io_uring can currently open/close regular files or fixed/direct
descriptors. Or you can instantiate a fixed descriptor from a regular
one, and then close the regular descriptor. But you currently can't turn
a purely fixed/direct descriptor into a regular file descriptor.

IORING_OP_FIXED_FD_INSTALL adds support for installing a direct
descriptor into the normal file table, just like receiving a file
descriptor or opening a new file would do. This is all nicely abstracted
into receive_fd(), and hence adding support for this is truly trivial.

Since direct descriptors are only usable within io_uring itself, it can
be useful to turn them into real file descriptors if they ever need to
be accessed via normal syscalls. This can either be a transitory thing,
or just a permanent transition for a given direct descriptor.

By default, new fds are installed with O_CLOEXEC set. The application
can disable O_CLOEXEC by setting IORING_FIXED_FD_NO_CLOEXEC in the
sqe->install_fd_flags member.
Suggested-by: Christian Brauner <brauner@kernel.org>
Reviewed-by: Christian Brauner <brauner@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

With io_uring_types.h we see all required definitions to inline
io_uring_cmd_get_task().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/aa8e317f09e651a5f3e72f8c0ad3902084c1f930.1701391955.git.asml.silence@gmail.comSigned-off-by: Jens Axboe <axboe@kernel.dk>

Now as we can easily include io_uring_types.h, move IOU_F_TWQ_LAZY_WAKE
and inline io_uring_cmd_do_in_task_lazy().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/2ec9fb31dd192d1c5cf26d0a2dec5657d88a8e48.1701391955.git.asml.silence@gmail.comSigned-off-by: Jens Axboe <axboe@kernel.dk>

linux/io_uring.h is slowly becoming a rubbish bin where we put
anything exposed to other subsystems. For instance, the task exit
hooks and io_uring cmd infra are completely orthogonal and don't need
each other's definitions. Start cleaning it up by splitting out all
command bits into a new header file.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/7ec50bae6e21f371d3850796e716917fc141225a.1701391955.git.asml.silence@gmail.comSigned-off-by: Jens Axboe <axboe@kernel.dk>

At one point in time we had an optimisation that would not spin up a
linked timeout timer when the master request successfully completes
inline (during the first nowait execution attempt). We somehow lost it,
so this patch restores it back.

Note, that it's fine using io_arm_ltimeout() after the io_issue_sqe()
completes the request because of delayed completion, but that that adds
unwanted overhead.
Reported-by: Christian Mazakas <christian.mazakas@gmail.com>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/8bf69c2a4beec14c565c85c86edb871ca8b8bcc8.1701390926.git.asml.silence@gmail.comSigned-off-by: Jens Axboe <axboe@kernel.dk>

IOPOLL request should never return IOU_OK, so the following iopoll
queueing check in io_issue_sqe() after getting IOU_OK doesn't make any
sense as would never turn true. Let's optimise on that and return a bit
earlier. It's also much more resilient to potential bugs from
mischieving iopoll implementations.

Cc:  <stable@vger.kernel.org>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/2f8690e2fa5213a2ff292fac29a7143c036cdd60.1701390926.git.asml.silence@gmail.comSigned-off-by: Jens Axboe <axboe@kernel.dk>

Merge branch 'vfs.file' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs into for-6.8/io_uring

Merge vfs.file from the VFS tree to avoid conflicts with receive_fd() now
having 3 arguments rather than just 2.

* 'vfs.file' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  file: remove __receive_fd()
  file: stop exposing receive_fd_user()
  fs: replace f_rcuhead with f_task_work
  file: remove pointless wrapper
  file: s/close_fd_get_file()/file_close_fd()/g
  Improve __fget_files_rcu() code generation (and thus __fget_light())
  file: massage cleanup of files that failed to open

Honestly, there's little value in having a helper with and without that
int __user *ufd argument. It's just messy and doesn't really give us
anything. Just expose receive_fd() with that argument and get rid of
that helper.

Link: https://lore.kernel.org/r/20231130-vfs-files-fixes-v1-5-e73ca6f4ea83@kernel.orgReviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Christian Brauner <brauner@kernel.org>

Not every subsystem needs to have their own specialized helper.
Just us the __receive_fd() helper.

Link: https://lore.kernel.org/r/20231130-vfs-files-fixes-v1-4-e73ca6f4ea83@kernel.orgReviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Christian Brauner <brauner@kernel.org>

The naming is actively misleading since we switched to
SLAB_TYPESAFE_BY_RCU. rcu_head is #define callback_head. Use
callback_head directly and rename f_rcuhead to f_task_work.

Add comments in there to explain what it's used for.

Link: https://lore.kernel.org/r/20231130-vfs-files-fixes-v1-3-e73ca6f4ea83@kernel.orgReviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Christian Brauner <brauner@kernel.org>

Only io_uring uses __close_fd_get_file(). All it does is hide
current->files but io_uring accesses files_struct directly right now
anyway so it's a bit pointless. Just rename pick_file() to
file_close_fd_locked() and let io_uring use it. Add a lockdep assert in
there that we expect the caller to hold file_lock while we're at it.

Link: https://lore.kernel.org/r/20231130-vfs-files-fixes-v1-2-e73ca6f4ea83@kernel.orgReviewed-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Christian Brauner <brauner@kernel.org>

That really shouldn't have "get" in there as that implies we're bumping
the reference count which we don't do at all. We used to but not anmore.
Now we're just closing the fd and pick that file from the fdtable
without bumping the reference count. Update the wrong documentation
while at it.

Link: https://lore.kernel.org/r/20231130-vfs-files-fixes-v1-1-e73ca6f4ea83@kernel.orgReviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Christian Brauner <brauner@kernel.org>

Commit 0ede61d8 ("file: convert to SLAB_TYPESAFE_BY_RCU") caused a
performance regression as reported by the kernel test robot.

The __fget_light() function is one of those critical ones for some
loads, and the code generation was unnecessarily impacted.  Let's just
write that function to better.
Reported-by: kernel test robot <oliver.sang@intel.com>
Cc: Christian Brauner <brauner@kernel.org>
Cc: Jann Horn <jannh@google.com>
Cc: Mateusz Guzik <mjguzik@gmail.com>
Closes: https://lore.kernel.org/oe-lkp/202311201406.2022ca3f-oliver.sang@intel.comSigned-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Link: https://lore.kernel.org/r/CAHk-=wiCJtLbFWNURB34b9a_R_unaH3CiMRXfkR0-iihB_z68A@mail.gmail.comSigned-off-by: Christian Brauner <brauner@kernel.org>

A file that has never gotten FMODE_OPENED will never have RCU-accessed
references, its final fput() is equivalent to file_free() and if it
doesn't have FMODE_BACKING either, it can be done from any context and
won't need task_work treatment.

Now that we have SLAB_TYPESAFE_BY_RCU we can simplify this and have
other callers benefit. All of that can be achieved easier is to make
fput() recoginze that case and call file_free() directly.

No need to introduce a special primitive for that. It also allowed
things like failing dentry_open() could benefit from that as well.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
[Christian Brauner <brauner@kernel.org>: massage commit message]
Link: https://lore.kernel.org/r/20231126020834.GC38156@ZenIVSigned-off-by: Christian Brauner <brauner@kernel.org>

Pull scheduler fix from Borislav Petkov:

 - Make sure tasks are thawed exactly and only once to avoid their state
   getting corrupted

* tag 'sched_urgent_for_v6.7_rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  freezer,sched: Do not restore saved_state of a thawed task

Pull perf event fix from Borislav Petkov:

 - Make sure perf event size validation is done on every event in the
   group

* tag 'perf_urgent_for_v6.7_rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf: Fix perf_event_validate_size()

Pull x86 fixes from Borislav Petkov:

 - Add a forgotten CPU vendor check in the AMD microcode post-loading
   callback so that the callback runs only on AMD

 - Make sure SEV-ES protocol negotiation happens only once and on the
   BSP

* tag 'x86_urgent_for_v6.7_rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/CPU/AMD: Check vendor in the AMD microcode callback
  x86/sev: Fix kernel crash due to late update to read-only ghcb_version

Pull kvm fixes from Paolo Bonzini:
 "Generic:

   - Set .owner for various KVM file_operations so that files refcount
     the KVM module until KVM is done executing _all_ code, including
     the last few instructions of kvm_put_kvm(). And then revert the
     misguided attempt to rely on "struct kvm" refcounts to pin
     KVM-the-module.

  ARM:

   - Do not redo the mapping of vLPIs, if they have already been mapped

  s390:

   - Do not leave bits behind in PTEs

   - Properly catch page invalidations that affect the prefix of a
     nested guest

  x86:

   - When checking if a _running_ vCPU is "in-kernel", i.e. running at
     CPL0, get the CPL directly instead of relying on
     preempted_in_kernel (which is valid if and only if the vCPU was
     preempted, i.e. NOT running).

   - Fix a benign "return void" that was recently introduced.

  Selftests:

   - Makefile tweak for dependency generation

   - '-Wformat' fix"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm:
  KVM: SVM: Update EFER software model on CR0 trap for SEV-ES
  KVM: selftests: add -MP to CFLAGS
  KVM: selftests: Actually print out magic token in NX hugepages skip message
  KVM: x86: Remove 'return void' expression for 'void function'
  Revert "KVM: Prevent module exit until all VMs are freed"
  KVM: Set file_operations.owner appropriately for all such structures
  KVM: x86: Get CPL directly when checking if loaded vCPU is in kernel mode
  KVM: arm64: GICv4: Do not perform a map to a mapped vLPI
  KVM: s390/mm: Properly reset no-dat
  KVM: s390: vsie: fix wrong VIR 37 when MSO is used

Pull powerpc fix from Michael Ellerman:

 - Fix stack teardown in ftrace_no_trace, seen as crashes doing CPU
   hotplug while ftrace is active.

   Thanks to Naveen N Rao.

* tag 'powerpc-6.7-4' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
  powerpc/ftrace: Fix stack teardown in ftrace_no_trace

Pull gpio fix from Bartosz Golaszewski:

 - fix an error path after a failed export in sysfs code

* tag 'gpio-fixes-for-v6.7-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/brgl/linux:
  gpiolib: sysfs: Fix error handling on failed export

Pull USB fixes from Greg KH:
 "Here are some small USB fixes for 6.7-rc5 to resolve some reported
  issues. Included in here are:

   - usb gadget f_hid, and uevent fix

   - xhci driver revert to resolve a much-reported issue

   - typec driver fix

  All of these have been in linux-next with no reported issues"

* tag 'usb-6.7-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
  usb: gadget: f_hid: fix report descriptor allocation
  Revert "xhci: Loosen RPM as default policy to cover for AMD xHC 1.1"
  usb: typec: class: fix typec_altmode_put_partner to put plugs
  USB: gadget: core: adjust uevent timing on gadget unbind

Pull serial driver fixes from Greg KH:
 "Here are some small serial driver fixes for 6.7-rc4 to resolve some
  reported issues. Included in here are:

   - pl011 dma support fix

   - sc16is7xx driver fix

   - ma35d1 console index fix

   - 8250 driver fixes for small issues

  All of these have been in linux-next with no reported issues"

* tag 'tty-6.7-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/tty:
  serial: 8250_dw: Add ACPI ID for Granite Rapids-D UART
  serial: ma35d1: Validate console index before assignment
  ARM: PL011: Fix DMA support
  serial: sc16is7xx: address RX timeout interrupt errata
  serial: 8250: 8250_omap: Clear UART_HAS_RHR_IT_DIS bit
  serial: 8250_omap: Add earlycon support for the AM654 UART controller
  serial: 8250: 8250_omap: Do not start RX DMA on THRI interrupt

Pull char / misc driver fixes from Greg KH:
 "Here are some small fixes for 6.7-rc5 for a variety of small driver
  subsystems. Included in here are:

   - debugfs revert for reported issue

   - greybus revert for reported issue

   - greybus fixup for endian build warning

   - coresight driver fixes

   - nvmem driver fixes

   - devcoredump fix

   - parport new device id

   - ndtest build fix

  All of these have ben in linux-next with no reported issues"

* tag 'char-misc-6.7-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
  nvmem: Do not expect fixed layouts to grab a layout driver
  parport: Add support for Brainboxes IX/UC/PX parallel cards
  Revert "greybus: gb-beagleplay: Ensure le for values in transport"
  greybus: gb-beagleplay: Ensure le for values in transport
  greybus: BeaglePlay driver needs CRC_CCITT
  Revert "debugfs: annotate debugfs handlers vs. removal with lockdep"
  devcoredump: Send uevent once devcd is ready
  ndtest: fix typo class_regster -> class_register
  misc: mei: client.c: fix problem of return '-EOVERFLOW' in mei_cl_write
  misc: mei: client.c: return negative error code in mei_cl_write
  mei: pxp: fix mei_pxp_send_message return value
  coresight: ultrasoc-smb: Fix uninitialized before use buf_hw_base
  coresight: ultrasoc-smb: Config SMB buffer before register sink
  coresight: ultrasoc-smb: Fix sleep while close preempt in enable_smb
  Documentation: coresight: fix `make refcheckdocs` warning
  hwtracing: hisi_ptt: Don't try to attach a task
  hwtracing: hisi_ptt: Handle the interrupt in hardirq context
  hwtracing: hisi_ptt: Add dummy callback pmu::read()
  coresight: Fix crash when Perf and sysfs modes are used concurrently
  coresight: etm4x: Remove bogous __exit annotation for some functions

Merge tag 'loongarch-fixes-6.7-2' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson

Pull LoongArch fixes from Huacai Chen:
 "Preserve syscall nr across execve(), slightly clean up drdtime(), fix
  the Clang built zboot kernel, fix a stack unwinder bug and several bpf
  jit bugs"

* tag 'loongarch-fixes-6.7-2' of git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson:
  LoongArch: BPF: Fix unconditional bswap instructions
  LoongArch: BPF: Fix sign-extension mov instructions
  LoongArch: BPF: Don't sign extend function return value
  LoongArch: BPF: Don't sign extend memory load operand
  LoongArch: Preserve syscall nr across execve()
  LoongArch: Set unwind stack type to unknown rather than set error flag
  LoongArch: Slightly clean up drdtime()
  LoongArch: Apply dynamic relocations for LLD

Pull MIPS fixes from Thomas Bogendoerfer:

 - Fixes for broken Loongson firmware

 - Fix lockdep splat

 - Fix FPU states when creating kernel threads

* tag 'mips-fixes_6.7_1' of git://git.kernel.org/pub/scm/linux/kernel/git/mips/linux:
  MIPS: kernel: Clear FPU states when setting up kernel threads
  MIPS: Loongson64: Handle more memory types passed from firmware
  MIPS: Loongson64: Enable DMA noncoherent support
  MIPS: Loongson64: Reserve vgabios memory on boot
  mips/smp: Call rcutree_report_cpu_starting() earlier

Merge tag 'perf-tools-fixes-for-v6.7-2-2023-12-08' of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools

Pull perf tools fixes from Namhyung Kim:
 "A random set of small bug fixes including:

   - Fix segfault on AmpereOne due to missing default metricgroup name

   - Fix segfault on `perf list --json` due to NULL pointer"

* tag 'perf-tools-fixes-for-v6.7-2-2023-12-08' of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools:
  perf list: Fix JSON segfault by setting the used skip_duplicate_pmus callback
  perf vendor events arm64: AmpereOne: Add missing DefaultMetricgroupName fields
  perf metrics: Avoid segv if default metricgroup isn't set

Pull smb client fixes from Steve French:
 "Six smb3 client fixes:

   - Fixes for copy_file_range and clone (cache invalidation and file
     size), also addresses an xfstest failure

   - Fix to return proper error if REMAP_FILE_DEDUP set (also fixes
     xfstest generic/304)

   - Fix potential null pointer reference with DFS

   - Multichannel fix addressing (reverting an earlier patch) some of
     the problems with enabling/disabling channels dynamically

  Still working on a followon multichannel fix to address another issue
  found in reconnect testing that will send next week"

* tag '6.7-rc4-smb3-client-fixes' of git://git.samba.org/sfrench/cifs-2.6:
  cifs: reconnect worker should take reference on server struct unconditionally
  Revert "cifs: reconnect work should have reference on server struct"
  cifs: Fix non-availability of dedup breaking generic/304
  smb: client: fix potential NULL deref in parse_dfs_referrals()
  cifs: Fix flushing, invalidation and file size with FICLONE
  cifs: Fix flushing, invalidation and file size with copy_file_range()

We can see that "bswap32: Takes an unsigned 32-bit number in either big-
or little-endian format and returns the equivalent number with the same
bit width but opposite endianness" in BPF Instruction Set Specification,
so it should clear the upper 32 bits in "case 32:" for both BPF_ALU and
BPF_ALU64.

[root@linux fedora]# echo 1 > /proc/sys/net/core/bpf_jit_enable
[root@linux fedora]# modprobe test_bpf

Before:
test_bpf: #313 BSWAP 32: 0x0123456789abcdef -> 0xefcdab89 jited:1 ret 1460850314 != -271733879 (0x5712ce8a != 0xefcdab89)FAIL (1 times)
test_bpf: #317 BSWAP 32: 0xfedcba9876543210 -> 0x10325476 jited:1 ret -1460850316 != 271733878 (0xa8ed3174 != 0x10325476)FAIL (1 times)

After:
test_bpf: #313 BSWAP 32: 0x0123456789abcdef -> 0xefcdab89 jited:1 4 PASS
test_bpf: #317 BSWAP 32: 0xfedcba9876543210 -> 0x10325476 jited:1 4 PASS

Fixes: 4ebf9216 ("LoongArch: BPF: Support unconditional bswap instructions")
Acked-by: Hengqi Chen <hengqi.chen@gmail.com>
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>

We can see that "Short form of movsx, dst_reg = (s8,s16,s32)src_reg" in
include/linux/filter.h, additionally, for BPF_ALU64 the value of the
destination register is unchanged whereas for BPF_ALU the upper 32 bits
of the destination register are zeroed, so it should clear the upper 32
bits for BPF_ALU.

[root@linux fedora]# echo 1 > /proc/sys/net/core/bpf_jit_enable
[root@linux fedora]# modprobe test_bpf

Before:
test_bpf: #81 ALU_MOVSX | BPF_B jited:1 ret 2 != 1 (0x2 != 0x1)FAIL (1 times)
test_bpf: #82 ALU_MOVSX | BPF_H jited:1 ret 2 != 1 (0x2 != 0x1)FAIL (1 times)

After:
test_bpf: #81 ALU_MOVSX | BPF_B jited:1 6 PASS
test_bpf: #82 ALU_MOVSX | BPF_H jited:1 6 PASS

By the way, the bpf selftest case "./test_progs -t verifier_movsx" can
also be fixed with this patch.

Fixes: f48012f1 ("LoongArch: BPF: Support sign-extension mov instructions")
Acked-by: Hengqi Chen <hengqi.chen@gmail.com>
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>

The `cls_redirect` test triggers a kernel panic like:

  # ./test_progs -t cls_redirect
  Can't find bpf_testmod.ko kernel module: -2
  WARNING! Selftests relying on bpf_testmod.ko will be skipped.
  [   30.938489] CPU 3 Unable to handle kernel paging request at virtual address fffffffffd814de0, era == ffff800002009fb8, ra == ffff800002009f9c
  [   30.939331] Oops[#1]:
  [   30.939513] CPU: 3 PID: 1260 Comm: test_progs Not tainted 6.7.0-rc2-loong-devel-g2f56bb0d2327 #35 a896aca3f4164f09cc346f89f2e09832e07be5f6
  [   30.939732] Hardware name: QEMU QEMU Virtual Machine, BIOS unknown 2/2/2022
  [   30.939901] pc ffff800002009fb8 ra ffff800002009f9c tp 9000000104da4000 sp 9000000104da7ab0
  [   30.940038] a0 fffffffffd814de0 a1 9000000104da7a68 a2 0000000000000000 a3 9000000104da7c10
  [   30.940183] a4 9000000104da7c14 a5 0000000000000002 a6 0000000000000021 a7 00005555904d7f90
  [   30.940321] t0 0000000000000110 t1 0000000000000000 t2 fffffffffd814de0 t3 0004c4b400000000
  [   30.940456] t4 ffffffffffffffff t5 00000000c3f63600 t6 0000000000000000 t7 0000000000000000
  [   30.940590] t8 000000000006d803 u0 0000000000000020 s9 9000000104da7b10 s0 900000010504c200
  [   30.940727] s1 fffffffffd814de0 s2 900000010504c200 s3 9000000104da7c10 s4 9000000104da7ad0
  [   30.940866] s5 0000000000000000 s6 90000000030e65bc s7 9000000104da7b44 s8 90000000044f6fc0
  [   30.941015]    ra: ffff800002009f9c bpf_prog_846803e5ae81417f_cls_redirect+0xa0/0x590
  [   30.941535]   ERA: ffff800002009fb8 bpf_prog_846803e5ae81417f_cls_redirect+0xbc/0x590
  [   30.941696]  CRMD: 000000b0 (PLV0 -IE -DA +PG DACF=CC DACM=CC -WE)
  [   30.942224]  PRMD: 00000004 (PPLV0 +PIE -PWE)
  [   30.942330]  EUEN: 00000003 (+FPE +SXE -ASXE -BTE)
  [   30.942453]  ECFG: 00071c1c (LIE=2-4,10-12 VS=7)
  [   30.942612] ESTAT: 00010000 [PIL] (IS= ECode=1 EsubCode=0)
  [   30.942764]  BADV: fffffffffd814de0
  [   30.942854]  PRID: 0014c010 (Loongson-64bit, Loongson-3A5000)
  [   30.942974] Modules linked in:
  [   30.943078] Process test_progs (pid: 1260, threadinfo=00000000ce303226, task=000000007d10bb76)
  [   30.943306] Stack : 900000010a064000 90000000044f6fc0 9000000104da7b48 0000000000000000
  [   30.943495]         0000000000000000 9000000104da7c14 9000000104da7c10 900000010504c200
  [   30.943626]         0000000000000001 ffff80001b88c000 9000000104da7b70 90000000030e6668
  [   30.943785]         0000000000000000 9000000104da7b58 ffff80001b88c048 9000000003d05000
  [   30.943936]         900000000303ac88 0000000000000000 0000000000000000 9000000104da7b70
  [   30.944091]         0000000000000000 0000000000000001 0000000731eeab00 0000000000000000
  [   30.944245]         ffff80001b88c000 0000000000000000 0000000000000000 54b99959429f83b8
  [   30.944402]         ffff80001b88c000 90000000044f6fc0 9000000101d70000 ffff80001b88c000
  [   30.944538]         000000000000005a 900000010504c200 900000010a064000 900000010a067000
  [   30.944697]         9000000104da7d88 0000000000000000 9000000003d05000 90000000030e794c
  [   30.944852]         ...
  [   30.944924] Call Trace:
  [   30.945120] [<ffff800002009fb8>] bpf_prog_846803e5ae81417f_cls_redirect+0xbc/0x590
  [   30.945650] [<90000000030e6668>] bpf_test_run+0x1ec/0x2f8
  [   30.945958] [<90000000030e794c>] bpf_prog_test_run_skb+0x31c/0x684
  [   30.946065] [<90000000026d4f68>] __sys_bpf+0x678/0x2724
  [   30.946159] [<90000000026d7288>] sys_bpf+0x20/0x2c
  [   30.946253] [<90000000032dd224>] do_syscall+0x7c/0x94
  [   30.946343] [<9000000002541c5c>] handle_syscall+0xbc/0x158
  [   30.946492]
  [   30.946549] Code: 0015030e  5c0009c0  5001d000 <28c00304> 02c00484  29c00304  00150009  2a42d2e4  0280200d
  [   30.946793]
  [   30.946971] ---[ end trace 0000000000000000 ]---
  [   32.093225] Kernel panic - not syncing: Fatal exception in interrupt
  [   32.093526] Kernel relocated by 0x2320000
  [   32.093630]  .text @ 0x9000000002520000
  [   32.093725]  .data @ 0x9000000003400000
  [   32.093792]  .bss  @ 0x9000000004413200
  [   34.971998] ---[ end Kernel panic - not syncing: Fatal exception in interrupt ]---

This is because we signed-extend function return values. When subprog
mode is enabled, we have:

  cls_redirect()
    -> get_global_metrics() returns pcpu ptr 0xfffffefffc00b480

The pointer returned is later signed-extended to 0xfffffffffc00b480 at
`BPF_JMP | BPF_EXIT`. During BPF prog run, this triggers unhandled page
fault and a kernel panic.

Drop the unnecessary signed-extension on return values like other
architectures do.

With this change, we have:

  # ./test_progs -t cls_redirect
  Can't find bpf_testmod.ko kernel module: -2
  WARNING! Selftests relying on bpf_testmod.ko will be skipped.
  #51/1    cls_redirect/cls_redirect_inlined:OK
  #51/2    cls_redirect/IPv4 TCP accept unknown (no hops, flags: SYN):OK
  #51/3    cls_redirect/IPv6 TCP accept unknown (no hops, flags: SYN):OK
  #51/4    cls_redirect/IPv4 TCP accept unknown (no hops, flags: ACK):OK
  #51/5    cls_redirect/IPv6 TCP accept unknown (no hops, flags: ACK):OK
  #51/6    cls_redirect/IPv4 TCP forward unknown (one hop, flags: ACK):OK
  #51/7    cls_redirect/IPv6 TCP forward unknown (one hop, flags: ACK):OK
  #51/8    cls_redirect/IPv4 TCP accept known (one hop, flags: ACK):OK
  #51/9    cls_redirect/IPv6 TCP accept known (one hop, flags: ACK):OK
  #51/10   cls_redirect/IPv4 UDP accept unknown (no hops, flags: none):OK
  #51/11   cls_redirect/IPv6 UDP accept unknown (no hops, flags: none):OK
  #51/12   cls_redirect/IPv4 UDP forward unknown (one hop, flags: none):OK
  #51/13   cls_redirect/IPv6 UDP forward unknown (one hop, flags: none):OK
  #51/14   cls_redirect/IPv4 UDP accept known (one hop, flags: none):OK
  #51/15   cls_redirect/IPv6 UDP accept known (one hop, flags: none):OK
  #51/16   cls_redirect/cls_redirect_subprogs:OK
  #51/17   cls_redirect/IPv4 TCP accept unknown (no hops, flags: SYN):OK
  #51/18   cls_redirect/IPv6 TCP accept unknown (no hops, flags: SYN):OK
  #51/19   cls_redirect/IPv4 TCP accept unknown (no hops, flags: ACK):OK
  #51/20   cls_redirect/IPv6 TCP accept unknown (no hops, flags: ACK):OK
  #51/21   cls_redirect/IPv4 TCP forward unknown (one hop, flags: ACK):OK
  #51/22   cls_redirect/IPv6 TCP forward unknown (one hop, flags: ACK):OK
  #51/23   cls_redirect/IPv4 TCP accept known (one hop, flags: ACK):OK
  #51/24   cls_redirect/IPv6 TCP accept known (one hop, flags: ACK):OK
  #51/25   cls_redirect/IPv4 UDP accept unknown (no hops, flags: none):OK
  #51/26   cls_redirect/IPv6 UDP accept unknown (no hops, flags: none):OK
  #51/27   cls_redirect/IPv4 UDP forward unknown (one hop, flags: none):OK
  #51/28   cls_redirect/IPv6 UDP forward unknown (one hop, flags: none):OK
  #51/29   cls_redirect/IPv4 UDP accept known (one hop, flags: none):OK
  #51/30   cls_redirect/IPv6 UDP accept known (one hop, flags: none):OK
  #51/31   cls_redirect/cls_redirect_dynptr:OK
  #51/32   cls_redirect/IPv4 TCP accept unknown (no hops, flags: SYN):OK
  #51/33   cls_redirect/IPv6 TCP accept unknown (no hops, flags: SYN):OK
  #51/34   cls_redirect/IPv4 TCP accept unknown (no hops, flags: ACK):OK
  #51/35   cls_redirect/IPv6 TCP accept unknown (no hops, flags: ACK):OK
  #51/36   cls_redirect/IPv4 TCP forward unknown (one hop, flags: ACK):OK
  #51/37   cls_redirect/IPv6 TCP forward unknown (one hop, flags: ACK):OK
  #51/38   cls_redirect/IPv4 TCP accept known (one hop, flags: ACK):OK
  #51/39   cls_redirect/IPv6 TCP accept known (one hop, flags: ACK):OK
  #51/40   cls_redirect/IPv4 UDP accept unknown (no hops, flags: none):OK
  #51/41   cls_redirect/IPv6 UDP accept unknown (no hops, flags: none):OK
  #51/42   cls_redirect/IPv4 UDP forward unknown (one hop, flags: none):OK
  #51/43   cls_redirect/IPv6 UDP forward unknown (one hop, flags: none):OK
  #51/44   cls_redirect/IPv4 UDP accept known (one hop, flags: none):OK
  #51/45   cls_redirect/IPv6 UDP accept known (one hop, flags: none):OK
  #51      cls_redirect:OK
  Summary: 1/45 PASSED, 0 SKIPPED, 0 FAILED

Fixes: 5dc61552 ("LoongArch: Add BPF JIT support")
Signed-off-by: Hengqi Chen <hengqi.chen@gmail.com>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>

The `cgrp_local_storage` test triggers a kernel panic like:

  # ./test_progs -t cgrp_local_storage
  Can't find bpf_testmod.ko kernel module: -2
  WARNING! Selftests relying on bpf_testmod.ko will be skipped.
  [  550.930632] CPU 1 Unable to handle kernel paging request at virtual address 0000000000000080, era == ffff80000200be34, ra == ffff80000200be00
  [  550.931781] Oops[#1]:
  [  550.931966] CPU: 1 PID: 1303 Comm: test_progs Not tainted 6.7.0-rc2-loong-devel-g2f56bb0d2327 #35 a896aca3f4164f09cc346f89f2e09832e07be5f6
  [  550.932215] Hardware name: QEMU QEMU Virtual Machine, BIOS unknown 2/2/2022
  [  550.932403] pc ffff80000200be34 ra ffff80000200be00 tp 9000000108350000 sp 9000000108353dc0
  [  550.932545] a0 0000000000000000 a1 0000000000000517 a2 0000000000000118 a3 00007ffffbb15558
  [  550.932682] a4 00007ffffbb15620 a5 90000001004e7700 a6 0000000000000021 a7 0000000000000118
  [  550.932824] t0 ffff80000200bdc0 t1 0000000000000517 t2 0000000000000517 t3 00007ffff1c06ee0
  [  550.932961] t4 0000555578ae04d0 t5 fffffffffffffff8 t6 0000000000000004 t7 0000000000000020
  [  550.933097] t8 0000000000000040 u0 00000000000007b8 s9 9000000108353e00 s0 90000001004e7700
  [  550.933241] s1 9000000004005000 s2 0000000000000001 s3 0000000000000000 s4 0000555555eb2ec8
  [  550.933379] s5 00007ffffbb15bb8 s6 00007ffff1dafd60 s7 000055555663f610 s8 00007ffff1db0050
  [  550.933520]    ra: ffff80000200be00 bpf_prog_98f1b9e767be2a84_on_enter+0x40/0x200
  [  550.933911]   ERA: ffff80000200be34 bpf_prog_98f1b9e767be2a84_on_enter+0x74/0x200
  [  550.934105]  CRMD: 000000b0 (PLV0 -IE -DA +PG DACF=CC DACM=CC -WE)
  [  550.934596]  PRMD: 00000004 (PPLV0 +PIE -PWE)
  [  550.934712]  EUEN: 00000003 (+FPE +SXE -ASXE -BTE)
  [  550.934836]  ECFG: 00071c1c (LIE=2-4,10-12 VS=7)
  [  550.934976] ESTAT: 00010000 [PIL] (IS= ECode=1 EsubCode=0)
  [  550.935097]  BADV: 0000000000000080
  [  550.935181]  PRID: 0014c010 (Loongson-64bit, Loongson-3A5000)
  [  550.935291] Modules linked in:
  [  550.935391] Process test_progs (pid: 1303, threadinfo=000000006c3b1c41, task=0000000061f84a55)
  [  550.935643] Stack : 00007ffffbb15bb8 0000555555eb2ec8 0000000000000000 0000000000000001
  [  550.935844]         9000000004005000 ffff80001b864000 00007ffffbb15450 90000000029aa034
  [  550.935990]         0000000000000000 9000000108353ec0 0000000000000118 d07d9dfb09721a09
  [  550.936175]         0000000000000001 0000000000000000 9000000108353ec0 0000000000000118
  [  550.936314]         9000000101d46ad0 900000000290abf0 000055555663f610 0000000000000000
  [  550.936479]         0000000000000003 9000000108353ec0 00007ffffbb15450 90000000029d7288
  [  550.936635]         00007ffff1dafd60 000055555663f610 0000000000000000 0000000000000003
  [  550.936779]         9000000108353ec0 90000000035dd1f0 00007ffff1dafd58 9000000002841c5c
  [  550.936939]         0000000000000119 0000555555eea5a8 00007ffff1d78780 00007ffffbb153e0
  [  550.937083]         ffffffffffffffda 00007ffffbb15518 0000000000000040 00007ffffbb15558
  [  550.937224]         ...
  [  550.937299] Call Trace:
  [  550.937521] [<ffff80000200be34>] bpf_prog_98f1b9e767be2a84_on_enter+0x74/0x200
  [  550.937910] [<90000000029aa034>] bpf_trace_run2+0x90/0x154
  [  550.938105] [<900000000290abf0>] syscall_trace_enter.isra.0+0x1cc/0x200
  [  550.938224] [<90000000035dd1f0>] do_syscall+0x48/0x94
  [  550.938319] [<9000000002841c5c>] handle_syscall+0xbc/0x158
  [  550.938477]
  [  550.938607] Code: 580009ae  50016000  262402e4 <28c20085> 14092084  03a00084  16000024  03240084  00150006
  [  550.938851]
  [  550.939021] ---[ end trace 0000000000000000 ]---

Further investigation shows that this panic is triggered by memory
load operations:

  ptr = bpf_cgrp_storage_get(&map_a, task->cgroups->dfl_cgrp, 0,
                             BPF_LOCAL_STORAGE_GET_F_CREATE);

The expression `task->cgroups->dfl_cgrp` involves two memory load.
Since the field offset fits in imm12 or imm14, we use ldd or ldptrd
instructions. But both instructions have the side effect that it will
signed-extended the imm operand. Finally, we got the wrong addresses
and panics is inevitable.

Use a generic ldxd instruction to avoid this kind of issues.

With this change, we have:

  # ./test_progs -t cgrp_local_storage
  Can't find bpf_testmod.ko kernel module: -2
  WARNING! Selftests relying on bpf_testmod.ko will be skipped.
  test_cgrp_local_storage:PASS:join_cgroup /cgrp_local_storage 0 nsec
  #48/1    cgrp_local_storage/tp_btf:OK
  test_attach_cgroup:PASS:skel_open 0 nsec
  test_attach_cgroup:PASS:prog_attach 0 nsec
  test_attach_cgroup:PASS:prog_attach 0 nsec
  libbpf: prog 'update_cookie_tracing': failed to attach: ERROR: strerror_r(-524)=22
  test_attach_cgroup:FAIL:prog_attach unexpected error: -524
  #48/2    cgrp_local_storage/attach_cgroup:FAIL
  test_recursion:PASS:skel_open_and_load 0 nsec
  libbpf: prog 'on_lookup': failed to attach: ERROR: strerror_r(-524)=22
  libbpf: prog 'on_lookup': failed to auto-attach: -524
  test_recursion:FAIL:skel_attach unexpected error: -524 (errno 524)
  #48/3    cgrp_local_storage/recursion:FAIL
  #48/4    cgrp_local_storage/negative:OK
  #48/5    cgrp_local_storage/cgroup_iter_sleepable:OK
  test_yes_rcu_lock:PASS:skel_open 0 nsec
  test_yes_rcu_lock:PASS:skel_load 0 nsec
  libbpf: prog 'yes_rcu_lock': failed to attach: ERROR: strerror_r(-524)=22
  libbpf: prog 'yes_rcu_lock': failed to auto-attach: -524
  test_yes_rcu_lock:FAIL:skel_attach unexpected error: -524 (errno 524)
  #48/6    cgrp_local_storage/yes_rcu_lock:FAIL
  #48/7    cgrp_local_storage/no_rcu_lock:OK
  #48      cgrp_local_storage:FAIL

  All error logs:
  test_cgrp_local_storage:PASS:join_cgroup /cgrp_local_storage 0 nsec
  test_attach_cgroup:PASS:skel_open 0 nsec
  test_attach_cgroup:PASS:prog_attach 0 nsec
  test_attach_cgroup:PASS:prog_attach 0 nsec
  libbpf: prog 'update_cookie_tracing': failed to attach: ERROR: strerror_r(-524)=22
  test_attach_cgroup:FAIL:prog_attach unexpected error: -524
  #48/2    cgrp_local_storage/attach_cgroup:FAIL
  test_recursion:PASS:skel_open_and_load 0 nsec
  libbpf: prog 'on_lookup': failed to attach: ERROR: strerror_r(-524)=22
  libbpf: prog 'on_lookup': failed to auto-attach: -524
  test_recursion:FAIL:skel_attach unexpected error: -524 (errno 524)
  #48/3    cgrp_local_storage/recursion:FAIL
  test_yes_rcu_lock:PASS:skel_open 0 nsec
  test_yes_rcu_lock:PASS:skel_load 0 nsec
  libbpf: prog 'yes_rcu_lock': failed to attach: ERROR: strerror_r(-524)=22
  libbpf: prog 'yes_rcu_lock': failed to auto-attach: -524
  test_yes_rcu_lock:FAIL:skel_attach unexpected error: -524 (errno 524)
  #48/6    cgrp_local_storage/yes_rcu_lock:FAIL
  #48      cgrp_local_storage:FAIL
  Summary: 0/4 PASSED, 0 SKIPPED, 1 FAILED

No panics any more (The test still failed because lack of BPF trampoline
which I am actively working on).

Fixes: 5dc61552 ("LoongArch: Add BPF JIT support")
Signed-off-by: Hengqi Chen <hengqi.chen@gmail.com>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>

Currently, we store syscall nr in pt_regs::regs[11] and syscall execve()
accidentally overrides it during its execution:

    sys_execve()
      -> do_execve()
        -> do_execveat_common()
          -> bprm_execve()
            -> exec_binprm()
              -> search_binary_handler()
                -> load_elf_binary()
                  -> ELF_PLAT_INIT()

ELF_PLAT_INIT() reset regs[11] to 0, so in syscall_exit_to_user_mode()
we later get a wrong syscall nr. This breaks tools like execsnoop since
it relies on execve() tracepoints.

Skip pt_regs::regs[11] reset in ELF_PLAT_INIT() to fix the issue.
Signed-off-by: Hengqi Chen <hengqi.chen@gmail.com>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>

During unwinding, unwind_done() is used as an end condition. Normally it
unwind to the user stack and then set the stack type to unknown, which
is a normal exit. When something unexpected happens in unwind process
and we cannot unwind anymore, we should set the error flag, and also set
the stack type to unknown to indicate that the unwind process can not
continue. The error flag emphasizes that the unwind process produce an
unexpected error. There is no unexpected things when we unwind the PT_REGS
in the top of IRQ stack and find out that is an user mode PT_REGS. Thus,
we should not set error flag and just set stack type to unknown.
Reported-by: Hengqi Chen <hengqi.chen@gmail.com>
Acked-by: Hengqi Chen <hengqi.chen@gmail.com>
Signed-off-by: Jinyang He <hejinyang@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>

As we are just discarding the stable clock ID, simply write it into
$zero instead of allocating a temporary register.
Signed-off-by: Xi Ruoyao <xry111@xry111.site>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>

For the following assembly code:

     .text
     .global func
 func:
     nop

     .data
 var:
     .dword func

When linked with `-pie`, GNU LD populates the `var` variable with the
pre-relocated value of `func`. However, LLVM LLD does not exhibit the
same behavior. This issue also arises with the `kernel_entry` in arch/
loongarch/kernel/head.S:

 _head:
     .word   MZ_MAGIC                /* "MZ", MS-DOS header */
     .org    0x8
     .dword  kernel_entry            /* Kernel entry point */

The correct kernel entry from the MS-DOS header is crucial for jumping
to vmlinux from zboot. This necessity is why the compressed relocatable
kernel compiled by Clang encounters difficulties in booting.

To address this problem, it is proposed to apply dynamic relocations to
place with `--apply-dynamic-relocs`.

Link: https://github.com/ClangBuiltLinux/linux/issues/1962Signed-off-by: WANG Rui <wangrui@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>

Pull SCSI fix from James Bottomley:
 "One tiny fix to the be2iscsi driver fixing a memory leak in an error
  leg"

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
  scsi: be2iscsi: Fix a memleak in beiscsi_init_wrb_handle()

Pull block fixes from Jens Axboe:
 "Nothing major in here, just miscellanous fixes for MD and NVMe:

   - NVMe pull request via Keith:
      - Proper nvme ctrl state setting (Keith)
      - Passthrough command optimization (Keith)
      - Spectre fix (Nitesh)
      - Kconfig clarifications (Shin'ichiro)
      - Frozen state deadlock fix (Bitao)
      - Power setting quirk (Georg)

   - MD pull requests via Song:
      - 6.7 regresisons with recovery/sync (Yu)
      - Reshape fix (David)"

* tag 'block-6.7-2023-12-08' of git://git.kernel.dk/linux:
  md: split MD_RECOVERY_NEEDED out of mddev_resume
  nvme-pci: Add sleep quirk for Kingston drives
  md: fix stopping sync thread
  md: don't leave 'MD_RECOVERY_FROZEN' in error path of md_set_readonly()
  md: fix missing flush of sync_work
  nvme: fix deadlock between reset and scan
  nvme: prevent potential spectre v1 gadget
  nvme: improve NVME_HOST_AUTH and NVME_TARGET_AUTH config descriptions
  nvme-ioctl: move capable() admin check to the end
  nvme: ensure reset state check ordering
  nvme: introduce helper function to get ctrl state
  md/raid6: use valid sector values to determine if an I/O should wait on the reshape

Pull io_uring fixes from Jens Axboe:
 "Two minor fixes for issues introduced in this release cycle, and two
  fixes for issues or potential issues that are heading to stable.

  One of these ends up disabling passing io_uring file descriptors via
  SCM_RIGHTS. There really shouldn't be an overlap between that kind of
  historic use case and modern usage of io_uring, which is why this was
  deemed appropriate"

* tag 'io_uring-6.7-2023-12-08' of git://git.kernel.dk/linux:
  io_uring/af_unix: disable sending io_uring over sockets
  io_uring/kbuf: check for buffer list readiness after NULL check
  io_uring/kbuf: Fix an NULL vs IS_ERR() bug in io_alloc_pbuf_ring()
  io_uring: fix mutex_unlock with unreferenced ctx

Pull rdma fixes from Jason Gunthorpe:
 "Primarily rtrs and irdma fixes:

   - Fix uninitialized value in ib_get_eth_speed()

   - Fix hns refusing to work if userspace doesn't select the correct
     congestion control algorithm

   - Several irdma fixes - unreliable Send Queue Drain, use after free,
     64k page size bugs, device removal races

   - Several rtrs bug fixes - crashes, memory leaks, use after free, bad
     credit accounting, bogus WARN_ON

   - Typos and a MAINTAINER update"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
  RDMA/irdma: Avoid free the non-cqp_request scratch
  RDMA/irdma: Fix support for 64k pages
  RDMA/irdma: Ensure iWarp QP queue memory is OS paged aligned
  RDMA/core: Fix umem iterator when PAGE_SIZE is greater then HCA pgsz
  RDMA/irdma: Fix UAF in irdma_sc_ccq_get_cqe_info()
  RDMA/bnxt_re: Correct module description string
  RDMA/rtrs-clt: Remove the warnings for req in_use check
  RDMA/rtrs-clt: Fix the max_send_wr setting
  RDMA/rtrs-srv: Destroy path files after making sure no IOs in-flight
  RDMA/rtrs-srv: Free srv_mr iu only when always_invalidate is true
  RDMA/rtrs-srv: Check return values while processing info request
  RDMA/rtrs-clt: Start hb after path_up
  RDMA/rtrs-srv: Do not unconditionally enable irq
  MAINTAINERS: Add Chengchang Tang as Hisilicon RoCE maintainer
  RDMA/irdma: Add wait for suspend on SQD
  RDMA/irdma: Do not modify to SQD on error
  RDMA/hns: Fix unnecessary err return when using invalid congest control algorithm
  RDMA/core: Fix uninit-value access in ib_get_eth_speed()