linux.git/lib, branch v6.2-rc5

Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma

2023-01-20T22:15:51+00:00

Pull rdma fixes from Jason Gunthorpe:

 - Several hfi1 patches fixing some long standing driver bugs

 - Overflow when working with sg lists with elements greater than 4G

 - An rxe regression with object numbering after the mrs reach their
   limit

 - A theoretical problem with the scatterlist merging code

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
  lib/scatterlist: Fix to calculate the last_pg properly
  IB/hfi1: Remove user expected buffer invalidate race
  IB/hfi1: Immediately remove invalid memory from hardware
  IB/hfi1: Fix expected receive setup error exit issues
  IB/hfi1: Reserve user expected TIDs
  IB/hfi1: Reject a zero-length user expected buffer
  RDMA/core: Fix ib block iterator counter overflow
  RDMA/rxe: Prevent faulty rkey generation
  RDMA/rxe: Fix inaccurate constants in rxe_type_info

Merge tag 'mm-hotfixes-stable-2023-01-16-15-23' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm

2023-01-17T00:36:39+00:00

Pull misc hotfixes from Andrew Morton:
 "21 hotfixes. Thirteen of these address pre-6.1 issues and hence have
  the cc:stable tag"

* tag 'mm-hotfixes-stable-2023-01-16-15-23' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (21 commits)
  init/Kconfig: fix typo (usafe -> unsafe)
  nommu: fix split_vma() map_count error
  nommu: fix do_munmap() error path
  nommu: fix memory leak in do_mmap() error path
  MAINTAINERS: update Robert Foss' email address
  proc: fix PIE proc-empty-vm, proc-pid-vm tests
  mm: update mmap_sem comments to refer to mmap_lock
  include/linux/mm: fix release_pages_arg kernel doc comment
  lib/win_minmax: use /* notation for regular comments
  kasan: mark kasan_kunit_executing as static
  nilfs2: fix general protection fault in nilfs_btree_insert()
  Docs/admin-guide/mm/zswap: remove zsmalloc's lack of writeback warning
  mm/hugetlb: pre-allocate pgtable pages for uffd wr-protects
  hugetlb: unshare some PMDs when splitting VMAs
  mm: fix vma->anon_name memory leak for anonymous shmem VMAs
  mm/shmem: restore SHMEM_HUGE_DENY precedence over MADV_COLLAPSE
  mm/MADV_COLLAPSE: don't expand collapse when vm_end is past requested end
  mm/userfaultfd: enable writenotify while userfaultfd-wp is enabled for a VMA
  mm/khugepaged: fix collapse_pte_mapped_thp() to allow anon_vma
  mm/hugetlb: fix uffd-wp handling for migration entries in hugetlb_change_protection()
  ...

lib/scatterlist: Fix to calculate the last_pg properly

2023-01-16T16:08:31+00:00

The last_pg is wrong, it is actually the first page of the last
scatterlist element. To get the last page of the last scatterlist element
we have to add prv->length. So it is checking mergability against the
wrong page, Further, a SG element is not guaranteed to end on a page
boundary, so we have to check the sub page location also for merge
eligibility.

Fix the above by checking physical contiguity based on PFNs, compute the
actual last page and then call pages_are_mergable().

Fixes: 1567b49d1a40 ("lib/scatterlist: add check when merging zone device pages")
Link: https://lore.kernel.org/r/20230111101054.188136-1-yishaih@nvidia.com
Reported-by: Jason Gunthorpe 
Signed-off-by: Yishai Hadas 
Reviewed-by: Chaitanya Kulkarni 
Signed-off-by: Jason Gunthorpe

lockref: stop doing cpu_relax in the cmpxchg loop

2023-01-13T20:35:38+00:00

On the x86-64 architecture even a failing cmpxchg grants exclusive
access to the cacheline, making it preferable to retry the failed op
immediately instead of stalling with the pause instruction.

To illustrate the impact, below are benchmark results obtained by
running various will-it-scale tests on top of the 6.2-rc3 kernel and
Cascade Lake (2 sockets * 24 cores * 2 threads) CPU.

All results in ops/s.  Note there is some variance in re-runs, but the
code is consistently faster when contention is present.

  open3 ("Same file open/close"):
  proc          stock       no-pause
     1         805603         814942       (+%1)
     2        1054980        1054781       (-0%)
     8        1544802        1822858      (+18%)
    24        1191064        2199665      (+84%)
    48         851582        1469860      (+72%)
    96         609481        1427170     (+134%)

  fstat2 ("Same file fstat"):
  proc          stock       no-pause
     1        3013872        3047636       (+1%)
     2        4284687        4400421       (+2%)
     8        3257721        5530156      (+69%)
    24        2239819        5466127     (+144%)
    48        1701072        5256609     (+209%)
    96        1269157        6649326     (+423%)

Additionally, a kernel with a private patch to help access() scalability:
access2 ("Same file access"):

  proc          stock        patched      patched
                                         +nopause
    24        2378041        2005501      5370335  (-15% / +125%)

That is, fixing the problems in access itself *reduces* scalability
after the cacheline ping-pong only happens in lockref with the pause
instruction.

Note that fstat and access benchmarks are not currently integrated into
will-it-scale, but interested parties can find them in pull requests to
said project.

Code at hand has a rather tortured history.  First modification showed
up in commit d472d9d98b46 ("lockref: Relax in cmpxchg loop"), written
with Itanium in mind.  Later it got patched up to use an arch-dependent
macro to stop doing it on s390 where it caused a significant regression.
Said macro had undergone revisions and was ultimately eliminated later,
going back to cpu_relax.

While I intended to only remove cpu_relax for x86-64, I got the
following comment from Linus:

    I would actually prefer just removing it entirely and see if
    somebody else hollers. You have the numbers to prove it hurts on
    real hardware, and I don't think we have any numbers to the
    contrary.

    So I think it's better to trust the numbers and remove it as a
    failure, than say "let's just remove it on x86-64 and leave
    everybody else with the potentially broken code"

Additionally, Will Deacon (maintainer of the arm64 port, one of the
architectures previously benchmarked):

    So, from the arm64 side of the fence, I'm perfectly happy just
    removing the cpu_relax() calls from lockref.

As such, come back full circle in history and whack it altogether.

Signed-off-by: Mateusz Guzik 
Link: https://lore.kernel.org/all/CAGudoHHx0Nqg6DE70zAVA75eV-HXfWyhVMWZ-aSeOofkA_=WdA@mail.gmail.com/
Acked-by: Tony Luck  # ia64
Acked-by: Nicholas Piggin  # powerpc
Acked-by: Will Deacon  # arm64
Acked-by: Peter Zijlstra 
Signed-off-by: Linus Torvalds

lib/win_minmax: use /* notation for regular comments

2023-01-12T00:14:21+00:00

Don't use kernel-doc "/**" notation for non-kernel-doc comments.
Prevents a kernel-doc warning:

lib/win_minmax.c:31: warning: expecting prototype for lib/minmax.c(). Prototype was for minmax_subwin_update() instead

Link: https://lkml.kernel.org/r/20230102211614.26343-1-rdunlap@infradead.org
Signed-off-by: Randy Dunlap 
Cc: Neal Cardwell 
Cc: Eric Dumazet 
Cc: David S. Miller 
Signed-off-by: Andrew Morton

Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma

2023-01-07T18:06:47+00:00

Pull rdma fixes from Jason Gunthorpe:
 "Most noticeable is that Yishai found a big data corruption regression
  due to a change in the scatterlist:

   - Do not wrongly combine non-contiguous pages in scatterlist

   - Fix compilation warnings on gcc 13

   - Oops when using some mlx5 stats

   - Bad enforcement of atomic responder resources in mlx5"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma:
  lib/scatterlist: Fix to merge contiguous pages into the last SG properly
  RDMA/mlx5: Fix validation of max_rd_atomic caps for DC
  RDMA/mlx5: Fix mlx5_ib_get_hw_stats when used for device
  RDMA/srp: Move large values to a new enum for gcc13

lib/scatterlist: Fix to merge contiguous pages into the last SG properly

2023-01-05T20:01:05+00:00

When sg_alloc_append_table_from_pages() calls to pages_are_mergeable() in
its 'sgt_append->prv' flow to check whether it can merge contiguous pages
into the last SG, it passes the page arguments in the wrong order.

The first parameter should be the next candidate page to be merged to
the last page and not the opposite.

The current code leads to a corrupted SG which resulted in OOPs and
unexpected errors when non-contiguous pages are merged wrongly.

Fix to pass the page parameters in the right order.

Fixes: 1567b49d1a40 ("lib/scatterlist: add check when merging zone device pages")
Link: https://lore.kernel.org/r/20230105112339.107969-1-yishaih@nvidia.com
Signed-off-by: Yishai Hadas 
Reviewed-by: Jason Gunthorpe 
Reviewed-by: Logan Gunthorpe 
Signed-off-by: Jason Gunthorpe

kunit: alloc_string_stream_fragment error handling bug fix

2022-12-26T23:01:36+00:00

When it fails to allocate fragment, it does not free and return error.
And check the pointer inappropriately.

Fixed merge conflicts with
commit 618887768bb7 ("kunit: update NULL vs IS_ERR() tests")
Shuah Khan 

Signed-off-by: YoungJun.park 
Reviewed-by: David Gow 
Signed-off-by: Shuah Khan

test_maple_tree: add test for mas_spanning_rebalance() on insufficient data

2022-12-21T22:31:52+00:00

Add a test to the maple tree test suite for the spanning rebalance
insufficient node issue does not go undetected again.

Link: https://lkml.kernel.org/r/20221219161922.2708732-3-Liam.Howlett@oracle.com
Fixes: 54a611b60590 ("Maple Tree: add new data structure")
Signed-off-by: Liam R. Howlett 
Cc: Andrei Vagin 
Cc: Mike Rapoport 
Cc: Muhammad Usama Anjum 
Cc: 
Signed-off-by: Andrew Morton

maple_tree: fix mas_spanning_rebalance() on insufficient data

2022-12-21T22:31:52+00:00

Mike Rapoport contacted me off-list with a regression in running criu. 
Periodic tests fail with an RCU stall during execution.  Although rare, it
is possible to hit this with other uses so this patch should be backported
to fix the regression.

This patchset adds the fix and a test case to the maple tree test
suite.


This patch (of 2):

An insufficient node was causing an out-of-bounds access on the node in
mas_leaf_max_gap().  The cause was the faulty detection of the new node
being a root node when overwriting many entries at the end of the tree.

Fix the detection of a new root and ensure there is sufficient data prior
to entering the spanning rebalance loop.

Link: https://lkml.kernel.org/r/20221219161922.2708732-1-Liam.Howlett@oracle.com
Link: https://lkml.kernel.org/r/20221219161922.2708732-2-Liam.Howlett@oracle.com
Fixes: 54a611b60590 ("Maple Tree: add new data structure")
Signed-off-by: Liam R. Howlett 
Reported-by: Mike Rapoport 
Tested-by: Mike Rapoport 
Cc: Andrei Vagin 
Cc: Mike Rapoport 
Cc: Muhammad Usama Anjum 
Cc: 
Signed-off-by: Andrew Morton