linux.git/fs/nfs/flexfilelayout, branch v6.12

pnfs/flexfiles: enable localio support

2024-09-23T19:03:30+00:00

If the DS is local to this client use localio to write the data.

Signed-off-by: Trond Myklebust 
Signed-off-by: Mike Snitzer 
Reviewed-by: NeilBrown 
Reviewed-by: Jeff Layton 
Signed-off-by: Anna Schumaker

nfs: pass struct nfsd_file to nfs_init_pgio and nfs_init_commit

2024-09-23T19:03:30+00:00

The nfsd_file will be passed, in future commits, by callers
that enable LOCALIO support (for both regular NFS and pNFS IO).

[Derived from patch authored by Weston Andros Adamson, but switched
 from passing struct file to struct nfsd_file]

Signed-off-by: Mike Snitzer 
Reviewed-by: NeilBrown 
Reviewed-by: Jeff Layton 
Signed-off-by: Anna Schumaker

nfs: factor out {encode,decode}_opaque_fixed to nfs_xdr.h

2024-09-23T19:03:29+00:00

Eliminates duplicate functions in various files to allow for
additional callers.

Signed-off-by: Mike Snitzer 
Reviewed-by: NeilBrown 
Reviewed-by: Jeff Layton 
Signed-off-by: Anna Schumaker

NFSv4/pNFS: Do layout state recovery upon reboot

2024-07-08T17:47:26+00:00

Some pNFS implementations, such as flexible files, want the client to
send the layout stats and layout errors that may have incurred while the
metadata server was booting. To do so, the client sends a layoutreturn
with an all-zero stateid while the server is in grace during reboot
recovery.

Signed-off-by: Trond Myklebust 
Reviewed-by: Jeff Layton 
Signed-off-by: Anna Schumaker

pNFS: rework pnfs_generic_pg_check_layout to check IO range

2024-05-21T12:34:15+00:00

All callers of pnfs_generic_pg_check_layout() also want to do a call to
check that the layout's range covers the IO range. Merge the functionality
of the pnfs_generic_pg_check_range() into that of
pnfs_generic_pg_check_layout().

Signed-off-by: Olga Kornievskaia 
Reviewed-by: Benjamin Coddington 
Signed-off-by: Trond Myklebust

nfs: fix panic when nfs4_ff_layout_prepare_ds() fails

2024-03-16T16:41:45+00:00

We've been seeing the following panic in production

BUG: kernel NULL pointer dereference, address: 0000000000000065
PGD 2f485f067 P4D 2f485f067 PUD 2cc5d8067 PMD 0
RIP: 0010:ff_layout_cancel_io+0x3a/0x90 [nfs_layout_flexfiles]
Call Trace:
 
 ? __die+0x78/0xc0
 ? page_fault_oops+0x286/0x380
 ? __rpc_execute+0x2c3/0x470 [sunrpc]
 ? rpc_new_task+0x42/0x1c0 [sunrpc]
 ? exc_page_fault+0x5d/0x110
 ? asm_exc_page_fault+0x22/0x30
 ? ff_layout_free_layoutreturn+0x110/0x110 [nfs_layout_flexfiles]
 ? ff_layout_cancel_io+0x3a/0x90 [nfs_layout_flexfiles]
 ? ff_layout_cancel_io+0x6f/0x90 [nfs_layout_flexfiles]
 pnfs_mark_matching_lsegs_return+0x1b0/0x360 [nfsv4]
 pnfs_error_mark_layout_for_return+0x9e/0x110 [nfsv4]
 ? ff_layout_send_layouterror+0x50/0x160 [nfs_layout_flexfiles]
 nfs4_ff_layout_prepare_ds+0x11f/0x290 [nfs_layout_flexfiles]
 ff_layout_pg_init_write+0xf0/0x1f0 [nfs_layout_flexfiles]
 __nfs_pageio_add_request+0x154/0x6c0 [nfs]
 nfs_pageio_add_request+0x26b/0x380 [nfs]
 nfs_do_writepage+0x111/0x1e0 [nfs]
 nfs_writepages_callback+0xf/0x30 [nfs]
 write_cache_pages+0x17f/0x380
 ? nfs_pageio_init_write+0x50/0x50 [nfs]
 ? nfs_writepages+0x6d/0x210 [nfs]
 ? nfs_writepages+0x6d/0x210 [nfs]
 nfs_writepages+0x125/0x210 [nfs]
 do_writepages+0x67/0x220
 ? generic_perform_write+0x14b/0x210
 filemap_fdatawrite_wbc+0x5b/0x80
 file_write_and_wait_range+0x6d/0xc0
 nfs_file_fsync+0x81/0x170 [nfs]
 ? nfs_file_mmap+0x60/0x60 [nfs]
 __x64_sys_fsync+0x53/0x90
 do_syscall_64+0x3d/0x90
 entry_SYSCALL_64_after_hwframe+0x46/0xb0

Inspecting the core with drgn I was able to pull this

  >>> prog.crashed_thread().stack_trace()[0]
  #0 at 0xffffffffa079657a (ff_layout_cancel_io+0x3a/0x84) in ff_layout_cancel_io at fs/nfs/flexfilelayout/flexfilelayout.c:2021:27
  >>> prog.crashed_thread().stack_trace()[0]['idx']
  (u32)1
  >>> prog.crashed_thread().stack_trace()[0]['flseg'].mirror_array[1].mirror_ds
  (struct nfs4_ff_layout_ds *)0xffffffffffffffed

This is clear from the stack trace, we call nfs4_ff_layout_prepare_ds()
which could error out initializing the mirror_ds, and then we go to
clean it all up and our check is only for if (!mirror->mirror_ds).  This
is inconsistent with the rest of the users of mirror_ds, which have

  if (IS_ERR_OR_NULL(mirror_ds))

to keep from tripping over this exact scenario.  Fix this up in
ff_layout_cancel_io() to make sure we don't panic when we get an error.
I also spot checked all the other instances of checking mirror_ds and we
appear to be doing the correct checks everywhere, only unconditionally
dereferencing mirror_ds when we know it would be valid.

Signed-off-by: Josef Bacik 
Fixes: b739a5bd9d9f ("NFSv4/flexfiles: Cancel I/O if the layout is recalled or revoked")
Reviewed-by: Jeff Layton 
Signed-off-by: Trond Myklebust

Merge tag 'hardening-v6.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux

2023-10-31T05:09:55+00:00

Pull hardening updates from Kees Cook:
 "One of the more voluminous set of changes is for adding the new
  __counted_by annotation[1] to gain run-time bounds checking of
  dynamically sized arrays with UBSan.

   - Add LKDTM test for stuck CPUs (Mark Rutland)

   - Improve LKDTM selftest behavior under UBSan (Ricardo Cañuelo)

   - Refactor more 1-element arrays into flexible arrays (Gustavo A. R.
     Silva)

   - Analyze and replace strlcpy and strncpy uses (Justin Stitt, Azeem
     Shaikh)

   - Convert group_info.usage to refcount_t (Elena Reshetova)

   - Add __counted_by annotations (Kees Cook, Gustavo A. R. Silva)

   - Add Kconfig fragment for basic hardening options (Kees Cook, Lukas
     Bulwahn)

   - Fix randstruct GCC plugin performance mode to stay in groups (Kees
     Cook)

   - Fix strtomem() compile-time check for small sources (Kees Cook)"

* tag 'hardening-v6.7-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/kees/linux: (56 commits)
  hwmon: (acpi_power_meter) replace open-coded kmemdup_nul
  reset: Annotate struct reset_control_array with __counted_by
  kexec: Annotate struct crash_mem with __counted_by
  virtio_console: Annotate struct port_buffer with __counted_by
  ima: Add __counted_by for struct modsig and use struct_size()
  MAINTAINERS: Include stackleak paths in hardening entry
  string: Adjust strtomem() logic to allow for smaller sources
  hardening: x86: drop reference to removed config AMD_IOMMU_V2
  randstruct: Fix gcc-plugin performance mode to stay in group
  mailbox: zynqmp: Annotate struct zynqmp_ipi_pdata with __counted_by
  drivers: thermal: tsens: Annotate struct tsens_priv with __counted_by
  irqchip/imx-intmux: Annotate struct intmux_data with __counted_by
  KVM: Annotate struct kvm_irq_routing_table with __counted_by
  virt: acrn: Annotate struct vm_memory_region_batch with __counted_by
  hwmon: Annotate struct gsc_hwmon_platform_data with __counted_by
  sparc: Annotate struct cpuinfo_tree with __counted_by
  isdn: kcapi: replace deprecated strncpy with strscpy_pad
  isdn: replace deprecated strncpy with strscpy
  NFS/flexfiles: Annotate struct nfs4_ff_layout_segment with __counted_by
  nfs41: Annotate struct nfs4_file_layout_dsaddr with __counted_by
  ...

pNFS/flexfiles: Check the layout validity in ff_layout_mirror_prepare_stats

2023-10-18T18:42:09+00:00

Ensure that we check the layout pointer and validity after dereferencing
it in ff_layout_mirror_prepare_stats.

Fixes: 08e2e5bc6c9a ("pNFS/flexfiles: Clean up layoutstats")
Signed-off-by: Trond Myklebust 
Signed-off-by: Anna Schumaker

NFS/flexfiles: Annotate struct nfs4_ff_layout_segment with __counted_by

2023-10-02T16:48:53+00:00

Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time checking via CONFIG_UBSAN_BOUNDS
(for array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).

As found with Coccinelle[1], add __counted_by for struct nfs4_ff_layout_segment.

[1] https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci

Cc: Trond Myklebust 
Cc: Anna Schumaker 
Cc: linux-nfs@vger.kernel.org
Reviewed-by: "Gustavo A. R. Silva" 
Link: https://lore.kernel.org/r/20230915201434.never.346-kees@kernel.org
Signed-off-by: Kees Cook

NFS/pNFS: Report EINVAL errors from connect() to the server

2023-09-13T15:51:11+00:00

With IPv6, connect() can occasionally return EINVAL if a route is
unavailable. If this happens during I/O to a data server, we want to
report it using LAYOUTERROR as an inability to connect.

Fixes: dd52128afdde ("NFSv4.1/pnfs Ensure flexfiles reports all connection related errors")
Signed-off-by: Trond Myklebust 
Signed-off-by: Anna Schumaker