From: Michael Ellerman <mpe@ellerman.id.au>
To: <linuxppc-dev@ozlabs.org>
Cc: aneesh.kumar@linux.vnet.ibm.com, bsingharora@gmail.com,
Paul Mackerras <paulus@samba.org>
Subject: [PATCH v3 04/70] powerpc/mm: Add pte_xchg() helper
Date: Fri, 29 Apr 2016 23:25:27 +1000 [thread overview]
Message-ID: <1461936393-10131-4-git-send-email-mpe@ellerman.id.au> (raw)
In-Reply-To: <1461936393-10131-1-git-send-email-mpe@ellerman.id.au>
We have five locations in 64-bit hash MMU code that do a cmpxchg() of a
PTE. Currently doing it inline OK, but in a future patch we will be
converting the PTEs to __be64 in some configs. In that case we will need
casts at every cmpxchg() site in order to keep sparse happy.
So move the logic into a helper, this is a reasonably nice cleanup on
its own.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
---
arch/powerpc/include/asm/kvm_book3s_64.h | 5 +----
arch/powerpc/include/asm/pgtable-types.h | 12 ++++++++++++
arch/powerpc/mm/hash64_4k.c | 4 ++--
arch/powerpc/mm/hash64_64k.c | 7 +++----
arch/powerpc/mm/hugetlbpage-hash64.c | 4 ++--
5 files changed, 20 insertions(+), 12 deletions(-)
diff --git a/arch/powerpc/include/asm/kvm_book3s_64.h b/arch/powerpc/include/asm/kvm_book3s_64.h
index 7529aab068f5..42da513fec2d 100644
--- a/arch/powerpc/include/asm/kvm_book3s_64.h
+++ b/arch/powerpc/include/asm/kvm_book3s_64.h
@@ -319,11 +319,8 @@ static inline pte_t kvmppc_read_update_linux_pte(pte_t *ptep, int writing)
if (writing && pte_write(old_pte))
new_pte = pte_mkdirty(new_pte);
- if (pte_val(old_pte) == __cmpxchg_u64((unsigned long *)ptep,
- pte_val(old_pte),
- pte_val(new_pte))) {
+ if (pte_xchg(ptep, old_pte, new_pte))
break;
- }
}
return new_pte;
}
diff --git a/arch/powerpc/include/asm/pgtable-types.h b/arch/powerpc/include/asm/pgtable-types.h
index 1464e74178d8..e7f4f3e0fcde 100644
--- a/arch/powerpc/include/asm/pgtable-types.h
+++ b/arch/powerpc/include/asm/pgtable-types.h
@@ -54,4 +54,16 @@ typedef struct { pte_t pte; unsigned long hidx; } real_pte_t;
#else
typedef struct { pte_t pte; } real_pte_t;
#endif
+
+#ifdef CONFIG_PPC_STD_MMU_64
+#include <asm/cmpxchg.h>
+
+static inline bool pte_xchg(pte_t *ptep, pte_t old, pte_t new)
+{
+ unsigned long *p = (unsigned long *)ptep;
+
+ return pte_val(old) == __cmpxchg_u64(p, pte_val(old), pte_val(new));
+}
+#endif
+
#endif /* _ASM_POWERPC_PGTABLE_TYPES_H */
diff --git a/arch/powerpc/mm/hash64_4k.c b/arch/powerpc/mm/hash64_4k.c
index 47d1b26effc6..dbd3133a8562 100644
--- a/arch/powerpc/mm/hash64_4k.c
+++ b/arch/powerpc/mm/hash64_4k.c
@@ -47,8 +47,8 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,
new_pte = old_pte | _PAGE_BUSY | _PAGE_ACCESSED;
if (access & _PAGE_RW)
new_pte |= _PAGE_DIRTY;
- } while (old_pte != __cmpxchg_u64((unsigned long *)ptep,
- old_pte, new_pte));
+ } while (!pte_xchg(ptep, __pte(old_pte), __pte(new_pte)));
+
/*
* PP bits. _PAGE_USER is already PP bit 0x2, so we only
* need to add in 0x1 if it's a read-only user page
diff --git a/arch/powerpc/mm/hash64_64k.c b/arch/powerpc/mm/hash64_64k.c
index b2d659cf51c6..6fbf983b3ae5 100644
--- a/arch/powerpc/mm/hash64_64k.c
+++ b/arch/powerpc/mm/hash64_64k.c
@@ -79,8 +79,8 @@ int __hash_page_4K(unsigned long ea, unsigned long access, unsigned long vsid,
new_pte = old_pte | _PAGE_BUSY | _PAGE_ACCESSED | _PAGE_COMBO;
if (access & _PAGE_RW)
new_pte |= _PAGE_DIRTY;
- } while (old_pte != __cmpxchg_u64((unsigned long *)ptep,
- old_pte, new_pte));
+ } while (!pte_xchg(ptep, __pte(old_pte), __pte(new_pte)));
+
/*
* Handle the subpage protection bits
*/
@@ -254,8 +254,7 @@ int __hash_page_64K(unsigned long ea, unsigned long access,
new_pte = old_pte | _PAGE_BUSY | _PAGE_ACCESSED;
if (access & _PAGE_RW)
new_pte |= _PAGE_DIRTY;
- } while (old_pte != __cmpxchg_u64((unsigned long *)ptep,
- old_pte, new_pte));
+ } while (!pte_xchg(ptep, __pte(old_pte), __pte(new_pte)));
rflags = htab_convert_pte_flags(new_pte);
diff --git a/arch/powerpc/mm/hugetlbpage-hash64.c b/arch/powerpc/mm/hugetlbpage-hash64.c
index 8555fce902fe..9c0518761279 100644
--- a/arch/powerpc/mm/hugetlbpage-hash64.c
+++ b/arch/powerpc/mm/hugetlbpage-hash64.c
@@ -57,8 +57,8 @@ int __hash_page_huge(unsigned long ea, unsigned long access, unsigned long vsid,
new_pte = old_pte | _PAGE_BUSY | _PAGE_ACCESSED;
if (access & _PAGE_RW)
new_pte |= _PAGE_DIRTY;
- } while(old_pte != __cmpxchg_u64((unsigned long *)ptep,
- old_pte, new_pte));
+ } while(!pte_xchg(ptep, __pte(old_pte), __pte(new_pte)));
+
rflags = htab_convert_pte_flags(new_pte);
sz = ((1UL) << shift);
--
2.5.0
next prev parent reply other threads:[~2016-04-29 13:26 UTC|newest]
Thread overview: 72+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-04-29 13:25 [PATCH v3 01/70] IB/qib: Use cache inhibitted and guarded mapping on powerpc Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 02/70] powerpc/mm: Always use STRICT_MM_TYPECHECKS Michael Ellerman
2016-05-01 13:02 ` [v3,02/70] " Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 03/70] powerpc/mm: Drop PTE_ATOMIC_UPDATES from pmd_hugepage_update() Michael Ellerman
2016-04-29 13:25 ` Michael Ellerman [this message]
2016-04-29 13:25 ` [PATCH v3 05/70] powerpc/mm: Use big endian Linux page tables for book3s 64 Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 06/70] powerpc/mm: Use pte_raw() in pte_same()/pmd_same() Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 07/70] powerpc/mm: Use _PAGE_READ to indicate Read access Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 08/70] powerpc/mm/subpage: Clear RWX bit to indicate no access Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 09/70] powerpc/mm: Convert pte_user() to static inline Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 10/70] powerpc/mm: Use pte_user() instead of open coding Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 11/70] powerpc/mm: Replace _PAGE_USER with _PAGE_PRIVILEGED Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 12/70] powerpc/mm: Remove RPN_SHIFT and RPN_SIZE Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 13/70] powerpc/mm: Update _PAGE_KERNEL_RO Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 14/70] powerpc/mm: Use a helper for finding pte bits mapping I/O area Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 15/70] powerpc/mm: Drop WIMG in favour of new constants Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 16/70] powerpc/mm: Use generic version of pmdp_clear_flush_young() Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 17/70] powerpc/mm: Use generic version of ptep_clear_flush_young() Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 18/70] powerpc/mm: Move radix/hash common data structures to book3s64 headers Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 19/70] powerpc/mm/radix: Add partition table format & callback Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 20/70] powerpc/mm/hash: Add support for Power9 Hash Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 21/70] powerpc/mm: Move hash and no hash code to separate files Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 22/70] powerpc/mm/book3s: Rename hash specific PTE bits to carry H_ prefix Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 23/70] powerpc/mm: Handle _PTE_NONE_MASK Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 24/70] powerpc/mm: Move common pte bits and accessors to book3s/64/pgtable.h Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 25/70] powerpc/mm: Move pte accessors that operate on common pte bits to pgtable.h Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 26/70] powerpc/mm: Make page table size a variable Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 27/70] powerpc/mm: Move page table index and and vaddr to pgtable.h Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 28/70] powerpc/mm: Move pte related functions together Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 29/70] powerpc/mm/radix: Add radix pte #defines Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 30/70] powerpc/mm/radix: Add dummy radix_enabled() Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 31/70] powerpc/mm: Add radix callbacks to pte accessors Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 32/70] powerpc/mm: Move hugetlb and THP related pmd accessors to pgtable.h Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 33/70] powerpc/mm/radix: Add radix callback for pmd accessors Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 34/70] powerpc/mm: Abstract early MMU init in preparation for radix Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 35/70] powerpc/mm/radix: Add radix callbacks for early init routines Michael Ellerman
2016-04-29 13:25 ` [PATCH v3 36/70] powerpc/mm: Abstraction for vmemmap and map_kernel_page() Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 37/70] powerpc/mm/radix: Add radix callbacks for vmemmap and map_kernel page() Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 38/70] powerpc/mm: Abstraction for switch_mmu_context() Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 39/70] powerpc/mm/radix: Add mmu context handling callback for radix Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 40/70] powerpc/mm: Rename mmu_context_hash64.c to mmu_context_book3s64.c Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 41/70] powerpc/mm: Hash abstraction for tlbflush routines Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 42/70] powerpc/mm/radix: Add " Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 43/70] powerpc/mm/radix: Add MMU_FTR_RADIX Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 44/70] powerpc/mm/radix: Use STD_MMU_64 to properly isolate hash related code Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 45/70] powerpc/mm/radix: Isolate hash table function from pseries guest code Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 46/70] powerpc/mm/radix: Add checks in slice code to catch radix usage Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 47/70] powerpc/mm/radix: Limit paca allocation in radix Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 48/70] powerpc/mm/radix: Pick the address layout for radix config Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 49/70] powerpc/mm/radix: Update PTCR on secondary CPUs Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 50/70] powerpc/mm: Make a copy of pgalloc.h for 32 and 64 book3s Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 51/70] powerpc/mm: Copy pgalloc (part 2) Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 52/70] powerpc/mm: Revert changes made to nohash pgalloc-64.h Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 53/70] powerpc/mm: Simplify the code dropping 4-level table #ifdef Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 54/70] powerpc/mm: Rename function to indicate we are allocating fragments Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 55/70] powerpc/mm: Make 4K and 64K use pte_t for pgtable_t Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 56/70] powerpc/mm: Add radix pgalloc details Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 57/70] powerpc/mm: Update pte filter for radix Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 58/70] powerpc/mm: vmalloc abstraction in preparation " Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 59/70] powerpc/radix: Update MMU cache Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 60/70] powerpc/mm: pte_frag abstraction Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 61/70] powerpc/mm: Fix vma_mmu_pagesize() for radix Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 62/70] powerpc/mm: Add radix support for hugetlb Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 63/70] powerpc/mm/radix: Make sure swapper pgdir is properly aligned Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 64/70] powerpc/mm/radix: Add hugetlb support 4K page size Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 65/70] powerpc/mm: THP is only available on hash64 as of now Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 66/70] powerpc/mm/thp: Abstraction for THP functions Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 67/70] powerpc/mm/radix: Add radix THP callbacks Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 68/70] powerpc/mm/radix: Add THP support for 4K linux page size Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 69/70] powerpc/mm/radix: Use firmware feature to enable Radix MMU Michael Ellerman
2016-04-29 13:26 ` [PATCH v3 70/70] powerpc/mm/radix: Document software bits for radix Michael Ellerman
2016-05-01 13:02 ` [v3, 01/70] IB/qib: Use cache inhibitted and guarded mapping on powerpc Michael Ellerman
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1461936393-10131-4-git-send-email-mpe@ellerman.id.au \
--to=mpe@ellerman.id.au \
--cc=aneesh.kumar@linux.vnet.ibm.com \
--cc=bsingharora@gmail.com \
--cc=linuxppc-dev@ozlabs.org \
--cc=paulus@samba.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).