linux-toradex.git/include/linux/percpu.h, branch v6.4-rc1

Merge branch 'for-6.2' of git://git.kernel.org/pub/scm/linux/kernel/git/dennis/percpu

2022-12-13T00:07:04+00:00

Pull percpu updates from Dennis Zhou:
 "Baoquan was nice enough to run some clean ups for percpu"

* 'for-6.2' of git://git.kernel.org/pub/scm/linux/kernel/git/dennis/percpu:
  mm/percpu: remove unused PERCPU_DYNAMIC_EARLY_SLOTS
  mm/percpu.c: remove the lcm code since block size is fixed at page size
  mm/percpu: replace the goto with break
  mm/percpu: add comment to state the empty populated pages accounting
  mm/percpu: Update the code comment when creating new chunk
  mm/percpu: use list_first_entry_or_null in pcpu_reclaim_populated()
  mm/percpu: remove unused pcpu_map_extend_chunks

percpu: adjust the value of PERCPU_DYNAMIC_EARLY_SIZE

2022-11-21T09:19:25+00:00

LKP reported a build failure as below on the following patch "mm/slub,
percpu: correct the calculation of early percpu allocation size"

~~~~~~
In file included from :
In function 'alloc_kmem_cache_cpus',
   inlined from 'kmem_cache_open' at mm/slub.c:4340:6:
>> >> include/linux/compiler_types.h:357:45: error: call to '__compiletime_assert_474' declared with attribute error:
BUILD_BUG_ON failed: PERCPU_DYNAMIC_EARLY_SIZE < NR_KMALLOC_TYPES * KMALLOC_SHIFT_HIGH * sizeof(struct kmem_cache_cpu)
     357 |         _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
~~~~~~

From the kernel config file provided by LKP, the building was made on
arm64 with below Kconfig item enabled:

  CONFIG_ZONE_DMA=y
  CONFIG_SLUB_CPU_PARTIAL=y
  CONFIG_DEBUG_LOCK_ALLOC=y
  CONFIG_SLUB_STATS=y
  CONFIG_ARM64_PAGE_SHIFT=16
  CONFIG_ARM64_64K_PAGES=y

Then we will have:
  NR_KMALLOC_TYPES:4
  KMALLOC_SHIFT_HIGH:17
  sizeof(struct kmem_cache_cpu):184

The product of them is 12512, which is bigger than PERCPU_DYNAMIC_EARLY_SIZE,
12K. Hence, the BUILD_BUG_ON in alloc_kmem_cache_cpus() is triggered.

Earlier, in commit 099a19d91ca4 ("percpu: allow limited allocation
before slab is online"), PERCPU_DYNAMIC_EARLY_SIZE was introduced and
set to 12K which is equal to the then PERPCU_DYNAMIC_RESERVE.
Later, in commit 1a4d76076cda ("percpu: implement asynchronous chunk
population"), PERPCU_DYNAMIC_RESERVE was increased by 8K, while
PERCPU_DYNAMIC_EARLY_SIZE was kept unchanged.

So, here increase PERCPU_DYNAMIC_EARLY_SIZE by 8K too to accommodate to
the slub's requirement.

Reported-by: kernel test robot 
Signed-off-by: Baoquan He 
Acked-by: Dennis Zhou 
Signed-off-by: Vlastimil Babka

mm/percpu: remove unused PERCPU_DYNAMIC_EARLY_SLOTS

2022-11-08T06:59:28+00:00

Since commit 40064aeca35c ("percpu: replace area map allocator with
bitmap"), there's no place to use PERCPU_DYNAMIC_EARLY_SLOTS. So
clean it up.

Signed-off-by: Baoquan He 
Signed-off-by: Dennis Zhou

mm: percpu: add generic pcpu_populate_pte() function

2022-01-20T06:52:52+00:00

With NEED_PER_CPU_PAGE_FIRST_CHUNK enabled, we need a function to
populate pte, this patch adds a generic pcpu populate pte function,
pcpu_populate_pte(), which is marked __weak and used on most
architectures, but it is overridden on x86, which has its own
implementation.

Link: https://lkml.kernel.org/r/20211216112359.103822-5-wangkefeng.wang@huawei.com
Signed-off-by: Kefeng Wang 
Cc: Michael Ellerman 
Cc: Benjamin Herrenschmidt 
Cc: Paul Mackerras 
Cc: "David S. Miller" 
Cc: Thomas Gleixner 
Cc: Ingo Molnar 
Cc: Borislav Petkov 
Cc: Dave Hansen 
Cc: "H. Peter Anvin" 
Cc: Greg Kroah-Hartman 
Cc: "Rafael J. Wysocki" 
Cc: Dennis Zhou 
Cc: Tejun Heo 
Cc: Christoph Lameter 
Cc: Albert Ou 
Cc: Catalin Marinas 
Cc: Palmer Dabbelt 
Cc: Paul Walmsley 
Cc: Thomas Bogendoerfer 
Cc: Will Deacon 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm: percpu: add generic pcpu_fc_alloc/free funciton

2022-01-20T06:52:52+00:00

With the previous patch, we could add a generic pcpu first chunk
allocate and free function to cleanup the duplicated definations on each
architecture.

Link: https://lkml.kernel.org/r/20211216112359.103822-4-wangkefeng.wang@huawei.com
Signed-off-by: Kefeng Wang 
Cc: Thomas Bogendoerfer 
Cc: Michael Ellerman 
Cc: Benjamin Herrenschmidt 
Cc: Paul Mackerras 
Cc: "David S. Miller" 
Cc: Thomas Gleixner 
Cc: Ingo Molnar 
Cc: Borislav Petkov 
Cc: Dave Hansen 
Cc: "H. Peter Anvin" 
Cc: Greg Kroah-Hartman 
Cc: Dennis Zhou 
Cc: Tejun Heo 
Cc: Christoph Lameter 
Cc: Albert Ou 
Cc: Catalin Marinas 
Cc: Palmer Dabbelt 
Cc: Paul Walmsley 
Cc: "Rafael J. Wysocki" 
Cc: Will Deacon 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm: percpu: add pcpu_fc_cpu_to_node_fn_t typedef

2022-01-20T06:52:52+00:00

Add pcpu_fc_cpu_to_node_fn_t and pass it into pcpu_fc_alloc_fn_t, pcpu
first chunk allocation will call it to alloc memblock on the
corresponding node by it, this is prepare for the next patch.

Link: https://lkml.kernel.org/r/20211216112359.103822-3-wangkefeng.wang@huawei.com
Signed-off-by: Kefeng Wang 
Cc: Thomas Bogendoerfer 
Cc: Michael Ellerman 
Cc: Benjamin Herrenschmidt 
Cc: Paul Mackerras 
Cc: "David S. Miller" 
Cc: Thomas Gleixner 
Cc: Ingo Molnar 
Cc: Borislav Petkov 
Cc: Dave Hansen 
Cc: "H. Peter Anvin" 
Cc: Greg Kroah-Hartman 
Cc: "Rafael J. Wysocki" 
Cc: Dennis Zhou 
Cc: Tejun Heo 
Cc: Christoph Lameter 
Cc: Albert Ou 
Cc: Catalin Marinas 
Cc: Palmer Dabbelt 
Cc: Paul Walmsley 
Cc: Will Deacon 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

Merge tag 'printk-for-5.16-fixup' of git://git.kernel.org/pub/scm/linux/kernel/git/printk/linux

2021-11-18T18:50:45+00:00

Pull printk fixes from Petr Mladek:

 - Try to flush backtraces from other CPUs also on the local one. This
   was a regression caused by printk_safe buffers removal.

 - Remove header dependency warning.

* tag 'printk-for-5.16-fixup' of git://git.kernel.org/pub/scm/linux/kernel/git/printk/linux:
  printk: Remove printk.h inclusion in percpu.h
  printk: restore flushing of NMI buffers on remote CPUs after NMI backtraces

printk: Remove printk.h inclusion in percpu.h

2021-11-15T14:03:32+00:00

After the commit 42a0bb3f7138 ("printk/nmi: generic solution for safe
printk in NMI") the printk.h is not needed anymore in percpu.h.

Moreover `make headerdep` complains (an excerpt)

In file included from linux/printk.h,
                 from linux/dynamic_debug.h:188
                 from linux/printk.h:559 <-- here
                 from linux/percpu.h:9
                 from linux/idr.h:17
include/net/9p/client.h:13: warning: recursive header inclusion

Yeah, it's not a root cause of this, but removing will help to reduce
the noise.

Fixes: 42a0bb3f7138 ("printk/nmi: generic solution for safe printk in NMI")
Signed-off-by: Andy Shevchenko 
Acked-by: Dennis Zhou 
Reviewed-by: Petr Mladek 
Signed-off-by: Petr Mladek 
Link: https://lore.kernel.org/r/20211112140749.80042-1-andriy.shevchenko@linux.intel.com

percpu: add __alloc_size attributes for better bounds checking

2021-11-06T20:30:34+00:00

As already done in GrapheneOS, add the __alloc_size attribute for
appropriate percpu allocator interfaces, to provide additional hinting
for better bounds checking, assisting CONFIG_FORTIFY_SOURCE and other
compiler optimizations.

Note that due to the implementation of the percpu API, this is unlikely
to ever actually provide compile-time checking beyond very simple
non-SMP builds.  But, since they are technically allocators, mark them
as such.

Link: https://lkml.kernel.org/r/20210930222704.2631604-9-keescook@chromium.org
Signed-off-by: Kees Cook 
Co-developed-by: Daniel Micay 
Signed-off-by: Daniel Micay 
Acked-by: Dennis Zhou 
Cc: Tejun Heo 
Cc: Christoph Lameter 
Cc: Andy Whitcroft 
Cc: David Rientjes 
Cc: Dwaipayan Ray 
Cc: Joe Perches 
Cc: Joonsoo Kim 
Cc: Lukas Bulwahn 
Cc: Miguel Ojeda 
Cc: Nathan Chancellor 
Cc: Nick Desaulniers 
Cc: Pekka Enberg 
Cc: Vlastimil Babka 
Cc: Alexandre Bounine 
Cc: Gustavo A. R. Silva 
Cc: Ira Weiny 
Cc: Jing Xiangfeng 
Cc: John Hubbard 
Cc: kernel test robot 
Cc: Matt Porter 
Cc: Randy Dunlap 
Cc: Souptick Joarder 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

percpu: Make pcpu_setup_first_chunk() void function

2019-07-04T15:05:52+00:00

pcpu_setup_first_chunk() will panic or BUG_ON if the are some
error and doesn't return any error, hence it can be defined to
return void.

Reported-by: kbuild test robot 
Signed-off-by: Kefeng Wang 
Signed-off-by: Dennis Zhou 
[Dennis: fixed kbuild warning for pcpu_page_first_chunk()]