riscv: Remove unused GENERATING_ASM_OFFSETS

[ Upstream commit 46d4e5ac6f2f801f97bcd0ec82365969197dc9b1 ]

The macro is not used in the current version of kernel, it looks like
can be removed to avoid a build warning:

../arch/riscv/kernel/asm-offsets.c: At top level:
../arch/riscv/kernel/asm-offsets.c:7: warning: macro "GENERATING_ASM_OFFSETS" is not used [-Wunused-macros]
    7 | #define GENERATING_ASM_OFFSETS

Fixes: 9639a44394b9 ("RISC-V: Provide a cleaner raw_smp_processor_id()")
Cc: stable@vger.kernel.org
Reviewed-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Tested-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Signed-off-by: Chunyan Zhang <zhangchunyan@iscas.ac.cn>
Link: https://lore.kernel.org/r/20241008094141.549248-2-zhangchunyan@iscas.ac.cn
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

riscv: Use '%u' to format the output of 'cpu'

[ Upstream commit e0872ab72630dada3ae055bfa410bf463ff1d1e0 ]

'cpu' is an unsigned integer, so its conversion specifier should
be %u, not %d.

Suggested-by: Wentao Guan <guanwentao@uniontech.com>
Suggested-by: Maciej W. Rozycki <macro@orcam.me.uk>
Link: https://lore.kernel.org/all/alpine.DEB.2.21.2409122309090.40372@angie.orcam.me.uk/
Signed-off-by: WangYuli <wangyuli@uniontech.com>
Reviewed-by: Charlie Jenkins <charlie@rivosinc.com>
Tested-by: Charlie Jenkins <charlie@rivosinc.com>
Fixes: f1e58583b9c7 ("RISC-V: Support cpu hotplug")
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/4C127DEECDA287C8+20241017032010.96772-1-wangyuli@uniontech.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

riscv: efi: Set NX compat flag in PE/COFF header

[ Upstream commit d41373a4b910961df5a5e3527d7bde6ad45ca438 ]

The IMAGE_DLLCHARACTERISTICS_NX_COMPAT informs the firmware that the
EFI binary does not rely on pages that are both executable and
writable.

The flag is used by some distro versions of GRUB to decide if the EFI
binary may be executed.

As the Linux kernel neither has RWX sections nor needs RWX pages for
relocation we should set the flag.

Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Emil Renner Berthing <emil.renner.berthing@canonical.com>
Fixes: cb7d2dd5612a ("RISC-V: Add PE/COFF header for EFI stub")
Acked-by: Ard Biesheuvel <ardb@kernel.org>
Link: https://lore.kernel.org/r/20240929140233.211800-1-heinrich.schuchardt@canonical.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ALSA: hda/realtek: Limit internal Mic boost on Dell platform

[ Upstream commit 78e7be018784934081afec77f96d49a2483f9188 ]

Dell want to limit internal Mic boost on all Dell platform.

Signed-off-by: Kailang Yang <kailang@realtek.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/561fc5f5eff04b6cbd79ed173cd1c1db@realtek.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>

Input: edt-ft5x06 - fix regmap leak when probe fails

[ Upstream commit bffdf9d7e51a7be8eeaac2ccf9e54a5fde01ff65 ]

The driver neglects to free the instance of I2C regmap constructed at
the beginning of the edt_ft5x06_ts_probe() method when probe fails.
Additionally edt_ft5x06_ts_remove() is freeing the regmap too early,
before the rest of the device resources that are managed by devm are
released.

Fix this by installing a custom devm action that will ensure that the
regmap is released at the right time during normal teardown as well as
in case of probe failure.

Note that devm_regmap_init_i2c() could not be used because the driver
may replace the original regmap with a regmap specific for M06 devices
in the middle of the probe, and using devm_regmap_init_i2c() would
result in releasing the M06 regmap too early.

Reported-by: Li Zetao <lizetao1@huawei.com>
Fixes: 9dfd9708ffba ("Input: edt-ft5x06 - convert to use regmap API")
Cc: stable@vger.kernel.org
Reviewed-by: Oliver Graute <oliver.graute@kococonnector.com>
Link: https://lore.kernel.org/r/ZxL6rIlVlgsAu-Jv@google.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

riscv: vdso: Prevent the compiler from inserting calls to memset()

[ Upstream commit bf40167d54d55d4b54d0103713d86a8638fb9290 ]

The compiler is smart enough to insert a call to memset() in
riscv_vdso_get_cpus(), which generates a dynamic relocation.

So prevent this by using -fno-builtin option.

Fixes: e2c0cdfba7f6 ("RISC-V: User-facing API")
Cc: stable@vger.kernel.org
Signed-off-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Reviewed-by: Guo Ren <guoren@kernel.org>
Link: https://lore.kernel.org/r/20241016083625.136311-2-alexghiti@rivosinc.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

spi: spi-fsl-dspi: Fix crash when not using GPIO chip select

[ Upstream commit 25f00a13dccf8e45441265768de46c8bf58e08f6 ]

Add check for the return value of spi_get_csgpiod() to avoid passing a NULL
pointer to gpiod_direction_output(), preventing a crash when GPIO chip
select is not used.

Fix below crash:
[    4.251960] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000000
[    4.260762] Mem abort info:
[    4.263556]   ESR = 0x0000000096000004
[    4.267308]   EC = 0x25: DABT (current EL), IL = 32 bits
[    4.272624]   SET = 0, FnV = 0
[    4.275681]   EA = 0, S1PTW = 0
[    4.278822]   FSC = 0x04: level 0 translation fault
[    4.283704] Data abort info:
[    4.286583]   ISV = 0, ISS = 0x00000004, ISS2 = 0x00000000
[    4.292074]   CM = 0, WnR = 0, TnD = 0, TagAccess = 0
[    4.297130]   GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0
[    4.302445] [0000000000000000] user address but active_mm is swapper
[    4.308805] Internal error: Oops: 0000000096000004 [#1] PREEMPT SMP
[    4.315072] Modules linked in:
[    4.318124] CPU: 2 UID: 0 PID: 1 Comm: swapper/0 Not tainted 6.12.0-rc4-next-20241023-00008-ga20ec42c5fc1 #359
[    4.328130] Hardware name: LS1046A QDS Board (DT)
[    4.332832] pstate: 40000005 (nZcv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[    4.339794] pc : gpiod_direction_output+0x34/0x5c
[    4.344505] lr : gpiod_direction_output+0x18/0x5c
[    4.349208] sp : ffff80008003b8f0
[    4.352517] x29: ffff80008003b8f0 x28: 0000000000000000 x27: ffffc96bcc7e9068
[    4.359659] x26: ffffc96bcc6e00b0 x25: ffffc96bcc598398 x24: ffff447400132810
[    4.366800] x23: 0000000000000000 x22: 0000000011e1a300 x21: 0000000000020002
[    4.373940] x20: 0000000000000000 x19: 0000000000000000 x18: ffffffffffffffff
[    4.381081] x17: ffff44740016e600 x16: 0000000500000003 x15: 0000000000000007
[    4.388221] x14: 0000000000989680 x13: 0000000000020000 x12: 000000000000001e
[    4.395362] x11: 0044b82fa09b5a53 x10: 0000000000000019 x9 : 0000000000000008
[    4.402502] x8 : 0000000000000002 x7 : 0000000000000007 x6 : 0000000000000000
[    4.409641] x5 : 0000000000000200 x4 : 0000000002000000 x3 : 0000000000000000
[    4.416781] x2 : 0000000000022202 x1 : 0000000000000000 x0 : 0000000000000000
[    4.423921] Call trace:
[    4.426362]  gpiod_direction_output+0x34/0x5c (P)
[    4.431067]  gpiod_direction_output+0x18/0x5c (L)
[    4.435771]  dspi_setup+0x220/0x334

Fixes: 9e264f3f85a5 ("spi: Replace all spi->chip_select and spi->cs_gpiod references with function call")
Cc: stable@vger.kernel.org
Signed-off-by: Frank Li <Frank.Li@nxp.com>
Link: https://patch.msgid.link/20241023203032.1388491-1-Frank.Li@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

phy: freescale: imx8m-pcie: Do CMN_RST just before PHY PLL lock check

[ Upstream commit f89263b69731e0144d275fff777ee0dd92069200 ]

When enable initcall_debug together with higher debug level below.
CONFIG_CONSOLE_LOGLEVEL_DEFAULT=9
CONFIG_CONSOLE_LOGLEVEL_QUIET=9
CONFIG_MESSAGE_LOGLEVEL_DEFAULT=7

The initialization of i.MX8MP PCIe PHY might be timeout failed randomly.
To fix this issue, adjust the sequence of the resets refer to the power
up sequence listed below.

i.MX8MP PCIe PHY power up sequence:
                          /---------------------------------------------
1.8v supply     ---------/
                    /---------------------------------------------------
0.8v supply     ---/

                ---\ /--------------------------------------------------
                    X        REFCLK Valid
Reference Clock ---/ \--------------------------------------------------
                             -------------------------------------------
                             |
i_init_restn    --------------
                                    ------------------------------------
                                    |
i_cmn_rstn      ---------------------
                                         -------------------------------
                                         |
o_pll_lock_done --------------------------

Logs:
imx6q-pcie 33800000.pcie: host bridge /soc@0/pcie@33800000 ranges:
imx6q-pcie 33800000.pcie:       IO 0x001ff80000..0x001ff8ffff -> 0x0000000000
imx6q-pcie 33800000.pcie:      MEM 0x0018000000..0x001fefffff -> 0x0018000000
probe of clk_imx8mp_audiomix.reset.0 returned 0 after 1052 usecs
probe of 30e20000.clock-controller returned 0 after 32971 usecs
phy phy-32f00000.pcie-phy.4: phy poweron failed --> -110
probe of 30e10000.dma-controller returned 0 after 10235 usecs
imx6q-pcie 33800000.pcie: waiting for PHY ready timeout!
dwhdmi-imx 32fd8000.hdmi: Detected HDMI TX controller v2.13a with HDCP (samsung_dw_hdmi_phy2)
imx6q-pcie 33800000.pcie: probe with driver imx6q-pcie failed with error -110

Fixes: dce9edff16ee ("phy: freescale: imx8m-pcie: Add i.MX8MP PCIe PHY support")
Cc: stable@vger.kernel.org
Signed-off-by: Richard Zhu <hongxing.zhu@nxp.com>
Signed-off-by: Frank Li <Frank.Li@nxp.com>
v2 changes:
- Rebase to latest fixes branch of linux-phy git repo.
- Richard's environment have problem and can't sent out patch. So I help
post this fix patch.

Link: https://lore.kernel.org/r/20241021155241.943665-1-Frank.Li@nxp.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

cgroup/bpf: use a dedicated workqueue for cgroup bpf destruction

[ Upstream commit 117932eea99b729ee5d12783601a4f7f5fd58a23 ]

A hung_task problem shown below was found:

INFO: task kworker/0:0:8 blocked for more than 327 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
Workqueue: events cgroup_bpf_release
Call Trace:
 <TASK>
 __schedule+0x5a2/0x2050
 ? find_held_lock+0x33/0x100
 ? wq_worker_sleeping+0x9e/0xe0
 schedule+0x9f/0x180
 schedule_preempt_disabled+0x25/0x50
 __mutex_lock+0x512/0x740
 ? cgroup_bpf_release+0x1e/0x4d0
 ? cgroup_bpf_release+0xcf/0x4d0
 ? process_scheduled_works+0x161/0x8a0
 ? cgroup_bpf_release+0x1e/0x4d0
 ? mutex_lock_nested+0x2b/0x40
 ? __pfx_delay_tsc+0x10/0x10
 mutex_lock_nested+0x2b/0x40
 cgroup_bpf_release+0xcf/0x4d0
 ? process_scheduled_works+0x161/0x8a0
 ? trace_event_raw_event_workqueue_execute_start+0x64/0xd0
 ? process_scheduled_works+0x161/0x8a0
 process_scheduled_works+0x23a/0x8a0
 worker_thread+0x231/0x5b0
 ? __pfx_worker_thread+0x10/0x10
 kthread+0x14d/0x1c0
 ? __pfx_kthread+0x10/0x10
 ret_from_fork+0x59/0x70
 ? __pfx_kthread+0x10/0x10
 ret_from_fork_asm+0x1b/0x30
 </TASK>

This issue can be reproduced by the following pressuse test:
1. A large number of cpuset cgroups are deleted.
2. Set cpu on and off repeatly.
3. Set watchdog_thresh repeatly.
The scripts can be obtained at LINK mentioned above the signature.

The reason for this issue is cgroup_mutex and cpu_hotplug_lock are
acquired in different tasks, which may lead to deadlock.
It can lead to a deadlock through the following steps:
1. A large number of cpusets are deleted asynchronously, which puts a
   large number of cgroup_bpf_release works into system_wq. The max_active
   of system_wq is WQ_DFL_ACTIVE(256). Consequently, all active works are
   cgroup_bpf_release works, and many cgroup_bpf_release works will be put
   into inactive queue. As illustrated in the diagram, there are 256 (in
   the acvtive queue) + n (in the inactive queue) works.
2. Setting watchdog_thresh will hold cpu_hotplug_lock.read and put
   smp_call_on_cpu work into system_wq. However step 1 has already filled
   system_wq, 'sscs.work' is put into inactive queue. 'sscs.work' has
   to wait until the works that were put into the inacvtive queue earlier
   have executed (n cgroup_bpf_release), so it will be blocked for a while.
3. Cpu offline requires cpu_hotplug_lock.write, which is blocked by step 2.
4. Cpusets that were deleted at step 1 put cgroup_release works into
   cgroup_destroy_wq. They are competing to get cgroup_mutex all the time.
   When cgroup_metux is acqured by work at css_killed_work_fn, it will
   call cpuset_css_offline, which needs to acqure cpu_hotplug_lock.read.
   However, cpuset_css_offline will be blocked for step 3.
5. At this moment, there are 256 works in active queue that are
   cgroup_bpf_release, they are attempting to acquire cgroup_mutex, and as
   a result, all of them are blocked. Consequently, sscs.work can not be
   executed. Ultimately, this situation leads to four processes being
   blocked, forming a deadlock.

system_wq(step1) WatchDog(step2) cpu offline(step3) cgroup_destroy_wq(step4)
...
2000+ cgroups deleted asyn
256 actives + n inactives
__lockup_detector_reconfigure
P(cpu_hotplug_lock.read)
put sscs.work into system_wq
256 + n + 1(sscs.work)
sscs.work wait to be executed
warting sscs.work finish
percpu_down_write
P(cpu_hotplug_lock.write)
...blocking...
css_killed_work_fn
P(cgroup_mutex)
cpuset_css_offline
P(cpu_hotplug_lock.read)
...blocking...
256 cgroup_bpf_release
mutex_lock(&cgroup_mutex);
..blocking...

To fix the problem, place cgroup_bpf_release works on a dedicated
workqueue which can break the loop and solve the problem. System wqs are
for misc things which shouldn't create a large number of concurrent work
items. If something is going to generate >WQ_DFL_ACTIVE(256) concurrent
work items, it should use its own dedicated workqueue.

Fixes: 4bfc0bb2c60e ("bpf: decouple the lifetime of cgroup_bpf from cgroup itself")
Cc: stable@vger.kernel.org # v5.3+
Link: https://lore.kernel.org/cgroups/e90c32d2-2a85-4f28-9154-09c7d320cb60@huawei.com/T/#t
Tested-by: Vishal Chourasia <vishalc@linux.ibm.com>
Signed-off-by: Chen Ridong <chenridong@huawei.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

block: fix sanity checks in blk_rq_map_user_bvec

[ Upstream commit 2ff949441802a8d076d9013c7761f63e8ae5a9bd ]

blk_rq_map_user_bvec contains a check bytes + bv->bv_len > nr_iter which
causes unnecessary failures in NVMe passthrough I/O, reproducible as
follows:

- register a 2 page, page-aligned buffer against a ring
- use that buffer to do a 1 page io_uring NVMe passthrough read

The second (i = 1) iteration of the loop in blk_rq_map_user_bvec will
then have nr_iter == 1 page, bytes == 1 page, bv->bv_len == 1 page, so
the check bytes + bv->bv_len > nr_iter will succeed, causing the I/O to
fail. This failure is unnecessary, as when the check succeeds, it means
we've checked the entire buffer that will be used by the request - i.e.
blk_rq_map_user_bvec should complete successfully. Therefore, terminate
the loop early and return successfully when the check bytes + bv->bv_len
> nr_iter succeeds.

While we're at it, also remove the check that all segments in the bvec
are single-page. While this seems to be true for all users of the
function, it doesn't appear to be required anywhere downstream.

CC: stable@vger.kernel.org
Signed-off-by: Xinyu Zhang <xizhang@purestorage.com>
Co-developed-by: Uday Shankar <ushankar@purestorage.com>
Signed-off-by: Uday Shankar <ushankar@purestorage.com>
Fixes: 37987547932c ("block: extend functionality to map bvec iterator")
Link: https://lore.kernel.org/r/20241023211519.4177873-1-ushankar@purestorage.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Sasha Levin <sashal@kernel.org>

mmc: sdhci-pci-gli: GL9767: Fix low power mode in the SD Express process

commit c4dedaaeb3f78d3718e9c1b1e4d972a6b99073cd upstream.

When starting the SD Express process, the low power negotiation mode will
be disabled, so we need to re-enable it after switching back to SD mode.

Fixes: 0e92aec2efa0 ("mmc: sdhci-pci-gli: Add support SD Express card for GL9767")
Signed-off-by: Ben Chuang <ben.chuang@genesyslogic.com.tw>
Cc: stable@vger.kernel.org
Message-ID: <20241025060017.1663697-2-benchuanggli@gmail.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mmc: sdhci-pci-gli: GL9767: Fix low power mode on the set clock function

commit 8c68b5656e55e9324875881f1000eb4ee3603a87 upstream.

On sdhci_gl9767_set_clock(), the vendor header space(VHS) is read-only
after calling gl9767_disable_ssc_pll() and gl9767_set_ssc_pll_205mhz().
So the low power negotiation mode cannot be enabled again.
Introduce gl9767_set_low_power_negotiation() function to fix it.

The explanation process is as below.

static void sdhci_gl9767_set_clock()
{
...
        gl9767_vhs_write();
        ...
value |= PCIE_GLI_9767_CFG_LOW_PWR_OFF;
        pci_write_config_dword(pdev, PCIE_GLI_9767_CFG, value); <--- (a)

        gl9767_disable_ssc_pll(); <--- (b)
        sdhci_writew(host, 0, SDHCI_CLOCK_CONTROL);

        if (clock == 0)
                return;  <-- (I)

...
        if (clock == 200000000 && ios->timing == MMC_TIMING_UHS_SDR104) {
...
                gl9767_set_ssc_pll_205mhz(); <--- (c)
        }
...
value &= ~PCIE_GLI_9767_CFG_LOW_PWR_OFF;
        pci_write_config_dword(pdev, PCIE_GLI_9767_CFG, value); <-- (II)
        gl9767_vhs_read();
}

(a) disable low power negotiation mode. When return on (I), the low power
mode is disabled.  After (b) and (c), VHS is read-only, the low power mode
cannot be enabled on (II).

Reported-by: Georg Gottleuber <ggo@tuxedocomputers.com>
Fixes: d2754355512e ("mmc: sdhci-pci-gli: Set SDR104's clock to 205MHz and enable SSC for GL9767")
Signed-off-by: Ben Chuang <ben.chuang@genesyslogic.com.tw>
Tested-by: Georg Gottleuber <ggo@tuxedocomputers.com>
Cc: stable@vger.kernel.org
Message-ID: <20241025060017.1663697-1-benchuanggli@gmail.com>
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

cxl/port: Fix use-after-free, permit out-of-order decoder shutdown

commit 101c268bd2f37e965a5468353e62d154db38838e upstream.

In support of investigating an initialization failure report [1],
cxl_test was updated to register mock memory-devices after the mock
root-port/bus device had been registered. That led to cxl_test crashing
with a use-after-free bug with the following signature:

    cxl_port_attach_region: cxl region3: cxl_host_bridge.0:port3 decoder3.0 add: mem0:decoder7.0 @ 0 next: cxl_switch_uport.0 nr_eps: 1 nr_targets: 1
    cxl_port_attach_region: cxl region3: cxl_host_bridge.0:port3 decoder3.0 add: mem4:decoder14.0 @ 1 next: cxl_switch_uport.0 nr_eps: 2 nr_targets: 1
    cxl_port_setup_targets: cxl region3: cxl_switch_uport.0:port6 target[0] = cxl_switch_dport.0 for mem0:decoder7.0 @ 0
1)  cxl_port_setup_targets: cxl region3: cxl_switch_uport.0:port6 target[1] = cxl_switch_dport.4 for mem4:decoder14.0 @ 1
    [..]
    cxld_unregister: cxl decoder14.0:
    cxl_region_decode_reset: cxl_region region3:
    mock_decoder_reset: cxl_port port3: decoder3.0 reset
2)  mock_decoder_reset: cxl_port port3: decoder3.0: out of order reset, expected decoder3.1
    cxl_endpoint_decoder_release: cxl decoder14.0:
    [..]
    cxld_unregister: cxl decoder7.0:
3)  cxl_region_decode_reset: cxl_region region3:
    Oops: general protection fault, probably for non-canonical address 0x6b6b6b6b6b6b6bc3: 0000 [#1] PREEMPT SMP PTI
    [..]
    RIP: 0010:to_cxl_port+0x8/0x60 [cxl_core]
    [..]
    Call Trace:
     <TASK>
     cxl_region_decode_reset+0x69/0x190 [cxl_core]
     cxl_region_detach+0xe8/0x210 [cxl_core]
     cxl_decoder_kill_region+0x27/0x40 [cxl_core]
     cxld_unregister+0x5d/0x60 [cxl_core]

At 1) a region has been established with 2 endpoint decoders (7.0 and
14.0). Those endpoints share a common switch-decoder in the topology
(3.0). At teardown, 2), decoder14.0 is the first to be removed and hits
the "out of order reset case" in the switch decoder. The effect though
is that region3 cleanup is aborted leaving it in-tact and
referencing decoder14.0. At 3) the second attempt to teardown region3
trips over the stale decoder14.0 object which has long since been
deleted.

The fix here is to recognize that the CXL specification places no
mandate on in-order shutdown of switch-decoders, the driver enforces
in-order allocation, and hardware enforces in-order commit. So, rather
than fail and leave objects dangling, always remove them.

In support of making cxl_region_decode_reset() always succeed,
cxl_region_invalidate_memregion() failures are turned into warnings.
Crashing the kernel is ok there since system integrity is at risk if
caches cannot be managed around physical address mutation events like
CXL region destruction.

A new device_for_each_child_reverse_from() is added to cleanup
port->commit_end after all dependent decoders have been disabled. In
other words if decoders are allocated 0->1->2 and disabled 1->2->0 then
port->commit_end only decrements from 2 after 2 has been disabled, and
it decrements all the way to zero since 1 was disabled previously.

Link: http://lore.kernel.org/20241004212504.1246-1-gourry@gourry.net
Cc: stable@vger.kernel.org
Fixes: 176baefb2eb5 ("cxl/hdm: Commit decoder state to hardware")
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Alison Schofield <alison.schofield@intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Zijun Hu <quic_zijuhu@quicinc.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Link: https://patch.msgid.link/172964782781.81806.17902885593105284330.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

thunderbolt: Honor TMU requirements in the domain when setting TMU mode

commit 3cea8af2d1a9ae5869b47c3dabe3b20f331f3bbd upstream.

Currently, when configuring TMU (Time Management Unit) mode of a given
router, we take into account only its own TMU requirements ignoring
other routers in the domain. This is problematic if the router we are
configuring has lower TMU requirements than what is already configured
in the domain.

In the scenario below, we have a host router with two USB4 ports: A and
B. Port A connected to device router #1 (which supports CL states) and
existing DisplayPort tunnel, thus, the TMU mode is HiFi uni-directional.

1. Initial topology

          [Host]
         A/
         /
 [Device #1]
   /
Monitor

2. Plug in device #2 (that supports CL states) to downstream port B of
   the host router

         [Host]
        A/    B\
        /       \
 [Device #1]    [Device #2]
   /
Monitor

The TMU mode on port B and port A will be configured to LowRes which is
not what we want and will cause monitor to start flickering.

To address this we first scan the domain and search for any router
configured to HiFi uni-directional mode, and if found, configure TMU
mode of the given router to HiFi uni-directional as well.

Cc: stable@vger.kernel.org
Signed-off-by: Gil Fine <gil.fine@linux.intel.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

tools/mm: -Werror fixes in page-types/slabinfo

commit ece5897e5a10fcd56a317e32f2dc7219f366a5a8 upstream.

Commit e6d2c436ff693 ("tools/mm: allow users to provide additional
cflags/ldflags") passes now CFLAGS to Makefile.  With this, build systems
with default -Werror enabled found:

slabinfo.c:1300:25: error: ignoring return value of 'chdir'
declared with attribute 'warn_unused_result' [-Werror=unused-result]
                         chdir("..");
                         ^~~~~~~~~~~
page-types.c:397:35: error: format '%lu' expects argument of type
'long unsigned int', but argument 2 has type 'uint64_t'
{aka 'long long unsigned int'} [-Werror=format=]
                         printf("%lu\t", mapcnt0);
                                 ~~^     ~~~~~~~
..

Fix page-types by using PRIu64 for uint64_t prints and check in slabinfo
for return code on chdir("..").

Link: https://lkml.kernel.org/r/c1ceb507-94bc-461c-934d-c19b77edd825@gmail.com
Fixes: e6d2c436ff69 ("tools/mm: allow users to provide additional cflags/ldflags")
Signed-off-by: Wladislav Wiebe <wladislav.kw@gmail.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Herton R. Krzesinski <herton@redhat.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

mm: shmem: fix data-race in shmem_getattr()

commit d949d1d14fa281ace388b1de978e8f2cd52875cf upstream.

I got the following KCSAN report during syzbot testing:

==================================================================
BUG: KCSAN: data-race in generic_fillattr / inode_set_ctime_current

write to 0xffff888102eb3260 of 4 bytes by task 6565 on cpu 1:
 inode_set_ctime_to_ts include/linux/fs.h:1638 [inline]
 inode_set_ctime_current+0x169/0x1d0 fs/inode.c:2626
 shmem_mknod+0x117/0x180 mm/shmem.c:3443
 shmem_create+0x34/0x40 mm/shmem.c:3497
 lookup_open fs/namei.c:3578 [inline]
 open_last_lookups fs/namei.c:3647 [inline]
 path_openat+0xdbc/0x1f00 fs/namei.c:3883
 do_filp_open+0xf7/0x200 fs/namei.c:3913
 do_sys_openat2+0xab/0x120 fs/open.c:1416
 do_sys_open fs/open.c:1431 [inline]
 __do_sys_openat fs/open.c:1447 [inline]
 __se_sys_openat fs/open.c:1442 [inline]
 __x64_sys_openat+0xf3/0x120 fs/open.c:1442
 x64_sys_call+0x1025/0x2d60 arch/x86/include/generated/asm/syscalls_64.h:258
 do_syscall_x64 arch/x86/entry/common.c:52 [inline]
 do_syscall_64+0x54/0x120 arch/x86/entry/common.c:83
 entry_SYSCALL_64_after_hwframe+0x76/0x7e

read to 0xffff888102eb3260 of 4 bytes by task 3498 on cpu 0:
 inode_get_ctime_nsec include/linux/fs.h:1623 [inline]
 inode_get_ctime include/linux/fs.h:1629 [inline]
 generic_fillattr+0x1dd/0x2f0 fs/stat.c:62
 shmem_getattr+0x17b/0x200 mm/shmem.c:1157
 vfs_getattr_nosec fs/stat.c:166 [inline]
 vfs_getattr+0x19b/0x1e0 fs/stat.c:207
 vfs_statx_path fs/stat.c:251 [inline]
 vfs_statx+0x134/0x2f0 fs/stat.c:315
 vfs_fstatat+0xec/0x110 fs/stat.c:341
 __do_sys_newfstatat fs/stat.c:505 [inline]
 __se_sys_newfstatat+0x58/0x260 fs/stat.c:499
 __x64_sys_newfstatat+0x55/0x70 fs/stat.c:499
 x64_sys_call+0x141f/0x2d60 arch/x86/include/generated/asm/syscalls_64.h:263
 do_syscall_x64 arch/x86/entry/common.c:52 [inline]
 do_syscall_64+0x54/0x120 arch/x86/entry/common.c:83
 entry_SYSCALL_64_after_hwframe+0x76/0x7e

value changed: 0x2755ae53 -> 0x27ee44d3

Reported by Kernel Concurrency Sanitizer on:
CPU: 0 UID: 0 PID: 3498 Comm: udevd Not tainted 6.11.0-rc6-syzkaller-00326-gd1f2d51b711a-dirty #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 08/06/2024
==================================================================

When calling generic_fillattr(), if you don't hold read lock, data-race
will occur in inode member variables, which can cause unexpected
behavior.

Since there is no special protection when shmem_getattr() calls
generic_fillattr(), data-race occurs by functions such as shmem_unlink()
or shmem_mknod(). This can cause unexpected results, so commenting it out
is not enough.

Therefore, when calling generic_fillattr() from shmem_getattr(), it is
appropriate to protect the inode using inode_lock_shared() and
inode_unlock_shared() to prevent data-race.

Link: https://lkml.kernel.org/r/20240909123558.70229-1-aha310510@gmail.com
Fixes: 44a30220bc0a ("shmem: recalculate file inode when fstat")
Signed-off-by: Jeongjun Park <aha310510@gmail.com>
Reported-by: syzbot <syzkaller@googlegroup.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Yu Zhao <yuzhao@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

RISC-V: ACPI: fix early_ioremap to early_memremap

commit 1966db682f064172891275cb951aa8c98a0a809b upstream.

When SVPBMT is enabled, __acpi_map_table() will directly access the
data in DDR through the IO attribute, rather than through hardware
cache consistency, resulting in incorrect data in the obtained ACPI
table.

The log: ACPI: [ACPI:0x18] Invalid zero length.

We do not assume whether the bootloader flushes or not. We should
access in a cacheable way instead of maintaining cache consistency
by software.

Fixes: 3b426d4b5b14 ("RISC-V: ACPI : Fix for usage of pointers in different address space")
Cc: stable@vger.kernel.org
Reviewed-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Signed-off-by: Yunhui Cui <cuiyunhui@bytedance.com>
Reviewed-by: Sunil V L <sunilvl@ventanamicro.com>
Link: https://lore.kernel.org/r/20241014130141.86426-1-cuiyunhui@bytedance.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

nilfs2: fix potential deadlock with newly created symlinks

commit b3a033e3ecd3471248d474ef263aadc0059e516a upstream.

Syzbot reported that page_symlink(), called by nilfs_symlink(), triggers
memory reclamation involving the filesystem layer, which can result in
circular lock dependencies among the reader/writer semaphore
nilfs->ns_segctor_sem, s_writers percpu_rwsem (intwrite) and the
fs_reclaim pseudo lock.

This is because after commit 21fc61c73c39 ("don't put symlink bodies in
pagecache into highmem"), the gfp flags of the page cache for symbolic
links are overwritten to GFP_KERNEL via inode_nohighmem().

This is not a problem for symlinks read from the backing device, because
the __GFP_FS flag is dropped after inode_nohighmem() is called.  However,
when a new symlink is created with nilfs_symlink(), the gfp flags remain
overwritten to GFP_KERNEL.  Then, memory allocation called from
page_symlink() etc.  triggers memory reclamation including the FS layer,
which may call nilfs_evict_inode() or nilfs_dirty_inode().  And these can
cause a deadlock if they are called while nilfs->ns_segctor_sem is held:

Fix this issue by dropping the __GFP_FS flag from the page cache GFP flags
of newly created symlinks in the same way that nilfs_new_inode() and
__nilfs_read_inode() do, as a workaround until we adopt nofs allocation
scope consistently or improve the locking constraints.

Link: https://lkml.kernel.org/r/20241020050003.4308-1-konishi.ryusuke@gmail.com
Fixes: 21fc61c73c39 ("don't put symlink bodies in pagecache into highmem")
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@gmail.com>
Reported-by: syzbot+9ef37ac20608f4836256@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=9ef37ac20608f4836256
Tested-by: syzbot+9ef37ac20608f4836256@syzkaller.appspotmail.com
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

iio: light: veml6030: fix microlux value calculation

commit 63dd163cd61dda6f38343776b42331cc6b7e56e0 upstream.

The raw value conversion to obtain a measurement in lux as
INT_PLUS_MICRO does not calculate the decimal part properly to display
it as micro (in this case microlux). It only calculates the module to
obtain the decimal part from a resolution that is 10000 times the
provided in the datasheet (0.5376 lux/cnt for the veml6030). The
resulting value must still be multiplied by 100 to make it micro.

This bug was introduced with the original implementation of the driver.

Only the illuminance channel is fixed becuase the scale is non sensical
for the intensity channels anyway.

Cc: stable@vger.kernel.org
Fixes: 7b779f573c48 ("iio: light: add driver for veml6030 ambient light sensor")
Signed-off-by: Javier Carrasco <javier.carrasco.cruz@gmail.com>
Link: https://patch.msgid.link/20241016-veml6030-fix-processed-micro-v1-1-4a5644796437@gmail.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

iio: gts-helper: Fix memory leaks in iio_gts_build_avail_scale_table()

commit 691e79ffc42154a9c91dc3b7e96a307037b4be74 upstream.

modprobe iio-test-gts and rmmod it, then the following memory leak
occurs:

unreferenced object 0xffffff80c810be00 (size 64):
  comm "kunit_try_catch", pid 1654, jiffies 4294913981
  hex dump (first 32 bytes):
    02 00 00 00 08 00 00 00 20 00 00 00 40 00 00 00  ........ ...@...
    80 00 00 00 00 02 00 00 00 04 00 00 00 08 00 00  ................
  backtrace (crc a63d875e):
    [<0000000028c1b3c2>] kmemleak_alloc+0x34/0x40
    [<000000001d6ecc87>] __kmalloc_noprof+0x2bc/0x3c0
    [<00000000393795c1>] devm_iio_init_iio_gts+0x4b4/0x16f4
    [<0000000071bb4b09>] 0xffffffdf052a62e0
    [<000000000315bc18>] 0xffffffdf052a6488
    [<00000000f9dc55b5>] kunit_try_run_case+0x13c/0x3ac
    [<00000000175a3fd4>] kunit_generic_run_threadfn_adapter+0x80/0xec
    [<00000000f505065d>] kthread+0x2e8/0x374
    [<00000000bbfb0e5d>] ret_from_fork+0x10/0x20
unreferenced object 0xffffff80cbfe9e70 (size 16):
  comm "kunit_try_catch", pid 1658, jiffies 4294914015
  hex dump (first 16 bytes):
    10 00 00 00 40 00 00 00 80 00 00 00 00 00 00 00  ....@...........
  backtrace (crc 857f0cb4):
    [<0000000028c1b3c2>] kmemleak_alloc+0x34/0x40
    [<000000001d6ecc87>] __kmalloc_noprof+0x2bc/0x3c0
    [<00000000393795c1>] devm_iio_init_iio_gts+0x4b4/0x16f4
    [<0000000071bb4b09>] 0xffffffdf052a62e0
    [<000000007d089d45>] 0xffffffdf052a6864
    [<00000000f9dc55b5>] kunit_try_run_case+0x13c/0x3ac
    [<00000000175a3fd4>] kunit_generic_run_threadfn_adapter+0x80/0xec
    [<00000000f505065d>] kthread+0x2e8/0x374
    [<00000000bbfb0e5d>] ret_from_fork+0x10/0x20
......

It includes 5*5 times "size 64" memory leaks, which correspond to 5 times
test_init_iio_gain_scale() calls with gts_test_gains size 10 (10*size(int))
and gts_test_itimes size 5. It also includes 5*1 times "size 16"
memory leak, which correspond to one time __test_init_iio_gain_scale()
call with gts_test_gains_gain_low size 3 (3*size(int)) and gts_test_itimes
size 5.

The reason is that the per_time_gains[i] is not freed which is allocated in
the "gts->num_itime" for loop in iio_gts_build_avail_scale_table().

Cc: stable@vger.kernel.org
Fixes: 38416c28e168 ("iio: light: Add gain-time-scale helpers")
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Reviewed-by: Matti Vaittinen <mazziesaccount@gmail.com>
Link: https://patch.msgid.link/20241011095512.3667549-1-ruanjinjie@huawei.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

iio: gts-helper: Fix memory leaks for the error path of iio_gts_build_avail_scale_table()

commit 369f05688911b05216cfcd6ca74473bec87948d7 upstream.

If per_time_scales[i] or per_time_gains[i] kcalloc fails in the for loop
of iio_gts_build_avail_scale_table(), the err_free_out will fail to call
kfree() each time when i is reduced to 0, so all the per_time_scales[0]
and per_time_gains[0] will not be freed, which will cause memory leaks.

Fix it by checking if i >= 0.

Cc: stable@vger.kernel.org
Fixes: 38416c28e168 ("iio: light: Add gain-time-scale helpers")
Reviewed-by: Matti Vaittinen <mazziesaccount@gmail.com>
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
Link: https://patch.msgid.link/20241016012453.2013302-1-ruanjinjie@huawei.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

iio: adc: ad7124: fix division by zero in ad7124_set_channel_odr()

commit efa353ae1b0541981bc96dbf2e586387d0392baa upstream.

In the ad7124_write_raw() function, parameter val can potentially
be zero. This may lead to a division by zero when DIV_ROUND_CLOSEST()
is called within ad7124_set_channel_odr(). The ad7124_write_raw()
function is invoked through the sequence: iio_write_channel_raw() ->
iio_write_channel_attribute() -> iio_channel_write(), with no checks
in place to ensure val is non-zero.

Cc: stable@vger.kernel.org
Fixes: 7b8d045e497a ("iio: adc: ad7124: allow more than 8 channels")
Signed-off-by: Zicheng Qu <quzicheng@huawei.com>
Reviewed-by: Nuno Sa <nuno.sa@analog.com>
Link: https://patch.msgid.link/20241022134330.574601-1-quzicheng@huawei.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

staging: iio: frequency: ad9832: fix division by zero in ad9832_calc_freqreg()

commit 6bd301819f8f69331a55ae2336c8b111fc933f3d upstream.

In the ad9832_write_frequency() function, clk_get_rate() might return 0.
This can lead to a division by zero when calling ad9832_calc_freqreg().
The check if (fout > (clk_get_rate(st->mclk) / 2)) does not protect
against the case when fout is 0. The ad9832_write_frequency() function
is called from ad9832_write(), and fout is derived from a text buffer,
which can contain any value.

Link: https://lore.kernel.org/all/2024100904-CVE-2024-47663-9bdc@gregkh/
Fixes: ea707584bac1 ("Staging: IIO: DDS: AD9832 / AD9835 driver")
Cc: stable@vger.kernel.org
Signed-off-by: Zicheng Qu <quzicheng@huawei.com>
Reviewed-by: Nuno Sa <nuno.sa@analog.com>
Reviewed-by: Dan Carpenter <dan.carpenter@linaro.org>
Link: https://patch.msgid.link/20241022134354.574614-1-quzicheng@huawei.com
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

wifi: iwlegacy: Clear stale interrupts before resuming device

commit 07c90acb071b9954e1fecb1e4f4f13d12c544b34 upstream.

iwl4965 fails upon resume from hibernation on my laptop. The reason
seems to be a stale interrupt which isn't being cleared out before
interrupts are enabled. We end up with a race beween the resume
trying to bring things back up, and the restart work (queued form
the interrupt handler) trying to bring things down. Eventually
the whole thing blows up.

Fix the problem by clearing out any stale interrupts before
interrupts get enabled during resume.

Here's a debug log of the indicent:
[   12.042589] ieee80211 phy0: il_isr ISR inta 0x00000080, enabled 0xaa00008b, fh 0x00000000
[   12.042625] ieee80211 phy0: il4965_irq_tasklet inta 0x00000080, enabled 0x00000000, fh 0x00000000
[   12.042651] iwl4965 0000:10:00.0: RF_KILL bit toggled to enable radio.
[   12.042653] iwl4965 0000:10:00.0: On demand firmware reload
[   12.042690] ieee80211 phy0: il4965_irq_tasklet End inta 0x00000000, enabled 0xaa00008b, fh 0x00000000, flags 0x00000282
[   12.052207] ieee80211 phy0: il4965_mac_start enter
[   12.052212] ieee80211 phy0: il_prep_station Add STA to driver ID 31: ff:ff:ff:ff:ff:ff
[   12.052244] ieee80211 phy0: il4965_set_hw_ready hardware  ready
[   12.052324] ieee80211 phy0: il_apm_init Init card's basic functions
[   12.052348] ieee80211 phy0: il_apm_init L1 Enabled; Disabling L0S
[   12.055727] ieee80211 phy0: il4965_load_bsm Begin load bsm
[   12.056140] ieee80211 phy0: il4965_verify_bsm Begin verify bsm
[   12.058642] ieee80211 phy0: il4965_verify_bsm BSM bootstrap uCode image OK
[   12.058721] ieee80211 phy0: il4965_load_bsm BSM write complete, poll 1 iterations
[   12.058734] ieee80211 phy0: __il4965_up iwl4965 is coming up
[   12.058737] ieee80211 phy0: il4965_mac_start Start UP work done.
[   12.058757] ieee80211 phy0: __il4965_down iwl4965 is going down
[   12.058761] ieee80211 phy0: il_scan_cancel_timeout Scan cancel timeout
[   12.058762] ieee80211 phy0: il_do_scan_abort Not performing scan to abort
[   12.058765] ieee80211 phy0: il_clear_ucode_stations Clearing ucode stations in driver
[   12.058767] ieee80211 phy0: il_clear_ucode_stations No active stations found to be cleared
[   12.058819] ieee80211 phy0: _il_apm_stop Stop card, put in low power state
[   12.058827] ieee80211 phy0: _il_apm_stop_master stop master
[   12.058864] ieee80211 phy0: il4965_clear_free_frames 0 frames on pre-allocated heap on clear.
[   12.058869] ieee80211 phy0: Hardware restart was requested
[   16.132299] iwl4965 0000:10:00.0: START_ALIVE timeout after 4000ms.
[   16.132303] ------------[ cut here ]------------
[   16.132304] Hardware became unavailable upon resume. This could be a software issue prior to suspend or a hardware issue.
[   16.132338] WARNING: CPU: 0 PID: 181 at net/mac80211/util.c:1826 ieee80211_reconfig+0x8f/0x14b0 [mac80211]
[   16.132390] Modules linked in: ctr ccm sch_fq_codel xt_tcpudp xt_multiport xt_state iptable_filter iptable_nat nf_nat nf_conntrack nf_defrag_ipv4 ip_tables x_tables binfmt_misc joydev mousedev btusb btrtl btintel btbcm bluetooth ecdh_generic ecc iTCO_wdt i2c_dev iwl4965 iwlegacy coretemp snd_hda_codec_analog pcspkr psmouse mac80211 snd_hda_codec_generic libarc4 sdhci_pci cqhci sha256_generic sdhci libsha256 firewire_ohci snd_hda_intel snd_intel_dspcfg mmc_core snd_hda_codec snd_hwdep firewire_core led_class iosf_mbi snd_hda_core uhci_hcd lpc_ich crc_itu_t cfg80211 ehci_pci ehci_hcd snd_pcm usbcore mfd_core rfkill snd_timer snd usb_common soundcore video parport_pc parport intel_agp wmi intel_gtt backlight e1000e agpgart evdev
[   16.132456] CPU: 0 UID: 0 PID: 181 Comm: kworker/u8:6 Not tainted 6.11.0-cl+ #143
[   16.132460] Hardware name: Hewlett-Packard HP Compaq 6910p/30BE, BIOS 68MCU Ver. F.19 07/06/2010
[   16.132463] Workqueue: async async_run_entry_fn
[   16.132469] RIP: 0010:ieee80211_reconfig+0x8f/0x14b0 [mac80211]
[   16.132501] Code: da 02 00 00 c6 83 ad 05 00 00 00 48 89 df e8 98 1b fc ff 85 c0 41 89 c7 0f 84 e9 02 00 00 48 c7 c7 a0 e6 48 a0 e8 d1 77 c4 e0 <0f> 0b eb 2d 84 c0 0f 85 8b 01 00 00 c6 87 ad 05 00 00 00 e8 69 1b
[   16.132504] RSP: 0018:ffffc9000029fcf0 EFLAGS: 00010282
[   16.132507] RAX: 0000000000000000 RBX: ffff8880072008e0 RCX: 0000000000000001
[   16.132509] RDX: ffffffff81f21a18 RSI: 0000000000000086 RDI: 0000000000000001
[   16.132510] RBP: ffff8880072003c0 R08: 0000000000000000 R09: 0000000000000003
[   16.132512] R10: 0000000000000000 R11: ffff88807e5b0000 R12: 0000000000000001
[   16.132514] R13: 0000000000000000 R14: 0000000000000000 R15: 00000000ffffff92
[   16.132515] FS:  0000000000000000(0000) GS:ffff88807c200000(0000) knlGS:0000000000000000
[   16.132517] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   16.132519] CR2: 000055dd43786c08 CR3: 000000000978f000 CR4: 00000000000006f0
[   16.132521] Call Trace:
[   16.132525]  <TASK>
[   16.132526]  ? __warn+0x77/0x120
[   16.132532]  ? ieee80211_reconfig+0x8f/0x14b0 [mac80211]
[   16.132564]  ? report_bug+0x15c/0x190
[   16.132568]  ? handle_bug+0x36/0x70
[   16.132571]  ? exc_invalid_op+0x13/0x60
[   16.132573]  ? asm_exc_invalid_op+0x16/0x20
[   16.132579]  ? ieee80211_reconfig+0x8f/0x14b0 [mac80211]
[   16.132611]  ? snd_hdac_bus_init_cmd_io+0x24/0x200 [snd_hda_core]
[   16.132617]  ? pick_eevdf+0x133/0x1c0
[   16.132622]  ? check_preempt_wakeup_fair+0x70/0x90
[   16.132626]  ? wakeup_preempt+0x4a/0x60
[   16.132628]  ? ttwu_do_activate.isra.0+0x5a/0x190
[   16.132632]  wiphy_resume+0x79/0x1a0 [cfg80211]
[   16.132675]  ? wiphy_suspend+0x2a0/0x2a0 [cfg80211]
[   16.132697]  dpm_run_callback+0x75/0x1b0
[   16.132703]  device_resume+0x97/0x200
[   16.132707]  async_resume+0x14/0x20
[   16.132711]  async_run_entry_fn+0x1b/0xa0
[   16.132714]  process_one_work+0x13d/0x350
[   16.132718]  worker_thread+0x2be/0x3d0
[   16.132722]  ? cancel_delayed_work_sync+0x70/0x70
[   16.132725]  kthread+0xc0/0xf0
[   16.132729]  ? kthread_park+0x80/0x80
[   16.132732]  ret_from_fork+0x28/0x40
[   16.132735]  ? kthread_park+0x80/0x80
[   16.132738]  ret_from_fork_asm+0x11/0x20
[   16.132741]  </TASK>
[   16.132742] ---[ end trace 0000000000000000 ]---
[   16.132930] ------------[ cut here ]------------
[   16.132932] WARNING: CPU: 0 PID: 181 at net/mac80211/driver-ops.c:41 drv_stop+0xe7/0xf0 [mac80211]
[   16.132957] Modules linked in: ctr ccm sch_fq_codel xt_tcpudp xt_multiport xt_state iptable_filter iptable_nat nf_nat nf_conntrack nf_defrag_ipv4 ip_tables x_tables binfmt_misc joydev mousedev btusb btrtl btintel btbcm bluetooth ecdh_generic ecc iTCO_wdt i2c_dev iwl4965 iwlegacy coretemp snd_hda_codec_analog pcspkr psmouse mac80211 snd_hda_codec_generic libarc4 sdhci_pci cqhci sha256_generic sdhci libsha256 firewire_ohci snd_hda_intel snd_intel_dspcfg mmc_core snd_hda_codec snd_hwdep firewire_core led_class iosf_mbi snd_hda_core uhci_hcd lpc_ich crc_itu_t cfg80211 ehci_pci ehci_hcd snd_pcm usbcore mfd_core rfkill snd_timer snd usb_common soundcore video parport_pc parport intel_agp wmi intel_gtt backlight e1000e agpgart evdev
[   16.133014] CPU: 0 UID: 0 PID: 181 Comm: kworker/u8:6 Tainted: G        W          6.11.0-cl+ #143
[   16.133018] Tainted: [W]=WARN
[   16.133019] Hardware name: Hewlett-Packard HP Compaq 6910p/30BE, BIOS 68MCU Ver. F.19 07/06/2010
[   16.133021] Workqueue: async async_run_entry_fn
[   16.133025] RIP: 0010:drv_stop+0xe7/0xf0 [mac80211]
[   16.133048] Code: 48 85 c0 74 0e 48 8b 78 08 89 ea 48 89 de e8 e0 87 04 00 65 ff 0d d1 de c4 5f 0f 85 42 ff ff ff e8 be 52 c2 e0 e9 38 ff ff ff <0f> 0b 5b 5d c3 0f 1f 40 00 41 54 49 89 fc 55 53 48 89 f3 2e 2e 2e
[   16.133050] RSP: 0018:ffffc9000029fc50 EFLAGS: 00010246
[   16.133053] RAX: 0000000000000000 RBX: ffff8880072008e0 RCX: ffff88800377f6c0
[   16.133054] RDX: 0000000000000001 RSI: 0000000000000000 RDI: ffff8880072008e0
[   16.133056] RBP: 0000000000000000 R08: ffffffff81f238d8 R09: 0000000000000000
[   16.133058] R10: ffff8880080520f0 R11: 0000000000000000 R12: ffff888008051c60
[   16.133060] R13: ffff8880072008e0 R14: 0000000000000000 R15: ffff8880072011d8
[   16.133061] FS:  0000000000000000(0000) GS:ffff88807c200000(0000) knlGS:0000000000000000
[   16.133063] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[   16.133065] CR2: 000055dd43786c08 CR3: 000000000978f000 CR4: 00000000000006f0
[   16.133067] Call Trace:
[   16.133069]  <TASK>
[   16.133070]  ? __warn+0x77/0x120
[   16.133075]  ? drv_stop+0xe7/0xf0 [mac80211]
[   16.133098]  ? report_bug+0x15c/0x190
[   16.133100]  ? handle_bug+0x36/0x70
[   16.133103]  ? exc_invalid_op+0x13/0x60
[   16.133105]  ? asm_exc_invalid_op+0x16/0x20
[   16.133109]  ? drv_stop+0xe7/0xf0 [mac80211]
[   16.133132]  ieee80211_do_stop+0x55a/0x810 [mac80211]
[   16.133161]  ? fq_codel_reset+0xa5/0xc0 [sch_fq_codel]
[   16.133164]  ieee80211_stop+0x4f/0x180 [mac80211]
[   16.133192]  __dev_close_many+0xa2/0x120
[   16.133195]  dev_close_many+0x90/0x150
[   16.133198]  dev_close+0x5d/0x80
[   16.133200]  cfg80211_shutdown_all_interfaces+0x40/0xe0 [cfg80211]
[   16.133223]  wiphy_resume+0xb2/0x1a0 [cfg80211]
[   16.133247]  ? wiphy_suspend+0x2a0/0x2a0 [cfg80211]
[   16.133269]  dpm_run_callback+0x75/0x1b0
[   16.133273]  device_resume+0x97/0x200
[   16.133277]  async_resume+0x14/0x20
[   16.133280]  async_run_entry_fn+0x1b/0xa0
[   16.133283]  process_one_work+0x13d/0x350
[   16.133287]  worker_thread+0x2be/0x3d0
[   16.133290]  ? cancel_delayed_work_sync+0x70/0x70
[   16.133294]  kthread+0xc0/0xf0
[   16.133296]  ? kthread_park+0x80/0x80
[   16.133299]  ret_from_fork+0x28/0x40
[   16.133302]  ? kthread_park+0x80/0x80
[   16.133304]  ret_from_fork_asm+0x11/0x20
[   16.133307]  </TASK>
[   16.133308] ---[ end trace 0000000000000000 ]---
[   16.133335] ieee80211 phy0: PM: dpm_run_callback(): wiphy_resume [cfg80211] returns -110
[   16.133360] ieee80211 phy0: PM: failed to restore async: error -110

Cc: stable@vger.kernel.org
Cc: Stanislaw Gruszka <stf_xl@wp.pl>
Cc: Kalle Valo <kvalo@kernel.org>
Cc: linux-wireless@vger.kernel.org
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Acked-by: Stanislaw Gruszka <stf_xl@wp.pl>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://patch.msgid.link/20241001200745.8276-1-ville.syrjala@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

wifi: cfg80211: clear wdev->cqm_config pointer on free

commit d5fee261dfd9e17b08b1df8471ac5d5736070917 upstream.

When we free wdev->cqm_config when unregistering, we also
need to clear out the pointer since the same wdev/netdev
may get re-registered in another network namespace, then
destroyed later, running this code again, which results in
a double-free.

Reported-by: syzbot+36218cddfd84b5cc263e@syzkaller.appspotmail.com
Fixes: 37c20b2effe9 ("wifi: cfg80211: fix cqm_config access race")
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20241022161742.7c34b2037726.I121b9cdb7eb180802eafc90b493522950d57ee18@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

wifi: ath10k: Fix memory leak in management tx

commit e15d84b3bba187aa372dff7c58ce1fd5cb48a076 upstream.

In the current logic, memory is allocated for storing the MSDU context
during management packet TX but this memory is not being freed during
management TX completion. Similar leaks are seen in the management TX
cleanup logic.

Kmemleak reports this problem as below,

unreferenced object 0xffffff80b64ed250 (size 16):
  comm "kworker/u16:7", pid 148, jiffies 4294687130 (age 714.199s)
  hex dump (first 16 bytes):
    00 2b d8 d8 80 ff ff ff c4 74 e9 fd 07 00 00 00  .+.......t......
  backtrace:
    [<ffffffe6e7b245dc>] __kmem_cache_alloc_node+0x1e4/0x2d8
    [<ffffffe6e7adde88>] kmalloc_trace+0x48/0x110
    [<ffffffe6bbd765fc>] ath10k_wmi_tlv_op_gen_mgmt_tx_send+0xd4/0x1d8 [ath10k_core]
    [<ffffffe6bbd3eed4>] ath10k_mgmt_over_wmi_tx_work+0x134/0x298 [ath10k_core]
    [<ffffffe6e78d5974>] process_scheduled_works+0x1ac/0x400
    [<ffffffe6e78d60b8>] worker_thread+0x208/0x328
    [<ffffffe6e78dc890>] kthread+0x100/0x1c0
    [<ffffffe6e78166c0>] ret_from_fork+0x10/0x20

Free the memory during completion and cleanup to fix the leak.

Protect the mgmt_pending_tx idr_remove() operation in
ath10k_wmi_tlv_op_cleanup_mgmt_tx_send() using ar->data_lock similar to
other instances.

Tested-on: WCN3990 hw1.0 SNOC WLAN.HL.2.0-01387-QCAHLSWMTPLZ-1

Fixes: dc405152bb64 ("ath10k: handle mgmt tx completion event")
Fixes: c730c477176a ("ath10k: Remove msdu from idr when management pkt send fails")
Cc: stable@vger.kernel.org
Signed-off-by: Manikanta Pubbisetty <quic_mpubbise@quicinc.com>
Link: https://patch.msgid.link/20241015064103.6060-1-quic_mpubbise@quicinc.com
Signed-off-by: Jeff Johnson <quic_jjohnson@quicinc.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

wifi: mac80211: do not pass a stopped vif to the driver in .get_txpower

commit 393b6bc174b0dd21bb2a36c13b36e62fc3474a23 upstream.

Avoid potentially crashing in the driver because of uninitialized private data

Fixes: 5b3dc42b1b0d ("mac80211: add support for driver tx power reporting")
Cc: stable@vger.kernel.org
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Link: https://patch.msgid.link/20241002095630.22431-1-nbd@nbd.name
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Revert "selftests/mm: replace atomic_bool with pthread_barrier_t"

commit 3673167a3a07f25b3f06754d69f406edea65543a upstream.

This reverts commit e61ef21e27e8deed8c474e9f47f4aa7bc37e138c.

uffd_poll_thread may be called by other tests that do not initialize the
pthread_barrier, so this approach is not correct.  This will revert to
using atomic_bool instead.

Link: https://lkml.kernel.org/r/20241018171734.2315053-3-edliaw@google.com
Fixes: e61ef21e27e8 ("selftests/mm: replace atomic_bool with pthread_barrier_t")
Signed-off-by: Edward Liaw <edliaw@google.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Revert "selftests/mm: fix deadlock for fork after pthread_create on ARM"

commit 5bb1f4c9340e01003b00b94d539eadb0da88f48e upstream.

Patch series "selftests/mm: revert pthread_barrier change"

On Android arm, pthread_create followed by a fork caused a deadlock in
the case where the fork required work to be completed by the created
thread.

The previous patches incorrectly assumed that the parent would
always initialize the pthread_barrier for the child thread.  This
reverts the change and replaces the fix for wp-fork-with-event with the
original use of atomic_bool.

This patch (of 3):

This reverts commit e142cc87ac4ec618f2ccf5f68aedcd6e28a59d9d.

fork_event_consumer may be called by other tests that do not initialize
the pthread_barrier, so this approach is not correct.  The subsequent
patch will revert to using atomic_bool instead.

Link: https://lkml.kernel.org/r/20241018171734.2315053-1-edliaw@google.com
Link: https://lkml.kernel.org/r/20241018171734.2315053-2-edliaw@google.com
Fixes: e142cc87ac4e ("fix deadlock for fork after pthread_create on ARM")
Signed-off-by: Edward Liaw <edliaw@google.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

Revert "driver core: Fix uevent_show() vs driver detach race"

commit 9a71892cbcdb9d1459c84f5a4c722b14354158a5 upstream.

This reverts commit 15fffc6a5624b13b428bb1c6e9088e32a55eb82c.

This commit causes a regression, so revert it for now until it can come
back in a way that works for everyone.

Link: https://lore.kernel.org/all/172790598832.1168608.4519484276671503678.stgit@dwillia2-xfh.jf.intel.com/
Fixes: 15fffc6a5624 ("driver core: Fix uevent_show() vs driver detach race")
Cc: stable <stable@kernel.org>
Cc: Ashish Sangwan <a.sangwan@samsung.com>
Cc: Namjae Jeon <namjae.jeon@samsung.com>
Cc: Dirk Behme <dirk.behme@de.bosch.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Rafael J. Wysocki <rafael@kernel.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: Use pm_runtime_get to prevent RPM on unsupported systems

commit 31004740e42846a6f0bb255e6348281df3eb8032 upstream.

Use pm_runtime_put in the remove function and pm_runtime_get to disable
RPM on platforms that don't support runtime D3, as re-enabling it through
sysfs auto power control may cause the controller to malfunction. This
can lead to issues such as hotplug devices not being detected due to
failed interrupt generation.

Fixes: a5d6264b638e ("xhci: Enable RPM on controllers that support low-power states")
Cc: stable <stable@kernel.org>
Signed-off-by: Basavaraj Natikar <Basavaraj.Natikar@amd.com>
Reviewed-by: Mario Limonciello <mario.limonciello@amd.com>
Link: https://lore.kernel.org/r/20241024133718.723846-1-Basavaraj.Natikar@amd.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

xhci: Fix Link TRB DMA in command ring stopped completion event

commit 075919f6df5dd82ad0b1894898b315fbb3c29b84 upstream.

During the aborting of a command, the software receives a command
completion event for the command ring stopped, with the TRB pointing
to the next TRB after the aborted command.

If the command we abort is located just before the Link TRB in the
command ring, then during the 'command ring stopped' completion event,
the xHC gives the Link TRB in the event's cmd DMA, which causes a
mismatch in handling command completion event.

To address this situation, move the 'command ring stopped' completion
event check slightly earlier, since the specific command it stopped
on isn't of significant concern.

Fixes: 7f84eef0dafb ("USB: xhci: No-op command queueing and irq handler.")
Cc: stable@vger.kernel.org
Signed-off-by: Faisal Hassan <quic_faisalh@quicinc.com>
Acked-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Link: https://lore.kernel.org/r/20241022155631.1185-1-quic_faisalh@quicinc.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

phy: qcom: qmp-usb-legacy: fix NULL-deref on runtime suspend

commit 29240130ab77c80bea1464317ae2a5fd29c16a0c upstream.

Commit 413db06c05e7 ("phy: qcom-qmp-usb: clean up probe initialisation")
removed most users of the platform device driver data from the
qcom-qmp-usb driver, but mistakenly also removed the initialisation
despite the data still being used in the runtime PM callbacks. This bug
was later reproduced when the driver was copied to create the
qmp-usb-legacy driver.

Restore the driver data initialisation at probe to avoid a NULL-pointer
dereference on runtime suspend.

Apparently no one uses runtime PM, which currently needs to be enabled
manually through sysfs, with these drivers.

Fixes: e464a3180a43 ("phy: qcom-qmp-usb: split off the legacy USB+dp_com support")
Cc: stable@vger.kernel.org # 6.6
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20240911115253.10920-3-johan+linaro@kernel.org
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

phy: qcom: qmp-usb: fix NULL-deref on runtime suspend

commit bd9e4d4a3b127686efc60096271b0a44c3100061 upstream.

Commit 413db06c05e7 ("phy: qcom-qmp-usb: clean up probe initialisation")
removed most users of the platform device driver data, but mistakenly
also removed the initialisation despite the data still being used in the
runtime PM callbacks.

Restore the driver data initialisation at probe to avoid a NULL-pointer
dereference on runtime suspend.

Apparently no one uses runtime PM, which currently needs to be enabled
manually through sysfs, with this driver.

Fixes: 413db06c05e7 ("phy: qcom-qmp-usb: clean up probe initialisation")
Cc: stable@vger.kernel.org # 6.2
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20240911115253.10920-2-johan+linaro@kernel.org
Signed-off-by: Vinod Koul <vkoul@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: typec: qcom-pmic-typec: use fwnode_handle_put() to release fwnodes

commit 7f02b8a5b602098f2901166e7e4d583acaed872a upstream.

The right function to release a fwnode acquired via
device_get_named_child_node() is fwnode_handle_put(), and not
fwnode_remove_software_node(), as no software node is being handled.

Replace the calls to fwnode_remove_software_node() with
fwnode_handle_put() in qcom_pmic_typec_probe() and
qcom_pmic_typec_remove().

Cc: stable@vger.kernel.org
Fixes: a4422ff22142 ("usb: typec: qcom: Add Qualcomm PMIC Type-C driver")
Suggested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Signed-off-by: Javier Carrasco <javier.carrasco.cruz@gmail.com>
Acked-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20241020-qcom_pmic_typec-fwnode_remove-v2-1-7054f3d2e215@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: typec: fix unreleased fwnode_handle in typec_port_register_altmodes()

commit 9581acb91eaf5bbe70086bbb6fca808220d358ba upstream.

The 'altmodes_node' fwnode_handle is never released after it is no
longer required, which leaks the resource.

Add the required call to fwnode_handle_put() when 'altmodes_node' is no
longer required.

Cc: stable@vger.kernel.org
Fixes: 7b458a4c5d73 ("usb: typec: Add typec_port_register_altmodes()")
Reviewed-by: Heikki Krogerus <heikki.krogerus@linux.intel.com>
Signed-off-by: Javier Carrasco <javier.carrasco.cruz@gmail.com>
Link: https://lore.kernel.org/r/20241021-typec-class-fwnode_handle_put-v2-1-3281225d3d27@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usb: phy: Fix API devm_usb_put_phy() can not release the phy

commit fdce49b5da6e0fb6d077986dec3e90ef2b094b50 upstream.

For devm_usb_put_phy(), its comment says it needs to invoke usb_put_phy()
to release the phy, but it does not do that actually, so it can not fully
undo what the API devm_usb_get_phy() does, that is wrong, fixed by using
devres_release() instead of devres_destroy() within the API.

Fixes: cedf8602373a ("usb: phy: move bulk of otg/otg.c to phy/phy.c")
Cc: stable@vger.kernel.org
Signed-off-by: Zijun Hu <quic_zijuhu@quicinc.com>
Link: https://lore.kernel.org/r/20241020-usb_phy_fix-v1-1-7f79243b8e1e@quicinc.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

usbip: tools: Fix detach_port() invalid port error path

commit e7cd4b811c9e019f5acbce85699c622b30194c24 upstream.

The detach_port() doesn't return error
when detach is attempted on an invalid port.

Fixes: 40ecdeb1a187 ("usbip: usbip_detach: fix to check for invalid ports")
Cc: stable@vger.kernel.org
Reviewed-by: Hongren Zheng <i@zenithal.me>
Reviewed-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: Zongmin Zhou <zhouzongmin@kylinos.cn>
Link: https://lore.kernel.org/r/20241024022700.1236660-1-min_halo@163.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

ALSA: usb-audio: Add quirks for Dell WD19 dock

commit 4413665dd6c528b31284119e3571c25f371e1c36 upstream.

The WD19 family of docks has the same audio chipset as the WD15. This
change enables jack detection on the WD19.

We don't need the dell_dock_mixer_init quirk for the WD19. It is only
needed because of the dell_alc4020_map quirk for the WD15 in
mixer_maps.c, which disables the volume controls. Even for the WD15,
this quirk was apparently only needed when the dock firmware was not
updated.

Signed-off-by: Jan Schär <jan@jschaer.ch>
Cc: <stable@vger.kernel.org>
Link: https://patch.msgid.link/20241029221249.15661-1-jan@jschaer.ch
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

rcu-tasks: Fix access non-existent percpu rtpcp variable in rcu_tasks_need_gpcb()

[ Upstream commit fd70e9f1d85f5323096ad313ba73f5fe3d15ea41 ]

For kernels built with CONFIG_FORCE_NR_CPUS=y, the nr_cpu_ids is
defined as NR_CPUS instead of the number of possible cpus, this
will cause the following system panic:

smpboot: Allowing 4 CPUs, 0 hotplug CPUs
...
setup_percpu: NR_CPUS:512 nr_cpumask_bits:512 nr_cpu_ids:512 nr_node_ids:1
...
BUG: unable to handle page fault for address: ffffffff9911c8c8
Oops: 0000 [#1] PREEMPT SMP PTI
CPU: 0 PID: 15 Comm: rcu_tasks_trace Tainted: G W
6.6.21 #1 5dc7acf91a5e8e9ac9dcfc35bee0245691283ea6
RIP: 0010:rcu_tasks_need_gpcb+0x25d/0x2c0
RSP: 0018:ffffa371c00a3e60 EFLAGS: 00010082
CR2: ffffffff9911c8c8 CR3: 000000040fa20005 CR4: 00000000001706f0
Call Trace:
<TASK>
? __die+0x23/0x80
? page_fault_oops+0xa4/0x180
? exc_page_fault+0x152/0x180
? asm_exc_page_fault+0x26/0x40
? rcu_tasks_need_gpcb+0x25d/0x2c0
? __pfx_rcu_tasks_kthread+0x40/0x40
rcu_tasks_one_gp+0x69/0x180
rcu_tasks_kthread+0x94/0xc0
kthread+0xe8/0x140
? __pfx_kthread+0x40/0x40
ret_from_fork+0x34/0x80
? __pfx_kthread+0x40/0x40
ret_from_fork_asm+0x1b/0x80
</TASK>

Considering that there may be holes in the CPU numbers, use the
maximum possible cpu number, instead of nr_cpu_ids, for configuring
enqueue and dequeue limits.

[ neeraj.upadhyay: Fix htmldocs build error reported by Stephen Rothwell ]

Closes: https://lore.kernel.org/linux-input/CALMA0xaTSMN+p4xUXkzrtR5r6k7hgoswcaXx7baR_z9r5jjskw@mail.gmail.com/T/#u
Reported-by: Zhixu Liu <zhixu.liu@gmail.com>
Signed-off-by: Zqiang <qiang.zhang1211@gmail.com>
Signed-off-by: Neeraj Upadhyay <neeraj.upadhyay@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

rcu-tasks: Initialize data to eliminate RCU-tasks/do_exit() deadlocks

[ Upstream commit 46faf9d8e1d52e4a91c382c6c72da6bd8e68297b ]

Holding a mutex across synchronize_rcu_tasks() and acquiring
that same mutex in code called from do_exit() after its call to
exit_tasks_rcu_start() but before its call to exit_tasks_rcu_stop()
results in deadlock.  This is by design, because tasks that are far
enough into do_exit() are no longer present on the tasks list, making
it a bit difficult for RCU Tasks to find them, let alone wait on them
to do a voluntary context switch.  However, such deadlocks are becoming
more frequent.  In addition, lockdep currently does not detect such
deadlocks and they can be difficult to reproduce.

In addition, if a task voluntarily context switches during that time
(for example, if it blocks acquiring a mutex), then this task is in an
RCU Tasks quiescent state.  And with some adjustments, RCU Tasks could
just as well take advantage of that fact.

This commit therefore initializes the data structures that will be needed
to rely on these quiescent states and to eliminate these deadlocks.

Link: https://lore.kernel.org/all/20240118021842.290665-1-chenzhongjin@huawei.com/
Reported-by: Chen Zhongjin <chenzhongjin@huawei.com>
Reported-by: Yang Jihong <yangjihong1@huawei.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Tested-by: Yang Jihong <yangjihong1@huawei.com>
Tested-by: Chen Zhongjin <chenzhongjin@huawei.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Stable-dep-of: fd70e9f1d85f ("rcu-tasks: Fix access non-existent percpu rtpcp variable in rcu_tasks_need_gpcb()")
Signed-off-by: Sasha Levin <sashal@kernel.org>

rcu-tasks: Add data to eliminate RCU-tasks/do_exit() deadlocks

[ Upstream commit bfe93930ea1ea3c6c115a7d44af6e4fea609067e ]

Holding a mutex across synchronize_rcu_tasks() and acquiring
that same mutex in code called from do_exit() after its call to
exit_tasks_rcu_start() but before its call to exit_tasks_rcu_stop()
results in deadlock.  This is by design, because tasks that are far
enough into do_exit() are no longer present on the tasks list, making
it a bit difficult for RCU Tasks to find them, let alone wait on them
to do a voluntary context switch.  However, such deadlocks are becoming
more frequent.  In addition, lockdep currently does not detect such
deadlocks and they can be difficult to reproduce.

In addition, if a task voluntarily context switches during that time
(for example, if it blocks acquiring a mutex), then this task is in an
RCU Tasks quiescent state.  And with some adjustments, RCU Tasks could
just as well take advantage of that fact.

This commit therefore adds the data structures that will be needed
to rely on these quiescent states and to eliminate these deadlocks.

Link: https://lore.kernel.org/all/20240118021842.290665-1-chenzhongjin@huawei.com/
Reported-by: Chen Zhongjin <chenzhongjin@huawei.com>
Reported-by: Yang Jihong <yangjihong1@huawei.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Tested-by: Yang Jihong <yangjihong1@huawei.com>
Tested-by: Chen Zhongjin <chenzhongjin@huawei.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>
Stable-dep-of: fd70e9f1d85f ("rcu-tasks: Fix access non-existent percpu rtpcp variable in rcu_tasks_need_gpcb()")
Signed-off-by: Sasha Levin <sashal@kernel.org>

rcu-tasks: Pull sampling of ->percpu_dequeue_lim out of loop

[ Upstream commit e62d8ae4620865411d1b2347980aa28ccf891a3d ]

The rcu_tasks_need_gpcb() samples ->percpu_dequeue_lim as part of the
condition clause of a "for" loop, which is a bit confusing.  This commit
therefore hoists this sampling out of the loop, using the result loaded
in the condition clause.

So why does this work in the face of a concurrent switch from single-CPU
queueing to per-CPU queueing?

o The call_rcu_tasks_generic() that makes the change has already
enqueued its callback, which means that all of the other CPU's
callback queues are empty.

o For the call_rcu_tasks_generic() that first notices
the switch to per-CPU queues, the smp_store_release()
used to update ->percpu_enqueue_lim pairs with the
raw_spin_trylock_rcu_node()'s full barrier that is
between the READ_ONCE(rtp->percpu_enqueue_shift) and the
rcu_segcblist_enqueue() that enqueues the callback.

o Because this CPU's queue is empty (unless it happens to
be the original single queue, in which case there is no
need for synchronization), this call_rcu_tasks_generic()
will do an irq_work_queue() to schedule a handler for the
needed rcuwait_wake_up() call. This call will be ordered
after the first call_rcu_tasks_generic() function's change to
->percpu_dequeue_lim.

o This rcuwait_wake_up() will either happen before or after the
set_current_state() in rcuwait_wait_event().  If it happens
before, the "condition" argument's call to rcu_tasks_need_gpcb()
will be ordered after the original change, and all callbacks on
all CPUs will be visible.  Otherwise, if it happens after, then
the grace-period kthread's state will be set back to running,
which will result in a later call to rcuwait_wait_event() and
thus to rcu_tasks_need_gpcb(), which will again see the change.

So it all works out.

Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Frederic Weisbecker <frederic@kernel.org>
Stable-dep-of: fd70e9f1d85f ("rcu-tasks: Fix access non-existent percpu rtpcp variable in rcu_tasks_need_gpcb()")
Signed-off-by: Sasha Levin <sashal@kernel.org>

USB: gadget: dummy-hcd: Fix "task hung" problem

[ Upstream commit 5189df7b8088268012882c220d6aca4e64981348 ]

The syzbot fuzzer has been encountering "task hung" problems ever
since the dummy-hcd driver was changed to use hrtimers instead of
regular timers.  It turns out that the problems are caused by a subtle
difference between the timer_pending() and hrtimer_active() APIs.

The changeover blindly replaced the first by the second.  However,
timer_pending() returns True when the timer is queued but not when its
callback is running, whereas hrtimer_active() returns True when the
hrtimer is queued _or_ its callback is running.  This difference
occasionally caused dummy_urb_enqueue() to think that the callback
routine had not yet started when in fact it was almost finished.  As a
result the hrtimer was not restarted, which made it impossible for the
driver to dequeue later the URB that was just enqueued.  This caused
usb_kill_urb() to hang, and things got worse from there.

Since hrtimers have no API for telling when they are queued and the
callback isn't running, the driver must keep track of this for itself.
That's what this patch does, adding a new "timer_pending" flag and
setting or clearing it at the appropriate times.

Reported-by: syzbot+f342ea16c9d06d80b585@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/linux-usb/6709234e.050a0220.3e960.0011.GAE@google.com/
Tested-by: syzbot+f342ea16c9d06d80b585@syzkaller.appspotmail.com
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Fixes: a7f3813e589f ("usb: gadget: dummy_hcd: Switch to hrtimer transfer scheduler")
Cc: Marcello Sylvester Bauer <sylv@sylv.io>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/2dab644e-ef87-4de8-ac9a-26f100b2c609@rowland.harvard.edu
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

usb: gadget: dummy_hcd: execute hrtimer callback in softirq context

[ Upstream commit 9313d139aa25e572d860f6f673b73a20f32d7f93 ]

Commit a7f3813e589f ("usb: gadget: dummy_hcd: Switch to hrtimer transfer
scheduler") switched dummy_hcd to use hrtimer and made the timer's
callback be executed in the hardirq context.

With that change, __usb_hcd_giveback_urb now gets executed in the hardirq
context, which causes problems for KCOV and KMSAN.

One problem is that KCOV now is unable to collect coverage from
the USB code that gets executed from the dummy_hcd's timer callback,
as KCOV cannot collect coverage in the hardirq context.

Another problem is that the dummy_hcd hrtimer might get triggered in the
middle of a softirq with KCOV remote coverage collection enabled, and that
causes a WARNING in KCOV, as reported by syzbot. (I sent a separate patch
to shut down this WARNING, but that doesn't fix the other two issues.)

Finally, KMSAN appears to ignore tracking memory copying operations
that happen in the hardirq context, which causes false positive
kernel-infoleaks, as reported by syzbot.

Change the hrtimer in dummy_hcd to execute the callback in the softirq
context.

Reported-by: syzbot+2388cdaeb6b10f0c13ac@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=2388cdaeb6b10f0c13ac
Reported-by: syzbot+17ca2339e34a1d863aad@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=17ca2339e34a1d863aad
Reported-by: syzbot+c793a7eca38803212c61@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=c793a7eca38803212c61
Reported-by: syzbot+1e6e0b916b211bee1bd6@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=1e6e0b916b211bee1bd6
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202406141323.413a90d2-lkp@intel.com
Fixes: a7f3813e589f ("usb: gadget: dummy_hcd: Switch to hrtimer transfer scheduler")
Cc: stable@vger.kernel.org
Acked-by: Marcello Sylvester Bauer <sylv@sylv.io>
Signed-off-by: Andrey Konovalov <andreyknvl@gmail.com>
Reported-by: syzbot+edd9fe0d3a65b14588d5@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=edd9fe0d3a65b14588d5
Link: https://lore.kernel.org/r/20240904013051.4409-1-andrey.konovalov@linux.dev
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

usb: gadget: dummy_hcd: Set transfer interval to 1 microframe

[ Upstream commit 0a723ed3baa941ca4f51d87bab00661f41142835 ]

Currently, the transfer polling interval is set to 1ms, which is the
frame rate of full-speed and low-speed USB. The USB 2.0 specification
introduces microframes (125 microseconds) to improve the timing
precision of data transfers.

Reducing the transfer interval to 1 microframe increases data throughput
for high-speed and super-speed USB communication

Signed-off-by: Marcello Sylvester Bauer <marcello.bauer@9elements.com>
Signed-off-by: Marcello Sylvester Bauer <sylv@sylv.io>
Link: https://lore.kernel.org/r/6295dbb84ca76884551df9eb157cce569377a22c.1712843963.git.sylv@sylv.io
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

usb: gadget: dummy_hcd: Switch to hrtimer transfer scheduler

[ Upstream commit a7f3813e589fd8e2834720829a47b5eb914a9afe ]

The dummy_hcd transfer scheduler assumes that the internal kernel timer
frequency is set to 1000Hz to give a polling interval of 1ms. Reducing
the timer frequency will result in an anti-proportional reduction in
transfer performance. Switch to a hrtimer to decouple this association.

Signed-off-by: Marcello Sylvester Bauer <marcello.bauer@9elements.com>
Signed-off-by: Marcello Sylvester Bauer <sylv@sylv.io>
Reviewed-by: Alan Stern <stern@rowland.harvard.edu>
Link: https://lore.kernel.org/r/57a1c2180ff74661600e010c234d1dbaba1d0d46.1712843963.git.sylv@sylv.io
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

misc: sgi-gru: Don't disable preemption in GRU driver

[ Upstream commit b983b271662bd6104d429b0fd97af3333ba760bf ]

Disabling preemption in the GRU driver is unnecessary, and clashes with
sleeping locks in several code paths.  Remove preempt_disable and
preempt_enable from the GRU driver.

Signed-off-by: Dimitri Sivanich <sivanich@hpe.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

NFS: remove revoked delegation from server's delegation list

[ Upstream commit 7ef60108069b7e3cc66432304e1dd197d5c0a9b5 ]

After the delegation is returned to the NFS server remove it
from the server's delegations list to reduce the time it takes
to scan this list.

Network trace captured while running the below script shows the
time taken to service the CB_RECALL increases gradually due to
the overhead of traversing the delegation list in
nfs_delegation_find_inode_server.

The NFS server in this test is a Solaris server which issues
CB_RECALL when receiving the all-zero stateid in the SETATTR.

mount=/mnt/data
for i in $(seq 1 20)
do
   echo $i
   mkdir $mount/testtarfile$i
   time  tar -C $mount/testtarfile$i -xf 5000_files.tar
done

Signed-off-by: Dai Ngo <dai.ngo@oracle.com>
Reviewed-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <anna.schumaker@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: amd: mvme147: Fix probe banner message

[ Upstream commit 82c5b53140faf89c31ea2b3a0985a2f291694169 ]

Currently this driver prints this line with what looks like
a rogue format specifier when the device is probed:
[    2.840000] eth%d: MVME147 at 0xfffe1800, irq 12, Hardware Address xx:xx:xx:xx:xx:xx

Change the printk() for netdev_info() and move it after the
registration has completed so it prints out the name of the
interface properly.

Signed-off-by: Daniel Palmer <daniel@0x0f.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>

thermal: intel: int340x: processor: Add MMIO RAPL PL4 support

[ Upstream commit 3fb0eea8a1c4be5884e0731ea76cbd3ce126e1f3 ]

Similar to the MSR RAPL interface, MMIO RAPL supports PL4 too, so add
MMIO RAPL PL4d support to the processor_thermal driver.

As a result, the powercap sysfs for MMIO RAPL will show a new "peak
power" constraint.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Reviewed-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://patch.msgid.link/20240930081801.28502-7-rui.zhang@intel.com
[ rjw: Subject and changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

thermal: intel: int340x: processor: Remove MMIO RAPL CPU hotplug support

[ Upstream commit bfc6819e4bf56a55df6178f93241b5845ad672eb ]

CPU0/package0 is always online and the MMIO RAPL driver runs on single
package systems only, so there is no need to handle CPU hotplug in it.

Always register a RAPL package device for package 0 and remove the
unnecessary CPU hotplug support.

Signed-off-by: Zhang Rui <rui.zhang@intel.com>
Reviewed-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Link: https://patch.msgid.link/20240930081801.28502-6-rui.zhang@intel.com
[ rjw: Subject edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

cifs: Fix creating native symlinks pointing to current or parent directory

[ Upstream commit 63271b7d569fbe924bccc7dadc17d3d07a4e5f7a ]

Calling 'ln -s . symlink' or 'ln -s .. symlink' creates symlink pointing to
some object name which ends with U+F029 unicode codepoint. This is because
trailing dot in the object name is replaced by non-ASCII unicode codepoint.

So Linux SMB client currently is not able to create native symlink pointing
to current or parent directory on Windows SMB server which can be read by
either on local Windows server or by any other SMB client which does not
implement compatible-reverse character replacement.

Fix this problem in cifsConvertToUTF16() function which is doing that
character replacement. Function comment already says that it does not need
to handle special cases '.' and '..', but after introduction of native
symlinks in reparse point form, this handling is needed.

Note that this change depends on the previous change
"cifs: Improve creating native symlinks pointing to directory".

Signed-off-by: Pali Rohár <pali@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

cifs: Improve creating native symlinks pointing to directory

[ Upstream commit 3eb40512530e4f64f819d8e723b6f41695dace5a ]

SMB protocol for native symlinks distinguish between symlink to directory
and symlink to file. These two symlink types cannot be exchanged, which
means that symlink of file type pointing to directory cannot be resolved at
all (and vice-versa).

Windows follows this rule for local filesystems (NTFS) and also for SMB.

Linux SMB client currenly creates all native symlinks of file type. Which
means that Windows (and some other SMB clients) cannot resolve symlinks
pointing to directory created by Linux SMB client.

As Linux system does not distinguish between directory and file symlinks,
its API does not provide enough information for Linux SMB client during
creating of native symlinks.

Add some heuristic into the Linux SMB client for choosing the correct
symlink type during symlink creation. Check if the symlink target location
ends with slash, or last path component is dot or dot-dot, and check if the
target location on SMB share exists and is a directory. If at least one
condition is truth then create a new SMB symlink of directory type.
Otherwise create it as file type symlink.

This change improves interoperability with Windows systems. Windows systems
would be able to resolve more SMB symlinks created by Linux SMB client
which points to existing directory.

Signed-off-by: Pali Rohár <pali@kernel.org>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

scsi: scsi_transport_fc: Allow setting rport state to current state

[ Upstream commit d539a871ae47a1f27a609a62e06093fa69d7ce99 ]

The only input fc_rport_set_marginal_state() currently accepts is
"Marginal" when port_state is "Online", and "Online" when the port_state
is "Marginal". It should also allow setting port_state to its current
state, either "Marginal or "Online".

Signed-off-by: Benjamin Marzinski <bmarzins@redhat.com>
Link: https://lore.kernel.org/r/20240917230643.966768-1-bmarzins@redhat.com
Reviewed-by: Ewan D. Milne <emilne@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

fs/ntfs3: Additional check in ntfs_file_release

[ Upstream commit 031d6f608290c847ba6378322d0986d08d1a645a ]

Reported-by: syzbot+8c652f14a0fde76ff11d@syzkaller.appspotmail.com
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

fs/ntfs3: Fix general protection fault in run_is_mapped_full

[ Upstream commit a33fb016e49e37aafab18dc3c8314d6399cb4727 ]

Fixed deleating of a non-resident attribute in ntfs_create_inode()
rollback.

Reported-by: syzbot+9af29acd8f27fbce94bc@syzkaller.appspotmail.com
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

fs/ntfs3: Additional check in ni_clear()

[ Upstream commit d178944db36b3369b78a08ba520de109b89bf2a9 ]

Checking of NTFS_FLAGS_LOG_REPLAYING added to prevent access to
uninitialized bitmap during replay process.

Reported-by: syzbot+3bfd2cc059ab93efcdb4@syzkaller.appspotmail.com
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

fs/ntfs3: Fix possible deadlock in mi_read

[ Upstream commit 03b097099eef255fbf85ea6a786ae3c91b11f041 ]

Mutex lock with another subclass used in ni_lock_dir().

Reported-by: syzbot+bc7ca0ae4591cb2550f9@syzkaller.appspotmail.com
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

fs/ntfs3: Add rough attr alloc_size check

[ Upstream commit c4a8ba334262e9a5c158d618a4820e1b9c12495c ]

Reported-by: syzbot+c6d94bedd910a8216d25@syzkaller.appspotmail.com
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

fs/ntfs3: Stale inode instead of bad

[ Upstream commit 1fd21919de6de245b63066b8ee3cfba92e36f0e9 ]

Fixed the logic of processing inode with wrong sequence number.

Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

fs/ntfs3: Fix warning possible deadlock in ntfs_set_state

[ Upstream commit 5b2db723455a89dc96743d34d8bdaa23a402db2f ]

Use non-zero subkey to skip analyzer warnings.

Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Reported-by: syzbot+c2ada45c23d98d646118@syzkaller.appspotmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

fs/ntfs3: Check if more than chunk-size bytes are written

[ Upstream commit 9931122d04c6d431b2c11b5bb7b10f28584067f0 ]

A incorrectly formatted chunk may decompress into
more than LZNT_CHUNK_SIZE bytes and a index out of bounds
will occur in s_max_off.

Signed-off-by: Andrew Ballance <andrewjballance@gmail.com>
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ntfs3: Add bounds checking to mi_enum_attr()

[ Upstream commit 556bdf27c2dd5c74a9caacbe524b943a6cd42d99 ]

Added bounds checking to make sure that every attr don't stray beyond
valid memory region.

Signed-off-by: lei lu <llfamsec@gmail.com>
Signed-off-by: Konstantin Komarov <almaz.alexandrovich@paragon-software.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

cxl/events: Fix Trace DRAM Event Record

[ Upstream commit 53ab8678e7180834be29cf56cd52825fc3427c02 ]

CXL spec rev 3.0 section 8.2.9.2.1.2 defines the DRAM Event Record.

Fix decode memory event type field of DRAM Event Record.
For e.g. if value is 0x1 it will be reported as an Invalid Address
(General Media Event Record - Memory Event Type) instead of Scrub Media
ECC Error (DRAM Event Record - Memory Event Type) and so on.

Fixes: 2d6c1e6d60ba ("cxl/mem: Trace DRAM Event Record")
Signed-off-by: Shiju Jose <shiju.jose@huawei.com>
Link: https://patch.msgid.link/20241014143003.1170-1-shiju.jose@huawei.com
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

smb: client: set correct device number on nfs reparse points

[ Upstream commit a9de67336a4aa3ff2e706ba023fb5f7ff681a954 ]

Fix major and minor numbers set on special files created with NFS
reparse points.

Signed-off-by: Paulo Alcantara (Red Hat) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

smb: client: fix parsing of device numbers

[ Upstream commit 663f295e35594f4c2584fc68c28546b747b637cd ]

Report correct major and minor numbers from special files created with
NFS reparse points.

Signed-off-by: Paulo Alcantara (Red Hat) <pc@manguebit.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ACPI: CPPC: Make rmw_lock a raw_spin_lock

[ Upstream commit 1c10941e34c5fdc0357e46a25bd130d9cf40b925 ]

The following BUG was triggered:

=============================
[ BUG: Invalid wait context ]
6.12.0-rc2-XXX #406 Not tainted
-----------------------------
kworker/1:1/62 is trying to lock:
ffffff8801593030 (&cpc_ptr->rmw_lock){+.+.}-{3:3}, at: cpc_write+0xcc/0x370
other info that might help us debug this:
context-{5:5}
2 locks held by kworker/1:1/62:
  #0: ffffff897ef5ec98 (&rq->__lock){-.-.}-{2:2}, at: raw_spin_rq_lock_nested+0x2c/0x50
  #1: ffffff880154e238 (&sg_policy->update_lock){....}-{2:2}, at: sugov_update_shared+0x3c/0x280
stack backtrace:
CPU: 1 UID: 0 PID: 62 Comm: kworker/1:1 Not tainted 6.12.0-rc2-g9654bd3e8806 #406
Workqueue:  0x0 (events)
Call trace:
  dump_backtrace+0xa4/0x130
  show_stack+0x20/0x38
  dump_stack_lvl+0x90/0xd0
  dump_stack+0x18/0x28
  __lock_acquire+0x480/0x1ad8
  lock_acquire+0x114/0x310
  _raw_spin_lock+0x50/0x70
  cpc_write+0xcc/0x370
  cppc_set_perf+0xa0/0x3a8
  cppc_cpufreq_fast_switch+0x40/0xc0
  cpufreq_driver_fast_switch+0x4c/0x218
  sugov_update_shared+0x234/0x280
  update_load_avg+0x6ec/0x7b8
  dequeue_entities+0x108/0x830
  dequeue_task_fair+0x58/0x408
  __schedule+0x4f0/0x1070
  schedule+0x54/0x130
  worker_thread+0xc0/0x2e8
  kthread+0x130/0x148
  ret_from_fork+0x10/0x20

sugov_update_shared() locks a raw_spinlock while cpc_write() locks a
spinlock.

To have a correct wait-type order, update rmw_lock to a raw spinlock and
ensure that interrupts will be disabled on the CPU holding it.

Fixes: 60949b7b8054 ("ACPI: CPPC: Fix MASK_VAL() usage")
Signed-off-by: Pierre Gondois <pierre.gondois@arm.com>
Link: https://patch.msgid.link/20241028125657.1271512-1-pierre.gondois@arm.com
[ rjw: Changelog edits ]
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

afs: Fix missing subdir edit when renamed between parent dirs

[ Upstream commit 247d65fb122ad560be1c8c4d87d7374fb28b0770 ]

When rename moves an AFS subdirectory between parent directories, the
subdir also needs a bit of editing: the ".." entry needs updating to point
to the new parent (though I don't make use of the info) and the DV needs
incrementing by 1 to reflect the change of content.  The server also sends
a callback break notification on the subdirectory if we have one, but we
can take care of recovering the promise next time we access the subdir.

This can be triggered by something like:

    mount -t afs %example.com:xfstest.test20 /xfstest.test/
    mkdir /xfstest.test/{aaa,bbb,aaa/ccc}
    touch /xfstest.test/bbb/ccc/d
    mv /xfstest.test/{aaa/ccc,bbb/ccc}
    touch /xfstest.test/bbb/ccc/e

When the pathwalk for the second touch hits "ccc", kafs spots that the DV
is incorrect and downloads it again (so the fix is not critical).

Fix this, if the rename target is a directory and the old and new
parents are different, by:

 (1) Incrementing the DV number of the target locally.

 (2) Editing the ".." entry in the target to refer to its new parent's
     vnode ID and uniquifier.

Link: https://lore.kernel.org/r/3340431.1729680010@warthog.procyon.org.uk
Fixes: 63a4681ff39c ("afs: Locally edit directory data for mkdir/create/unlink/...")
cc: David Howells <dhowells@redhat.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

afs: Automatically generate trace tag enums

[ Upstream commit 2daa6404fd2f00985d5bfeb3c161f4630b46b6bf ]

Automatically generate trace tag enums from the symbol -> string mapping
tables rather than having the enums as well, thereby reducing duplicated
data.

Signed-off-by: David Howells <dhowells@redhat.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: Jeff Layton <jlayton@kernel.org>
cc: linux-afs@lists.infradead.org
cc: linux-fsdevel@vger.kernel.org
Stable-dep-of: 247d65fb122a ("afs: Fix missing subdir edit when renamed between parent dirs")
Signed-off-by: Sasha Levin <sashal@kernel.org>

firmware: arm_sdei: Fix the input parameter of cpuhp_remove_state()

[ Upstream commit c83212d79be2c9886d3e6039759ecd388fd5fed1 ]

In sdei_device_freeze(), the input parameter of cpuhp_remove_state() is
passed as 'sdei_entry_point' by mistake. Change it to 'sdei_hp_state'.

Fixes: d2c48b2387eb ("firmware: arm_sdei: Fix sleep from invalid context BUG")
Signed-off-by: Xiongfeng Wang <wangxiongfeng2@huawei.com>
Reviewed-by: James Morse <james.morse@arm.com>
Link: https://lore.kernel.org/r/20241016084740.183353-1-wangxiongfeng2@huawei.com
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

kasan: Fix Software Tag-Based KASAN with GCC

[ Upstream commit 894b00a3350c560990638bdf89bdf1f3d5491950 ]

Per [1], -fsanitize=kernel-hwaddress with GCC currently does not disable
instrumentation in functions with __attribute__((no_sanitize_address)).

However, __attribute__((no_sanitize("hwaddress"))) does correctly
disable instrumentation. Use it instead.

Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117196
Link: https://lore.kernel.org/r/000000000000f362e80620e27859@google.com
Link: https://lore.kernel.org/r/ZvFGwKfoC4yVjN_X@J2N7QTR9R3
Link: https://bugzilla.kernel.org/show_bug.cgi?id=218854
Reported-by: syzbot+908886656a02769af987@syzkaller.appspotmail.com
Tested-by: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Andrew Pinski <pinskia@gmail.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Marco Elver <elver@google.com>
Reviewed-by: Andrey Konovalov <andreyknvl@gmail.com>
Fixes: 7b861a53e46b ("kasan: Bump required compiler version")
Link: https://lore.kernel.org/r/20241021120013.3209481-1-elver@google.com
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

iomap: turn iomap_want_unshare_iter into an inline function

[ Upstream commit 6db388585e486c0261aeef55f8bc63a9b45756c0 ]

iomap_want_unshare_iter currently sits in fs/iomap/buffered-io.c, which
depends on CONFIG_BLOCK.  It is also in used in fs/dax.c whіch has no
such dependency.  Given that it is a trivial check turn it into an inline
in include/linux/iomap.h to fix the DAX && !BLOCK build.

Fixes: 6ef6a0e821d3 ("iomap: share iomap_unshare_iter predicate code with fsdax")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20241015041350.118403-1-hch@lst.de
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

fsdax: dax_unshare_iter needs to copy entire blocks

[ Upstream commit 50793801fc7f6d08def48754fb0f0706b0cfc394 ]

The code that copies data from srcmap to iomap in dax_unshare_iter is
very very broken, which bfoster's recent fsx changes have exposed.

If the pos and len passed to dax_file_unshare are not aligned to an
fsblock boundary, the iter pos and length in the _iter function will
reflect this unalignment.

dax_iomap_direct_access always returns a pointer to the start of the
kmapped fsdax page, even if its pos argument is in the middle of that
page.  This is catastrophic for data integrity when iter->pos is not
aligned to a page, because daddr/saddr do not point to the same byte in
the file as iter->pos.  Hence we corrupt user data by copying it to the
wrong place.

If iter->pos + iomap_length() in the _iter function not aligned to a
page, then we fail to copy a full block, and only partially populate the
destination block.  This is catastrophic for data confidentiality
because we expose stale pmem contents.

Fix both of these issues by aligning copy_pos/copy_len to a page
boundary (remember, this is fsdax so 1 fsblock == 1 base page) so that
we always copy full blocks.

We're not done yet -- there's no call to invalidate_inode_pages2_range,
so programs that have the file range mmap'd will continue accessing the
old memory mapping after the file metadata updates have completed.

Be careful with the return value -- if the unshare succeeds, we still
need to return the number of bytes that the iomap iter thinks we're
operating on.

Cc: ruansy.fnst@fujitsu.com
Fixes: d984648e428b ("fsdax,xfs: port unshare to fsdax")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Link: https://lore.kernel.org/r/172796813328.1131942.16777025316348797355.stgit@frogsfrogsfrogs
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

fsdax: remove zeroing code from dax_unshare_iter

[ Upstream commit 95472274b6fed8f2d30fbdda304e12174b3d4099 ]

Remove the code in dax_unshare_iter that zeroes the destination memory
because it's not necessary.

If srcmap is unwritten, we don't have to do anything because that
unwritten extent came from the regular file mapping, and unwritten
extents cannot be shared.  The same applies to holes.

Furthermore, zeroing to unshare a mapping is just plain wrong because
unsharing means copy on write, and we should be copying data.

This is effectively a revert of commit 13dd4e04625f ("fsdax: unshare:
zero destination if srcmap is HOLE or UNWRITTEN")

Cc: ruansy.fnst@fujitsu.com
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Link: https://lore.kernel.org/r/172796813311.1131942.16033376284752798632.stgit@frogsfrogsfrogs
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Stable-dep-of: 50793801fc7f ("fsdax: dax_unshare_iter needs to copy entire blocks")
Signed-off-by: Sasha Levin <sashal@kernel.org>

iomap: share iomap_unshare_iter predicate code with fsdax

[ Upstream commit 6ef6a0e821d3dad6bf8a5d5508762dba9042c84b ]

The predicate code that iomap_unshare_iter uses to decide if it's really
needs to unshare a file range mapping should be shared with the fsdax
version, because right now they're opencoded and inconsistent.

Note that we simplify the predicate logic a bit -- we no longer allow
unsharing of inline data mappings, but there aren't any filesystems that
allow shared inline data currently.

This is a fix in the sense that it should have been ported to fsdax.

Fixes: b53fdb215d13 ("iomap: improve shared block detection in iomap_unshare_iter")
Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Link: https://lore.kernel.org/r/172796813294.1131942.15762084021076932620.stgit@frogsfrogsfrogs
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Stable-dep-of: 50793801fc7f ("fsdax: dax_unshare_iter needs to copy entire blocks")
Signed-off-by: Sasha Levin <sashal@kernel.org>

iomap: don't bother unsharing delalloc extents

[ Upstream commit f7a4874d977bf4202ad575031222e78809a36292 ]

If unshare encounters a delalloc reservation in the srcmap, that means
that the file range isn't shared because delalloc reservations cannot be
reflinked.  Therefore, don't try to unshare them.

Signed-off-by: Darrick J. Wong <djwong@kernel.org>
Link: https://lore.kernel.org/r/20241002150040.GB21853@frogsfrogsfrogs
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Stable-dep-of: 50793801fc7f ("fsdax: dax_unshare_iter needs to copy entire blocks")
Signed-off-by: Sasha Levin <sashal@kernel.org>

iomap: improve shared block detection in iomap_unshare_iter

[ Upstream commit b53fdb215d13f8e9c29541434bf2d14dac8bcbdc ]

Currently iomap_unshare_iter relies on the IOMAP_F_SHARED flag to detect
blocks to unshare.  This is reasonable, but IOMAP_F_SHARED is also useful
for the file system to do internal book keeping for out of place writes.
XFS used to that, until it got removed in commit 72a048c1056a
("xfs: only set IOMAP_F_SHARED when providing a srcmap to a write")
because unshare for incorrectly unshare such blocks.

Add an extra safeguard by checking the explicitly provided srcmap instead
of the fallback to the iomap for valid data, as that catches the case
where we'd just copy from the same place we'd write to easily, allowing
to reinstate setting IOMAP_F_SHARED for all XFS writes that go to the
COW fork.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20240910043949.3481298-3-hch@lst.de
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Christian Brauner <brauner@kernel.org>
Stable-dep-of: 50793801fc7f ("fsdax: dax_unshare_iter needs to copy entire blocks")
Signed-off-by: Sasha Levin <sashal@kernel.org>

bpf, test_run: Fix LIVE_FRAME frame update after a page has been recycled

[ Upstream commit c40dd8c4732551605712985bc5b7045094c6458d ]

The test_run code detects whether a page has been modified and
re-initialises the xdp_frame structure if it has, using
xdp_update_frame_from_buff(). However, xdp_update_frame_from_buff()
doesn't touch frame->mem, so that wasn't correctly re-initialised, which
led to the pages from page_pool not being returned correctly. Syzbot
noticed this as a memory leak.

Fix this by also copying the frame->mem structure when re-initialising
the frame, like we do on initialisation of a new page from page_pool.

Fixes: e5995bc7e2ba ("bpf, test_run: fix crashes due to XDP frame overwriting/corruption")
Fixes: b530e9e1063e ("bpf: Add "live packet" mode for XDP in BPF_PROG_RUN")
Reported-by: syzbot+d121e098da06af416d23@syzkaller.appspotmail.com
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Tested-by: syzbot+d121e098da06af416d23@syzkaller.appspotmail.com
Reviewed-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Acked-by: Stanislav Fomichev <sdf@fomichev.me>
Link: https://lore.kernel.org/bpf/20241030-test-run-mem-fix-v1-1-41e88e8cae43@redhat.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

netfilter: nft_payload: sanitize offset and length before calling skb_checksum()

[ Upstream commit d5953d680f7e96208c29ce4139a0e38de87a57fe ]

If access to offset + length is larger than the skbuff length, then
skb_checksum() triggers BUG_ON().

skb_checksum() internally subtracts the length parameter while iterating
over skbuff, BUG_ON(len) at the end of it checks that the expected
length to be included in the checksum calculation is fully consumed.

Fixes: 7ec3f7b47b8d ("netfilter: nft_payload: add packet mangling support")
Reported-by: Slavin Liu <slavin-ayu@qq.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

mlxsw: spectrum_ipip: Fix memory leak when changing remote IPv6 address

[ Upstream commit 12ae97c531fcd3bfd774d4dfeaeac23eafe24280 ]

The device stores IPv6 addresses that are used for encapsulation in
linear memory that is managed by the driver.

Changing the remote address of an ip6gre net device never worked
properly, but since cited commit the following reproducer [1] would
result in a warning [2] and a memory leak [3]. The problem is that the
new remote address is never added by the driver to its hash table (and
therefore the device) and the old address is never removed from it.

Fix by programming the new address when the configuration of the ip6gre
net device changes and removing the old one. If the address did not
change, then the above would result in increasing the reference count of
the address and then decreasing it.

[1]
 # ip link add name bla up type ip6gre local 2001:db8:1::1 remote 2001:db8:2::1 tos inherit ttl inherit
 # ip link set dev bla type ip6gre remote 2001:db8:3::1
 # ip link del dev bla
 # devlink dev reload pci/0000:01:00.0

[2]
WARNING: CPU: 0 PID: 1682 at drivers/net/ethernet/mellanox/mlxsw/spectrum.c:3002 mlxsw_sp_ipv6_addr_put+0x140/0x1d0
Modules linked in:
CPU: 0 UID: 0 PID: 1682 Comm: ip Not tainted 6.12.0-rc3-custom-g86b5b55bc835 #151
Hardware name: Nvidia SN5600/VMOD0013, BIOS 5.13 05/31/2023
RIP: 0010:mlxsw_sp_ipv6_addr_put+0x140/0x1d0
[...]
Call Trace:
 <TASK>
 mlxsw_sp_router_netdevice_event+0x55f/0x1240
 notifier_call_chain+0x5a/0xd0
 call_netdevice_notifiers_info+0x39/0x90
 unregister_netdevice_many_notify+0x63e/0x9d0
 rtnl_dellink+0x16b/0x3a0
 rtnetlink_rcv_msg+0x142/0x3f0
 netlink_rcv_skb+0x50/0x100
 netlink_unicast+0x242/0x390
 netlink_sendmsg+0x1de/0x420
 ____sys_sendmsg+0x2bd/0x320
 ___sys_sendmsg+0x9a/0xe0
 __sys_sendmsg+0x7a/0xd0
 do_syscall_64+0x9e/0x1a0
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

[3]
unreferenced object 0xffff898081f597a0 (size 32):
  comm "ip", pid 1626, jiffies 4294719324
  hex dump (first 32 bytes):
    20 01 0d b8 00 02 00 00 00 00 00 00 00 00 00 01   ...............
    21 49 61 83 80 89 ff ff 00 00 00 00 01 00 00 00  !Ia.............
  backtrace (crc fd9be911):
    [<00000000df89c55d>] __kmalloc_cache_noprof+0x1da/0x260
    [<00000000ff2a1ddb>] mlxsw_sp_ipv6_addr_kvdl_index_get+0x281/0x340
    [<000000009ddd445d>] mlxsw_sp_router_netdevice_event+0x47b/0x1240
    [<00000000743e7757>] notifier_call_chain+0x5a/0xd0
    [<000000007c7b9e13>] call_netdevice_notifiers_info+0x39/0x90
    [<000000002509645d>] register_netdevice+0x5f7/0x7a0
    [<00000000c2e7d2a9>] ip6gre_newlink_common.isra.0+0x65/0x130
    [<0000000087cd6d8d>] ip6gre_newlink+0x72/0x120
    [<000000004df7c7cc>] rtnl_newlink+0x471/0xa20
    [<0000000057ed632a>] rtnetlink_rcv_msg+0x142/0x3f0
    [<0000000032e0d5b5>] netlink_rcv_skb+0x50/0x100
    [<00000000908bca63>] netlink_unicast+0x242/0x390
    [<00000000cdbe1c87>] netlink_sendmsg+0x1de/0x420
    [<0000000011db153e>] ____sys_sendmsg+0x2bd/0x320
    [<000000003b6d53eb>] ___sys_sendmsg+0x9a/0xe0
    [<00000000cae27c62>] __sys_sendmsg+0x7a/0xd0

Fixes: cf42911523e0 ("mlxsw: spectrum_ipip: Use common hash table for IPv6 address mapping")
Reported-by: Maksym Yaremchuk <maksymy@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Link: https://patch.msgid.link/e91012edc5a6cb9df37b78fd377f669381facfcb.1729866134.git.petrm@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

mlxsw: spectrum_ptp: Add missing verification before pushing Tx header

[ Upstream commit 0a66e5582b5102c4d7b866b977ff7c850c1174ce ]

Tx header should be pushed for each packet which is transmitted via
Spectrum ASICs. The cited commit moved the call to skb_cow_head() from
mlxsw_sp_port_xmit() to functions which handle Tx header.

In case that mlxsw_sp->ptp_ops->txhdr_construct() is used to handle Tx
header, and txhdr_construct() is mlxsw_sp_ptp_txhdr_construct(), there is
no call for skb_cow_head() before pushing Tx header size to SKB. This flow
is relevant for Spectrum-1 and Spectrum-4, for PTP packets.

Add the missing call to skb_cow_head() to make sure that there is both
enough room to push the Tx header and that the SKB header is not cloned and
can be modified.

An additional set will be sent to net-next to centralize the handling of
the Tx header by pushing it to every packet just before transmission.

Cc: Richard Cochran <richardcochran@gmail.com>
Fixes: 24157bc69f45 ("mlxsw: Send PTP packets as data packets to overcome a limitation")
Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Link: https://patch.msgid.link/5145780b07ebbb5d3b3570f311254a3a2d554a44.1729866134.git.petrm@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: skip offload for NETIF_F_IPV6_CSUM if ipv6 header contains extension

[ Upstream commit 04c20a9356f283da623903e81e7c6d5df7e4dc3c ]

As documented in skbuff.h, devices with NETIF_F_IPV6_CSUM capability
can only checksum TCP and UDP over IPv6 if the IP header does not
contains extension.

This is enforced for UDP packets emitted from user-space to an IPv6
address as they go through ip6_make_skb(), which calls
__ip6_append_data() where a check is done on the header size before
setting CHECKSUM_PARTIAL.

But the introduction of UDP encapsulation with fou6 added a code-path
where it is possible to get an skb with a partial UDP checksum and an
IPv6 header with extension:
* fou6 adds a UDP header with a partial checksum if the inner packet
does not contains a valid checksum.
* ip6_tunnel adds an IPv6 header with a destination option extension
header if encap_limit is non-zero (the default value is 4).

The thread linked below describes in more details how to reproduce the
problem with GRE-in-UDP tunnel.

Add a check on the network header size in skb_csum_hwoffload_help() to
make sure no IPv6 packet with extension header is handed to a network
device with NETIF_F_IPV6_CSUM capability.

Link: https://lore.kernel.org/netdev/26548921.1r3eYUQgxm@benoit.monin/T/#u
Fixes: aa3463d65e7b ("fou: Add encap ops for IPv6 tunnels")
Signed-off-by: Benoît Monin <benoit.monin@gmx.fr>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://patch.msgid.link/5fbeecfc311ea182aa1d1c771725ab8b4cac515e.1729778144.git.benoit.monin@gmx.fr
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

Bluetooth: hci: fix null-ptr-deref in hci_read_supported_codecs

[ Upstream commit 1e67d8641813f1876a42eeb4f532487b8a7fb0a8 ]

Fix __hci_cmd_sync_sk() to return not NULL for unknown opcodes.

__hci_cmd_sync_sk() returns NULL if a command returns a status event.
However, it also returns NULL where an opcode doesn't exist in the
hci_cc table because hci_cmd_complete_evt() assumes status = skb->data[0]
for unknown opcodes.
This leads to null-ptr-deref in cmd_sync for HCI_OP_READ_LOCAL_CODECS as
there is no hci_cc for HCI_OP_READ_LOCAL_CODECS, which always assumes
status = skb->data[0].

KASAN: null-ptr-deref in range [0x0000000000000070-0x0000000000000077]
CPU: 1 PID: 2000 Comm: kworker/u9:5 Not tainted 6.9.0-ga6bcb805883c-dirty #10
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.15.0-1 04/01/2014
Workqueue: hci7 hci_power_on
RIP: 0010:hci_read_supported_codecs+0xb9/0x870 net/bluetooth/hci_codec.c:138
Code: 08 48 89 ef e8 b8 c1 8f fd 48 8b 75 00 e9 96 00 00 00 49 89 c6 48 ba 00 00 00 00 00 fc ff df 4c 8d 60 70 4c 89 e3 48 c1 eb 03 <0f> b6 04 13 84 c0 0f 85 82 06 00 00 41 83 3c 24 02 77 0a e8 bf 78
RSP: 0018:ffff888120bafac8 EFLAGS: 00010212
RAX: 0000000000000000 RBX: 000000000000000e RCX: ffff8881173f0040
RDX: dffffc0000000000 RSI: ffffffffa58496c0 RDI: ffff88810b9ad1e4
RBP: ffff88810b9ac000 R08: ffffffffa77882a7 R09: 1ffffffff4ef1054
R10: dffffc0000000000 R11: fffffbfff4ef1055 R12: 0000000000000070
R13: 0000000000000000 R14: 0000000000000000 R15: ffff88810b9ac000
FS:  0000000000000000(0000) GS:ffff8881f6c00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f6ddaa3439e CR3: 0000000139764003 CR4: 0000000000770ef0
PKRU: 55555554
Call Trace:
 <TASK>
 hci_read_local_codecs_sync net/bluetooth/hci_sync.c:4546 [inline]
 hci_init_stage_sync net/bluetooth/hci_sync.c:3441 [inline]
 hci_init4_sync net/bluetooth/hci_sync.c:4706 [inline]
 hci_init_sync net/bluetooth/hci_sync.c:4742 [inline]
 hci_dev_init_sync net/bluetooth/hci_sync.c:4912 [inline]
 hci_dev_open_sync+0x19a9/0x2d30 net/bluetooth/hci_sync.c:4994
 hci_dev_do_open net/bluetooth/hci_core.c:483 [inline]
 hci_power_on+0x11e/0x560 net/bluetooth/hci_core.c:1015
 process_one_work kernel/workqueue.c:3267 [inline]
 process_scheduled_works+0x8ef/0x14f0 kernel/workqueue.c:3348
 worker_thread+0x91f/0xe50 kernel/workqueue.c:3429
 kthread+0x2cb/0x360 kernel/kthread.c:388
 ret_from_fork+0x4d/0x80 arch/x86/kernel/process.c:147
 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244

Fixes: abfeea476c68 ("Bluetooth: hci_sync: Convert MGMT_OP_START_DISCOVERY")
Signed-off-by: Sungwoo Kim <iam@sung-woo.kim>
Signed-off-by: Luiz Augusto von Dentz <luiz.von.dentz@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

netfilter: nf_reject_ipv6: fix potential crash in nf_send_reset6()

[ Upstream commit 4ed234fe793f27a3b151c43d2106df2ff0d81aac ]

I got a syzbot report without a repro [1] crashing in nf_send_reset6()

I think the issue is that dev->hard_header_len is zero, and we attempt
later to push an Ethernet header.

Use LL_MAX_HEADER, as other functions in net/ipv6/netfilter/nf_reject_ipv6.c.

[1]

skbuff: skb_under_panic: text:ffffffff89b1d008 len:74 put:14 head:ffff88803123aa00 data:ffff88803123a9f2 tail:0x3c end:0x140 dev:syz_tun
 kernel BUG at net/core/skbuff.c:206 !
Oops: invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI
CPU: 0 UID: 0 PID: 7373 Comm: syz.1.568 Not tainted 6.12.0-rc2-syzkaller-00631-g6d858708d465 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 09/13/2024
 RIP: 0010:skb_panic net/core/skbuff.c:206 [inline]
 RIP: 0010:skb_under_panic+0x14b/0x150 net/core/skbuff.c:216
Code: 0d 8d 48 c7 c6 60 a6 29 8e 48 8b 54 24 08 8b 0c 24 44 8b 44 24 04 4d 89 e9 50 41 54 41 57 41 56 e8 ba 30 38 02 48 83 c4 20 90 <0f> 0b 0f 1f 00 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 f3
RSP: 0018:ffffc900045269b0 EFLAGS: 00010282
RAX: 0000000000000088 RBX: dffffc0000000000 RCX: cd66dacdc5d8e800
RDX: 0000000000000000 RSI: 0000000000000200 RDI: 0000000000000000
RBP: ffff88802d39a3d0 R08: ffffffff8174afec R09: 1ffff920008a4ccc
R10: dffffc0000000000 R11: fffff520008a4ccd R12: 0000000000000140
R13: ffff88803123aa00 R14: ffff88803123a9f2 R15: 000000000000003c
FS:  00007fdbee5ff6c0(0000) GS:ffff8880b8600000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 0000000000000000 CR3: 000000005d322000 CR4: 00000000003526f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 <TASK>
  skb_push+0xe5/0x100 net/core/skbuff.c:2636
  eth_header+0x38/0x1f0 net/ethernet/eth.c:83
  dev_hard_header include/linux/netdevice.h:3208 [inline]
  nf_send_reset6+0xce6/0x1270 net/ipv6/netfilter/nf_reject_ipv6.c:358
  nft_reject_inet_eval+0x3b9/0x690 net/netfilter/nft_reject_inet.c:48
  expr_call_ops_eval net/netfilter/nf_tables_core.c:240 [inline]
  nft_do_chain+0x4ad/0x1da0 net/netfilter/nf_tables_core.c:288
  nft_do_chain_inet+0x418/0x6b0 net/netfilter/nft_chain_filter.c:161
  nf_hook_entry_hookfn include/linux/netfilter.h:154 [inline]
  nf_hook_slow+0xc3/0x220 net/netfilter/core.c:626
  nf_hook include/linux/netfilter.h:269 [inline]
  NF_HOOK include/linux/netfilter.h:312 [inline]
  br_nf_pre_routing_ipv6+0x63e/0x770 net/bridge/br_netfilter_ipv6.c:184
  nf_hook_entry_hookfn include/linux/netfilter.h:154 [inline]
  nf_hook_bridge_pre net/bridge/br_input.c:277 [inline]
  br_handle_frame+0x9fd/0x1530 net/bridge/br_input.c:424
  __netif_receive_skb_core+0x13e8/0x4570 net/core/dev.c:5562
  __netif_receive_skb_one_core net/core/dev.c:5666 [inline]
  __netif_receive_skb+0x12f/0x650 net/core/dev.c:5781
  netif_receive_skb_internal net/core/dev.c:5867 [inline]
  netif_receive_skb+0x1e8/0x890 net/core/dev.c:5926
  tun_rx_batched+0x1b7/0x8f0 drivers/net/tun.c:1550
  tun_get_user+0x3056/0x47e0 drivers/net/tun.c:2007
  tun_chr_write_iter+0x10d/0x1f0 drivers/net/tun.c:2053
  new_sync_write fs/read_write.c:590 [inline]
  vfs_write+0xa6d/0xc90 fs/read_write.c:683
  ksys_write+0x183/0x2b0 fs/read_write.c:736
  do_syscall_x64 arch/x86/entry/common.c:52 [inline]
  do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
 entry_SYSCALL_64_after_hwframe+0x77/0x7f
RIP: 0033:0x7fdbeeb7d1ff
Code: 89 54 24 18 48 89 74 24 10 89 7c 24 08 e8 c9 8d 02 00 48 8b 54 24 18 48 8b 74 24 10 41 89 c0 8b 7c 24 08 b8 01 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 31 44 89 c7 48 89 44 24 08 e8 1c 8e 02 00 48
RSP: 002b:00007fdbee5ff000 EFLAGS: 00000293 ORIG_RAX: 0000000000000001
RAX: ffffffffffffffda RBX: 00007fdbeed36058 RCX: 00007fdbeeb7d1ff
RDX: 000000000000008e RSI: 0000000020000040 RDI: 00000000000000c8
RBP: 00007fdbeebf12be R08: 0000000000000000 R09: 0000000000000000
R10: 000000000000008e R11: 0000000000000293 R12: 0000000000000000
R13: 0000000000000000 R14: 00007fdbeed36058 R15: 00007ffc38de06e8
 </TASK>

Fixes: c8d7b98bec43 ("netfilter: move nf_send_resetX() code to nf_reject_ipvX modules")
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

netfilter: Fix use-after-free in get_info()

[ Upstream commit f48d258f0ac540f00fa617dac496c4c18b5dc2fa ]

ip6table_nat module unload has refcnt warning for UAF. call trace is:

WARNING: CPU: 1 PID: 379 at kernel/module/main.c:853 module_put+0x6f/0x80
Modules linked in: ip6table_nat(-)
CPU: 1 UID: 0 PID: 379 Comm: ip6tables Not tainted 6.12.0-rc4-00047-gc2ee9f594da8-dirty #205
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996),
BIOS rel-1.13.0-0-gf21b5a4aeb02-prebuilt.qemu.org 04/01/2014
RIP: 0010:module_put+0x6f/0x80
Call Trace:
 <TASK>
 get_info+0x128/0x180
 do_ip6t_get_ctl+0x6a/0x430
 nf_getsockopt+0x46/0x80
 ipv6_getsockopt+0xb9/0x100
 rawv6_getsockopt+0x42/0x190
 do_sock_getsockopt+0xaa/0x180
 __sys_getsockopt+0x70/0xc0
 __x64_sys_getsockopt+0x20/0x30
 do_syscall_64+0xa2/0x1a0
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

Concurrent execution of module unload and get_info() trigered the warning.
The root cause is as follows:

cpu0       cpu1
module_exit
//mod->state = MODULE_STATE_GOING
  ip6table_nat_exit
    xt_unregister_template
kfree(t)
//removed from templ_list
      getinfo()
  t = xt_find_table_lock
list_for_each_entry(tmpl, &xt_templates[af]...)
if (strcmp(tmpl->name, name))
continue;  //table not found
try_module_get
list_for_each_entry(t, &xt_net->tables[af]...)
return t;  //not get refcnt
  module_put(t->me) //uaf
    unregister_pernet_subsys
    //remove table from xt_net list

While xt_table module was going away and has been removed from
xt_templates list, we couldnt get refcnt of xt_table->me. Check
module in xt_net->tables list re-traversal to fix it.

Fixes: fdacd57c79b7 ("netfilter: x_tables: never register tables by default")
Signed-off-by: Dong Chenchen <dongchenchen2@huawei.com>
Reviewed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: fix crash when config small gso_max_size/gso_ipv4_max_size

[ Upstream commit 9ab5cf19fb0e4680f95e506d6c544259bf1111c4 ]

Config a small gso_max_size/gso_ipv4_max_size will lead to an underflow
in sk_dst_gso_max_size(), which may trigger a BUG_ON crash,
because sk->sk_gso_max_size would be much bigger than device limits.
Call Trace:
tcp_write_xmit
    tso_segs = tcp_init_tso_segs(skb, mss_now);
        tcp_set_skb_tso_segs
            tcp_skb_pcount_set
                // skb->len = 524288, mss_now = 8
                // u16 tso_segs = 524288/8 = 65535 -> 0
                tso_segs = DIV_ROUND_UP(skb->len, mss_now)
    BUG_ON(!tso_segs)
Add check for the minimum value of gso_max_size and gso_ipv4_max_size.

Fixes: 46e6b992c250 ("rtnetlink: allow GSO maximums to be set on device creation")
Fixes: 9eefedd58ae1 ("net: add gso_ipv4_max_size and gro_ipv4_max_size per device")
Signed-off-by: Wang Liang <wangliang74@huawei.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://patch.msgid.link/20241023035213.517386-1-wangliang74@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

bpf: Fix out-of-bounds write in trie_get_next_key()

[ Upstream commit 13400ac8fb80c57c2bfb12ebd35ee121ce9b4d21 ]

trie_get_next_key() allocates a node stack with size trie->max_prefixlen,
while it writes (trie->max_prefixlen + 1) nodes to the stack when it has
full paths from the root to leaves. For example, consider a trie with
max_prefixlen is 8, and the nodes with key 0x00/0, 0x00/1, 0x00/2, ...
0x00/8 inserted. Subsequent calls to trie_get_next_key with _key with
.prefixlen = 8 make 9 nodes be written on the node stack with size 8.

Fixes: b471f2f1de8b ("bpf: implement MAP_GET_NEXT_KEY command for LPM_TRIE map")
Signed-off-by: Byeonguk Jeong <jungbu2855@gmail.com>
Reviewed-by: Toke Høiland-Jørgensen <toke@kernel.org>
Tested-by: Hou Tao <houtao1@huawei.com>
Acked-by: Hou Tao <houtao1@huawei.com>
Link: https://lore.kernel.org/r/Zxx384ZfdlFYnz6J@localhost.localdomain
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

netdevsim: Add trailing zero to terminate the string in nsim_nexthop_bucket_activity_write()

[ Upstream commit 4ce1f56a1eaced2523329bef800d004e30f2f76c ]

This was found by a static analyzer.
We should not forget the trailing zero after copy_from_user()
if we will further do some string operations, sscanf() in this
case. Adding a trailing zero will ensure that the function
performs properly.

Fixes: c6385c0b67c5 ("netdevsim: Allow reporting activity on nexthop buckets")
Signed-off-by: Zichen Xie <zichenxie0106@gmail.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Link: https://patch.msgid.link/20241022171907.8606-1-zichenxie0106@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

bpf: Force checkpoint when jmp history is too long

[ Upstream commit aa30eb3260b2dea3a68d3c42a39f9a09c5e99cee ]

A specifically crafted program might trick verifier into growing very
long jump history within a single bpf_verifier_state instance.
Very long jump history makes mark_chain_precision() unreasonably slow,
especially in case if verifier processes a loop.

Mitigate this by forcing new state in is_state_visited() in case if
current state's jump history is too long.

Use same constant as in `skip_inf_loop_check`, but multiply it by
arbitrarily chosen value 2 to account for jump history containing not
only information about jumps, but also information about stack access.

For an example of problematic program consider the code below,
w/o this patch the example is processed by verifier for ~15 minutes,
before failing to allocate big-enough chunk for jmp_history.

    0: r7 = *(u16 *)(r1 +0);"
    1: r7 += 0x1ab064b9;"
    2: if r7 & 0x702000 goto 1b;
    3: r7 &= 0x1ee60e;"
    4: r7 += r1;"
    5: if r7 s> 0x37d2 goto +0;"
    6: r0 = 0;"
    7: exit;"

Perf profiling shows that most of the time is spent in
mark_chain_precision() ~95%.

The easiest way to explain why this program causes problems is to
apply the following patch:

    diff --git a/include/linux/bpf.h b/include/linux/bpf.h
    index 0c216e71cec7..4b4823961abe 100644
    \--- a/include/linux/bpf.h
    \+++ b/include/linux/bpf.h
    \@@ -1926,7 +1926,7 @@ struct bpf_array {
            };
     };

    -#define BPF_COMPLEXITY_LIMIT_INSNS      1000000 /* yes. 1M insns */
    +#define BPF_COMPLEXITY_LIMIT_INSNS      256 /* yes. 1M insns */
     #define MAX_TAIL_CALL_CNT 33

     /* Maximum number of loops for bpf_loop and bpf_iter_num.
    diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
    index f514247ba8ba..75e88be3bb3e 100644
    \--- a/kernel/bpf/verifier.c
    \+++ b/kernel/bpf/verifier.c
    \@@ -18024,8 +18024,13 @@ static int is_state_visited(struct bpf_verifier_env *env, int insn_idx)
     skip_inf_loop_check:
                            if (!force_new_state &&
                                env->jmps_processed - env->prev_jmps_processed < 20 &&
    -                           env->insn_processed - env->prev_insn_processed < 100)
    +                           env->insn_processed - env->prev_insn_processed < 100) {
    +                               verbose(env, "is_state_visited: suppressing checkpoint at %d, %d jmps processed, cur->jmp_history_cnt is %d\n",
    +                                       env->insn_idx,
    +                                       env->jmps_processed - env->prev_jmps_processed,
    +                                       cur->jmp_history_cnt);
                                    add_new_state = false;
    +                       }
                            goto miss;
                    }
                    /* If sl->state is a part of a loop and this loop's entry is a part of
    \@@ -18142,6 +18147,9 @@ static int is_state_visited(struct bpf_verifier_env *env, int insn_idx)
            if (!add_new_state)
                    return 0;

    +       verbose(env, "is_state_visited: new checkpoint at %d, resetting env->jmps_processed\n",
    +               env->insn_idx);
    +
            /* There were no equivalent states, remember the current one.
             * Technically the current state is not proven to be safe yet,
             * but it will either reach outer most bpf_exit (which means it's safe)

And observe verification log:

    ...
    is_state_visited: new checkpoint at 5, resetting env->jmps_processed
    5: R1=ctx() R7=ctx(...)
    5: (65) if r7 s> 0x37d2 goto pc+0     ; R7=ctx(...)
    6: (b7) r0 = 0                        ; R0_w=0
    7: (95) exit

    from 5 to 6: R1=ctx() R7=ctx(...) R10=fp0
    6: R1=ctx() R7=ctx(...) R10=fp0
    6: (b7) r0 = 0                        ; R0_w=0
    7: (95) exit
    is_state_visited: suppressing checkpoint at 1, 3 jmps processed, cur->jmp_history_cnt is 74

    from 2 to 1: R1=ctx() R7_w=scalar(...) R10=fp0
    1: R1=ctx() R7_w=scalar(...) R10=fp0
    1: (07) r7 += 447767737
    is_state_visited: suppressing checkpoint at 2, 3 jmps processed, cur->jmp_history_cnt is 75
    2: R7_w=scalar(...)
    2: (45) if r7 & 0x702000 goto pc-2
    ... mark_precise 152 steps for r7 ...
    2: R7_w=scalar(...)
    is_state_visited: suppressing checkpoint at 1, 4 jmps processed, cur->jmp_history_cnt is 75
    1: (07) r7 += 447767737
    is_state_visited: suppressing checkpoint at 2, 4 jmps processed, cur->jmp_history_cnt is 76
    2: R7_w=scalar(...)
    2: (45) if r7 & 0x702000 goto pc-2
    ...
    BPF program is too large. Processed 257 insn

The log output shows that checkpoint at label (1) is never created,
because it is suppressed by `skip_inf_loop_check` logic:
a. When 'if' at (2) is processed it pushes a state with insn_idx (1)
   onto stack and proceeds to (3);
b. At (5) checkpoint is created, and this resets
   env->{jmps,insns}_processed.
c. Verification proceeds and reaches `exit`;
d. State saved at step (a) is popped from stack and is_state_visited()
   considers if checkpoint needs to be added, but because
   env->{jmps,insns}_processed had been just reset at step (b)
   the `skip_inf_loop_check` logic forces `add_new_state` to false.
e. Verifier proceeds with current state, which slowly accumulates
   more and more entries in the jump history.

The accumulation of entries in the jump history is a problem because
of two factors:
- it eventually exhausts memory available for kmalloc() allocation;
- mark_chain_precision() traverses the jump history of a state,
  meaning that if `r7` is marked precise, verifier would iterate
  ever growing jump history until parent state boundary is reached.

(note: the log also shows a REG INVARIANTS VIOLATION warning
       upon jset processing, but that's another bug to fix).

With this patch applied, the example above is rejected by verifier
under 1s of time, reaching 1M instructions limit.

The program is a simplified reproducer from syzbot report.
Previous discussion could be found at [1].
The patch does not cause any changes in verification performance,
when tested on selftests from veristat.cfg and cilium programs taken
from [2].

[1] https://lore.kernel.org/bpf/20241009021254.2805446-1-eddyz87@gmail.com/
[2] https://github.com/anakryiko/cilium

Changelog:
- v1 -> v2:
  - moved patch to bpf tree;
  - moved force_new_state variable initialization after declaration and
    shortened the comment.
v1: https://lore.kernel.org/bpf/20241018020307.1766906-1-eddyz87@gmail.com/

Fixes: 2589726d12a1 ("bpf: introduce bounded loops")
Reported-by: syzbot+7e46cdef14bf496a3ab4@syzkaller.appspotmail.com
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20241029172641.1042523-1-eddyz87@gmail.com
Closes: https://lore.kernel.org/bpf/670429f6.050a0220.49194.0517.GAE@google.com/
Signed-off-by: Sasha Levin <sashal@kernel.org>

selftests/bpf: Add bpf_percpu_obj_{new,drop}() macro in bpf_experimental.h

[ Upstream commit 968c76cb3dc6cc86e8099ecaa5c30dc0d4738a30 ]

The new macro bpf_percpu_obj_{new/drop}() is very similar to bpf_obj_{new,drop}()
as they both take a type as the argument.

Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/r/20230827152805.1999417-1-yonghong.song@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Stable-dep-of: aa30eb3260b2 ("bpf: Force checkpoint when jmp history is too long")
Signed-off-by: Sasha Levin <sashal@kernel.org>

net/sched: stop qdisc_tree_reduce_backlog on TC_H_ROOT

[ Upstream commit 2e95c4384438adeaa772caa560244b1a2efef816 ]

In qdisc_tree_reduce_backlog, Qdiscs with major handle ffff: are assumed
to be either root or ingress. This assumption is bogus since it's valid
to create egress qdiscs with major handle ffff:
Budimir Markovic found that for qdiscs like DRR that maintain an active
class list, it will cause a UAF with a dangling class pointer.

In 066a3b5b2346, the concern was to avoid iterating over the ingress
qdisc since its parent is itself. The proper fix is to stop when parent
TC_H_ROOT is reached because the only way to retrieve ingress is when a
hierarchy which does not contain a ffff: major handle call into
qdisc_lookup with TC_H_MAJ(TC_H_ROOT).

In the scenario where major ffff: is an egress qdisc in any of the tree
levels, the updates will also propagate to TC_H_ROOT, which then the
iteration must stop.

Fixes: 066a3b5b2346 ("[NET_SCHED] sch_api: fix qdisc_tree_decrease_qlen() loop")
Reported-by: Budimir Markovic <markovicbudimir@gmail.com>
Suggested-by: Jamal Hadi Salim <jhs@mojatatu.com>
Tested-by: Victor Nogueira <victor@mojatatu.com>
Signed-off-by: Pedro Tammela <pctammela@mojatatu.com>
Signed-off-by: Jamal Hadi Salim <jhs@mojatatu.com>
 net/sched/sch_api.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20241024165547.418570-1-jhs@mojatatu.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

gtp: allow -1 to be specified as file description from userspace

[ Upstream commit 7515e37bce5c428a56a9b04ea7e96b3f53f17150 ]

Existing user space applications maintained by the Osmocom project are
breaking since a recent fix that addresses incorrect error checking.

Restore operation for user space programs that specify -1 as file
descriptor to skip GTPv0 or GTPv1 only sockets.

Fixes: defd8b3c37b0 ("gtp: fix a potential NULL pointer dereference")
Reported-by: Pau Espin Pedrol <pespin@sysmocom.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Tested-by: Oliver Smith <osmith@sysmocom.de>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20241022144825.66740-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ipv4: ip_tunnel: Fix suspicious RCU usage warning in ip_tunnel_init_flow()

[ Upstream commit ad4a3ca6a8e886f6491910a3ae5d53595e40597d ]

There are code paths from which the function is called without holding
the RCU read lock, resulting in a suspicious RCU usage warning [1].

Fix by using l3mdev_master_upper_ifindex_by_index() which will acquire
the RCU read lock before calling
l3mdev_master_upper_ifindex_by_index_rcu().

[1]
WARNING: suspicious RCU usage
6.12.0-rc3-custom-gac8f72681cf2 #141 Not tainted
-----------------------------
net/core/dev.c:876 RCU-list traversed in non-reader section!!

other info that might help us debug this:

rcu_scheduler_active = 2, debug_locks = 1
1 lock held by ip/361:
 #0: ffffffff86fc7cb0 (rtnl_mutex){+.+.}-{3:3}, at: rtnetlink_rcv_msg+0x377/0xf60

stack backtrace:
CPU: 3 UID: 0 PID: 361 Comm: ip Not tainted 6.12.0-rc3-custom-gac8f72681cf2 #141
Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
Call Trace:
 <TASK>
 dump_stack_lvl+0xba/0x110
 lockdep_rcu_suspicious.cold+0x4f/0xd6
 dev_get_by_index_rcu+0x1d3/0x210
 l3mdev_master_upper_ifindex_by_index_rcu+0x2b/0xf0
 ip_tunnel_bind_dev+0x72f/0xa00
 ip_tunnel_newlink+0x368/0x7a0
 ipgre_newlink+0x14c/0x170
 __rtnl_newlink+0x1173/0x19c0
 rtnl_newlink+0x6c/0xa0
 rtnetlink_rcv_msg+0x3cc/0xf60
 netlink_rcv_skb+0x171/0x450
 netlink_unicast+0x539/0x7f0
 netlink_sendmsg+0x8c1/0xd80
 ____sys_sendmsg+0x8f9/0xc20
 ___sys_sendmsg+0x197/0x1e0
 __sys_sendmsg+0x122/0x1f0
 do_syscall_64+0xbb/0x1d0
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

Fixes: db53cd3d88dc ("net: Handle l3mdev in ip_tunnel_init_flow")
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://patch.msgid.link/20241022063822.462057-1-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: stmmac: TSO: Fix unbalanced DMA map/unmap for non-paged SKB data

[ Upstream commit 66600fac7a984dea4ae095411f644770b2561ede ]

In case the non-paged data of a SKB carries protocol header and protocol
payload to be transmitted on a certain platform that the DMA AXI address
width is configured to 40-bit/48-bit, or the size of the non-paged data
is bigger than TSO_MAX_BUFF_SIZE on a certain platform that the DMA AXI
address width is configured to 32-bit, then this SKB requires at least
two DMA transmit descriptors to serve it.

For example, three descriptors are allocated to split one DMA buffer
mapped from one piece of non-paged data:
    dma_desc[N + 0],
    dma_desc[N + 1],
    dma_desc[N + 2].
Then three elements of tx_q->tx_skbuff_dma[] will be allocated to hold
extra information to be reused in stmmac_tx_clean():
    tx_q->tx_skbuff_dma[N + 0],
    tx_q->tx_skbuff_dma[N + 1],
    tx_q->tx_skbuff_dma[N + 2].
Now we focus on tx_q->tx_skbuff_dma[entry].buf, which is the DMA buffer
address returned by DMA mapping call. stmmac_tx_clean() will try to
unmap the DMA buffer _ONLY_IF_ tx_q->tx_skbuff_dma[entry].buf
is a valid buffer address.

The expected behavior that saves DMA buffer address of this non-paged
data to tx_q->tx_skbuff_dma[entry].buf is:
    tx_q->tx_skbuff_dma[N + 0].buf = NULL;
    tx_q->tx_skbuff_dma[N + 1].buf = NULL;
    tx_q->tx_skbuff_dma[N + 2].buf = dma_map_single();
Unfortunately, the current code misbehaves like this:
    tx_q->tx_skbuff_dma[N + 0].buf = dma_map_single();
    tx_q->tx_skbuff_dma[N + 1].buf = NULL;
    tx_q->tx_skbuff_dma[N + 2].buf = NULL;

On the stmmac_tx_clean() side, when dma_desc[N + 0] is closed by the
DMA engine, tx_q->tx_skbuff_dma[N + 0].buf is a valid buffer address
obviously, then the DMA buffer will be unmapped immediately.
There may be a rare case that the DMA engine does not finish the
pending dma_desc[N + 1], dma_desc[N + 2] yet. Now things will go
horribly wrong, DMA is going to access a unmapped/unreferenced memory
region, corrupted data will be transmited or iommu fault will be
triggered :(

In contrast, the for-loop that maps SKB fragments behaves perfectly
as expected, and that is how the driver should do for both non-paged
data and paged frags actually.

This patch corrects DMA map/unmap sequences by fixing the array index
for tx_q->tx_skbuff_dma[entry].buf when assigning DMA buffer address.

Tested and verified on DWXGMAC CORE 3.20a

Reported-by: Suraj Jaiswal <quic_jsuraj@quicinc.com>
Fixes: f748be531d70 ("stmmac: support new GMAC4")
Signed-off-by: Furong Xu <0x1207@gmail.com>
Reviewed-by: Hariprasad Kelam <hkelam@marvell.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20241021061023.2162701-1-0x1207@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

net: stmmac: dwmac4: Fix high address display by updating reg_space[] from register values

[ Upstream commit f84ef58e553206b02d06e02158c98fbccba25d19 ]

The high address will display as 0 if the driver does not set the
reg_space[]. To fix this, read the high address registers and
update the reg_space[] accordingly.

Fixes: fbf68229ffe7 ("net: stmmac: unify registers dumps methods")
Signed-off-by: Ley Foon Tan <leyfoon.tan@starfivetech.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20241021054625.1791965-1-leyfoon.tan@starfivetech.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

macsec: Fix use-after-free while sending the offloading packet

[ Upstream commit f1e54d11b210b53d418ff1476c6b58a2f434dfc0 ]

KASAN reports the following UAF. The metadata_dst, which is used to
store the SCI value for macsec offload, is already freed by
metadata_dst_free() in macsec_free_netdev(), while driver still use it
for sending the packet.

To fix this issue, dst_release() is used instead to release
metadata_dst. So it is not freed instantly in macsec_free_netdev() if
still referenced by skb.

 BUG: KASAN: slab-use-after-free in mlx5e_xmit+0x1e8f/0x4190 [mlx5_core]
 Read of size 2 at addr ffff88813e42e038 by task kworker/7:2/714
 [...]
 Workqueue: mld mld_ifc_work
 Call Trace:
  <TASK>
  dump_stack_lvl+0x51/0x60
  print_report+0xc1/0x600
  kasan_report+0xab/0xe0
  mlx5e_xmit+0x1e8f/0x4190 [mlx5_core]
  dev_hard_start_xmit+0x120/0x530
  sch_direct_xmit+0x149/0x11e0
  __qdisc_run+0x3ad/0x1730
  __dev_queue_xmit+0x1196/0x2ed0
  vlan_dev_hard_start_xmit+0x32e/0x510 [8021q]
  dev_hard_start_xmit+0x120/0x530
  __dev_queue_xmit+0x14a7/0x2ed0
  macsec_start_xmit+0x13e9/0x2340
  dev_hard_start_xmit+0x120/0x530
  __dev_queue_xmit+0x14a7/0x2ed0
  ip6_finish_output2+0x923/0x1a70
  ip6_finish_output+0x2d7/0x970
  ip6_output+0x1ce/0x3a0
  NF_HOOK.constprop.0+0x15f/0x190
  mld_sendpack+0x59a/0xbd0
  mld_ifc_work+0x48a/0xa80
  process_one_work+0x5aa/0xe50
  worker_thread+0x79c/0x1290
  kthread+0x28f/0x350
  ret_from_fork+0x2d/0x70
  ret_from_fork_asm+0x11/0x20
  </TASK>

 Allocated by task 3922:
  kasan_save_stack+0x20/0x40
  kasan_save_track+0x10/0x30
  __kasan_kmalloc+0x77/0x90
  __kmalloc_noprof+0x188/0x400
  metadata_dst_alloc+0x1f/0x4e0
  macsec_newlink+0x914/0x1410
  __rtnl_newlink+0xe08/0x15b0
  rtnl_newlink+0x5f/0x90
  rtnetlink_rcv_msg+0x667/0xa80
  netlink_rcv_skb+0x12c/0x360
  netlink_unicast+0x551/0x770
  netlink_sendmsg+0x72d/0xbd0
  __sock_sendmsg+0xc5/0x190
  ____sys_sendmsg+0x52e/0x6a0
  ___sys_sendmsg+0xeb/0x170
  __sys_sendmsg+0xb5/0x140
  do_syscall_64+0x4c/0x100
  entry_SYSCALL_64_after_hwframe+0x4b/0x53

 Freed by task 4011:
  kasan_save_stack+0x20/0x40
  kasan_save_track+0x10/0x30
  kasan_save_free_info+0x37/0x50
  poison_slab_object+0x10c/0x190
  __kasan_slab_free+0x11/0x30
  kfree+0xe0/0x290
  macsec_free_netdev+0x3f/0x140
  netdev_run_todo+0x450/0xc70
  rtnetlink_rcv_msg+0x66f/0xa80
  netlink_rcv_skb+0x12c/0x360
  netlink_unicast+0x551/0x770
  netlink_sendmsg+0x72d/0xbd0
  __sock_sendmsg+0xc5/0x190
  ____sys_sendmsg+0x52e/0x6a0
  ___sys_sendmsg+0xeb/0x170
  __sys_sendmsg+0xb5/0x140
  do_syscall_64+0x4c/0x100
  entry_SYSCALL_64_after_hwframe+0x4b/0x53

Fixes: 0a28bfd4971f ("net/macsec: Add MACsec skb_metadata_dst Tx Data path support")
Signed-off-by: Jianbo Liu <jianbol@nvidia.com>
Reviewed-by: Patrisious Haddad <phaddad@nvidia.com>
Reviewed-by: Chris Mi <cmi@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Reviewed-by: Sabrina Dubroca <sd@queasysnail.net>
Link: https://patch.msgid.link/20241021100309.234125-1-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

ASoC: cs42l51: Fix some error handling paths in cs42l51_probe()

[ Upstream commit d221b844ee79823ffc29b7badc4010bdb0960224 ]

If devm_gpiod_get_optional() fails, we need to disable previously enabled
regulators, as done in the other error handling path of the function.

Also, gpiod_set_value_cansleep(, 1) needs to be called to undo a
potential gpiod_set_value_cansleep(, 0).
If the "reset" gpio is not defined, this additional call is just a no-op.

This behavior is the same as the one already in the .remove() function.

Fixes: 11b9cd748e31 ("ASoC: cs42l51: add reset management")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://patch.msgid.link/a5e5f4b9fb03f46abd2c93ed94b5c395972ce0d1.1729975570.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>

wifi: iwlwifi: mvm: don't add default link in fw restart flow

[ Upstream commit 734a377e1eacc5153bae0ccd4423365726876e93 ]

When we add the vif (and its default link) in fw restart we may
override the link that already exists. We take care of this but if
link 0 is a valid MLO link, then we will re-create a default link on
mvmvif->link[0] and we'll loose the real link we had there.

In non-MLO, we need to re-create the default link upon the interface
creation, this is fine. In MLO, we'll just wait for change_vif_links()
to re-build the links.

Fixes: bf976c814c86 ("wifi: iwlwifi: mvm: implement link change ops")
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20241010140328.385bfea1b2e9.I4a127312285ccb529cc95cc4edf6fbe1e0a136ad@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>

wifi: iwlwifi: mvm: Fix response handling in iwl_mvm_send_recovery_cmd()

[ Upstream commit 07a6e3b78a65f4b2796a8d0d4adb1a15a81edead ]

1. The size of the response packet is not validated.
2. The response buffer is not freed.

Resolve these issues by switching to iwl_mvm_send_cmd_status(),
which handles both size validation and frees the buffer.

Fixes: f130bb75d881 ("iwlwifi: add FW recovery flow")
Signed-off-by: Daniel Gabay <daniel.gabay@intel.com>
Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
Link: https://patch.msgid.link/20241010140328.76c73185951e.Id3b6ca82ced2081f5ee4f33c997491d0ebda83f7@changeid
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>