Skip to content

C
clone-Linux

gsplhtlxg / clone-Linux

通知 2
Star 0
Fork 1
C
clone-Linux

体验新版 GitCode，发现更多精彩内容 >>

切换分支/标签
  1. 15 2月, 2016 1 次提交
    • A
      powerpc/mm: Fix Multi hit ERAT cause by recent THP update · c777e2a8
      由 Aneesh Kumar K.V 提交于 
      With ppc64 we use the deposited pgtable_t to store the hash pte slot
information. We should not withdraw the deposited pgtable_t without
marking the pmd none. This ensure that low level hash fault handling
will skip this huge pte and we will handle them at upper levels.

Recent change to pmd splitting changed the above in order to handle the
race between pmd split and exit_mmap. The race is explained below.

Consider following race:

		CPU0				CPU1
shrink_page_list()
  add_to_swap()
    split_huge_page_to_list()
      __split_huge_pmd_locked()
        pmdp_huge_clear_flush_notify()
	// pmd_none() == true
					exit_mmap()
					  unmap_vmas()
					    zap_pmd_range()
					      // no action on pmd since pmd_none() == true
	pmd_populate()

As result the THP will not be freed. The leak is detected by check_mm():

	BUG: Bad rss-counter state mm:ffff880058d2e580 idx:1 val:512

The above required us to not mark pmd none during a pmd split.

The fix for ppc is to clear the huge pte of _PAGE_USER, so that low
level fault handling code skip this pte. At higher level we do take ptl
lock. That should serialze us against the pmd split. Once the lock is
acquired we do check the pmd again using pmd_same. That should always
return false for us and hence we should retry the access. We do the
pmd_same check in all case after taking plt with
THP (do_huge_pmd_wp_page, do_huge_pmd_numa_page and
huge_pmd_set_accessed)

Also make sure we wait for irq disable section in other cpus to finish
before flipping a huge pte entry with a regular pmd entry. Code paths
like find_linux_pte_or_hugepte depend on irq disable to get
a stable pte_t pointer. A parallel thp split need to make sure we
don't convert a pmd pte to a regular pmd entry without waiting for the
irq disable section to finish.

Fixes: eef1b3ba ("thp: implement split_huge_pmd()")
Acked-by: NKirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: NMichael Ellerman <mpe@ellerman.id.au>
      c777e2a8
  3. 25 1月, 2016 1 次提交
  5. 16 1月, 2016 3 次提交
  7. 09 1月, 2016 1 次提交
  9. 27 12月, 2015 1 次提交
  11. 23 12月, 2015 1 次提交
  13. 19 12月, 2015 1 次提交
  15. 14 12月, 2015 15 次提交
  17. 06 11月, 2015 1 次提交
    • R
      arch/powerpc/mm/numa.c: do not allocate bootmem memory for non existing nodes · c118baf8
      由 Raghavendra K T 提交于 
      With the setup_nr_nodes(), we have already initialized
node_possible_map.  So it is safe to use for_each_node here.

There are many places in the kernel that use hardcoded 'for' loop with
nr_node_ids, because all other architectures have numa nodes populated
serially.  That should be reason we had maintained the same for
powerpc.

But, since sparse numa node ids possible on powerpc, we unnecessarily
allocate memory for non existent numa nodes.

For e.g., on a system with 0,1,16,17 as numa nodes nr_node_ids=18 and
we allocate memory for nodes 2-14.  This patch we allocate memory for
only existing numa nodes.

The patch is boot tested on a 4 node tuleta, confirming with printks
that it works as expected.
Signed-off-by: NRaghavendra K T <raghavendra.kt@linux.vnet.ibm.com>
Cc: Vladimir Davydov <vdavydov@parallels.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Anton Blanchard <anton@samba.org>
Cc: Nishanth Aravamudan <nacc@linux.vnet.ibm.com>
Cc: Greg Kurz <gkurz@linux.vnet.ibm.com>
Cc: Grant Likely <grant.likely@linaro.org>
Cc: Nikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      c118baf8
  19. 28 10月, 2015 2 次提交
    • K
      powerpc/e6500: hw tablewalk: make sure we invalidate and write to the same tlb entry · e1f580e8
      由 Kevin Hao 提交于 
      In order to workaround Erratum A-008139, we have to invalidate the
tlb entry with tlbilx before overwriting. Due to the performance
consideration, we don't add any memory barrier when acquire/release
the tcd lock. This means the two load instructions for esel_next do
have the possibility to return different value. This is definitely
not acceptable due to the Erratum A-008139. We have two options to
fix this issue:
  a) Add memory barrier when acquire/release tcd lock to order the
     load/store to esel_next.
  b) Just make sure to invalidate and write to the same tlb entry and
     tolerate the race that we may get the wrong value and overwrite
     the tlb entry just updated by the other thread.

We observe better performance using option b. So reserve an additional
register to save the value of the esel_next.
Signed-off-by: NKevin Hao <haokexin@gmail.com>
Signed-off-by: NScott Wood <scottwood@freescale.com>
      e1f580e8
    • S
      powerpc/fsl-booke-64: Don't limit ppc64_rma_size to one TLB entry · eba5de8d
      由 Scott Wood 提交于 
      This is required for kdump to work when loaded at at an address that
does not fall within the first TLB entry -- which can easily happen
because while the lower limit is enforced via reserved memory, which
doesn't affect how much is mapped, the upper limit is enforced via a
different mechanism that does.  Thus, more TLB entries are needed than
would normally be used, as the total memory to be mapped might not be a
power of two.
Signed-off-by: NScott Wood <scottwood@freescale.com>
      eba5de8d
  21. 23 10月, 2015 1 次提交
    • S
      powerpc/85xx: Load all early TLB entries at once · d9e1831a
      由 Scott Wood 提交于 
      Use an AS=1 trampoline TLB entry to allow all normal TLB1 entries to
be loaded at once.  This avoids the need to keep the translation that
code is executing from in the same TLB entry in the final TLB
configuration as during early boot, which in turn is helpful for
relocatable kernels (e.g. kdump) where the kernel is not running from
what would be the first TLB entry.

On e6500, we limit map_mem_in_cams() to the primary hwthread of a
core (the boot cpu is always considered primary, as a kdump kernel
can be entered on any cpu).  Each TLB only needs to be set up once,
and when we do, we don't want another thread to be running when we
create a temporary trampoline TLB1 entry.
Signed-off-by: NScott Wood <scottwood@freescale.com>
      d9e1831a
  23. 15 10月, 2015 1 次提交
  25. 12 10月, 2015 2 次提交
  27. 09 10月, 2015 1 次提交
    • C
      powerpc: Fix checkstop in native_hpte_clear() with lockdep · fdf880a6
      由 Cyril Bur 提交于 
      native_hpte_clear() is called in real mode from two places:
- Early in boot during htab initialisation if firmware assisted dump is
  active.
- Late in the kexec path.

In both contexts there is no need to disable interrupts are they are
already disabled. Furthermore, locking around the tlbie() is only required
for pre POWER5 hardware.

On POWER5 or newer hardware concurrent tlbie()s work as expected and on pre
POWER5 hardware concurrent tlbie()s could result in deadlock. This code
would only be executed at crashdump time, during which all bets are off,
concurrent tlbie()s are unlikely and taking locks is unsafe therefore the
best course of action is to simply do nothing. Concurrent tlbie()s are not
possible in the first case as secondary CPUs have not come up yet.
Signed-off-by: NCyril Bur <cyrilbur@gmail.com>
Signed-off-by: NMichael Ellerman <mpe@ellerman.id.au>
      fdf880a6
  29. 01 10月, 2015 2 次提交
  31. 16 9月, 2015 1 次提交
  33. 28 8月, 2015 1 次提交
    • D
      mm: ZONE_DEVICE for "device memory" · 033fbae9
      由 Dan Williams 提交于 
      While pmem is usable as a block device or via DAX mappings to userspace
there are several usage scenarios that can not target pmem due to its
lack of struct page coverage. In preparation for "hot plugging" pmem
into the vmemmap add ZONE_DEVICE as a new zone to tag these pages
separately from the ones that are subject to standard page allocations.
Importantly "device memory" can be removed at will by userspace
unbinding the driver of the device.

Having a separate zone prevents allocation and otherwise marks these
pages that are distinct from typical uniform memory.  Device memory has
different lifetime and performance characteristics than RAM.  However,
since we have run out of ZONES_SHIFT bits this functionality currently
depends on sacrificing ZONE_DMA.

Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Jerome Glisse <j.glisse@gmail.com>
[hch: various simplifications in the arch interface]
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
      033fbae9
  35. 18 8月, 2015 4 次提交
    • N
      powerpc/numa: initialize distance lookup table from drconf path · 1d805440
      由 Nikunj A Dadhania 提交于 
      In some situations, a NUMA guest that supports
ibm,dynamic-memory-reconfiguration node will end up having flat NUMA
distances between nodes. This is because of two problems in the
current code.

1) Different representations of associativity lists.

   There is an assumption about the associativity list in
   initialize_distance_lookup_table(). Associativity list has two forms:

   a) [cpu,memory]@x/ibm,associativity has following
      format:
           <N> <N integers>

   b) ibm,dynamic-reconfiguration-memory/ibm,associativity-lookup-arrays

           <M> <N> <M associativity lists each having N integers>
           M = the number of associativity lists
           N = the number of entries per associativity list

   Fix initialize_distance_lookup_table() so that it does not assume
   "case a". And update the caller to skip the length field before
   sending the associativity list.

2) Distance table not getting updated from drconf path.

   Node distance table will not get initialized in certain cases as
   ibm,dynamic-reconfiguration-memory path does not initialize the
   lookup table.

   Call initialize_distance_lookup_table() from drconf path with
   appropriate associativity list.
Reported-by: NBharata B Rao <bharata@linux.vnet.ibm.com>
Signed-off-by: NNikunj A Dadhania <nikunj@linux.vnet.ibm.com>
Acked-by: NAnton Blanchard <anton@samba.org>
Signed-off-by: NMichael Ellerman <mpe@ellerman.id.au>
      1d805440
    • M
      powerpc/mm: Drop CONFIG_PPC_HAS_HASH_64K · 73b341ef
      由 Michael Ellerman 提交于 
      The relation between CONFIG_PPC_HAS_HASH_64K and CONFIG_PPC_64K_PAGES is
painfully complicated.

But if we rearrange it enough we can see that PPC_HAS_HASH_64K
essentially depends on PPC_STD_MMU_64 && PPC_64K_PAGES.

We can then notice that PPC_HAS_HASH_64K is used in files that are only
built for PPC_STD_MMU_64, meaning it's equivalent to PPC_64K_PAGES.

So replace all uses and drop it.
Signed-off-by: NMichael Ellerman <mpe@ellerman.id.au>
Reviewed-by: NAneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
      73b341ef
    • M
      powerpc/cell: Drop support for 64K local store on 4K kernels · f444f1f8
      由 Michael Ellerman 提交于 
      Back in the olden days we added support for using 64K pages to map the
SPU (Synergistic Processing Unit) local store on Cell, when the main
kernel was using 4K pages.

This was useful at the time because distros were using 4K pages, but
using 64K pages on the SPUs could reduce TLB pressure there.

However these days the number of Cell users is approaching zero, and
supporting this option adds unpleasant complexity to the memory
management code.

So drop the option, CONFIG_SPU_FS_64K_LS, and all related code.
Signed-off-by: NMichael Ellerman <mpe@ellerman.id.au>
Acked-by: NJeremy Kerr <jk@ozlabs.org>
      f444f1f8
    • K
      powerpc/e6500: hw tablewalk: optimize a bit for tcd lock acquiring codes · 69399ee9
      由 Kevin Hao 提交于 
      It makes no sense to put the instructions for calculating the lock
value (cpu number + 1) and the clearing of eq bit of cr1 in lbarx/stbcx
loop. And when the lock is acquired by the other thread, the current
lock value has no chance to equal with the lock value used by current
cpu. So we can skip the comparing for these two lock values in the
lbz/bne loop.
Signed-off-by: NKevin Hao <haokexin@gmail.com>
Signed-off-by: NScott Wood <scottwood@freescale.com>
      69399ee9