summaryrefslogtreecommitdiff
path: root/include
diff options
context:
space:
mode:
authorPeter Zijlstra <peterz@infradead.org>2020-02-04 04:37:11 +0300
committerLinus Torvalds <torvalds@linux-foundation.org>2020-02-04 06:05:26 +0300
commit0d6e24d430ef23280d8dea0ba1faeefc66c26a57 (patch)
tree6fb47b4bee252781c30105392a143e9a9a688903 /include
parent580a586c409ab3040b7284a19cd9e281692c40c7 (diff)
downloadlinux-0d6e24d430ef23280d8dea0ba1faeefc66c26a57.tar.xz
asm-generic/tlb: provide MMU_GATHER_TABLE_FREE
As described in the comment, the correct order for freeing pages is: 1) unhook page 2) TLB invalidate page 3) free page This order equally applies to page directories. Currently there are two correct options: - use tlb_remove_page(), when all page directores are full pages and there are no futher contraints placed by things like software walkers (HAVE_FAST_GUP). - use MMU_GATHER_RCU_TABLE_FREE and tlb_remove_table() when the architecture does not do IPI based TLB invalidate and has HAVE_FAST_GUP (or software TLB fill). This however leaves architectures that don't have page based directories but don't need RCU in a bind. For those, provide MMU_GATHER_TABLE_FREE, which provides the independent batching for directories without the additional RCU freeing. Link: http://lkml.kernel.org/r/20200116064531.483522-10-aneesh.kumar@linux.ibm.com Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> Cc: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'include')
-rw-r--r--include/asm-generic/tlb.h72
1 files changed, 37 insertions, 35 deletions
diff --git a/include/asm-generic/tlb.h b/include/asm-generic/tlb.h
index ca0fe75b5355..f391f6b500b4 100644
--- a/include/asm-generic/tlb.h
+++ b/include/asm-generic/tlb.h
@@ -56,6 +56,15 @@
* Defaults to flushing at tlb_end_vma() to reset the range; helps when
* there's large holes between the VMAs.
*
+ * - tlb_remove_table()
+ *
+ * tlb_remove_table() is the basic primitive to free page-table directories
+ * (__p*_free_tlb()). In it's most primitive form it is an alias for
+ * tlb_remove_page() below, for when page directories are pages and have no
+ * additional constraints.
+ *
+ * See also MMU_GATHER_TABLE_FREE and MMU_GATHER_RCU_TABLE_FREE.
+ *
* - tlb_remove_page() / __tlb_remove_page()
* - tlb_remove_page_size() / __tlb_remove_page_size()
*
@@ -129,17 +138,24 @@
* This might be useful if your architecture has size specific TLB
* invalidation instructions.
*
- * MMU_GATHER_RCU_TABLE_FREE
+ * MMU_GATHER_TABLE_FREE
*
* This provides tlb_remove_table(), to be used instead of tlb_remove_page()
- * for page directores (__p*_free_tlb()). This provides separate freeing of
- * the page-table pages themselves in a semi-RCU fashion (see comment below).
- * Useful if your architecture doesn't use IPIs for remote TLB invalidates
- * and therefore doesn't naturally serialize with software page-table walkers.
+ * for page directores (__p*_free_tlb()).
+ *
+ * Useful if your architecture has non-page page directories.
*
* When used, an architecture is expected to provide __tlb_remove_table()
* which does the actual freeing of these pages.
*
+ * MMU_GATHER_RCU_TABLE_FREE
+ *
+ * Like MMU_GATHER_TABLE_FREE, and adds semi-RCU semantics to the free (see
+ * comment below).
+ *
+ * Useful if your architecture doesn't use IPIs for remote TLB invalidates
+ * and therefore doesn't naturally serialize with software page-table walkers.
+ *
* MMU_GATHER_NO_RANGE
*
* Use this if your architecture lacks an efficient flush_tlb_range().
@@ -155,37 +171,12 @@
* various ptep_get_and_clear() functions.
*/
-#ifdef CONFIG_MMU_GATHER_RCU_TABLE_FREE
-/*
- * Semi RCU freeing of the page directories.
- *
- * This is needed by some architectures to implement software pagetable walkers.
- *
- * gup_fast() and other software pagetable walkers do a lockless page-table
- * walk and therefore needs some synchronization with the freeing of the page
- * directories. The chosen means to accomplish that is by disabling IRQs over
- * the walk.
- *
- * Architectures that use IPIs to flush TLBs will then automagically DTRT,
- * since we unlink the page, flush TLBs, free the page. Since the disabling of
- * IRQs delays the completion of the TLB flush we can never observe an already
- * freed page.
- *
- * Architectures that do not have this (PPC) need to delay the freeing by some
- * other means, this is that means.
- *
- * What we do is batch the freed directory pages (tables) and RCU free them.
- * We use the sched RCU variant, as that guarantees that IRQ/preempt disabling
- * holds off grace periods.
- *
- * However, in order to batch these pages we need to allocate storage, this
- * allocation is deep inside the MM code and can thus easily fail on memory
- * pressure. To guarantee progress we fall back to single table freeing, see
- * the implementation of tlb_remove_table_one().
- *
- */
+#ifdef CONFIG_MMU_GATHER_TABLE_FREE
+
struct mmu_table_batch {
+#ifdef CONFIG_MMU_GATHER_RCU_TABLE_FREE
struct rcu_head rcu;
+#endif
unsigned int nr;
void *tables[0];
};
@@ -195,6 +186,17 @@ struct mmu_table_batch {
extern void tlb_remove_table(struct mmu_gather *tlb, void *table);
+#else /* !CONFIG_MMU_GATHER_HAVE_TABLE_FREE */
+
+/*
+ * Without MMU_GATHER_TABLE_FREE the architecture is assumed to have page based
+ * page directories and we can use the normal page batching to free them.
+ */
+#define tlb_remove_table(tlb, page) tlb_remove_page((tlb), (page))
+
+#endif /* CONFIG_MMU_GATHER_TABLE_FREE */
+
+#ifdef CONFIG_MMU_GATHER_RCU_TABLE_FREE
/*
* This allows an architecture that does not use the linux page-tables for
* hardware to skip the TLBI when freeing page tables.
@@ -248,7 +250,7 @@ extern bool __tlb_remove_page_size(struct mmu_gather *tlb, struct page *page,
struct mmu_gather {
struct mm_struct *mm;
-#ifdef CONFIG_MMU_GATHER_RCU_TABLE_FREE
+#ifdef CONFIG_MMU_GATHER_TABLE_FREE
struct mmu_table_batch *batch;
#endif