提交 12124b81 编写于 作者: T Thomas Gleixner 提交者: Shile Zhang

x86/mm/cpa: Avoid the 4k pages check completely

commit 585948f4f695b07204702cfee0f828424af32aa7 upstream.

The extra loop which tries hard to preserve large pages in case of conflicts
with static protection regions turns out to be not preserving anything, at
least not in the experiments which have been conducted.

There might be corner cases in which the code would be able to preserve a
large page oaccsionally, but it's really not worth the extra code and the
cycles wasted in the common case.

Before:

 1G pages checked:                    2
 1G pages sameprot:                   0
 1G pages preserved:                  0
 2M pages checked:                  541
 2M pages sameprot:                 466
 2M pages preserved:                 47
 4K pages checked:                  514
 4K pages set-checked:             7668

After:
 1G pages checked:                    2
 1G pages sameprot:                   0
 1G pages preserved:                  0
 2M pages checked:                  538
 2M pages sameprot:                 466
 2M pages preserved:                 47
 4K pages set-checked:             7668
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Reviewed-by: NDave Hansen <dave.hansen@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Bin Yang <bin.yang@intel.com>
Cc: Mark Gross <mark.gross@intel.com>
Link: https://lkml.kernel.org/r/20180917143546.589642503@linutronix.deSigned-off-by: NShile Zhang <shile.zhang@linux.alibaba.com>
Acked-by: NJoseph Qi <joseph.qi@linux.alibaba.com>
上级 861a2ee6
...@@ -111,7 +111,6 @@ static unsigned long cpa_1g_preserved; ...@@ -111,7 +111,6 @@ static unsigned long cpa_1g_preserved;
static unsigned long cpa_2m_checked; static unsigned long cpa_2m_checked;
static unsigned long cpa_2m_sameprot; static unsigned long cpa_2m_sameprot;
static unsigned long cpa_2m_preserved; static unsigned long cpa_2m_preserved;
static unsigned long cpa_4k_checked;
static unsigned long cpa_4k_install; static unsigned long cpa_4k_install;
static inline void cpa_inc_1g_checked(void) static inline void cpa_inc_1g_checked(void)
...@@ -124,11 +123,6 @@ static inline void cpa_inc_2m_checked(void) ...@@ -124,11 +123,6 @@ static inline void cpa_inc_2m_checked(void)
cpa_2m_checked++; cpa_2m_checked++;
} }
static inline void cpa_inc_4k_checked(void)
{
cpa_4k_checked++;
}
static inline void cpa_inc_4k_install(void) static inline void cpa_inc_4k_install(void)
{ {
cpa_4k_install++; cpa_4k_install++;
...@@ -158,7 +152,6 @@ static int cpastats_show(struct seq_file *m, void *p) ...@@ -158,7 +152,6 @@ static int cpastats_show(struct seq_file *m, void *p)
seq_printf(m, "2M pages checked: %16lu\n", cpa_2m_checked); seq_printf(m, "2M pages checked: %16lu\n", cpa_2m_checked);
seq_printf(m, "2M pages sameprot: %16lu\n", cpa_2m_sameprot); seq_printf(m, "2M pages sameprot: %16lu\n", cpa_2m_sameprot);
seq_printf(m, "2M pages preserved: %16lu\n", cpa_2m_preserved); seq_printf(m, "2M pages preserved: %16lu\n", cpa_2m_preserved);
seq_printf(m, "4K pages checked: %16lu\n", cpa_4k_checked);
seq_printf(m, "4K pages set-checked: %16lu\n", cpa_4k_install); seq_printf(m, "4K pages set-checked: %16lu\n", cpa_4k_install);
return 0; return 0;
} }
...@@ -185,7 +178,6 @@ late_initcall(cpa_stats_init); ...@@ -185,7 +178,6 @@ late_initcall(cpa_stats_init);
#else #else
static inline void cpa_inc_1g_checked(void) { } static inline void cpa_inc_1g_checked(void) { }
static inline void cpa_inc_2m_checked(void) { } static inline void cpa_inc_2m_checked(void) { }
static inline void cpa_inc_4k_checked(void) { }
static inline void cpa_inc_4k_install(void) { } static inline void cpa_inc_4k_install(void) { }
static inline void cpa_inc_lp_sameprot(int level) { } static inline void cpa_inc_lp_sameprot(int level) { }
static inline void cpa_inc_lp_preserved(int level) { } static inline void cpa_inc_lp_preserved(int level) { }
...@@ -745,11 +737,10 @@ static pgprot_t pgprot_clear_protnone_bits(pgprot_t prot) ...@@ -745,11 +737,10 @@ static pgprot_t pgprot_clear_protnone_bits(pgprot_t prot)
static int __should_split_large_page(pte_t *kpte, unsigned long address, static int __should_split_large_page(pte_t *kpte, unsigned long address,
struct cpa_data *cpa) struct cpa_data *cpa)
{ {
unsigned long numpages, pmask, psize, lpaddr, addr, pfn, old_pfn; unsigned long numpages, pmask, psize, lpaddr, pfn, old_pfn;
pgprot_t old_prot, new_prot, req_prot, chk_prot; pgprot_t old_prot, new_prot, req_prot, chk_prot;
pte_t new_pte, old_pte, *tmp; pte_t new_pte, old_pte, *tmp;
enum pg_level level; enum pg_level level;
int i;
/* /*
* Check for races, another CPU might have split this page * Check for races, another CPU might have split this page
...@@ -854,53 +845,30 @@ static int __should_split_large_page(pte_t *kpte, unsigned long address, ...@@ -854,53 +845,30 @@ static int __should_split_large_page(pte_t *kpte, unsigned long address,
} }
/* /*
* Optimization: Check whether the requested pgprot is conflicting * If the requested range does not cover the full page, split it up
* with a static protection requirement in the large page. If not,
* then checking whether the requested range is fully covering the
* large page can be done right here.
*/ */
new_prot = static_protections(req_prot, lpaddr, old_pfn, numpages, if (address != lpaddr || cpa->numpages != numpages)
CPA_DETECT); return 1;
if (pgprot_val(req_prot) == pgprot_val(new_prot)) {
if (address != lpaddr || cpa->numpages != numpages)
return 1;
goto setlp;
}
/* /*
* Slow path. The full large page check above established that the * Check whether the requested pgprot is conflicting with a static
* requested pgprot cannot be applied to the full large page due to * protection requirement in the large page.
* conflicting requirements of static protection regions. It might
* turn out that the whole requested range is covered by the
* modified protection of the first 4k segment at @address. This
* might result in the ability to preserve the large page
* nevertheless.
*/ */
new_prot = static_protections(req_prot, address, pfn, 1, CPA_DETECT); new_prot = static_protections(req_prot, lpaddr, old_pfn, numpages,
pfn = old_pfn; CPA_DETECT);
for (i = 0, addr = lpaddr; i < numpages; i++, addr += PAGE_SIZE, pfn++) {
chk_prot = static_protections(req_prot, addr, pfn, 1,
CPA_DETECT);
cpa_inc_4k_checked();
if (pgprot_val(chk_prot) != pgprot_val(new_prot))
return 1;
}
/* If there are no changes, return. */
if (pgprot_val(new_prot) == pgprot_val(old_prot)) {
cpa_inc_lp_sameprot(level);
return 0;
}
/* /*
* Verify that the address is aligned and the number of pages * If there is a conflict, split the large page.
* covers the full page. *
* There used to be a 4k wise evaluation trying really hard to
* preserve the large pages, but experimentation has shown, that this
* does not help at all. There might be corner cases which would
* preserve one large page occasionally, but it's really not worth the
* extra code and cycles for the common case.
*/ */
if (address != lpaddr || cpa->numpages != numpages) if (pgprot_val(req_prot) != pgprot_val(new_prot))
return 1; return 1;
setlp:
/* All checks passed. Update the large page mapping. */ /* All checks passed. Update the large page mapping. */
new_pte = pfn_pte(old_pfn, new_prot); new_pte = pfn_pte(old_pfn, new_prot);
__set_pmd_pte(kpte, address, new_pte); __set_pmd_pte(kpte, address, new_pte);
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册