another mount option, "errors=x" to specify
  how to handle when it detects an unexpected behavior at runtime.

  Enhancements:
   - support 'errors=remount-ro|continue|panic' mount option
   - enforce some inode flag policies
   - allow .tmp compression given extensions
   - add some ioctls to manage the f2fs compression
   - improve looped node chain flow
   - avoid issuing small-sized discard commands during checkpoint
   - implement an asynchronous zone reset

  Bug fixes:
   - fix deadlock in xattr and inode page lock
   - fix and add sanity check in some error paths
   - fix to avoid NULL pointer dereference f2fs_write_end_io() along
     with put_super
   - set proper flags to quota files
   - fix potential deadlock due to unpaired node_write lock use
   - fix over-estimating free section during FG GC
   - fix the wrong condition to determine atomic context

  As usual, also there are a number of patches with code refactoring and
  minor clean-ups"

* tag 'f2fs-for-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs: (46 commits)
  f2fs: fix to do sanity check on direct node in truncate_dnode()
  f2fs: only set release for file that has compressed data
  f2fs: fix compile warning in f2fs_destroy_node_manager()
  f2fs: fix error path handling in truncate_dnode()
  f2fs: fix deadlock in i_xattr_sem and inode page lock
  f2fs: remove unneeded page uptodate check/set
  f2fs: update mtime and ctime in move file range method
  f2fs: compress tmp files given extension
  f2fs: refactor struct f2fs_attr macro
  f2fs: convert to use sbi directly
  f2fs: remove redundant assignment to variable err
  f2fs: do not issue small discard commands during checkpoint
  f2fs: check zone write pointer points to the end of zone
  f2fs: add f2fs_ioc_get_compress_blocks
  f2fs: cleanup MIN_INLINE_XATTR_SIZE
  f2fs: add helper to check compression level
  f2fs: set FMODE_CAN_ODIRECT instead of a dummy direct_IO method
  f2fs: do more sanity check on inode
  f2fs: compress: fix to check validity of i_compress_flag field
  f2fs: add sanity compress level check for compressed file
  ...

17 months agousb: ch9: Replace bmSublinkSpeedAttr 1-element array with flexible array
Kees Cook [Thu, 29 Jun 2023 19:09:00 +0000 (12:09 -0700)]
usb: ch9: Replace bmSublinkSpeedAttr 1-element array with flexible array

Since commit 2d47c6956ab3 ("ubsan: Tighten UBSAN_BOUNDS on GCC"),
UBSAN_BOUNDS no longer pretends 1-element arrays are unbounded. Walking
bmSublinkSpeedAttr will trigger a warning, so make it a proper flexible
array. Add a union to keep the struct size identical for userspace in
case anything was depending on the old size.

False positive warning was:

UBSAN: array-index-out-of-bounds in drivers/usb/host/xhci-hub.c:231:31 index 1 is out of range for type '__le32 [1]'

for this line of code:

ssp_cap->bmSublinkSpeedAttr[offset++] = cpu_to_le32(attr);

Reported-by: Borislav Petkov <bp@alien8.de>
Closes: https://lore.kernel.org/lkml/2023062945-fencing-pebble-0411@gregkh/
Reported-by: Mirsad Todorovac <mirsad.todorovac@alu.unizg.hr>
Closes: https://lore.kernel.org/lkml/9a8e34ad-8a8b-3830-4878-3c2c82e69dd9@alu.unizg.hr/
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: "Gustavo A. R. Silva" <gustavoars@kernel.org>
Tested-by: "Borislav Petkov (AMD)" <bp@alien8.de>
Tested-by: Mirsad Todorovac <mirsad.todorovac@alu.unizg.hr>
Reviewed-by: "Gustavo A. R. Silva" <gustavoars@kernel.org>
Link: https://lore.kernel.org/r/20230629190900.never.787-kees@kernel.org
Signed-off-by: Kees Cook <keescook@chromium.org>
17 months agoMerge tag 'xfs-6.5-merge-5' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux
Linus Torvalds [Wed, 5 Jul 2023 21:08:03 +0000 (14:08 -0700)]
Merge tag 'xfs-6.5-merge-5' of git://git./fs/xfs/xfs-linux

Pull more xfs updates from Darrick Wong:

 - Fix some ordering problems with log items during log recovery

 - Don't deadlock the system by trying to flush busy freed extents while
   holding on to busy freed extents

 - Improve validation of log geometry parameters when reading the
   primary superblock

 - Validate the length field in the AGF header

 - Fix recordset filtering bugs when re-calling GETFSMAP to return more
   results when the resultset didn't previously fit in the caller's
   buffer

 - Fix integer overflows in GETFSMAP when working with rt volumes larger
   than 2^32 fsblocks

 - Fix GETFSMAP reporting the undefined space beyond the last rtextent

 - Fix filtering bugs in GETFSMAP's log device backend if the log ever
   becomes longer than 2^32 fsblocks

 - Improve validation of file offsets in the GETFSMAP range parameters

 - Fix an off by one bug in the pmem media failure notification
   computation

 - Validate the length field in the AGI header too

* tag 'xfs-6.5-merge-5' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
  xfs: Remove unneeded semicolon
  xfs: AGI length should be bounds checked
  xfs: fix the calculation for "end" and "length"
  xfs: fix xfs_btree_query_range callers to initialize btree rec fully
  xfs: validate fsmap offsets specified in the query keys
  xfs: fix logdev fsmap query result filtering
  xfs: clean up the rtbitmap fsmap backend
  xfs: fix getfsmap reporting past the last rt extent
  xfs: fix integer overflows in the fsmap rtbitmap and logdev backends
  xfs: fix interval filtering in multi-step fsmap queries
  xfs: fix bounds check in xfs_defer_agfl_block()
  xfs: AGF length has never been bounds checked
  xfs: journal geometry is not properly bounds checked
  xfs: don't block in busy flushing when freeing extents
  xfs: allow extent free intents to be retried
  xfs: pass alloc flags through to xfs_extent_busy_flush()
  xfs: use deferred frees for btree block freeing
  xfs: don't reverse order of items in bulk AIL insertion
  xfs: remove redundant initializations of pointers drop_leaf and save_leaf

17 months agobus: ixp4xx: fix IXP4XX_EXP_T1_MASK
Jonas Gorski [Sat, 24 Jun 2023 12:21:39 +0000 (14:21 +0200)]
bus: ixp4xx: fix IXP4XX_EXP_T1_MASK

The IXP4XX_EXP_T1_MASK was shifted one bit to the right, overlapping
IXP4XX_EXP_T2_MASK and leaving bit 29 unused. The offset being wrong is
also confirmed at least by the datasheet of IXP45X/46X [1].

Fix this by aligning it to IXP4XX_EXP_T1_SHIFT.

[1] https://www.intel.com/content/dam/www/public/us/en/documents/manuals/ixp45x-ixp46x-developers-manual.pdf

Cc: stable@vger.kernel.org
Fixes: 1c953bda90ca ("bus: ixp4xx: Add a driver for IXP4xx expansion bus")
Signed-off-by: Jonas Gorski <jonas.gorski@gmail.com>
Link: https://lore.kernel.org/r/20230624112958.27727-1-jonas.gorski@gmail.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20230624122139.3229642-1-linus.walleij@linaro.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
17 months agoARM: dts: st: add missing space before {
Krzysztof Kozlowski [Wed, 5 Jul 2023 15:00:33 +0000 (17:00 +0200)]
ARM: dts: st: add missing space before {

Add missing whitespace between node name/label and opening {.

Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20230705150033.293832-1-krzysztof.kozlowski@linaro.org
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
17 months agoRISC-V: make ARCH_THEAD preclude XIP_KERNEL
Conor Dooley [Wed, 28 Jun 2023 07:12:30 +0000 (08:12 +0100)]
RISC-V: make ARCH_THEAD preclude XIP_KERNEL

Randy reported build errors in linux-next where XIP_KERNEL was enabled.
ARCH_THEAD requires alternatives to support the non-standard ISA
extensions used by the THEAD cores, which are mutually exclusive with
XIP kernels. Clone the dependency list from the Allwinner entry, since
Allwinner's D1 uses T-Head cores with the same non-standard extensions.

Reported-by: Randy Dunlap <rdunlap@infradead.org>
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
Link: https://lore.kernel.org/all/ab38f6af-cb68-a918-1a63-2e7c927a8ffc@infradead.org/
Fixes: da47ce003963 ("riscv: Add the T-HEAD SoC family Kconfig option")
Reviewed-by: Palmer Dabbelt <palmer@rivosinc.com>
Acked-by: Palmer Dabbelt <palmer@rivosinc.com>
Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20230628-left-attractor-94b7bd5fbb83@wendy
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
17 months agoMerge tag 'pwm/for-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/thierry...
Linus Torvalds [Wed, 5 Jul 2023 19:55:06 +0000 (12:55 -0700)]
Merge tag 'pwm/for-6.5-rc1' of git://git./linux/kernel/git/thierry.reding/linux-pwm

Pull pwm updates from Thierry Reding:
 "There's a little bit of everything in here: we've got various
  improvements and cleanups to drivers, some fixes across the board and
  a bit of new hardware support"

* tag 'pwm/for-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/thierry.reding/linux-pwm: (22 commits)
  dt-bindings: pwm: convert pwm-bcm2835 bindings to YAML
  pwm: Add Renesas RZ/G2L MTU3a PWM driver
  pwm: mtk_disp: Fix the disable flow of disp_pwm
  dt-bindings: pwm: restrict node name suffixes
  pwm: pca9685: Switch i2c driver back to use .probe()
  pwm: ab8500: Fix error code in probe()
  MAINTAINERS: add pwm to PolarFire SoC entry
  pwm: add microchip soft ip corePWM driver
  pwm: sysfs: Do not apply state to already disabled PWMs
  pwm: imx-tpm: force 'real_period' to be zero in suspend
  pwm: meson: make full use of common clock framework
  pwm: meson: don't use hdmi/video clock as mux parent
  pwm: meson: switch to using struct clk_parent_data for mux parents
  pwm: meson: remove not needed check in meson_pwm_calc
  pwm: meson: fix handling of period/duty if greater than UINT_MAX
  pwm: meson: modify and simplify calculation in meson_pwm_get_state
  dt-bindings: pwm: Add R-Car V3U device tree bindings
  dt-bindings: pwm: imx: add i.MX8QXP compatible
  pwm: mediatek: Add support for MT7981
  dt-bindings: pwm: mediatek: Add mediatek,mt7981 compatible
  ...

17 months agoMerge tag 'devicetree-for-6.5-2' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Wed, 5 Jul 2023 19:50:27 +0000 (12:50 -0700)]
Merge tag 'devicetree-for-6.5-2' of git://git./linux/kernel/git/robh/linux

Pull more devicetree updates from Rob Herring:

 - Whitespace clean-ups in binding examples

 - Restrict node name suffixes to "-[0-9]+" for cases of multiple
   instances which don't have unit-addresses

 - Convert brcm,kona-wdt and cdns,wdt-r1p2 watchdog bindings to DT
   schema

* tag 'devicetree-for-6.5-2' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux:
  dt-bindings: soc: qcom: stats: Update maintainer email
  dt-bindings: cleanup DTS example whitespaces
  dt-bindings: timestamp: restrict node name suffixes
  dt-bindings: slimbus: restrict node name suffixes
  dt-bindings: watchdog: restrict node name suffixes
  dt-bindings: watchdog: brcm,kona-wdt: convert txt file to yaml
  dt-bindings: watchdog: cdns,wdt-r1p2: Convert cadence watchdog to yaml

17 months agoperf test: Fix event parsing test when PERF_PMU_CAP_EXTENDED_HW_TYPE isn't supported.
James Clark [Wed, 5 Jul 2023 08:26:52 +0000 (09:26 +0100)]
perf test: Fix event parsing test when PERF_PMU_CAP_EXTENDED_HW_TYPE isn't supported.

Arm has multiple PMU types for heterogeneous systems, but doesn't
currently support PERF_PMU_CAP_EXTENDED_HW_TYPE. Make the tests
support both scenarios so that they pass on Arm, and will still pass
once PERF_PMU_CAP_EXTENDED_HW_TYPE support is added.

Fixes: 27c9fcfc1e14 ("perf test: Update parse-events expectations to test for multiple events")
Acked-by: Ian Rogers <irogers@google.com>
Signed-off-by: James Clark <james.clark@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Link: https://lore.kernel.org/r/20230705082653.23566-3-james.clark@arm.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
17 months agoperf test: Fix event parsing test on Arm
James Clark [Wed, 5 Jul 2023 08:26:51 +0000 (09:26 +0100)]
perf test: Fix event parsing test on Arm

The test looks for a PMU from sysfs with type = PERF_TYPE_RAW when
opening a raw event. Arm doesn't have a real raw PMU, only core PMUs
with unique types other than raw.

Instead of looking for a matching PMU, just test that the event type
was parsed as raw and skip the PMU search on Arm. The raw event type
test should also apply to all platforms so add it outside of the ifdef.

Fixes: aefde50a446b ("perf test: Fix parse-events tests for >1 core PMU")
Acked-by: Ian Rogers <irogers@google.com>
Signed-off-by: James Clark <james.clark@arm.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Link: https://lore.kernel.org/r/20230705082653.23566-2-james.clark@arm.com
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
17 months agoigc: Handle PPS start time programming for past time values
Aravindhan Gunasekaran [Thu, 15 Jun 2023 06:30:43 +0000 (12:00 +0530)]
igc: Handle PPS start time programming for past time values

I225/6 hardware can be programmed to start PPS output once
the time in Target Time registers is reached. The time
programmed in these registers should always be into future.
Only then PPS output is triggered when SYSTIM register
reaches the programmed value. There are two modes in i225/6
hardware to program PPS, pulse and clock mode.

There were issues reported where PPS is not generated when
start time is in past.

Example 1, "echo 0 0 0 2 0 > /sys/class/ptp/ptp0/period"

In the current implementation, a value of '0' is programmed
into Target time registers and PPS output is in pulse mode.
Eventually an interrupt which is triggered upon SYSTIM
register reaching Target time is not fired. Thus no PPS
output is generated.

Example 2, "echo 0 0 0 1 0 > /sys/class/ptp/ptp0/period"

Above case, a value of '0' is programmed into Target time
registers and PPS output is in clock mode. Here, HW tries to
catch-up the current time by incrementing Target Time
register. This catch-up time seem to vary according to
programmed PPS period time as per the HW design. In my
experiments, the delay ranged between few tens of seconds to
few minutes. The PPS output is only generated after the
Target time register reaches current time.

In my experiments, I also observed PPS stopped working with
below test and could not recover until module is removed and
loaded again.

1) echo 0 <future time> 0 1 0 > /sys/class/ptp/ptp1/period
2) echo 0 0 0 1 0 > /sys/class/ptp/ptp1/period
3) echo 0 0 0 1 0 > /sys/class/ptp/ptp1/period

After this PPS did not work even if i re-program with proper
values. I could only get this back working by reloading the
driver.

This patch takes care of calculating and programming
appropriate future time value into Target Time registers.

Fixes: 5e91c72e560c ("igc: Fix PPS delta between two synchronized end-points")
Signed-off-by: Aravindhan Gunasekaran <aravindhan.gunasekaran@intel.com>
Reviewed-by: Muhammad Husaini Zulkifli <muhammad.husaini.zulkifli@intel.com>
Tested-by: Naama Meir <naamax.meir@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
17 months agoigc: Include the length/type field and VLAN tag in queueMaxSDU
Tan Tee Min [Fri, 9 Jun 2023 03:28:42 +0000 (11:28 +0800)]
igc: Include the length/type field and VLAN tag in queueMaxSDU

IEEE 802.1Q does not have clear definitions of what constitutes an
SDU (Service Data Unit), but IEEE Std 802.3 clause 3.1.2 does define
the MAC service primitives and clause 3.2.7 does define the MAC Client
Data for Q-tagged frames.

It shows that the mac_service_data_unit (MSDU) does NOT contain the
preamble, destination and source address, or FCS. The MSDU does contain
the length/type field, MAC client data, VLAN tag and any padding
data (prior to the FCS).

Thus, the maximum 802.3 frame size that is allowed to be transmitted
should be QueueMaxSDU (MSDU) + 16 (6 byte SA + 6 byte DA + 4 byte FCS).

Fixes: 92a0dcb8427d ("igc: offload queue max SDU from tc-taprio")
Signed-off-by: Tan Tee Min <tee.min.tan@linux.intel.com>
Reviewed-by: Muhammad Husaini Zulkifli <muhammad.husaini.zulkifli@intel.com>
Tested-by: Naama Meir <naamax.meir@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
17 months agoigc: set TP bit in 'supported' and 'advertising' fields of ethtool_link_ksettings
Prasad Koya [Mon, 5 Jun 2023 18:09:01 +0000 (11:09 -0700)]
igc: set TP bit in 'supported' and 'advertising' fields of ethtool_link_ksettings

set TP bit in the 'supported' and 'advertising' fields. i225/226 parts
only support twisted pair copper.

Fixes: 8c5ad0dae93c ("igc: Add ethtool support")
Signed-off-by: Prasad Koya <prasad@arista.com>
Acked-by: Sasha Neftin <sasha.neftin@intel.com>
Tested-by: Naama Meir <naamax.meir@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
17 months agonfp: clean mc addresses in application firmware when closing port
Yinjun Zhang [Wed, 5 Jul 2023 05:28:18 +0000 (07:28 +0200)]
nfp: clean mc addresses in application firmware when closing port

When moving devices from one namespace to another, mc addresses are
cleaned in software while not removed from application firmware. Thus
the mc addresses are remained and will cause resource leak.

Now use `__dev_mc_unsync` to clean mc addresses when closing port.

Fixes: e20aa071cd95 ("nfp: fix schedule in atomic context when sync mc address")
Cc: stable@vger.kernel.org
Signed-off-by: Yinjun Zhang <yinjun.zhang@corigine.com>
Acked-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Louis Peens <louis.peens@corigine.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Message-ID: <20230705052818.7122-1-louis.peens@corigine.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agoMerge tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Jakub Kicinski [Wed, 5 Jul 2023 17:57:13 +0000 (10:57 -0700)]
Merge tag 'for-netdev' of https://git./linux/kernel/git/bpf/bpf

Daniel Borkmann says:

====================
pull-request: bpf 2023-07-05

We've added 2 non-merge commits during the last 1 day(s) which contain
a total of 3 files changed, 16 insertions(+), 4 deletions(-).

The main changes are:

1) Fix BTF to warn but not returning an error for a NULL BTF to still be
   able to load modules under CONFIG_DEBUG_INFO_BTF, from SeongJae Park.

2) Fix xsk sockets to honor SO_BINDTODEVICE in bind(), from Ilya Maximets.

* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
  xsk: Honor SO_BINDTODEVICE on bind
  bpf, btf: Warn but return no error for NULL btf from __register_btf_kfunc_id_set()
====================

Link: https://lore.kernel.org/r/20230705171716.6494-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
17 months agonet/mlx5e: RX, Fix page_pool page fragment tracking for XDP
Dragos Tatulea [Wed, 31 May 2023 18:18:49 +0000 (21:18 +0300)]
net/mlx5e: RX, Fix page_pool page fragment tracking for XDP

Currently mlx5e releases pages directly to the page_pool for XDP_TX and
does page fragment counting for XDP_REDIRECT. RX pages from the
page_pool are leaking on XDP_REDIRECT because the xdp core will release
only one fragment out of MLX5E_PAGECNT_BIAS_MAX and subsequently the page
is marked as "skip release" which avoids the driver release.

A fix would be to take an extra fragment for XDP_REDIRECT and not set the
"skip release" bit so that the release on the driver side can handle the
remaining bias fragments. But this would be a shortsighted solution.
Instead, this patch converges the two XDP paths (XDP_TX and XDP_REDIRECT) to
always do fragment tracking. The "skip release" bit is no longer
necessary for XDP.

Fixes: 6f5742846053 ("net/mlx5e: RX, Enable skb page recycling through the page_pool")
Signed-off-by: Dragos Tatulea <dtatulea@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: Query hca_cap_2 only when supported
Maher Sanalla [Tue, 20 Jun 2023 11:07:03 +0000 (14:07 +0300)]
net/mlx5: Query hca_cap_2 only when supported

On vport enable, where fw's hca caps are queried, the driver queries
hca_caps_2 without checking if fw truly supports them, causing a false
failure of vfs vport load and blocking SRIOV enablement on old devices
such as CX4 where hca_caps_2 support is missing.

Thus, add a check for the said caps support before accessing them.

Fixes: e5b9642a33be ("net/mlx5: E-Switch, Implement devlink port function cmds to control migratable")
Signed-off-by: Maher Sanalla <msanalla@nvidia.com>
Reviewed-by: Shay Drory <shayd@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5e: TC, CT: Offload ct clear only once
Yevgeny Kliteynik [Sun, 4 Jun 2023 09:45:38 +0000 (12:45 +0300)]
net/mlx5e: TC, CT: Offload ct clear only once

Non-clear CT action causes a flow rule split, while CT clear action
doesn't and is just a header-rewrite to the current flow rule.
But ct offload is done in post_parse and is per ct action instance,
so ct clear offload is parsed multiple times, while its deleted once.

Fix this by post_parsing the ct action only once per flow attribute
(which is per flow rule) by using a offloaded ct_attr flag.

Fixes: 08fe94ec5f77 ("net/mlx5e: TC, Remove special handling of CT action")
Signed-off-by: Paul Blakey <paulb@nvidia.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5e: Check for NOT_READY flag state after locking
Vlad Buslov [Thu, 8 Jun 2023 07:32:10 +0000 (09:32 +0200)]
net/mlx5e: Check for NOT_READY flag state after locking

Currently the check for NOT_READY flag is performed before obtaining the
necessary lock. This opens a possibility for race condition when the flow
is concurrently removed from unready_flows list by the workqueue task,
which causes a double-removal from the list and a crash[0]. Fix the issue
by moving the flag check inside the section protected by
uplink_priv->unready_flows_lock mutex.

[0]:
[44376.389654] general protection fault, probably for non-canonical address 0xdead000000000108: 0000 [#1] SMP
[44376.391665] CPU: 7 PID: 59123 Comm: tc Not tainted 6.4.0-rc4+ #1
[44376.392984] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
[44376.395342] RIP: 0010:mlx5e_tc_del_fdb_flow+0xb3/0x340 [mlx5_core]
[44376.396857] Code: 00 48 8b b8 68 ce 02 00 e8 8a 4d 02 00 4c 8d a8 a8 01 00 00 4c 89 ef e8 8b 79 88 e1 48 8b 83 98 06 00 00 48 8b 93 90 06 00 00 <48> 89 42 08 48 89 10 48 b8 00 01 00 00 00 00 ad de 48 89 83 90 06
[44376.399167] RSP: 0018:ffff88812cc97570 EFLAGS: 00010246
[44376.399680] RAX: dead000000000122 RBX: ffff8881088e3800 RCX: ffff8881881bac00
[44376.400337] RDX: dead000000000100 RSI: ffff88812cc97500 RDI: ffff8881242f71b0
[44376.401001] RBP: ffff88811cbb0940 R08: 0000000000000400 R09: 0000000000000001
[44376.401663] R10: 0000000000000001 R11: 0000000000000000 R12: ffff88812c944000
[44376.402342] R13: ffff8881242f71a8 R14: ffff8881222b4000 R15: 0000000000000000
[44376.402999] FS:  00007f0451104800(0000) GS:ffff88852cb80000(0000) knlGS:0000000000000000
[44376.403787] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[44376.404343] CR2: 0000000000489108 CR3: 0000000123a79003 CR4: 0000000000370ea0
[44376.405004] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[44376.405665] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[44376.406339] Call Trace:
[44376.406651]  <TASK>
[44376.406939]  ? die_addr+0x33/0x90
[44376.407311]  ? exc_general_protection+0x192/0x390
[44376.407795]  ? asm_exc_general_protection+0x22/0x30
[44376.408292]  ? mlx5e_tc_del_fdb_flow+0xb3/0x340 [mlx5_core]
[44376.408876]  __mlx5e_tc_del_fdb_peer_flow+0xbc/0xe0 [mlx5_core]
[44376.409482]  mlx5e_tc_del_flow+0x42/0x210 [mlx5_core]
[44376.410055]  mlx5e_flow_put+0x25/0x50 [mlx5_core]
[44376.410529]  mlx5e_delete_flower+0x24b/0x350 [mlx5_core]
[44376.411043]  tc_setup_cb_reoffload+0x22/0x80
[44376.411462]  fl_reoffload+0x261/0x2f0 [cls_flower]
[44376.411907]  ? mlx5e_rep_indr_setup_ft_cb+0x160/0x160 [mlx5_core]
[44376.412481]  ? mlx5e_rep_indr_setup_ft_cb+0x160/0x160 [mlx5_core]
[44376.413044]  tcf_block_playback_offloads+0x76/0x170
[44376.413497]  tcf_block_unbind+0x7b/0xd0
[44376.413881]  tcf_block_setup+0x17d/0x1c0
[44376.414269]  tcf_block_offload_cmd.isra.0+0xf1/0x130
[44376.414725]  tcf_block_offload_unbind+0x43/0x70
[44376.415153]  __tcf_block_put+0x82/0x150
[44376.415532]  ingress_destroy+0x22/0x30 [sch_ingress]
[44376.415986]  qdisc_destroy+0x3b/0xd0
[44376.416343]  qdisc_graft+0x4d0/0x620
[44376.416706]  tc_get_qdisc+0x1c9/0x3b0
[44376.417074]  rtnetlink_rcv_msg+0x29c/0x390
[44376.419978]  ? rep_movs_alternative+0x3a/0xa0
[44376.420399]  ? rtnl_calcit.isra.0+0x120/0x120
[44376.420813]  netlink_rcv_skb+0x54/0x100
[44376.421192]  netlink_unicast+0x1f6/0x2c0
[44376.421573]  netlink_sendmsg+0x232/0x4a0
[44376.421980]  sock_sendmsg+0x38/0x60
[44376.422328]  ____sys_sendmsg+0x1d0/0x1e0
[44376.422709]  ? copy_msghdr_from_user+0x6d/0xa0
[44376.423127]  ___sys_sendmsg+0x80/0xc0
[44376.423495]  ? ___sys_recvmsg+0x8b/0xc0
[44376.423869]  __sys_sendmsg+0x51/0x90
[44376.424226]  do_syscall_64+0x3d/0x90
[44376.424587]  entry_SYSCALL_64_after_hwframe+0x46/0xb0
[44376.425046] RIP: 0033:0x7f045134f887
[44376.425403] Code: 0a 00 f7 d8 64 89 02 48 c7 c0 ff ff ff ff eb b9 0f 1f 00 f3 0f 1e fa 64 8b 04 25 18 00 00 00 85 c0 75 10 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 51 c3 48 83 ec 28 89 54 24 1c 48 89 74 24 10
[44376.426914] RSP: 002b:00007ffd63a82b98 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
[44376.427592] RAX: ffffffffffffffda RBX: 000000006481955f RCX: 00007f045134f887
[44376.428195] RDX: 0000000000000000 RSI: 00007ffd63a82c00 RDI: 0000000000000003
[44376.428796] RBP: 0000000000000000 R08: 0000000000000001 R09: 0000000000000000
[44376.429404] R10: 00007f0451208708 R11: 0000000000000246 R12: 0000000000000001
[44376.430039] R13: 0000000000409980 R14: 000000000047e538 R15: 0000000000485400
[44376.430644]  </TASK>
[44376.430907] Modules linked in: mlx5_ib mlx5_core act_mirred act_tunnel_key cls_flower vxlan dummy sch_ingress openvswitch nsh rpcrdma rdma_ucm ib_iser libiscsi scsi_transport_iscsi ib_umad rdma_cm ib_ipoib iw_cm ib_cm ib_uverbs ib_core xt_conntrack xt_MASQUERADE nf_conntrack_netlink nfnetlink xt_addrtype iptable_nat nf_nat br_netfilter rpcsec_g
ss_krb5 auth_rpcgss oid_registry overlay zram zsmalloc fuse [last unloaded: mlx5_core]
[44376.433936] ---[ end trace 0000000000000000 ]---
[44376.434373] RIP: 0010:mlx5e_tc_del_fdb_flow+0xb3/0x340 [mlx5_core]
[44376.434951] Code: 00 48 8b b8 68 ce 02 00 e8 8a 4d 02 00 4c 8d a8 a8 01 00 00 4c 89 ef e8 8b 79 88 e1 48 8b 83 98 06 00 00 48 8b 93 90 06 00 00 <48> 89 42 08 48 89 10 48 b8 00 01 00 00 00 00 ad de 48 89 83 90 06
[44376.436452] RSP: 0018:ffff88812cc97570 EFLAGS: 00010246
[44376.436924] RAX: dead000000000122 RBX: ffff8881088e3800 RCX: ffff8881881bac00
[44376.437530] RDX: dead000000000100 RSI: ffff88812cc97500 RDI: ffff8881242f71b0
[44376.438179] RBP: ffff88811cbb0940 R08: 0000000000000400 R09: 0000000000000001
[44376.438786] R10: 0000000000000001 R11: 0000000000000000 R12: ffff88812c944000
[44376.439393] R13: ffff8881242f71a8 R14: ffff8881222b4000 R15: 0000000000000000
[44376.439998] FS:  00007f0451104800(0000) GS:ffff88852cb80000(0000) knlGS:0000000000000000
[44376.440714] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[44376.441225] CR2: 0000000000489108 CR3: 0000000123a79003 CR4: 0000000000370ea0
[44376.441843] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[44376.442471] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400

Fixes: ad86755b18d5 ("net/mlx5e: Protect unready flows with dedicated lock")
Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5: Register a unique thermal zone per device
Saeed Mahameed [Tue, 27 Jun 2023 03:36:41 +0000 (20:36 -0700)]
net/mlx5: Register a unique thermal zone per device

Prior to this patch only one "mlx5" thermal zone could have been
registered regardless of the number of individual mlx5 devices in the
system.

To fix this setup a unique name per device to register its own thermal
zone.

In order to not register a thermal zone for a virtual device (VF/SF) add
a check for PF device type.

The new name is a concatenation between "mlx5_" and "<PCI_DEV_BDF>", which
will also help associating a thermal zone with its PCI device.

$ lspci | grep ConnectX
00:04.0 Ethernet controller: Mellanox Technologies MT2892 Family [ConnectX-6 Dx]
00:05.0 Ethernet controller: Mellanox Technologies MT2892 Family [ConnectX-6 Dx]

$ cat /sys/devices/virtual/thermal/thermal_zone0/type
mlx5_0000:00:04.0
$ cat /sys/devices/virtual/thermal/thermal_zone1/type
mlx5_0000:00:05.0

Fixes: c1fef618d611 ("net/mlx5: Implement thermal zone")
CC: Sandipan Patra <spatra@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5e: RX, Fix flush and close release flow of regular rq for legacy rq
Dragos Tatulea [Mon, 22 May 2023 18:18:53 +0000 (21:18 +0300)]
net/mlx5e: RX, Fix flush and close release flow of regular rq for legacy rq

Regular (non-XSK) RQs get flushed on XSK setup and re-activated on XSK
close. If the same regular RQ is closed (a config change for example)
soon after the XSK close, a double release occurs because the missing
wqes get released a second time.

Fixes: 3f93f82988bc ("net/mlx5e: RX, Defer page release in legacy rq for better recycling")
Signed-off-by: Dragos Tatulea <dtatulea@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5e: fix memory leak in mlx5e_ptp_open
Zhengchao Shao [Fri, 30 Jun 2023 01:49:03 +0000 (09:49 +0800)]
net/mlx5e: fix memory leak in mlx5e_ptp_open

When kvzalloc_node or kvzalloc failed in mlx5e_ptp_open, the memory
pointed by "c" or "cparams" is not freed, which can lead to a memory
leak. Fix by freeing the array in the error path.

Fixes: 145e5637d941 ("net/mlx5e: Add TX PTP port object support")
Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com>
Reviewed-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Reviewed-by: Gal Pressman <gal@nvidia.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5e: fix memory leak in mlx5e_fs_tt_redirect_any_create
Zhengchao Shao [Fri, 30 Jun 2023 01:49:02 +0000 (09:49 +0800)]
net/mlx5e: fix memory leak in mlx5e_fs_tt_redirect_any_create

The memory pointed to by the fs->any pointer is not freed in the error
path of mlx5e_fs_tt_redirect_any_create, which can lead to a memory leak.
Fix by freeing the memory in the error path, thereby making the error path
identical to mlx5e_fs_tt_redirect_any_destroy().

Fixes: 0f575c20bf06 ("net/mlx5e: Introduce Flow Steering ANY API")
Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Reviewed-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agonet/mlx5e: fix double free in mlx5e_destroy_flow_table
Zhengchao Shao [Wed, 28 Jun 2023 00:59:34 +0000 (08:59 +0800)]
net/mlx5e: fix double free in mlx5e_destroy_flow_table

In function accel_fs_tcp_create_groups(), when the ft->g memory is
successfully allocated but the 'in' memory fails to be allocated, the
memory pointed to by ft->g is released once. And in function
accel_fs_tcp_create_table, mlx5e_destroy_flow_table is called to release
the memory pointed to by ft->g again. This will cause double free problem.

Fixes: c062d52ac24c ("net/mlx5e: Receive flow steering framework for accelerated TCP flows")
Signed-off-by: Zhengchao Shao <shaozhengchao@huawei.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
17 months agoMerge tag 'soundwire-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul...
Linus Torvalds [Wed, 5 Jul 2023 17:54:43 +0000 (10:54 -0700)]
Merge tag 'soundwire-6.5-rc1' of git://git./linux/kernel/git/vkoul/soundwire

Pull soundwire updates from Vinod Koul:

 - Stream handling and slave alert handling

 - Qualcomm Soundwire v2.0.0 controller support

 - Intel ACE2.x initial support and code reorganization

* tag 'soundwire-6.5-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/soundwire: (55 commits)
  soundwire: stream: Make master_list ordered to prevent deadlocks
  soundwire: bus: Prevent lockdep asserts when stream has multiple buses
  soundwire: qcom: fix storing port config out-of-bounds
  soundwire: intel_ace2x: fix SND_SOC_SOF_HDA_MLINK dependency
  soundwire: debugfs: Add missing SCP registers
  soundwire: stream: Remove unnecessary gotos
  soundwire: stream: Invert logic on runtime alloc flags
  soundwire: stream: Remove unneeded checks for NULL bus
  soundwire: bandwidth allocation: Remove pointless variable
  soundwire: cadence: revisit parity injection
  soundwire: intel/cadence: update hardware reset sequence
  soundwire: intel_bus_common: enable interrupts last
  soundwire: intel_bus_common: update error log
  soundwire: amd: Improve error message in remove callback
  soundwire: debugfs: fix unbalanced pm_runtime_put()
  soundwire: qcom: fix unbalanced pm_runtime_put()
  soundwire: qcom: set clk stop need reset flag at runtime
  soundwire: qcom: add software workaround for bus clash interrupt assertion
  soundwire: qcom: wait for fifo to be empty before suspend
  soundwire: qcom: drop unused struct qcom_swrm_ctrl members
  ...

17 months agoMerge tag 'media/v6.5-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab...
Linus Torvalds [Wed, 5 Jul 2023 17:42:32 +0000 (10:42 -0700)]
Merge tag 'media/v6.5-1' of git://git./linux/kernel/git/mchehab/linux-media

Pull media updates from Mauro Carvalho Chehab:

 - Lots of improvement at atomisp driver, which is starting to look in
   good shape

 - Mediatek vcodec driver has gained support for av1 and hevc stateless
   codecs

 - New sensor driver: ov01a10

 - verisilicon driver has gained AV1 entropy helpers

 - tegra-video has gained support for Tegra20 parallel input

 - dvb core has gained an extra property to better support DVB-S2X

 - as usual, lots of cleanups, fixes and improvements on media drivers

* tag 'media/v6.5-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media: (253 commits)
  media: wl128x: fix a clang warning
  media: dvb: mb86a20s: get rid of a clang-15 warning
  media: cec: i2c: ch7322: also select REGMAP
  media: add HAS_IOPORT dependencies
  media: tc358746: select CONFIG_GENERIC_PHY
  media: mediatek: vcodec: Add dbgfs help function
  media: mediatek: vcodec: Add encode to support dbgfs
  media: mediatek: vcodec: Change dbgfs interface to support encode
  media: mediatek: vcodec: Get each instance format type
  media: mediatek: vcodec: Get each context resolution information
  media: mediatek: vcodec: Add a debugfs file to get different useful information
  media: mediatek: vcodec: Add debug params to control different log level
  media: mediatek: vcodec: Add debugfs interface to get debug information
  media: mediatek: vcodec: support stateless AV1 decoder
  media: verisilicon: Conditionally ignore native formats
  media: verisilicon: Enable AV1 decoder on rk3588
  media: verisilicon: Add film grain feature to AV1 driver
  media: verisilicon: Add Rockchip AV1 decoder
  media: verisilicon: Add AV1 entropy helpers
  media: verisilicon: Compute motion vectors size for AV1 frames
  ...

17 months agoigc: Fix TX Hang issue when QBV Gate is closed
Muhammad Husaini Zulkifli [Sat, 3 Jun 2023 12:59:34 +0000 (20:59 +0800)]
igc: Fix TX Hang issue when QBV Gate is closed

If a user schedules a Gate Control List (GCL) to close one of
the QBV gates while also transmitting a packet to that closed gate,
TX Hang will be happen. HW would not drop any packet when the gate
is closed and keep queuing up in HW TX FIFO until the gate is re-opened.
This patch implements the solution to drop the packet for the closed
gate.

This patch will also reset the adapter to perform SW initialization
for each 1st Gate Control List (GCL) to avoid hang.
This is due to the HW design, where changing to TSN transmit mode
requires SW initialization. Intel Discrete I225/6 transmit mode
cannot be changed when in dynamic mode according to Software User
Manual Section 7.5.2.1. Subsequent Gate Control List (GCL) operations
will proceed without a reset, as they already are in TSN Mode.

Step to reproduce:

DUT:
1) Configure GCL List with certain gate close.

BASE=$(date +%s%N)
tc qdisc replace dev $IFACE parent root handle 100 taprio \
    num_tc 4 \
    map 0 1 2 3 3 3 3 3 3 3 3 3 3 3 3 3 \
    queues 1@0 1@1 1@2 1@3 \
    base-time $BASE \
    sched-entry S 0x8 500000 \
    sched-entry S 0x4 500000 \
    flags 0x2

2) Transmit the packet to closed gate. You may use udp_tai
application to transmit UDP packet to any of the closed gate.

./udp_tai -i <interface> -P 100000 -p 90 -c 1 -t <0/1> -u 30004

Fixes: ec50a9d437f0 ("igc: Add support for taprio offloading")
Co-developed-by: Tan Tee Min <tee.min.tan@linux.intel.com>
Signed-off-by: Tan Tee Min <tee.min.tan@linux.intel.com>
Tested-by: Chwee Lin Choong <chwee.lin.choong@intel.com>
Signed-off-by: Muhammad Husaini Zulkifli <muhammad.husaini.zulkifli@intel.com>
Tested-by: Naama Meir <naamax.meir@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
17 months agoMerge tag 'trace-tools-v6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/trace...
Linus Torvalds [Wed, 5 Jul 2023 17:34:30 +0000 (10:34 -0700)]
Merge tag 'trace-tools-v6.5' of git://git./linux/kernel/git/trace/linux-trace

Pull tracing tooling updates from Steven Rostedt:

 - Add cgroup support for rtla via the -C option

 - Add --house-keeping option that tells rtla where to place the
   housekeeping threads

 - Have rtla/timerlat have its own tracing instance instead of using the
   top level tracing instance that is the default for other tracing
   users to use

 - Add auto analysis to timerlat_hist

 - Have rtla start the tracers after creating the instances

 - Reduce rtla hwnoise down to 75% from 100% as it runs with preemption
   disabled and can cause system instability at 100%

 - Add support to run timerlat_top and timerlat_hist threads in
   user-space instead of just using the kernel tasks

 - Some minor clean ups and documentation changes

* tag 'trace-tools-v6.5' of git://git.kernel.org/pub/scm/linux/kernel/git/trace/linux-trace:
  Documentation: Add tools/rtla timerlat -u option documentation
  rtla/timerlat_hist: Add timerlat user-space support
  rtla/timerlat_top: Add timerlat user-space support
  rtla/hwnoise: Reduce runtime to 75%
  rtla: Start the tracers after creating all instances
  rtla/timerlat_hist: Add auto-analysis support
  rtla/timerlat: Give timerlat auto analysis its own instance
  rtla: Automatically move rtla to a house-keeping cpu
  rtla: Change monitored_cpus from char * to cpu_set_t
  rtla: Add --house-keeping option
  rtla: Add -C cgroup support

17 months agoMerge tag 'parisc-for-6.5-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Wed, 5 Jul 2023 17:28:38 +0000 (10:28 -0700)]
Merge tag 'parisc-for-6.5-rc1-2' of git://git./linux/kernel/git/deller/parisc-linux

Pull more parisc architecture updates from Helge Deller:

 -  Fix all compiler warnings in arch/parisc and drivers/parisc when
    compiled with W=1

* tag 'parisc-for-6.5-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:
  parisc: syscalls: Avoid compiler warnings with W=1
  parisc: math-emu: Avoid compiler warnings with W=1
  parisc: Raise minimal GCC version to 12.0.0
  parisc: unwind: Avoid missing prototype warning for handle_interruption()
  parisc: smp: Add declaration for start_cpu_itimer()
  parisc: pdt: Get prototype for arch_report_meminfo()

17 months agoigc: Remove delay during TX ring configuration
Muhammad Husaini Zulkifli [Wed, 17 May 2023 00:18:12 +0000 (08:18 +0800)]
igc: Remove delay during TX ring configuration

Remove unnecessary delay during the TX ring configuration.
This will cause delay, especially during link down and
link up activity.

Furthermore, old SKUs like as I225 will call the reset_adapter
to reset the controller during TSN mode Gate Control List (GCL)
setting. This will add more time to the configuration of the
real-time use case.

It doesn't mentioned about this delay in the Software User Manual.
It might have been ported from legacy code I210 in the past.

Fixes: 13b5b7fd6a4a ("igc: Add support for Tx/Rx rings")
Signed-off-by: Muhammad Husaini Zulkifli <muhammad.husaini.zulkifli@intel.com>
Acked-by: Sasha Neftin <sasha.neftin@intel.com>
Tested-by: Naama Meir <naamax.meir@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
17 months agoigc: Add condition for qbv_config_change_errors counter
Muhammad Husaini Zulkifli [Mon, 15 May 2023 06:03:36 +0000 (14:03 +0800)]
igc: Add condition for qbv_config_change_errors counter

Add condition to increase the qbv counter during taprio qbv
configuration only.

There might be a case when TC already been setup then user configure
the ETF/CBS qdisc and this counter will increase if no condition above.

Fixes: ae4fe4698300 ("igc: Add qbv_config_change_errors counter")
Signed-off-by: Muhammad Husaini Zulkifli <muhammad.husaini.zulkifli@intel.com>
Tested-by: Naama Meir <naamax.meir@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
17 months agosh: Provide unxlate_dev_mem_ptr() in asm/io.h
Guenter Roeck [Tue, 4 Jul 2023 19:01:44 +0000 (12:01 -0700)]
sh: Provide unxlate_dev_mem_ptr() in asm/io.h

The unxlate_dev_mem_ptr() function has no prototype on the sh architecture
which does not include asm-generic/io.h. This results in the following
build failure:

drivers/char/mem.c: In function 'read_mem':
drivers/char/mem.c:164:25: error: implicit declaration of function 'unxlate_dev_mem_ptr'

This compile error is now seen because commit 99b619b37ae1 ("mips: provide
unxlate_dev_mem_ptr() in asm/io.h") removed the weak function which was
previously in place to handle this problem.

Add a trivial macro to the sh header to provide the now missing dummy
function.

Fixes: 99b619b37ae1 ("mips: provide unxlate_dev_mem_ptr() in asm/io.h")
Cc: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Link: https://lore.kernel.org/r/20230704190144.2888679-1-linux@roeck-us.net
Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
17 months agosh: dma: Correct the number of DMA channels for SH7709
Artur Rojek [Sat, 27 May 2023 16:44:52 +0000 (18:44 +0200)]
sh: dma: Correct the number of DMA channels for SH7709

According to the hardware manual [1], the DMAC found in the SH7709 SoC
features only 4 channels. While at it, also sort the existing targets.

[1] https://www.renesas.com/us/en/document/mah/sh7709s-group-hardware-manual (p. 373)

Signed-off-by: Artur Rojek <contact@artur-rojek.eu>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Link: https://lore.kernel.org/r/20230527164452.64797-4-contact@artur-rojek.eu
Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
17 months agosh: dma: Drop incorrect SH_DMAC_BASE1 definition for SH4
Artur Rojek [Sat, 27 May 2023 16:44:51 +0000 (18:44 +0200)]
sh: dma: Drop incorrect SH_DMAC_BASE1 definition for SH4

None of the supported SH4 family SoCs features a second DMAC module. As
this definition negatively impacts DMA channel calculation for the above
targets, remove it from the code.

Signed-off-by: Artur Rojek <contact@artur-rojek.eu>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Link: https://lore.kernel.org/r/20230527164452.64797-3-contact@artur-rojek.eu
Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
17 months agosh: dma: Fix DMA channel offset calculation
Artur Rojek [Sat, 27 May 2023 16:44:50 +0000 (18:44 +0200)]
sh: dma: Fix DMA channel offset calculation

Various SoCs of the SH3, SH4 and SH4A family, which use this driver,
feature a differing number of DMA channels, which can be distributed
between up to two DMAC modules. The existing implementation fails to
correctly accommodate for all those variations, resulting in wrong
channel offset calculations and leading to kernel panics.

Rewrite dma_base_addr() in order to properly calculate channel offsets
in a DMAC module. Fix dmaor_read_reg() and dmaor_write_reg(), so that
the correct DMAC module base is selected for the DMAOR register.

Fixes: 7f47c7189b3e8f19 ("sh: dma: More legacy cpu dma chainsawing.")
Signed-off-by: Artur Rojek <contact@artur-rojek.eu>
Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Link: https://lore.kernel.org/r/20230527164452.64797-2-contact@artur-rojek.eu
Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
17 months agosh: Remove compiler flag duplication
Masahiro Yamada [Sun, 19 Feb 2023 14:15:55 +0000 (23:15 +0900)]
sh: Remove compiler flag duplication

Every compiler flag added by arch/sh/Makefile is passed to the
compiler twice:

$(KBUILD_CPPFLAGS) + $(KBUILD_CFLAGS) is used for compiling *.c
$(KBUILD_CPPFLAGS) + $(KBUILD_AFLAGS) is used for compiling *.S

Given the above, adding $(cflags-y) to all of KBUILD_{CPP/C/A}FLAGS
ends up with duplication.

Add -I options to $(KBUILD_CPPFLAGS), and the rest of $(cflags-y)
to KBUILD_{C,A}FLAGS.

Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Tested-by: Randy Dunlap <rdunlap@infradead.org>
Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Link: https://lore.kernel.org/r/20230219141555.2308306-4-masahiroy@kernel.org
Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
17 months agosh: Refactor header include path addition
Masahiro Yamada [Sun, 19 Feb 2023 14:15:54 +0000 (23:15 +0900)]
sh: Refactor header include path addition

Shorten the code. No functional change intended.

Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Tested-by: Randy Dunlap <rdunlap@infradead.org>
Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Link: https://lore.kernel.org/r/20230219141555.2308306-3-masahiroy@kernel.org
Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
17 months agosh: Move build rule for cchips/hd6446x/ to arch/sh/Kbuild
Masahiro Yamada [Sun, 19 Feb 2023 14:15:53 +0000 (23:15 +0900)]
sh: Move build rule for cchips/hd6446x/ to arch/sh/Kbuild

This is the last user of core-y in arch/sh.

Use the standard obj-y syntax.

Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Tested-by: Randy Dunlap <rdunlap@infradead.org>
Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Link: https://lore.kernel.org/r/20230219141555.2308306-2-masahiroy@kernel.org
Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
17 months agosh: Fix -Wmissing-include-dirs warnings for various platforms
Masahiro Yamada [Sun, 19 Feb 2023 14:15:52 +0000 (23:15 +0900)]
sh: Fix -Wmissing-include-dirs warnings for various platforms

The 0day bot reports a lot of warnings (or errors due to CONFIG_WERROR)
like this:

  cc1: error: arch/sh/include/mach-hp6xx: No such file or directory [-Werror=missing-include-dirs]

Indeed, arch/sh/include/mach-hp6xx does not exist.

While -Wmissing-include-dirs is only a W=1 warning, it may be
annoying when CONFIG_BTRFS_FS is enabled because fs/btrfs/Makefile
unconditionally adds this warning option.

arch/sh/Makefile defines machdir-y for two purposes:

 - Build platform code in arch/sh/boards/mach-*/
 - Add arch/sh/include/mach-*/ to the header search path

For the latter, some platforms use arch/sh/include/mach-common/
instead of having its own arch/sh/include/mach-*/.

Drop unneeded machdir-y to omit non-existing include directories.

To build arch/sh/boards/mach-*/, use the standard obj-y syntax in
arch/sh/boards/Makefile.

Reported-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/oe-kbuild-all/202302190641.30VVXnPb-lkp@intel.com/
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Acked-by: Randy Dunlap <rdunlap@infradead.org>
Tested-by: Randy Dunlap <rdunlap@infradead.org>
Reviewed-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
Link: https://lore.kernel.org/r/20230219141555.2308306-1-masahiroy@kernel.org
Signed-off-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de>
17 months agogup: make the stack expansion warning a bit more targeted
Linus Torvalds [Wed, 5 Jul 2023 16:33:31 +0000 (09:33 -0700)]
gup: make the stack expansion warning a bit more targeted

I added a warning about about GUP no longer expanding the stack in
commit a425ac5365f6 ("gup: add warning if some caller would seem to want
stack expansion"), but didn't really expect anybody to hit it.

And it's true that nobody seems to have hit a _real_ case yet, but we
certainly have a number of reports of false positives.  Which not only
causes extra noise in itself, but might also end up hiding any real
cases if they do exist.

So let's tighten up the warning condition, and replace the simplistic

vma = find_vma(mm, start);
if (vma && (start < vma->vm_start)) {
WARN_ON_ONCE(vma->vm_flags & VM_GROWSDOWN);

with a

vma = gup_vma_lookup(mm, start);

helper function which works otherwise like just "vma_lookup()", but with
some heuristics for when to warn about gup no longer causing stack
expansion.

In particular, don't just warn for "below the stack", but warn if it's
_just_ below the stack (with "just below" arbitrarily defined as 64kB,
because why not?).  And rate-limit it to at most once per hour, which
means that any false positives shouldn't completely hide subsequent
reports, but we won't be flooding the logs about it either.

The previous code triggered when some GUP user (chromium crashpad)
accessing past the end of the previous vma, for example.  That has never
expanded the stack, it just causes GUP to return early, and as such we
shouldn't be warning about it.

This is still going trigger the randomized testers, but to mitigate the
noise from that, use "dump_stack()" instead of "WARN_ON_ONCE()" to get
the kernel call chain.  We'll get the relevant information, but syzbot
shouldn't get too upset about it.

Also, don't even bother with the GROWSUP case, which would be using
different heuristics entirely, but only happens on parisc.

Reported-by: kernel test robot <oliver.sang@intel.com>
Reported-by: John Hubbard <jhubbard@nvidia.com>
Reported-by: syzbot+6cf44e127903fdf9d929@syzkaller.appspotmail.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
17 months agoice: Fix tx queue rate limit when TCs are configured
Sridhar Samudrala [Sat, 10 Jun 2023 00:40:24 +0000 (17:40 -0700)]
ice: Fix tx queue rate limit when TCs are configured

Configuring tx_maxrate via sysfs interface
/sys/class/net/eth0/queues/tx-1/tx_maxrate was not working when
TCs are configured because always main VSI was being used. Fix by
using correct VSI in ice_set_tx_maxrate when TCs are configured.

Fixes: 1ddef455f4a8 ("ice: Add NDO callback to set the maximum per-queue bitrate")
Signed-off-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
Signed-off-by: Sudheer Mogilappagari <sudheer.mogilappagari@intel.com>
Tested-by: Bharathi Sreenivas <bharathi.sreenivas@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
17 months agoice: Fix max_rate check while configuring TX rate limits
Sridhar Samudrala [Sat, 10 Jun 2023 00:40:23 +0000 (17:40 -0700)]
ice: Fix max_rate check while configuring TX rate limits

Remove incorrect check in ice_validate_mqprio_opt() that limits
filter configuration when sum of max_rates of all TCs exceeds
the link speed. The max rate of each TC is unrelated to value
used by other TCs and is valid as long as it is less than link
speed.

Fixes: fbc7b27af0f9 ("ice: enable ndo_setup_tc support for mqprio_qdisc")
Signed-off-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
Signed-off-by: Sudheer Mogilappagari <sudheer.mogilappagari@intel.com>
Tested-by: Bharathi Sreenivas <bharathi.sreenivas@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
17 months agodt-bindings: soc: qcom: stats: Update maintainer email
Maulik Shah [Mon, 3 Jul 2023 09:20:26 +0000 (14:50 +0530)]
dt-bindings: soc: qcom: stats: Update maintainer email

Replace my email.

Cc: devicetree@vger.kernel.org
Signed-off-by: Maulik Shah <quic_mkshah@quicinc.com>
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20230703092026.4923-1-quic_mkshah@quicinc.com
Signed-off-by: Rob Herring <robh@kernel.org>
17 months agodt-bindings: cleanup DTS example whitespaces
Krzysztof Kozlowski [Sun, 2 Jul 2023 18:23:08 +0000 (20:23 +0200)]
dt-bindings: cleanup DTS example whitespaces

The DTS code coding style expects spaces around '=' sign.

Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Matthias Brugger <matthias.bgg@gmail.com>
Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Conor Dooley <conor.dooley@microchip.com>
Acked-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org> #display/msm
Acked-by: Neil Armstrong <neil.armstrong@linaro.org>
Acked-by: Mike Leach <mike.leach@linaro.org>
Reviewed-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Acked-by: Vinod Koul <vkoul@kernel.org>
Link: https://lore.kernel.org/r/20230702182308.7583-1-krzysztof.kozlowski@linaro.org
Signed-off-by: Rob Herring <robh@kernel.org>
17 months agotracing/boot: Test strscpy() against less than zero for error
Steven Rostedt (Google) [Tue, 4 Jul 2023 14:08:07 +0000 (10:08 -0400)]
tracing/boot: Test strscpy() against less than zero for error

Instead of checking for -E2BIG, it is better to just check for less than
zero of strscpy() for error. Testing for -E2BIG is not very robust, and
the calling code does not really care about the error code, just that
there was an error.

One of the updates to convert strlcpy() to strscpy() had a v2 version
that changed the test from testing against -E2BIG to less than zero, but I
took the v1 version that still tested for -E2BIG.

Link: https://lore.kernel.org/linux-trace-kernel/20230615180420.400769-1-azeemshaikh38@gmail.com/
Link: https://lore.kernel.org/linux-trace-kernel/20230704100807.707d1605@rorschach.local.home
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Azeem Shaikh <azeemshaikh38@gmail.com>
Cc: Kees Cook <keescook@chromium.org>
Acked-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
17 months agorisc-v: Fix order of IPI enablement vs RCU startup
Marc Zyngier [Mon, 3 Jul 2023 18:31:26 +0000 (19:31 +0100)]
risc-v: Fix order of IPI enablement vs RCU startup

Conor reports that risc-v tries to enable IPIs before telling the
core code to enable RCU. With the introduction of the mapple tree
as a backing store for the irq descriptors, this results in
a very shouty boot sequence, as RCU is legitimately upset.

Restore some sanity by moving the risc_ipi_enable() call after
notify_cpu_starting(), which explicitly enables RCU on the calling
CPU.

Fixes: 832f15f42646 ("RISC-V: Treat IPIs as normal Linux IRQs")
Reported-by: Conor Dooley <conor@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20230703-dupe-frying-79ae2ccf94eb@spud
Cc: Anup Patel <apatel@ventanamicro.com>
Cc: Palmer Dabbelt <palmer@rivosinc.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Tested-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20230703183126.1567625-1-maz@kernel.org
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
17 months agomm: riscv: fix an unsafe pte read in huge_pte_alloc()
John Hubbard [Mon, 3 Jul 2023 19:00:44 +0000 (12:00 -0700)]
mm: riscv: fix an unsafe pte read in huge_pte_alloc()

The WARN_ON_ONCE() statement in riscv's huge_pte_alloc() is susceptible
to false positives, because the pte is read twice at the C language
level, locklessly, within the same conditional statement. Depending on
compiler behavior, this can lead to generated machine code that actually
reads the pte just once, or twice. Reading twice will expose the code to
changing pte values and cause incorrect behavior.

In [1], similar code actually caused a kernel crash on 64-bit x86, when
using clang to build the kernel, but only after the conversion from *pte
reads, to ptep_get(pte). The latter uses READ_ONCE(), which forced a
double read of *pte.

Rather than waiting for the upcoming ptep_get() conversion, just convert
this part of the code now, but in a way that avoids the above problem:
take a single snapshot of the pte before using it in the WARN
conditional.

As expected, this preparatory step does not actually change the
generated code ("make mm/hugetlbpage.s"), on riscv64, when using a gcc
12.2 cross compiler.

[1] https://lore.kernel.org/20230630013203.1955064-1-jhubbard@nvidia.com

Suggested-by: James Houghton <jthoughton@google.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Reviewed-by: Andrew Jones <ajones@ventanamicro.com>
Reviewed-by: Ryan Roberts <ryan.roberts@arm.com>
Link: https://lore.kernel.org/r/20230703190044.311730-1-jhubbard@nvidia.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
17 months agodt-bindings: riscv: deprecate riscv,isa
Conor Dooley [Sat, 1 Jul 2023 23:10:01 +0000 (00:10 +0100)]
dt-bindings: riscv: deprecate riscv,isa

intro
=====

When the RISC-V dt-bindings were accepted upstream in Linux, the base
ISA etc had yet to be ratified. By the ratification of the base ISA,
incompatible changes had snuck into the specifications - for example the
Zicsr and Zifencei extensions were spun out of the base ISA.

Fast forward to today, and the reason for this patch.
Currently the riscv,isa dt property permits only a specific subset of
the ISA string - in particular it excludes version numbering.
With the current constraints, it is not possible to discern whether
"rv64i" means that the hart supports the fence.i instruction, for
example.
Future systems may choose to implement their own instruction fencing,
perhaps using a vendor extension, or they may not implement the optional
counter extensions. Software needs a way to determine this.

versioning schemes
==================

"Use the extension versions that are described in the ISA manual" you
may say, and it's not like this has not been considered.
Firstly, software that parses the riscv,isa property at runtime will
need to contain a lookup table of some sort that maps arbitrary versions
to versions it understands. There is not a consistent application of
version number applied to extensions, with a higgledy-piggledy
collection of tags, "bare" and versioned documents awaiting the reader
on the "recently ratified extensions" page:
https://wiki.riscv.org/display/HOME/Recently+Ratified+Extensions

As an aside, and this is reflected in the patch too, since many
extensions have yet to appear in a release of the ISA specs,
they are defined by commits in their respective "working draft"
repositories.

Secondly, there is an issue of backwards compatibility, whereby allowing
numbers in the ISA string, some parsers may be broken. This would
require an additional property to be created to even use the versions in
this manner.

~boolean properties~ string array property
==========================================

If a new property is needed, the whole approach may as well be looked at
from the bottom up. A string with limited character choices etc is
hardly the best approach for communicating extension information to
software.

Switching to using properties that are defined on a per extension basis,
allows us to define explicit meanings for the DT representation of each
extension - rather than the current situation where different operating
systems or other bits of software may impart different meanings to
characters in the string.
Clearly the best source of meanings is the specifications themselves,
this just provides us the ability to choose at what point in time the
meaning is set. If an extension changes incompatibility in the future,
a new property will be required.

Off-list, some of the RVI folks have committed to shoring up the wording
in either the ISA specifications, the riscv-isa-manual or
so that in the future, modifications to and additions or removals of
features will require a new extension. Codifying that assertion
somewhere would make it quite unlikely that compatibility would be
broken, but we have the tools required to deal with it, if & when it
crops up.
It is in our collective interest, as consumers of extension meanings, to
define a scheme that enforces compatibility.

The use of individual elements, rather than a single string, will also
permit validation that the properties have a meaning, as well as
potentially reject mutually exclusive combinations, or enforce
dependencies between extensions. That would not have be possible with
the current dt-schema infrastructure for arbitrary strings, as we would
need to add a riscv,isa parser to dt-validate!
That's not implemented in this patch, but rather left as future work (for
the brave, or the foolish).

parser simplicity
=================

Many systems that parse DT at runtime already implement an function that
can check for the presence of a string in an array of string, as it is
similar to the process for parsing a list of compatible strings, so a
bunch of new, custom, DT parsing should not be needed.
Getting rid of "riscv,isa" parsing would be a nice simplification, but
unfortunately for backwards compatibility with old dtbs, existing
parsers may not be removable - which may greatly simplify
dt parsing code. In Linux, for example, checking for whether a hart
supports an extension becomes as simple as:
of_property_match_string(node, "riscv,isa-extensions", "zicbom")

vendor extensions
=================

Compared to riscv,isa, this proposed scheme promotes vendor extensions,
oft touted as the strength of RISC-V, to first-class citizens.
At present, extensions are defined as meaning what the RISC-V ISA
specifications say they do. There is no realistic way of using that
interface to provide cross-platform definitions for what vendor
extensions mean. Vendor extensions may also have even less consistency
than RVI do in terms of versioning, or no care about backwards
compatibility.
The new property allows us to assign explicit meanings on a per vendor
extension basis, backed up by a description of their meanings.

fin
===

Create a new file to store the extension meanings and a new
riscv,isa-base property to replace the aspect of riscv,isa that is
not represented by the new property - the base ISA implemented by a hart.

As a starting point, add properties for extensions currently used in
Linux.

Finally, mark riscv,isa as deprecated, as removing support for it in
existing programs would be an ABI break.

CC: Palmer Dabbelt <palmer@dabbelt.com>
CC: Paul Walmsley <paul.walmsley@sifive.com>
CC: Rob Herring <robh+dt@kernel.org>
CC: Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>
CC: Alistair Francis <alistair.francis@wdc.com>
CC: Andrew Jones <ajones@ventanamicro.com>
CC: Anup Patel <apatel@ventanamicro.com>
CC: Atish Patra <atishp@atishpatra.org>
CC: Jessica Clarke <jrtc27@jrtc27.com>
CC: Rick Chen <rick@andestech.com>
CC: Leo <ycliang@andestech.com>
CC: Oleksii <oleksii.kurochko@gmail.com>
CC: linux-riscv@lists.infradead.org
CC: qemu-riscv@nongnu.org
CC: u-boot@lists.denx.de
CC: devicetree@vger.kernel.org
CC: linux-kernel@vger.kernel.org
Reviewed-by: Palmer Dabbelt <palmer@rivosinc.com>
Acked-by: Palmer Dabbelt <palmer@rivosinc.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Conor Dooley <conor.dooley@microchip.com>
Link: https://lore.kernel.org/r/20230702-eats-scorebook-c951f170d29f@spud
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
17 months agoarm64: ftrace: fix build error with CONFIG_FUNCTION_GRAPH_TRACER=n
Arnd Bergmann [Fri, 23 Jun 2023 15:21:55 +0000 (17:21 +0200)]
arm64: ftrace: fix build error with CONFIG_FUNCTION_GRAPH_TRACER=n

It appears that a merge conflict ended up hiding a newly added constant
in some configurations:

arch/arm64/kernel/entry-ftrace.S: Assembler messages:
arch/arm64/kernel/entry-ftrace.S:59: Error: undefined symbol FTRACE_OPS_DIRECT_CALL used as an immediate value

FTRACE_OPS_DIRECT_CALL is still used when CONFIG_DYNAMIC_FTRACE_WITH_DIRECT_CALLS
is enabled, even if CONFIG_FUNCTION_GRAPH_TRACER is disabled, so change the
ifdef accordingly.

Link: https://lkml.kernel.org/r/20230623152204.2216297-1-arnd@kernel.org
Cc: Will Deacon <will@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Donglin Peng <pengdonglin@sangfor.com.cn>
Fixes: 3646970322464 ("arm64: ftrace: Enable HAVE_FUNCTION_GRAPH_RETVAL")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Florent Revest <revest@chromium.org>
Reviewed-by: Randy Dunlap <rdunlap@infradead.org>
Tested-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
17 months agotracing: Fix null pointer dereference in tracing_err_log_open()
Mateusz Stachyra [Tue, 4 Jul 2023 10:27:06 +0000 (12:27 +0200)]
tracing: Fix null pointer dereference in tracing_err_log_open()

Fix an issue in function 'tracing_err_log_open'.
The function doesn't call 'seq_open' if the file is opened only with
write permissions, which results in 'file->private_data' being left as null.
If we then use 'lseek' on that opened file, 'seq_lseek' dereferences
'file->private_data' in 'mutex_lock(&m->lock)', resulting in a kernel panic.
Writing to this node requires root privileges, therefore this bug
has very little security impact.

Tracefs node: /sys/kernel/tracing/error_log

Example Kernel panic:

Unable to handle kernel NULL pointer dereference at virtual address 0000000000000038
Call trace:
 mutex_lock+0x30/0x110
 seq_lseek+0x34/0xb8
 __arm64_sys_lseek+0x6c/0xb8
 invoke_syscall+0x58/0x13c
 el0_svc_common+0xc4/0x10c
 do_el0_svc+0x24/0x98
 el0_svc+0x24/0x88
 el0t_64_sync_handler+0x84/0xe4
 el0t_64_sync+0x1b4/0x1b8
Code: d503201f aa0803e0 aa1f03e1 aa0103e9 (c8e97d02)
---[ end trace 561d1b49c12cf8a5 ]---
Kernel panic - not syncing: Oops: Fatal exception

Link: https://lore.kernel.org/linux-trace-kernel/20230703155237eucms1p4dfb6a19caa14c79eb6c823d127b39024@eucms1p4
Link: https://lore.kernel.org/linux-trace-kernel/20230704102706eucms1p30d7ecdcc287f46ad67679fc8491b2e0f@eucms1p3
Cc: stable@vger.kernel.org
Fixes: 8a062902be725 ("tracing: Add tracing error log")
Signed-off-by: Mateusz Stachyra <m.stachyra@samsung.com>
Suggested-by: Steven Rostedt <rostedt@goodmis.org>
Acked-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>
17 months agonetfilter: nf_tables: do not ignore genmask when looking up chain by id
Thadeu Lima de Souza Cascardo [Wed, 5 Jul 2023 12:12:55 +0000 (09:12 -0300)]
netfilter: nf_tables: do not ignore genmask when looking up chain by id

When adding a rule to a chain referring to its ID, if that chain had been
deleted on the same batch, the rule might end up referring to a deleted
chain.

This will lead to a WARNING like following:

[   33.098431] ------------[ cut here ]------------
[   33.098678] WARNING: CPU: 5 PID: 69 at net/netfilter/nf_tables_api.c:2037 nf_tables_chain_destroy+0x23d/0x260
[   33.099217] Modules linked in:
[   33.099388] CPU: 5 PID: 69 Comm: kworker/5:1 Not tainted 6.4.0+ #409
[   33.099726] Workqueue: events nf_tables_trans_destroy_work
[   33.100018] RIP: 0010:nf_tables_chain_destroy+0x23d/0x260
[   33.100306] Code: 8b 7c 24 68 e8 64 9c ed fe 4c 89 e7 e8 5c 9c ed fe 48 83 c4 08 5b 41 5c 41 5d 41 5e 41 5f 5d 31 c0 89 c6 89 c7 c3 cc cc cc cc <0f> 0b 48 83 c4 08 5b 41 5c 41 5d 41 5e 41 5f 5d 31 c0 89 c6 89 c7
[   33.101271] RSP: 0018:ffffc900004ffc48 EFLAGS: 00010202
[   33.101546] RAX: 0000000000000001 RBX: ffff888006fc0a28 RCX: 0000000000000000
[   33.101920] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000000
[   33.102649] RBP: ffffc900004ffc78 R08: 0000000000000000 R09: 0000000000000000
[   33.103018] R10: 0000000000000000 R11: 0000000000000000 R12: ffff8880135ef500
[   33.103385] R13: 0000000000000000 R14: dead000000000122 R15: ffff888006fc0a10
[   33.103762] FS:  0000000000000000(0000) GS:ffff888024c80000(0000) knlGS:0000000000000000
[   33.104184] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   33.104493] CR2: 00007fe863b56a50 CR3: 00000000124b0001 CR4: 0000000000770ee0
[   33.104872] PKRU: 55555554
[   33.104999] Call Trace:
[   33.105113]  <TASK>
[   33.105214]  ? show_regs+0x72/0x90
[   33.105371]  ? __warn+0xa5/0x210
[   33.105520]  ? nf_tables_chain_destroy+0x23d/0x260
[   33.105732]  ? report_bug+0x1f2/0x200
[   33.105902]  ? handle_bug+0x46/0x90
[   33.106546]  ? exc_invalid_op+0x19/0x50
[   33.106762]  ? asm_exc_invalid_op+0x1b/0x20
[   33.106995]  ? nf_tables_chain_destroy+0x23d/0x260
[   33.107249]  ? nf_tables_chain_destroy+0x30/0x260
[   33.107506]  nf_tables_trans_destroy_work+0x669/0x680
[   33.107782]  ? mark_held_locks+0x28/0xa0
[   33.107996]  ? __pfx_nf_tables_trans_destroy_work+0x10/0x10
[   33.108294]  ? _raw_spin_unlock_irq+0x28/0x70
[   33.108538]  process_one_work+0x68c/0xb70
[   33.108755]  ? lock_acquire+0x17f/0x420
[   33.108977]  ? __pfx_process_one_work+0x10/0x10
[   33.109218]  ? do_raw_spin_lock+0x128/0x1d0
[   33.109435]  ? _raw_spin_lock_irq+0x71/0x80
[   33.109634]  worker_thread+0x2bd/0x700
[   33.109817]  ? __pfx_worker_thread+0x10/0x10
[   33.110254]  kthread+0x18b/0x1d0
[   33.110410]  ? __pfx_kthread+0x10/0x10
[   33.110581]  ret_from_fork+0x29/0x50
[   33.110757]  </TASK>
[   33.110866] irq event stamp: 1651
[   33.111017] hardirqs last  enabled at (1659): [<ffffffffa206a209>] __up_console_sem+0x79/0xa0
[   33.111379] hardirqs last disabled at (1666): [<ffffffffa206a1ee>] __up_console_sem+0x5e/0xa0
[   33.111740] softirqs last  enabled at (1616): [<ffffffffa1f5d40e>] __irq_exit_rcu+0x9e/0xe0
[   33.112094] softirqs last disabled at (1367): [<ffffffffa1f5d40e>] __irq_exit_rcu+0x9e/0xe0
[   33.112453] ---[ end trace 0000000000000000 ]---

This is due to the nft_chain_lookup_byid ignoring the genmask. After this
change, adding the new rule will fail as it will not find the chain.

Fixes: 837830a4b439 ("netfilter: nf_tables: add NFTA_RULE_CHAIN_ID attribute")
Cc: stable@vger.kernel.org
Reported-by: Mingi Cho of Theori working with ZDI
Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@canonical.com>
Reviewed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
17 months agonetfilter: conntrack: don't fold port numbers into addresses before hashing
Florian Westphal [Tue, 4 Jul 2023 10:25:23 +0000 (12:25 +0200)]
netfilter: conntrack: don't fold port numbers into addresses before hashing

Originally this used jhash2() over tuple and folded the zone id,
the pernet hash value, destination port and l4 protocol number into the
32bit seed value.

When the switch to siphash was done, I used an on-stack temporary
buffer to build a suitable key to be hashed via siphash().

But this showed up as performance regression, so I got rid of
the temporary copy and collected to-be-hashed data in 4 u64 variables.

This makes it easy to build tuples that produce the same hash, which isn't
desirable even though chain lengths are limited.

Switch back to plain siphash, but just like with jhash2(), take advantage
of the fact that most of to-be-hashed data is already in a suitable order.

Use an empty struct as annotation in 'struct nf_conntrack_tuple' to mark
last member that can be used as hash input.

The only remaining data that isn't present in the tuple structure are the
zone identifier and the pernet hash: fold those into the key.

Fixes: d2c806abcf0b ("netfilter: conntrack: use siphash_4u64")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
17 months agonetfilter: conntrack: Avoid nf_ct_helper_hash uses after free
Florent Revest [Mon, 3 Jul 2023 14:52:16 +0000 (16:52 +0200)]
netfilter: conntrack: Avoid nf_ct_helper_hash uses after free

If nf_conntrack_init_start() fails (for example due to a
register_nf_conntrack_bpf() failure), the nf_conntrack_helper_fini()
clean-up path frees the nf_ct_helper_hash map.

When built with NF_CONNTRACK=y, further netfilter modules (e.g:
netfilter_conntrack_ftp) can still be loaded and call
nf_conntrack_helpers_register(), independently of whether nf_conntrack
initialized correctly. This accesses the nf_ct_helper_hash dangling
pointer and causes a uaf, possibly leading to random memory corruption.

This patch guards nf_conntrack_helper_register() from accessing a freed
or uninitialized nf_ct_helper_hash pointer and fixes possible
uses-after-free when loading a conntrack module.

Cc: stable@vger.kernel.org
Fixes: 12f7a505331e ("netfilter: add user-space connection tracking helper infrastructure")
Signed-off-by: Florent Revest <revest@chromium.org>
Reviewed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
17 months agonetfilter: conntrack: gre: don't set assured flag for clash entries
Florian Westphal [Mon, 3 Jul 2023 11:43:18 +0000 (13:43 +0200)]
netfilter: conntrack: gre: don't set assured flag for clash entries

Now that conntrack core is allowd to insert clashing entries, make sure
GRE won't set assured flag on NAT_CLASH entries, just like UDP.

Doing so prevents early_drop logic for these entries.

Fixes: d671fd82eaa9 ("netfilter: conntrack: allow insertion clash of gre protocol")
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
17 months agonetfilter: nf_tables: report use refcount overflow
Pablo Neira Ayuso [Wed, 28 Jun 2023 14:24:27 +0000 (16:24 +0200)]
netfilter: nf_tables: report use refcount overflow

Overflow use refcount checks are not complete.

Add helper function to deal with object reference counter tracking.
Report -EMFILE in case UINT_MAX is reached.

nft_use_dec() splats in case that reference counter underflows,
which should not ever happen.

Add nft_use_inc_restore() and nft_use_dec_restore() which are used
to restore reference counter from error and abort paths.

Use u32 in nft_flowtable and nft_object since helper functions cannot
work on bitfields.

Remove the few early incomplete checks now that the helper functions
are in place and used to check for refcount overflow.

Fixes: 96518518cc41 ("netfilter: add nftables")
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
17 months agoMerge branch 'mptcp-fixes'
David S. Miller [Wed, 5 Jul 2023 09:51:14 +0000 (10:51 +0100)]
Merge branch 'mptcp-fixes'

Matthieu Baerts says:

====================
mptcp: fixes for v6.5

Here is a first batch of fixes for v6.5 and older.

The fixes are not linked to each others.

Patch 1 ensures subflows are unhashed before cleaning the backlog to
avoid races. This fixes another recent fix from v6.4.

Patch 2 does not rely on implicit state check in mptcp_listen() to avoid
races when receiving an MP_FASTCLOSE. A regression from v5.17.

The rest fixes issues in the selftests.

Patch 3 makes sure errors when setting up the environment are no longer
ignored. For v5.17+.

Patch 4 uses 'iptables-legacy' if available to be able to run on older
kernels. A fix for v5.13 and newer.

Patch 5 catches errors when issues are detected with packet marks. Also
for v5.13+.

Patch 6 uses the correct variable instead of an undefined one. Even if
there was no visible impact, it can help to find regressions later. An
issue visible in v5.19+.

Patch 7 makes sure errors with some sub-tests are reported to have the
selftest marked as failed as expected. Also for v5.19+.

Patch 8 adds a kernel config that is required to execute MPTCP
selftests. It is valid for v5.9+.

Patch 9 fixes issues when validating the userspace path-manager with
32-bit arch, an issue affecting v5.19+.
====================

Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
17 months agoselftests: mptcp: pm_nl_ctl: fix 32-bit support
Matthieu Baerts [Tue, 4 Jul 2023 20:44:41 +0000 (22:44 +0200)]
selftests: mptcp: pm_nl_ctl: fix 32-bit support

When using pm_nl_ctl to validate userspace path-manager's behaviours, it
was failing on 32-bit architectures ~half of the time.

pm_nl_ctl was not reporting any error but the command was not doing what
it was expected to do. As a result, the expected linked event was not
triggered after and the test failed.

This is due to the fact the token given in argument to the application
was parsed as an integer with atoi(): in a 32-bit arch, if the number
was bigger than INT_MAX, 2147483647 was used instead.

This can simply be fixed by using strtoul() instead of atoi().

The errors have been seen "by chance" when manually looking at the
results from LKFT.

Fixes: 9a0b36509df0 ("selftests: mptcp: support MPTCP_PM_CMD_ANNOUNCE")
Cc: stable@vger.kernel.org
Fixes: ecd2a77d672f ("selftests: mptcp: support MPTCP_PM_CMD_REMOVE")
Fixes: cf8d0a6dfd64 ("selftests: mptcp: support MPTCP_PM_CMD_SUBFLOW_CREATE")
Fixes: 57cc361b8d38 ("selftests: mptcp: support MPTCP_PM_CMD_SUBFLOW_DESTROY")
Fixes: ca188a25d43f ("selftests: mptcp: userspace PM support for MP_PRIO signals")
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agoselftests: mptcp: depend on SYN_COOKIES
Matthieu Baerts [Tue, 4 Jul 2023 20:44:40 +0000 (22:44 +0200)]
selftests: mptcp: depend on SYN_COOKIES

MPTCP selftests are using TCP SYN Cookies for quite a while now, since
v5.9.

Some CIs don't have this config option enabled and this is causing
issues in the tests:

  # ns1 MPTCP -> ns1 (10.0.1.1:10000      ) MPTCP     (duration   167ms) sysctl: cannot stat /proc/sys/net/ipv4/tcp_syncookies: No such file or directory
  # [ OK ]./mptcp_connect.sh: line 554: [: -eq: unary operator expected

There is no impact in the results but the test is not doing what it is
supposed to do.

Fixes: fed61c4b584c ("selftests: mptcp: make 2nd net namespace use tcp syn cookies unconditionally")
Cc: stable@vger.kernel.org
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agoselftests: mptcp: userspace_pm: report errors with 'remove' tests
Matthieu Baerts [Tue, 4 Jul 2023 20:44:39 +0000 (22:44 +0200)]
selftests: mptcp: userspace_pm: report errors with 'remove' tests

A message was mentioning an issue with the "remove" tests but the
selftest was not marked as failed.

Directly exit with an error like it is done everywhere else in this
selftest.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 259a834fadda ("selftests: mptcp: functional tests for the userspace PM type")
Cc: stable@vger.kernel.org
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agoselftests: mptcp: userspace_pm: use correct server port
Matthieu Baerts [Tue, 4 Jul 2023 20:44:38 +0000 (22:44 +0200)]
selftests: mptcp: userspace_pm: use correct server port

"server4_port" variable is not set but "app4_port" is the server port in
v4 and the correct variable name to use.

The port is optional so there was no visible impact.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: ca188a25d43f ("selftests: mptcp: userspace PM support for MP_PRIO signals")
Cc: stable@vger.kernel.org
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agoselftests: mptcp: sockopt: return error if wrong mark
Matthieu Baerts [Tue, 4 Jul 2023 20:44:37 +0000 (22:44 +0200)]
selftests: mptcp: sockopt: return error if wrong mark

When an error was detected when checking the marks, a message was
correctly printed mentioning the error but followed by another one
saying everything was OK and the selftest was not marked as failed as
expected.

Now the 'ret' variable is directly set to 1 in order to make sure the
exit is done with an error, similar to what is done in other functions.
While at it, the error is correctly propagated to the caller.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: dc65fe82fb07 ("selftests: mptcp: add packet mark test case")
Cc: stable@vger.kernel.org
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agoselftests: mptcp: sockopt: use 'iptables-legacy' if available
Matthieu Baerts [Tue, 4 Jul 2023 20:44:36 +0000 (22:44 +0200)]
selftests: mptcp: sockopt: use 'iptables-legacy' if available

IPTables commands using 'iptables-nft' fail on old kernels, at least
on v5.15 because it doesn't see the default IPTables chains:

  $ iptables -L
  iptables/1.8.2 Failed to initialize nft: Protocol not supported

As a first step before switching to NFTables, we can use iptables-legacy
if available.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: dc65fe82fb07 ("selftests: mptcp: add packet mark test case")
Cc: stable@vger.kernel.org
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agoselftests: mptcp: connect: fail if nft supposed to work
Matthieu Baerts [Tue, 4 Jul 2023 20:44:35 +0000 (22:44 +0200)]
selftests: mptcp: connect: fail if nft supposed to work

In case of "external" errors when preparing the environment for the
TProxy tests, the subtests were marked as skipped.

This is fine but it means these errors are ignored. On MPTCP Public CI,
we do want to catch such issues and mark the selftest as failed if there
are such issues. We can then use mptcp_lib_fail_if_expected_feature()
helper that has been recently added to fail if needed.

Link: https://github.com/multipath-tcp/mptcp_net-next/issues/368
Fixes: 5fb62e9cd3ad ("selftests: mptcp: add tproxy test case")
Cc: stable@vger.kernel.org
Acked-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agomptcp: do not rely on implicit state check in mptcp_listen()
Paolo Abeni [Tue, 4 Jul 2023 20:44:34 +0000 (22:44 +0200)]
mptcp: do not rely on implicit state check in mptcp_listen()

Since the blamed commit, closing the first subflow resets the first
subflow socket state to SS_UNCONNECTED.

The current mptcp listen implementation relies only on such
state to prevent touching not-fully-disconnected sockets.

Incoming mptcp fastclose (or paired endpoint removal) unconditionally
closes the first subflow.

All the above allows an incoming fastclose followed by a listen() call
to successfully race with a blocking recvmsg(), potentially causing the
latter to hit a divide by zero bug in cleanup_rbuf/__tcp_select_window().

Address the issue explicitly checking the msk socket state in
mptcp_listen(). An alternative solution would be moving the first
subflow socket state update into mptcp_disconnect(), but in the long
term the first subflow socket should be removed: better avoid relaying
on it for internal consistency check.

Fixes: b29fcfb54cd7 ("mptcp: full disconnect implementation")
Cc: stable@vger.kernel.org
Reported-by: Christoph Paasch <cpaasch@apple.com>
Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/414
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agomptcp: ensure subflow is unhashed before cleaning the backlog
Paolo Abeni [Tue, 4 Jul 2023 20:44:33 +0000 (22:44 +0200)]
mptcp: ensure subflow is unhashed before cleaning the backlog

While tacking care of the mptcp-level listener I unintentionally
moved the subflow level unhash after the subflow listener backlog
cleanup.

That could cause some nasty race and makes the code harder to read.

Address the issue restoring the proper order of operations.

Fixes: 57fc0f1ceaa4 ("mptcp: ensure listener is unhashed before updating the sk status")
Cc: stable@vger.kernel.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Reviewed-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
17 months agoMerge branch 'topic/midi20' into for-linus
Takashi Iwai [Wed, 5 Jul 2023 09:50:51 +0000 (11:50 +0200)]
Merge branch 'topic/midi20' into for-linus

Pull yet more a typo fix in the converter code.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
17 months agoALSA: seq: ump: fix typo in system_2p_ev_to_ump_midi1()
Minjie Du [Wed, 5 Jul 2023 09:35:45 +0000 (17:35 +0800)]
ALSA: seq: ump: fix typo in system_2p_ev_to_ump_midi1()

Fix data->system.parm2 typo.

Fixes: e9e02819a98a ("ALSA: seq: Automatic conversion of UMP events")
Signed-off-by: Minjie Du <duminjie@vivo.com>
Link: https://lore.kernel.org/r/20230705093545.14695-1-duminjie@vivo.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>