The linux-5.4.y commit 8781011a302b ("bpf: Test_progs, add test to catch
retval refine error handling") fails to build when libbpf headers are
not installed, as it tries to include <bpf/bpf_helpers.h>:
progs/test_get_stack_rawtp_err.c:4:10:
fatal error: 'bpf/bpf_helpers.h' file not found
For 5.4-stable (only) the new test prog needs to include "bpf_helpers.h"
instead (like all the rest of progs/*.c do) because 5.4-stable does not
carry commit e01a75c15969 ("libbpf: Move bpf_{helpers, helper_defs,
endian, tracing}.h into libbpf").
Signed-off-by: Kamal Mostafa <kamal@canonical.com> Fixes: 8781011a302b ("bpf: Test_progs, add test to catch retval refine error handling") Cc: <stable@vger.kernel.org> # v5.4 Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Fix up RET_IF as CHECK macro to make selftests compile again.
Fixes: b911c5e8686a ("selftests: bpf: Reset global state between reuseport test runs") Signed-off-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This logic is re-used for parsing a set of online CPUs. Having it as an
isolated piece of code working with input string makes it conveninent to test
this logic as well. While refactoring, also improve the robustness of original
implementation.
tcp_bpf_recvmsg() invokes sk_psock_get(), which returns a reference of
the specified sk_psock object to "psock" with increased refcnt.
When tcp_bpf_recvmsg() returns, local variable "psock" becomes invalid,
so the refcount should be decreased to keep refcount balanced.
The reference counting issue happens in several exception handling paths
of tcp_bpf_recvmsg(). When those error scenarios occur such as "flags"
includes MSG_ERRQUEUE, the function forgets to decrease the refcnt
increased by sk_psock_get(), causing a refcnt leak.
Fix this issue by calling sk_psock_put() or pulling up the error queue
read handling when those error scenarios occur.
Fixes: e7a5f1f1cd000 ("bpf/sockmap: Read psock ingress_msg before sk_receive_queue") Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn> Signed-off-by: Xin Tan <tanxin.ctf@gmail.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com> Link: https://lore.kernel.org/bpf/1587872115-42805-1-git-send-email-xiyuyang19@fudan.edu.cn Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
I've noticed that when krb5i or krb5p security is in use,
retransmitted requests are missing the server's duplicate reply
cache. The computed checksum on the retransmitted request does not
match the cached checksum, resulting in the server performing the
retransmitted request again instead of returning the cached reply.
The assumptions made when removing xdr_buf_trim() were not correct.
In the send paths, the upper layer has already set the segment
lengths correctly, and shorting the buffer's content is simply a
matter of reducing buf->len.
xdr_buf_trim() is the right answer in the receive/unwrap path on
both the client and the server. The buffer segment lengths have to
be shortened one-by-one.
On the server side in particular, head.iov_len needs to be updated
correctly to enable nfsd_cache_csum() to work correctly. The simple
buf->len computation doesn't do that, and that results in
checksumming stale data in the buffer.
The problem isn't noticed until there's significant instability of
the RPC transport. At that point, the reliability of retransmit
detection on the server becomes crucial.
The Clock Pulse Generator (CPG) device node lacks the extal2 clock.
This may lead to a failure registering the "r" clock, or to a wrong
parent for the "usb24s" clock, depending on MD_CK2 pin configuration and
boot loader CPG_USBCKCR register configuration.
This went unnoticed, as this does not affect the single upstream board
configuration, which relies on the first clock input only.
The R-Mobile APE6 Compare Match Timer 1 generates 8 interrupts, one for
each channel, but currently only 1 is described.
Fix this by adding the missing interrupts.
Using SDMA1 with UART1 is causing a "Timeout waiting for CH0" error.
This patch changes to ahb clock from SDMA1_ROOT to AHB which fixes the
timeout error.
The device tree compiler complains that the dwc3 nodes have regs
properties but no matching unit addresses.
Add the unit addresses to the device node name. While at it, also rename
the nodes from "dwc3" to "usb", as guidelines require device nodes have
generic names.
Fixes: 7144224f2c2b ("arm64: dts: rockchip: support dwc3 USB for rk3399") Signed-off-by: Chen-Yu Tsai <wens@csie.org> Link: https://lore.kernel.org/r/20200327030414.5903-7-wens@kernel.org Signed-off-by: Heiko Stuebner <heiko@sntech.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In some board device tree files, "rk805" was used for the RK805 PMIC's
node name. However the policy for device trees is that generic names
should be used.
Replace the "rk805" node name with the generic "pmic" name.
In the process of moving the VIM3 audio nodes to a G12B specific dtsi
for enabling the SM1 based VIM3L, the frddr_a status = "okay" property
got dropped.
This re-enables the frddr_a node to fix audio support.
Fixes: 4f26cc1c96c9 ("arm64: dts: khadas-vim3: move common nodes into meson-khadas-vim3.dtsi") Reported-by: Christian Hewitt <christianshewitt@gmail.com> Signed-off-by: Neil Armstrong <narmstrong@baylibre.com> Signed-off-by: Kevin Hilman <khilman@baylibre.com> Reviewed-by: Jerome Brunet <jbrunet@baylibre.com> Tested-by: Jerome Brunet <jbrunet@baylibre.com> Link: https://lore.kernel.org/r/20191018140216.4257-1-narmstrong@baylibre.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
'dev' is allocated in 'net2272_probe_init()'. It must be freed in the error
handling path, as already done in the remove function (i.e.
'net2272_plat_remove()')
Change the logic of FAN_ONDIR in two ways that are similar to the logic
of FAN_EVENT_ON_CHILD, that was fixed in commit 54a307ba8d3c ("fanotify:
fix logic of events on child"):
1. The flag is meaningless in ignore mask
2. The flag refers only to events in the mask of the mark where it is set
This is what the fanotify_mark.2 man page says about FAN_ONDIR:
"Without this flag, only events for files are created." It doesn't
say anything about setting this flag in ignore mask to stop getting
events on directories nor can I think of any setup where this capability
would be useful.
Currently, when marks masks are merged, the FAN_ONDIR flag set in one
mark affects the events that are set in another mark's mask and this
behavior causes unexpected results. For example, a user adds a mark on a
directory with mask FAN_ATTRIB | FAN_ONDIR and a mount mark with mask
FAN_OPEN (without FAN_ONDIR). An opendir() of that directory (which is
inside that mount) generates a FAN_OPEN event even though neither of the
marks requested to get open events on directories.
The check for the HWO flag in dwc3_gadget_ep_reclaim_trb_sg()
causes us to break out of the loop before we call
dwc3_gadget_ep_reclaim_completed_trb(), which is what likely
should be clearing the HWO flag.
This can cause odd behavior where we never reclaim all the trbs
in the sg list, so we never call giveback on a usb req, and that
will causes transfer stalls.
This effectively resovles the adb stalls seen on HiKey960
after userland changes started only using AIO in adbd.
Cc: YongQin Liu <yongqin.liu@linaro.org> Cc: Anurag Kumar Vulisha <anurag.kumar.vulisha@xilinx.com> Cc: Yang Fei <fei.yang@intel.com> Cc: Thinh Nguyen <thinhn@synopsys.com> Cc: Tejas Joglekar <tejas.joglekar@synopsys.com> Cc: Andrzej Pietrasiewicz <andrzej.p@collabora.com> Cc: Jack Pham <jackp@codeaurora.org> Cc: Josh Gao <jmgao@google.com> Cc: Todd Kjos <tkjos@google.com> Cc: Felipe Balbi <balbi@kernel.org> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: linux-usb@vger.kernel.org Cc: stable@vger.kernel.org #4.20+ Signed-off-by: John Stultz <john.stultz@linaro.org> Signed-off-by: Felipe Balbi <balbi@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The following changes prevent the unrecoverable freezes and rcu_sched
stall warnings experienced in each of my attempts to take advantage of
lima.
Replace the COMPOSITE_NOGATE definition of aclk_gpu_pre with a
COMPOSITE that retains the selection of HDMIPHY as the PLL source, but
instead makes uses of the aclk_gpu PLL source gate and parent names
defined by mux_pll_src_4plls_p rather than mux_aclk_gpu_pre_p.
Remove the now unused mux_aclk_gpu_pre_p and the four named but also
unused definitions (cpll_gpu, gpll_gpu, hdmiphy_gpu and usb480m_gpu)
of the aclk_gpu PLL source gate.
Use the correct gate offset for aclk_gpu and aclk_gpu_noc.
I goofed when I added mm->user_ns support to would_dump. I missed the
fact that in the case of binfmt_loader, binfmt_em86, binfmt_misc, and
binfmt_script bprm->file is reassigned. Which made the move of
would_dump from setup_new_exec to __do_execve_file before exec_binprm
incorrect as it can result in would_dump running on the script instead
of the interpreter of the script.
The net result is that the code stopped making unreadable interpreters
undumpable. Which allows them to be ptraced and written to disk
without special permissions. Oops.
The move was necessary because the call in set_new_exec was after
bprm->mm was no longer valid.
To correct this mistake move the misplaced would_dump from
__do_execve_file into flos_old_exec, before exec_mmap is called.
I tested and confirmed that without this fix I can attach with gdb to
a script with an unreadable interpreter, and with this fix I can not.
Cc: stable@vger.kernel.org Fixes: f84df2a6f268 ("exec: Ensure mm->user_ns contains the execed files") Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The unwind_state 'error' field is used to inform the reliable unwinding
code that the stack trace can't be trusted. Set this field for all
errors in __unwind_start().
Also, move the zeroing out of the unwind_state struct to before the ORC
table initialization check, to prevent the caller from reading
uninitialized data if the ORC table is corrupted.
Fixes: af085d9084b4 ("stacktrace/x86: add function for detecting reliable stack traces") Fixes: d3a09104018c ("x86/unwinder/orc: Dont bail on stack overflow") Fixes: 98d0c8ebf77e ("x86/unwind/orc: Prevent unwinding before ORC initialization") Reported-by: Pavel Machek <pavel@denx.de> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Link: https://lkml.kernel.org/r/d6ac7215a84ca92b895fdd2e1aa546729417e6e6.1589487277.git.jpoimboe@redhat.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
... or the odyssey of trying to disable the stack protector for the
function which generates the stack canary value.
The whole story started with Sergei reporting a boot crash with a kernel
built with gcc-10:
Kernel panic — not syncing: stack-protector: Kernel stack is corrupted in: start_secondary
CPU: 1 PID: 0 Comm: swapper/1 Not tainted 5.6.0-rc5—00235—gfffb08b37df9 #139
Hardware name: Gigabyte Technology Co., Ltd. To be filled by O.E.M./H77M—D3H, BIOS F12 11/14/2013
Call Trace:
dump_stack
panic
? start_secondary
__stack_chk_fail
start_secondary
secondary_startup_64
-—-[ end Kernel panic — not syncing: stack—protector: Kernel stack is corrupted in: start_secondary
This happens because gcc-10 tail-call optimizes the last function call
in start_secondary() - cpu_startup_entry() - and thus emits a stack
canary check which fails because the canary value changes after the
boot_init_stack_canary() call.
To fix that, the initial attempt was to mark the one function which
generates the stack canary with:
however, using the optimize attribute doesn't work cumulatively
as the attribute does not add to but rather replaces previously
supplied optimization options - roughly all -fxxx options.
The key one among them being -fno-omit-frame-pointer and thus leading to
not present frame pointer - frame pointer which the kernel needs.
The next attempt to prevent compilers from tail-call optimizing
the last function call cpu_startup_entry(), shy of carving out
start_secondary() into a separate compilation unit and building it with
-fno-stack-protector, was to add an empty asm("").
This current solution was short and sweet, and reportedly, is supported
by both compilers but we didn't get very far this time: future (LTO?)
optimization passes could potentially eliminate this, which leads us
to the third attempt: having an actual memory barrier there which the
compiler cannot ignore or move around etc.
That should hold for a long time, but hey we said that about the other
two solutions too so...
Reported-by: Sergei Trofimovich <slyfox@gentoo.org> Signed-off-by: Borislav Petkov <bp@suse.de> Tested-by: Kalle Valo <kvalo@codeaurora.org> Cc: <stable@vger.kernel.org> Link: https://lkml.kernel.org/r/20200314164451.346497-1-slyfox@gentoo.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Failed async writes that are requeued may not clean up a refcount
on the file, which can result in a leaked open. This scenario arises
very reliably when using persistent handles and a reconnect occurs
while writing.
cifs_writev_requeue only releases the reference if the write fails
(rc != 0). The server->ops->async_writev operation will take its own
reference, so the initial reference can always be released.
Signed-off-by: Adam McCoy <adam@forsedomani.com> Signed-off-by: Steve French <stfrench@microsoft.com> CC: Stable <stable@vger.kernel.org> Reviewed-by: Pavel Shilovsky <pshilov@microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This patch adds a basic cursor check when an atomic test-only commit is
performed. The position and size of the cursor plane is checked.
This should fix user-space relying on atomic checks to assign buffers to
planes.
Signed-off-by: Simon Ser <contact@emersion.fr> Reported-by: Roman Gilg <subdiff@gmail.com>
References: https://github.com/emersion/libliftoff/issues/46 Cc: Alex Deucher <alexander.deucher@amd.com> Cc: Harry Wentland <hwentlan@amd.com> Reviewed-by: Nicholas Kazlauskas <nicholas.kazlauskas@amd.com> Signed-off-by: Alex Deucher <alexander.deucher@amd.com> Cc: stable@vger.kernel.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The Y Soft yapp4 platform supports up to two Ethernet ports.
The Ursa board though has only one Ethernet port populated and that is
the port@2. Since the introduction of this platform into mainline a wrong
port was deleted and the Ethernet could never work. Fix this by deleting
the correct port node.
Fixes: 87489ec3a77f ("ARM: dts: imx: Add Y Soft IOTA Draco, Hydra and Ursa boards") Cc: stable@vger.kernel.org Signed-off-by: Michal Vokáč <michal.vokac@ysoft.com> Signed-off-by: Shawn Guo <shawnguo@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The I2C2 pins are already used and the following errors are seen:
imx27-pinctrl 10015000.iomuxc: pin MX27_PAD_I2C2_SDA already requested by 10012000.i2c; cannot claim for 1001d000.i2c
imx27-pinctrl 10015000.iomuxc: pin-69 (1001d000.i2c) status -22
imx27-pinctrl 10015000.iomuxc: could not request pin 69 (MX27_PAD_I2C2_SDA) from group i2c2grp on device 10015000.iomuxc
imx-i2c 1001d000.i2c: Error applying setting, reverse things back
imx-i2c: probe of 1001d000.i2c failed with error -22
Fix it by adding the correct I2C1 IOMUX entries for the pinctrl_i2c1 group.
Even though commit cfb5d65f2595 ("ARM: dts: dra7: Add bus_dma_limit
for L3 bus") added bus_dma_limit for L3 bus, the PCIe controller
gets incorrect value of bus_dma_limit.
Fix it by adding empty dma-ranges property to axi@0 and axi@1
(parent device tree node of PCIe controller).
Cc: stable@kernel.org Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com> Signed-off-by: Tony Lindgren <tony@atomide.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
random:~$ dmesg | tail -1
[523884.039227] [drm] Reducing the compressed framebuffer size. This may lead to less power savings than a non-reduced-size. Try to increase stolen memory size if available in BIOS.
random:~$ dmesg | grep -c "Reducing the compressed"
47
This patch makes it DRM_INFO_ONCE() just like the similar message
farther down in that function is pr_info_once().
Cc: stable@vger.kernel.org Signed-off-by: Peter Jones <pjones@redhat.com> Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/1745 Link: https://patchwork.freedesktop.org/patch/msgid/20180706190424.29194-1-pjones@redhat.com
[vsyrjala: Rebase due to per-device logging] Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
(cherry picked from commit 6b7fc6a3e6af4ff5773949d0fed70d8e7f68d5ce)
[Rodrigo: port back to DRM_INFO_ONCE] Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On platforms with IOMMU enabled, multiple SGs can be coalesced into one
by the IOMMU driver. In that case the SG list processing as part of the
completion of a urb on a bulk endpoint can result into a NULL pointer
dereference with the below stack dump.
<6> Unable to handle kernel NULL pointer dereference at virtual address 0000000c
<6> pgd = c0004000
<6> [0000000c] *pgd=00000000
<6> Internal error: Oops: 5 [#1] PREEMPT SMP ARM
<2> PC is at xhci_queue_bulk_tx+0x454/0x80c
<2> LR is at xhci_queue_bulk_tx+0x44c/0x80c
<2> pc : [<c08907c4>] lr : [<c08907bc>] psr: 000000d3
<2> sp : ca337c80 ip : 00000000 fp : ffffffff
<2> r10: 00000000 r9 : 50037000 r8 : 00004000
<2> r7 : 00000000 r6 : 00004000 r5 : 00000000 r4 : 00000000
<2> r3 : 00000000 r2 : 00000082 r1 : c2c1a200 r0 : 00000000
<2> Flags: nzcv IRQs off FIQs off Mode SVC_32 ISA ARM Segment none
<2> Control: 10c0383d Table: b412c06a DAC: 00000051
<6> Process usb-storage (pid: 5961, stack limit = 0xca336210)
<snip>
<2> [<c08907c4>] (xhci_queue_bulk_tx)
<2> [<c0881b3c>] (xhci_urb_enqueue)
<2> [<c0831068>] (usb_hcd_submit_urb)
<2> [<c08350b4>] (usb_sg_wait)
<2> [<c089f384>] (usb_stor_bulk_transfer_sglist)
<2> [<c089f2c0>] (usb_stor_bulk_srb)
<2> [<c089fe38>] (usb_stor_Bulk_transport)
<2> [<c089f468>] (usb_stor_invoke_transport)
<2> [<c08a11b4>] (usb_stor_control_thread)
<2> [<c014a534>] (kthread)
The above NULL pointer dereference is the result of block_len and the
sent_len set to zero after the first SG of the list when IOMMU driver
is enabled. Because of this the loop of processing the SGs has run
more than num_sgs which resulted in a sg_next on the last SG of the
list which has SG_END set.
Fix this by check for the sg before any attributes of the sg are
accessed.
[modified reason for null pointer dereference in commit message subject -Mathias] Fixes: f9c589e142d04 ("xhci: TD-fragment, align the unsplittable case with a bounce buffer") Cc: stable@vger.kernel.org Signed-off-by: Sriharsha Allenki <sallenki@codeaurora.org> Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com> Link: https://lore.kernel.org/r/20200514110432.25564-2-mathias.nyman@linux.intel.com Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This bug occurs when a size variable used for a buffer
is misused to access its strcpy-ed buffer.
Given a buffer along with its size variable (taken from user input),
from which, a new buffer is created using kstrdup().
Due to the original buffer containing 0 value in the middle,
the size of the kstrdup-ed buffer becomes smaller than that of the original.
So accessing the kstrdup-ed buffer with the same size variable
triggers memory access violation.
The fix makes sure no zero value in the buffer,
by comparing the strlen() of the orignal buffer with the size variable,
so that the access to the kstrdup-ed buffer is safe.
BUG: KASAN: slab-out-of-bounds in gadget_dev_desc_UDC_store+0x1ba/0x200
drivers/usb/gadget/configfs.c:266
Read of size 1 at addr ffff88806a55dd7e by task syz-executor.0/17208
While removing the host (e.g. for USB role switch from host to device),
if runtime pm is enabled by user, below oops occurs on dwc3 and cdns3
platforms.
Keeping the xhci-plat device active during host removal, and disabling
runtime pm before calling pm_runtime_set_suspended() fixes them.
On Tue, May 12, 2020 at 09:36:07PM +0800, Kai-Heng Feng wrote [1]:
> This patch prevents my Raven Ridge xHCI from getting runtime suspend.
The problem described in v5.6 commit 1208f9e1d758c9 ("USB: hub: Fix the
broken detection of USB3 device in SMSC hub") applies solely to the
USB5534B hub [2] present on the Kingfisher Infotainment Carrier Board,
manufactured by Shimafuji Electric Inc [3].
Despite that, the aforementioned commit applied the quirk to _all_ hubs
carrying vendor ID 0x424 (i.e. SMSC), of which there are more [4] than
initially expected. Consequently, the quirk is now enabled on platforms
carrying SMSC/Microchip hub models which potentially don't exhibit the
original issue.
To avoid reports like [1], further limit the quirk's scope to
USB5534B [2], by employing both Vendor and Product ID checks.
The rawmidi core allows user to resize the runtime buffer via ioctl,
and this may lead to UAF when performed during concurrent reads or
writes: the read/write functions unlock the runtime lock temporarily
during copying form/to user-space, and that's the race window.
This patch fixes the hole by introducing a reference counter for the
runtime buffer read/write access and returns -EBUSY error when the
resize is performed concurrently against read/write.
Note that the ref count field is a simple integer instead of
refcount_t here, since the all contexts accessing the buffer is
basically protected with a spinlock, hence we need no expensive atomic
ops. Also, note that this busy check is needed only against read /
write functions, and not in receive/transmit callbacks; the race can
happen only at the spinlock hole mentioned in the above, while the
whole function is protected for receive / transmit callbacks.
ASUS ZenBook UX431DA requires an additional COEF setup when booted
from the recent Windows 10, otherwise it produces the noisy output.
The quirk turns on COEF 0x1b bit 10 that has been cleared supposedly
due to the pop noise reduction.
Lenovo Thinkpad T530 seems to have a sensitive internal mic capture
that needs to limit the mic boost like a few other Thinkpad models.
Although we may change the quirk for ALC269_FIXUP_LENOVO_DOCK, this
hits way too many other laptop models, so let's add a new fixup model
that limits the internal mic boost on top of the existing quirk and
apply to only T530.
In commit 2bef9aed6f0e ("usb: usbfs: correct kernel->user page attribute
mismatch") we switched from always calling remap_pfn_range() to call
dma_mmap_coherent() to handle issues with systems with non-coherent USB host
controller drivers. Unfortunatly, as syzbot quickly told us, not all the world
is host controllers with DMA support, so we need to check what host controller
we are attempting to talk to before doing this type of allocation.
Thanks to Christoph for the quick idea of how to fix this.
Fixes: 2bef9aed6f0e ("usb: usbfs: correct kernel->user page attribute mismatch") Cc: Christoph Hellwig <hch@lst.de> Cc: Hillf Danton <hdanton@sina.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Jeremy Linton <jeremy.linton@arm.com> Cc: stable <stable@vger.kernel.org> Reported-by: syzbot+353be47c9ce21b68b7ed@syzkaller.appspotmail.com Reviewed-by: Jeremy Linton <jeremy.linton@arm.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Link: https://lore.kernel.org/r/20200514112711.1858252-1-gregkh@linuxfoundation.org Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On some architectures (e.g. arm64) requests for
IO coherent memory may use non-cachable attributes if
the relevant device isn't cache coherent. If these
pages are then remapped into userspace as cacheable,
they may not be coherent with the non-cacheable mappings.
In particular this happens with libusb, when it attempts
to create zero-copy buffers for use by rtl-sdr
(https://github.com/osmocom/rtl-sdr/). On low end arm
devices with non-coherent USB ports, the application will
be unexpectedly killed, while continuing to work fine on
arm machines with coherent USB controllers.
This bug has been discovered/reported a few times over
the last few years. In the case of rtl-sdr a compile time
option to enable/disable zero copy was implemented to
work around it.
Rather than relaying on application specific workarounds,
dma_mmap_coherent() can be used instead of remap_pfn_range().
The page cache/etc attributes will then be correctly set in
userspace to match the kernel mapping.
gcc-10 has started warning about conflicting types for a few new
built-in functions, particularly 'free()'.
This results in warnings like:
crypto/xts.c:325:13: warning: conflicting types for built-in function ‘free’; expected ‘void(void *)’ [-Wbuiltin-declaration-mismatch]
because the crypto layer had its local freeing functions called
'free()'.
Gcc-10 is in the wrong here, since that function is marked 'static', and
thus there is no chance of confusion with any standard library function
namespace.
But the simplest thing to do is to just use a different name here, and
avoid this gcc mis-feature.
[ Side note: gcc knowing about 'free()' is in itself not the
mis-feature: the semantics of 'free()' are special enough that a
compiler can validly do special things when seeing it.
So the mis-feature here is that gcc thinks that 'free()' is some
restricted name, and you can't shadow it as a local static function.
Making the special 'free()' semantics be a function attribute rather
than tied to the name would be the much better model ]
It seems that for whatever reason, gcc-10 ends up not inlining a couple
of functions that used to be inlined before. Even if they only have one
single callsite - it looks like gcc may have decided that the code was
unlikely, and not worth inlining.
The code generation difference is harmless, but caused a few new section
mismatch errors, since the (now no longer inlined) function wasn't in
the __init section, but called other init functions:
Section mismatch in reference from the function kexec_free_initrd() to the function .init.text:free_initrd_mem()
Section mismatch in reference from the function tpm2_calc_event_log_size() to the function .init.text:early_memremap()
Section mismatch in reference from the function tpm2_calc_event_log_size() to the function .init.text:early_memunmap()
So add the appropriate __init annotation to make modpost not complain.
In both cases there were trivially just a single callsite from another
__init function.
Due to a bug-report that was compiler-dependent, I updated one of my
machines to gcc-10. That shows a lot of new warnings. Happily they
seem to be mostly the valid kind, but it's going to cause a round of
churn for getting rid of them..
This is the really low-hanging fruit of removing a couple of zero-sized
arrays in some core code. We have had a round of these patches before,
and we'll have many more coming, and there is nothing special about
these except that they were particularly trivial, and triggered more
warnings than most.
gcc-10 now warns about passing aliasing pointers to functions that take
restricted pointers.
That's actually a great warning, and if we ever start using 'restrict'
in the kernel, it might be quite useful. But right now we don't, and it
turns out that the only thing this warns about is an idiom where we have
declared a few functions to be "printf-like" (which seems to make gcc
pick up the restricted pointer thing), and then we print to the same
buffer that we also use as an input.
And people do that as an odd concatenation pattern, with code like this:
where we have 'buffer' as both the destination of the final result, and
as the initial argument.
Yes, it's a bit questionable. And outside of the kernel, people do have
standard declarations like
int snprintf( char *restrict buffer, size_t bufsz,
const char *restrict format, ... );
where that output buffer is marked as a restrict pointer that cannot
alias with any other arguments.
But in the context of the kernel, that 'use snprintf() to concatenate to
the end result' does work, and the pattern shows up in multiple places.
And we have not marked our own version of snprintf() as taking restrict
pointers, so the warning is incorrect for now, and gcc picks it up on
its own.
If we do start using 'restrict' in the kernel (and it might be a good
idea if people find places where it matters), we'll need to figure out
how to avoid this issue for snprintf and friends. But in the meantime,
this warning is not useful.
This is the final array bounds warning removal for gcc-10 for now.
Again, the warning is good, and we should re-enable all these warnings
when we have converted all the legacy array declaration cases to
flexible arrays. But in the meantime, it's just noise.
This is another fine warning, related to the 'zero-length-bounds' one,
but hitting the same historical code in the kernel.
Because C didn't historically support flexible array members, we have
code that instead uses a one-sized array, the same way we have cases of
zero-sized arrays.
The one-sized arrays come from either not wanting to use the gcc
zero-sized array extension, or from a slight convenience-feature, where
particularly for strings, the size of the structure now includes the
allocation for the final NUL character.
So with a "char name[1];" at the end of a structure, you can do things
like
v = my_malloc(sizeof(struct vendor) + strlen(name));
and avoid the "+1" for the terminator.
Yes, the modern way to do that is with a flexible array, and using
'offsetof()' instead of 'sizeof()', and adding the "+1" by hand. That
also technically gets the size "more correct" in that it avoids any
alignment (and thus padding) issues, but this is another long-term
cleanup thing that will not happen for 5.7.
So disable the warning for now, even though it's potentially quite
useful. Having a slew of warnings that then hide more urgent new issues
is not an improvement.
This is a fine warning, but we still have a number of zero-length arrays
in the kernel that come from the traditional gcc extension. Yes, they
are getting converted to flexible arrays, but in the meantime the gcc-10
warning about zero-length bounds is very verbose, and is hiding other
issues.
I missed one actual build failure because it was hidden among hundreds
of lines of warning. Thankfully I caught it on the second go before
pushing things out, but it convinced me that I really need to disable
the new warnings for now.
We'll hopefully be all done with our conversion to flexible arrays in
the not too distant future, and we can then re-enable this warning.
We have some rather random rules about when we accept the
"maybe-initialized" warnings, and when we don't.
For example, we consider it unreliable for gcc versions < 4.9, but also
if -O3 is enabled, or if optimizing for size. And then various kernel
config options disabled it, because they know that they trigger that
warning by confusing gcc sufficiently (ie PROFILE_ALL_BRANCHES).
And now gcc-10 seems to be introducing a lot of those warnings too, so
it falls under the same heading as 4.9 did.
At the same time, we have a very straightforward way to _enable_ that
warning when wanted: use "W=2" to enable more warnings.
So stop playing these ad-hoc games, and just disable that warning by
default, with the known and straight-forward "if you want to work on the
extra compiler warnings, use W=123".
Would it be great to have code that is always so obvious that it never
confuses the compiler whether a variable is used initialized or not?
Yes, it would. In a perfect world, the compilers would be smarter, and
our source code would be simpler.
That's currently not the world we live in, though.
Returning the error code via a 'int *ret' when the function returns a
pointer is very un-kernely and causes gcc 10's static analysis to choke:
net/rds/message.c: In function ‘rds_message_map_pages’:
net/rds/message.c:358:10: warning: ‘ret’ may be used uninitialized in this function [-Wmaybe-uninitialized]
358 | return ERR_PTR(ret);
Use a typical ERR_PTR return instead.
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Acked-by: Santosh Shilimkar <santosh.shilimkar@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Prior to commit e3d3ab64dd66 ("SUNRPC: Use au_rslack when
computing reply buffer size"), there was enough slack in the reply
buffer to commodate filehandles of size 60bytes. However, the real
problem was that the reply buffer size for the MOUNT operation was
not correctly calculated. Received buffer size used the filehandle
size for NFSv2 (32bytes) which is much smaller than the allowed
filehandle size for the v3 mounts.
Fix the reply buffer size (decode arguments size) for the MNT command.
Fixes: 2c94b8eca1a2 ("SUNRPC: Use au_rslack when computing reply buffer size") Signed-off-by: Olga Kornievskaia <kolga@netapp.com> Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
A recent commit 9852ae3fe529 ("mm, memcg: consider subtrees in
memory.events") changed the behavior of memcg events, which will now
consider subtrees in memory.events.
But oom_kill event is a special one as it is used in both cgroup1 and
cgroup2. In cgroup1, it is displayed in memory.oom_control. The file
memory.oom_control is in both root memcg and non root memcg, that is
different with memory.event as it only in non-root memcg. That commit
is okay for cgroup2, but it is not okay for cgroup1 as it will cause
inconsistent behavior between root memcg and non-root memcg.
Here's an example on why this behavior is inconsistent in cgroup1.
root memcg
/
memcg foo
/
memcg bar
Suppose there's an oom_kill in memcg bar, then the oon_kill will be
For the non-root memcg, its memory.oom_control(oom_kill) includes its
descendants' oom_kill, but for root memcg, it doesn't include its
descendants' oom_kill. That means, memory.oom_control(oom_kill) has
different meanings in different memcgs. That is inconsistent. Then the
user has to know whether the memcg is root or not.
If we can't fully support it in cgroup1, for example by adding
memory.events.local into cgroup1 as well, then let's don't touch its
original behavior.
Fixes: 9852ae3fe529 ("mm, memcg: consider subtrees in memory.events") Reported-by: Randy Dunlap <rdunlap@infradead.org> Signed-off-by: Yafang Shao <laoar.shao@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Reviewed-by: Shakeel Butt <shakeelb@google.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Chris Down <chris@chrisdown.name> Acked-by: Michal Hocko <mhocko@suse.com> Cc: <stable@vger.kernel.org> Link: http://lkml.kernel.org/r/20200502141055.7378-1-laoar.shao@gmail.com Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
When tsi-as-adc is configured it is possible for in7[0123]_input read to
return an incorrect value if a concurrent read to in[456]_input is
performed. This is caused by a concurrent manipulation of the mux
channel without proper locking as hwmon and mfd use different locks for
synchronization.
Switch hwmon to use the same lock as mfd when accessing the TSI channel.
Fixes: 4f16cab19a3d5 ("hwmon: da9052: Add support for TSI channel") Signed-off-by: Samu Nuutamo <samu.nuutamo@vincit.fi>
[rebase to current master, reword commit message slightly] Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com> Signed-off-by: Guenter Roeck <linux@roeck-us.net> Signed-off-by: Sasha Levin <sashal@kernel.org>
While reading the TCB field in t4_tcb_get_field32() the wrong mask is
passed as a parameter which leads the driver eventually to a kernel
panic/app segfault from access to an illegal SRQ index while flushing the
SRQ completions during connection teardown.
The IB core pkey cache is populated by procedure ib_cache_update().
Initially, the pkey cache pointer is NULL. ib_cache_update allocates a
buffer and populates it with the device's pkeys, via repeated calls to
procedure ib_query_pkey().
If there is a failure in populating the pkey buffer via ib_query_pkey(),
ib_cache_update does not replace the old pkey buffer cache with the
updated one -- it leaves the old cache as is.
Since initially the pkey buffer cache is NULL, when calling
ib_cache_update the first time, a failure in ib_query_pkey() will cause
the pkey buffer cache pointer to remain NULL.
In this situation, any calls subsequent to ib_get_cached_pkey(),
ib_find_cached_pkey(), or ib_find_cached_pkey_exact() will try to
dereference the NULL pkey cache pointer, causing a kernel panic.
Fix this by checking the ib_cache_update() return value.
Fixes: 8faea9fd4a39 ("RDMA/cache: Move the cache per-port data into the main ib_port_data") Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2") Link: https://lore.kernel.org/r/20200507071012.100594-1-leon@kernel.org Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
In the mlx4_ib_post_send() flow, some functions call ib_get_cached_pkey()
without checking its return value. If ib_get_cached_pkey() returns an
error code, these functions should return failure.
Fixes: 1ffeb2eb8be9 ("IB/mlx4: SR-IOV IB context objects and proxy/tunnel SQP support") Fixes: 225c7b1feef1 ("IB/mlx4: Add a driver Mellanox ConnectX InfiniBand adapters") Fixes: e622f2f4ad21 ("IB: split struct ib_send_wr") Link: https://lore.kernel.org/r/20200426075921.130074-1-leon@kernel.org Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il> Signed-off-by: Leon Romanovsky <leonro@mellanox.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The commit below modified rxe_create_mmap_info() to return ERR_PTR's but
didn't update the callers to handle them. Modify rxe_create_mmap_info() to
only return ERR_PTR and fix all error checking after
rxe_create_mmap_info() is called.
Ensure that all other exit paths properly set the error return.
Expired intervals would still match and be dumped to user space until
garbage collection wiped them out. Make sure they stop matching and
disappear (from users' perspective) as soon as they expire.
Fixes: 8d8540c4f5e03 ("netfilter: nft_set_rbtree: add timeout support") Signed-off-by: Phil Sutter <phil@nwl.cc> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
Replace negations of nft_rbtree_interval_end() with a new helper,
nft_rbtree_interval_start(), wherever this helps to visualise the
problem at hand, that is, for all the occurrences except for the
comparison against given flags in __nft_rbtree_get().
We add the new state to the nfsi->open_states list, making it
potentially visible to other threads, before we've finished initializing
it.
That wasn't a problem when all the readers were also taking the i_lock
(as we do here), but since we switched to RCU, there's now a possibility
that a reader could see the partially initialized state.
Symptoms observed were a crash when another thread called
nfs4_get_valid_delegation() on a NULL inode, resulting in an oops like:
As i915 won't allocate extra PDP for current default PML4 table,
so for 3-level ppgtt guest, we would hit kernel pointer access
failure on extra PDP pointers. So this trys to bypass that now.
It won't impact real shadow PPGTT setup, so guest context still
works.
This is verified on 4.15 guest kernel with i915.enable_ppgtt=1
to force on old aliasing ppgtt behavior.
gcc-10 warns around a suspicious access to an empty struct member:
net/netfilter/nf_conntrack_core.c: In function '__nf_conntrack_alloc':
net/netfilter/nf_conntrack_core.c:1522:9: warning: array subscript 0 is outside the bounds of an interior zero-length array 'u8[0]' {aka 'unsigned char[0]'} [-Wzero-length-bounds]
1522 | memset(&ct->__nfct_init_offset[0], 0,
| ^~~~~~~~~~~~~~~~~~~~~~~~~~
In file included from net/netfilter/nf_conntrack_core.c:37:
include/net/netfilter/nf_conntrack.h:90:5: note: while referencing '__nfct_init_offset'
90 | u8 __nfct_init_offset[0];
| ^~~~~~~~~~~~~~~~~~
The code is correct but a bit unusual. Rework it slightly in a way that
does not trigger the warning, using an empty struct instead of an empty
array. There are probably more elegant ways to do this, but this is the
smallest change.
Commit 402cb8dda949 ("fscache: Attach the index key and aux data to
the cookie") added the aux_data and aux_data_len to parameters to
fscache_acquire_cookie(), and updated the callers in the NFS client.
In the process it modified the aux_data to include the change_attr,
but missed adding change_attr to a couple places where aux_data was
used. Specifically, when opening a file and the change_attr is not
added, the following attempt to lookup an object will fail inside
cachefiles_check_object_xattr() = -116 due to
nfs_fscache_inode_check_aux() failing memcmp on auxdata and returning
FSCACHE_CHECKAUX_OBSOLETE.
Fix this by adding nfs_fscache_update_auxdata() to set the auxdata
from all relevant fields in the inode, including the change_attr.
Fixes: 402cb8dda949 ("fscache: Attach the index key and aux data to the cookie") Signed-off-by: Dave Wysochanski <dwysocha@redhat.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
nfs currently behaves differently on 32-bit and 64-bit kernels regarding
the on-disk format of nfs_fscache_inode_auxdata.
That format should really be the same on any kernel, and we should avoid
the 'timespec' type in order to remove that from the kernel later on.
Using plain 'timespec64' would not be good here, since that includes
implied padding and would possibly leak kernel stack data to the on-disk
format on 32-bit architectures.
struct __kernel_timespec would work as a replacement, but open-coding
the two struct members in nfs_fscache_inode_auxdata makes it more
obvious what's going on here, and keeps the current format for 64-bit
architectures.
Commit 402cb8dda949 ("fscache: Attach the index key and aux data to
the cookie") added the index_key and index_key_len parameters to
fscache_acquire_cookie(), and updated the callers in the NFS client.
One of the callers was inside nfs_fscache_get_super_cookie()
and was changed to use the full struct nfs_fscache_key as the
index_key. However, a couple members of this structure contain
pointers and thus will change each time the same NFS share is
remounted. Since index_key is used for fscache_cookie->key_hash
and this subsequently is used to compare cookies, the effectiveness
of fscache with NFS is reduced to the point at which a umount
occurs. Any subsequent remount of the same share will cause a
unique NFS super_block index_key and key_hash to be generated for
the same data, rendering any prior fscache data unable to be
found. A simple reproducer demonstrates the problem.
1. Mount share with 'fsc', create a file, drop page cache
systemctl start cachefilesd
mount -o vers=3,fsc 127.0.0.1:/export /mnt
dd if=/dev/zero of=/mnt/file1.bin bs=4096 count=1
echo 3 > /proc/sys/vm/drop_caches
2. Read file into page cache and fscache, then unmount
dd if=/mnt/file1.bin of=/dev/null bs=4096 count=1
umount /mnt
3. Remount and re-read which should come from fscache
mount -o vers=3,fsc 127.0.0.1:/export /mnt
echo 3 > /proc/sys/vm/drop_caches
dd if=/mnt/file1.bin of=/dev/null bs=4096 count=1
4. Check for READ ops in mountstats - there should be none
grep READ: /proc/self/mountstats
Looking at the history and the removed function, nfs_super_get_key(),
we should only use nfs_fscache_key.key plus any uniquifier, for
the fscache index_key.
Fixes: 402cb8dda949 ("fscache: Attach the index key and aux data to the cookie") Signed-off-by: Dave Wysochanski <dwysocha@redhat.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
Jan reported an issue where an interaction between sign-extending clone's
flag argument on ppc64le and the new CLONE_INTO_CGROUP feature causes
clone() to consistently fail with EBADF.
The whole story is a little longer. The legacy clone() syscall is odd in a
bunch of ways and here two things interact. First, legacy clone's flag
argument is word-size dependent, i.e. it's an unsigned long whereas most
system calls with flag arguments use int or unsigned int. Second, legacy
clone() ignores unknown and deprecated flags. The two of them taken
together means that users on 64bit systems can pass garbage for the upper
32bit of the clone() syscall since forever and things would just work fine.
Just try this on a 64bit kernel prior to v5.7-rc1 where this will succeed
and on v5.7-rc1 where this will fail with EBADF:
int main(int argc, char *argv[])
{
pid_t pid;
/* Note that legacy clone() has different argument ordering on
* different architectures so this won't work everywhere.
*
* Only set the upper 32 bits.
*/
pid = syscall(__NR_clone, 0xffffffff00000000 | SIGCHLD,
NULL, NULL, NULL, NULL);
if (pid < 0)
exit(EXIT_FAILURE);
if (pid == 0)
exit(EXIT_SUCCESS);
if (wait(NULL) != pid)
exit(EXIT_FAILURE);
exit(EXIT_SUCCESS);
}
Since legacy clone() couldn't be extended this was not a problem so far and
nobody really noticed or cared since nothing in the kernel ever bothered to
look at the upper 32 bits.
But once we introduced clone3() and expanded the flag argument in struct
clone_args to 64 bit we opened this can of worms. With the first flag-based
extension to clone3() making use of the upper 32 bits of the flag argument
we've effectively made it possible for the legacy clone() syscall to reach
clone3() only flags. The sign extension scenario is just the odd
corner-case that we needed to figure this out.
The reason we just realized this now and not already when we introduced
CLONE_CLEAR_SIGHAND was that CLONE_INTO_CGROUP assumes that a valid cgroup
file descriptor has been given. So the sign extension (or the user
accidently passing garbage for the upper 32 bits) caused the
CLONE_INTO_CGROUP bit to be raised and the kernel to error out when it
didn't find a valid cgroup file descriptor.
Let's fix this by always capping the upper 32 bits for all codepaths that
are not aware of clone3() features. This ensures that we can't reach
clone3() only features by accident via legacy clone as with the sign
extension case and also that legacy clone() works exactly like before, i.e.
ignoring any unknown flags. This solution risks no regressions and is also
pretty clean.
It turns out that when extending an existing bio, gfs2_find_jhead fails to
check if the block number is consecutive, which leads to incorrect reads for
fragmented journals.
In addition, limit the maximum bio size to an arbitrary value of 2 megabytes:
since commit 07173c3ec276 ("block: enable multipage bvecs"), if we just keep
adding pages until bio_add_page fails, bios will grow much larger than useful,
which pins more memory than necessary with barely any additional performance
gains.
Fixes: f4686c26ecc3 ("gfs2: read journal in large chunks") Cc: stable@vger.kernel.org # v5.2+ Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com> Signed-off-by: Bob Peterson <rpeterso@redhat.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
First, it should be noted that the CQE timeout (60 seconds) is substantial
so a CQE request that times out is really stuck, and the race between
timeout and completion is extremely unlikely. Nevertheless this patch
fixes an issue with it.
Commit ad73d6feadbd7b ("mmc: complete requests from ->timeout")
preserved the existing functionality, to complete the request.
However that had only been necessary because the block layer
timeout handler had been marking the request to prevent it from being
completed normally. That restriction was removed at the same time, the
result being that a request that has gone will have been completed anyway.
That is, the completion was unnecessary.
At the time, the unnecessary completion was harmless because the block
layer would ignore it, although that changed in kernel v5.0.
Note for stable, this patch will not apply cleanly without patch "mmc:
core: Fix recursive locking issue in CQE recovery path"
mmc_mq_timed_out acquires the queue_lock for the first
time. The mmc_blk_cqe_complete_rq function also tries to acquire
the same queue lock resulting in recursive locking where the task
is spinning for the same lock which it has already acquired leading
to watchdog bark.
Fix this issue with the lock only for the required critical section.
In the request completion path with CQE, request type is being checked
after the request is getting completed. This is resulting in returning
the wrong request type and leading to the IO hang issue.
ASYNC request type is getting returned for DCMD type requests.
Because of this mismatch, mq->cqe_busy flag is never getting cleared
and the driver is not invoking blk_mq_hw_run_queue. So requests are not
getting dispatched to the LLD from the block layer.
All these eventually leading to IO hang issues.
So, get the request type before completing the request.
In bpf_tcp_ingress we used apply_bytes to subtract bytes from sg.size
which is used to track total bytes in a message. But this is not
correct because apply_bytes is itself modified in the main loop doing
the mem_charge.
Then at the end of this we have sg.size incorrectly set and out of
sync with actual sk values. Then we can get a splat if we try to
cork the data later and again try to redirect the msg to ingress. To
fix instead of trying to track msg.size do the easy thing and include
it as part of the sk_msg_xfer logic so that when the msg is moved the
sg.size is always correct.
To reproduce the below users will need ingress + cork and hit an
error path that will then try to 'free' the skmsg.
[ 173.699981] BUG: KASAN: null-ptr-deref in sk_msg_free_elem+0xdd/0x120
[ 173.699987] Read of size 8 at addr 0000000000000008 by task test_sockmap/5317
When sk_msg_pop() is called where the pop operation is working on
the end of a sge element and there is no additional trailing data
and there _is_ data in front of pop, like the following case,
|____________a_____________|__pop__|
We have out of order operations where we incorrectly set the pop
variable so that instead of zero'ing pop we incorrectly leave it
untouched, effectively. This can cause later logic to shift the
buffers around believing it should pop extra space. The result is
we have 'popped' more data then we expected potentially breaking
program logic.
It took us a while to hit this case because typically we pop headers
which seem to rarely be at the end of a scatterlist elements but
we can't rely on this.
Fixes: 7246d8ed4dcce ("bpf: helper to pop data from messages") Signed-off-by: John Fastabend <john.fastabend@gmail.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com> Acked-by: Martin KaFai Lau <kafai@fb.com> Link: https://lore.kernel.org/bpf/158861288359.14306.7654891716919968144.stgit@john-Precision-5820-Tower Signed-off-by: Sasha Levin <sashal@kernel.org>
In commit 5a7d202b1574, a logical AND was erroneously changed to an OR,
causing WaIncreaseLatencyIPCEnabled to be enabled unconditionally for
kabylake and coffeelake, even when IPC is disabled. Fix the logic so
that WaIncreaseLatencyIPCEnabled is only used when IPC is enabled.
The snd-firewire-lib.ko has 'amdtp-packet' event of tracepoints. Current
printk format for the event includes 'sizeof(u8)' macro expected to be
extended in compilation time. However, this is not done. As a result,
perf tools cannot parse the event for printing:
According to Braswell NDA Specification Update (#557593),
concurrent read accesses may result in returning 0xffffffff and write
instructions may be dropped. We have an established format for the
commit references, i.e. cdca06e4e859 ("pinctrl: baytrail: Add missing spinlock usage in
byt_gpio_irq_handler")
Fixes: 0bd50d719b00 ("pinctrl: cherryview: prevent concurrent access to GPIO controllers") Signed-off-by: Grace Kao <grace.kao@intel.com> Reported-by: Brian Norris <briannorris@chromium.org> Reviewed-by: Brian Norris <briannorris@chromium.org> Acked-by: Mika Westerberg <mika.westerberg@linux.intel.com> Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
If the EC GPE status is not set after checking all of the other GPEs,
acpi_s2idle_wake() returns 'false', to indicate that the SCI event
that has just triggered is not a system wakeup one, but it does that
without canceling the pending wakeup and re-arming the SCI for system
wakeup which is a mistake, because it may cause s2idle_loop() to busy
spin until the next valid wakeup event. [If that happens, the first
spurious wakeup is still pending after acpi_s2idle_wake() has
returned, so s2idle_enter() does nothing, acpi_s2idle_wake()
is called again and it sees that the SCI has triggered, but no GPEs
are active, so 'false' is returned again, and so on.]
Fix that by moving all of the GPE checking logic from
acpi_s2idle_wake() to acpi_ec_dispatch_gpe() and making the
latter return 'true' only if a non-EC GPE has triggered and
'false' otherwise, which will cause acpi_s2idle_wake() to
cancel the pending SCI wakeup and re-arm the SCI for system
wakeup regardless of the EC GPE status.
This also addresses a lockup observed on an Elitegroup EF20EA laptop
after attempting to wake it up from suspend-to-idle by a key press.
Fixes: d5406284ff80 ("ACPI: PM: s2idle: Refine active GPEs check") Link: https://bugzilla.kernel.org/show_bug.cgi?id=207603 Reported-by: Todd Brandt <todd.e.brandt@linux.intel.com> Fixes: fdde0ff8590b ("ACPI: PM: s2idle: Prevent spurious SCIs from waking up the system") Link: https://lore.kernel.org/linux-acpi/CAB4CAwdqo7=MvyG_PE+PGVfeA17AHF5i5JucgaKqqMX6mjArbQ@mail.gmail.com/ Reported-by: Chris Chiu <chiu@endlessm.com> Tested-by: Chris Chiu <chiu@endlessm.com> Cc: 5.4+ <stable@vger.kernel.org> # 5.4+ Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The commit noted below fixed a case where a pq is left on the sdma wait
list.
It however missed another case.
user_sdma_send_pkts() has two calls from hfi1_user_sdma_process_request().
If the first one fails as indicated by -EBUSY, the pq will be placed on
the waitlist as by design.
If the second call then succeeds, the pq is still on the waitlist setting
up a race with the interrupt handler if a subsequent request uses a
different SDMA engine
Fix by deleting the first call.
The use of pcount and the intent to send a short burst of packets followed
by the larger balance of packets was never correctly implemented, because
the two calls always send pcount packets no matter what. A subsequent
patch will correct that issue.
Fixes: 9a293d1e21a6 ("IB/hfi1: Ensure pq is not left on waitlist") Link: https://lore.kernel.org/r/20200504130917.175613.43231.stgit@awfm-01.aw.intel.com Cc: <stable@vger.kernel.org> Reviewed-by: Kaike Wan <kaike.wan@intel.com> Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com> Signed-off-by: Sasha Levin <sashal@kernel.org>
The kernel prints a message similar to
"[ 28.881959] do_IRQ: 5.36 No irq handler for vector"
when GL975x resumes from suspend. Implement a resume callback to fix this.
Fixes: 31e43f31890c ("mmc: sdhci-pci-gli: Enable MSI interrupt for GL975x") Co-developed-by: Renius Chen <renius.chen@genesyslogic.com.tw> Signed-off-by: Renius Chen <renius.chen@genesyslogic.com.tw> Tested-by: Dave Flogeras <dflogeras2@gmail.com> Signed-off-by: Ben Chuang <ben.chuang@genesyslogic.com.tw> Tested-by: Vineeth Pillai <vineethrp@gmail.com> Acked-by: Adrian Hunter <adrian.hunter@intel.com> Link: https://lore.kernel.org/r/20200427103048.20785-1-benchuanggli@gmail.com Cc: stable@vger.kernel.org Signed-off-by: Samuel Zou <zou_wei@huawei.com>
[Samuel Zou: Make sdhci_pci_gli_resume() static] Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> Signed-off-by: Sasha Levin <sashal@kernel.org>