OpenVZ Forum


Home » Mailing lists » Devel » [RFC][PATCH 0/7] Resource controllers based on process containers
[RFC][PATCH 4/7] RSS accounting hooks over the code [message #10892 is a reply to message #10888] Tue, 06 March 2007 14:57 Go to previous messageGo to previous message
xemul is currently offline  xemul
Messages: 248
Registered: November 2005
Senior Member
Pages are charged to their first touchers which are
determined using pages' mapcount manipulations in
rmap calls.

diff -upr linux-2.6.20.orig/fs/exec.c linux-2.6.20-0/fs/exec.c
--- linux-2.6.20.orig/fs/exec.c 2007-02-04 21:44:54.000000000 +0300
+++ linux-2.6.20-0/fs/exec.c 2007-03-06 13:33:28.000000000 +0300
@@ -58,6 +58,8 @@
#include <linux/kmod.h>
#endif

+#include <linux/rss_container.h>
+
int core_uses_pid;
char core_pattern[128] = "core";
int suid_dumpable = 0;
@@ -309,27 +311,34 @@ void install_arg_page(struct vm_area_str
struct mm_struct *mm = vma->vm_mm;
pte_t * pte;
spinlock_t *ptl;
+ struct page_container *pcont;

if (unlikely(anon_vma_prepare(vma)))
goto out;

+ if (container_rss_prepare(page, vma, &pcont))
+ goto out;
+
flush_dcache_page(page);
pte = get_locked_pte(mm, address, &ptl);
if (!pte)
- goto out;
+ goto out_release;
if (!pte_none(*pte)) {
pte_unmap_unlock(pte, ptl);
- goto out;
+ goto out_release;
}
inc_mm_counter(mm, anon_rss);
lru_cache_add_active(page);
set_pte_at(mm, address, pte, pte_mkdirty(pte_mkwrite(mk_pte(
page, vma->vm_page_prot))));
- page_add_new_anon_rmap(page, vma, address);
+ page_add_new_anon_rmap(page, vma, address, pcont);
pte_unmap_unlock(pte, ptl);

/* no need for flush_tlb */
return;
+
+out_release:
+ container_rss_release(pcont);
out:
__free_page(page);
force_sig(SIGKILL, current);
diff -upr linux-2.6.20.orig/include/linux/rmap.h linux-2.6.20-0/include/linux/rmap.h
--- linux-2.6.20.orig/include/linux/rmap.h 2007-02-04 21:44:54.000000000 +0300
+++ linux-2.6.20-0/include/linux/rmap.h 2007-03-06 13:33:28.000000000 +0300
@@ -69,9 +69,13 @@ void __anon_vma_link(struct vm_area_stru
/*
* rmap interfaces called when adding or removing pte of page
*/
-void page_add_anon_rmap(struct page *, struct vm_area_struct *, unsigned long);
-void page_add_new_anon_rmap(struct page *, struct vm_area_struct *, unsigned long);
-void page_add_file_rmap(struct page *);
+struct page_container;
+
+void page_add_anon_rmap(struct page *, struct vm_area_struct *,
+ unsigned long, struct page_container *);
+void page_add_new_anon_rmap(struct page *, struct vm_area_struct *,
+ unsigned long, struct page_container *);
+void page_add_file_rmap(struct page *, struct page_container *);
void page_remove_rmap(struct page *, struct vm_area_struct *);

/**
diff -upr linux-2.6.20.orig/mm/fremap.c linux-2.6.20-0/mm/fremap.c
--- linux-2.6.20.orig/mm/fremap.c 2007-02-04 21:44:54.000000000 +0300
+++ linux-2.6.20-0/mm/fremap.c 2007-03-06 13:33:28.000000000 +0300
@@ -20,6 +20,8 @@
#include <asm/cacheflush.h>
#include <asm/tlbflush.h>

+#include <linux/rss_container.h>
+
static int zap_pte(struct mm_struct *mm, struct vm_area_struct *vma,
unsigned long addr, pte_t *ptep)
{
@@ -57,6 +59,10 @@ int install_page(struct mm_struct *mm, s
pte_t *pte;
pte_t pte_val;
spinlock_t *ptl;
+ struct page_container *pcont;
+
+ if (container_rss_prepare(page, vma, &pcont))
+ goto out_release;

pte = get_locked_pte(mm, addr, &ptl);
if (!pte)
@@ -81,13 +87,16 @@ int install_page(struct mm_struct *mm, s
flush_icache_page(vma, page);
pte_val = mk_pte(page, prot);
set_pte_at(mm, addr, pte, pte_val);
- page_add_file_rmap(page);
+ page_add_file_rmap(page, pcont);
update_mmu_cache(vma, addr, pte_val);
lazy_mmu_prot_update(pte_val);
err = 0;
unlock:
pte_unmap_unlock(pte, ptl);
out:
+ if (err != 0)
+ container_rss_release(pcont);
+out_release:
return err;
}
EXPORT_SYMBOL(install_page);
diff -upr linux-2.6.20.orig/mm/memory.c linux-2.6.20-0/mm/memory.c
--- linux-2.6.20.orig/mm/memory.c 2007-02-04 21:44:54.000000000 +0300
+++ linux-2.6.20-0/mm/memory.c 2007-03-06 13:33:28.000000000 +0300
@@ -60,6 +60,8 @@
#include <linux/swapops.h>
#include <linux/elf.h>

+#include <linux/rss_container.h>
+
#ifndef CONFIG_NEED_MULTIPLE_NODES
/* use the per-pgdat data instead for discontigmem - mbligh */
unsigned long max_mapnr;
@@ -1126,7 +1128,7 @@ static int zeromap_pte_range(struct mm_s
break;
}
page_cache_get(page);
- page_add_file_rmap(page);
+ page_add_file_rmap(page, NULL);
inc_mm_counter(mm, file_rss);
set_pte_at(mm, addr, pte, zero_pte);
} while (pte++, addr += PAGE_SIZE, addr != end);
@@ -1234,7 +1236,7 @@ static int insert_page(struct mm_struct
/* Ok, finally just insert the thing.. */
get_page(page);
inc_mm_counter(mm, file_rss);
- page_add_file_rmap(page);
+ page_add_file_rmap(page, NULL);
set_pte_at(mm, addr, pte, mk_pte(page, prot));

retval = 0;
@@ -1495,6 +1497,7 @@ static int do_wp_page(struct mm_struct *
pte_t entry;
int reuse = 0, ret = VM_FAULT_MINOR;
struct page *dirty_page = NULL;
+ struct page_container *pcont;

old_page = vm_normal_page(vma, address, orig_pte);
if (!old_page)
@@ -1580,6 +1583,9 @@ gotten:
cow_user_page(new_page, old_page, address, vma);
}

+ if (container_rss_prepare(new_page, vma, &pcont))
+ goto oom;
+
/*
* Re-check the pte - we dropped the lock
*/
@@ -1607,12 +1613,14 @@ gotten:
set_pte_at(mm, address, page_table, entry);
update_mmu_cache(vma, address, entry);
lru_cache_add_active(new_page);
- page_add_new_anon_rmap(new_page, vma, address);
+ page_add_new_anon_rmap(new_page, vma, address, pcont);

/* Free the old page.. */
new_page = old_page;
ret |= VM_FAULT_WRITE;
- }
+ } else
+ container_rss_release(pcont);
+
if (new_page)
page_cache_release(new_page);
if (old_page)
@@ -1988,6 +1996,7 @@ static int do_swap_page(struct mm_struct
swp_entry_t entry;
pte_t pte;
int ret = VM_FAULT_MINOR;
+ struct page_container *pcont;

if (!pte_unmap_same(mm, pmd, page_table, orig_pte))
goto out;
@@ -2020,6 +2029,11 @@ static int do_swap_page(struct mm_struct
count_vm_event(PGMAJFAULT);
}

+ if (container_rss_prepare(page, vma, &pcont)) {
+ ret = VM_FAULT_OOM;
+ goto out;
+ }
+
delayacct_clear_flag(DELAYACCT_PF_SWAPIN);
mark_page_accessed(page);
lock_page(page);
@@ -2033,6 +2047,7 @@ static int do_swap_page(struct mm_struct

if (unlikely(!PageUptodate(page))) {
ret = VM_FAULT_SIGBUS;
+ container_rss_release(pcont);
goto out_nomap;
}

@@ -2047,7 +2062,7 @@ static int do_swap_page(struct mm_struct

flush_icache_page(vma, page);
set_pte_at(mm, address, page_table, pte);
- page_add_anon_rmap(page, vma, address);
+ page_add_anon_rmap(page, vma, address, pcont);

swap_free(entry);
if (vm_swap_full())
@@ -2069,6 +2084,7 @@ unlock:
out:
return ret;
out_nomap:
+ container_rss_release(pcont);
pte_unmap_unlock(page_table, ptl);
unlock_page(page);
page_cache_release(page);
@@ -2087,6 +2103,7 @@ static int do_anonymous_page(struct mm_s
struct page *page;
spinlock_t *ptl;
pte_t entry;
+ struct page_container *pcont;

if (write_access) {
/* Allocate our own private page. */
@@ -2098,15 +2115,19 @@ static int do_anonymous_page(struct mm_s
if (!page)
goto oom;

+ if (container_rss_prepare(page, vma, &pcont))
+ goto oom_release;
+
entry = mk_pte(page, vma->vm_page_prot);
entry = maybe_mkwrite(pte_mkdirty(entry), vma);

page_table = pte_offset_map_lock(mm, pmd, address, &ptl);
if (!pte_none(*page_table))
- goto release;
+ goto release_container;
+
inc_mm_counter(mm, anon_rss);
lru_cache_add_active(page);
- page_add_new_anon_rmap(page, vma, address);
+ page_add_new_anon_rmap(page, vma, address, pcont);
} else {
/* Map the ZERO_PAGE - vm_page_prot is readonly */
page = ZERO_PAGE(address);
@@ -2118,7 +2139,7 @@ static int do_anonymous_page(struct mm_s
if (!pte_none(*page_table))
goto release;
inc_mm_counter(mm, file_rss);
- page_add_file_rmap(page);
+ page_add_file_rmap(page, NULL);
}

set_pte_at(mm, address, page_table, entry);
@@ -2129,9 +2150,14 @@ static int do_anonymous_page(struct mm_s
unlock:
pte_unmap_unlock(page_table, ptl);
return VM_FAULT_MINOR;
+release_container:
+ container_rss_release(pcont);
release:
page_cache_release(page);
goto unlock;
+
+oom_release:
+ page_cache_release(page);
oom:
return VM_FAULT_OOM;
}
@@ -2161,6 +2187,7 @@ static int do_no_page(struct mm_struct *
int ret = VM_FAULT_MINOR;
int anon = 0;
struct page *dirty_page = NULL;
+ struct page_container *pcont;

pte_unmap(page_table);
BUG_ON(vma->vm_flags & VM_PFNMAP);
@@ -2218,6 +2245,9 @@ retry:
}
}

+ if (container_rss_prepare(new_page, vma, &pcont))
+ goto oom;
+
page_table = pte_offset_map_lock(mm, pmd, address, &ptl);
/*
* For a file-backed vma, someone could have truncated or otherwise
@@ -2226,6 +2256,7 @@ retry:
*/
if (mapping && unlikely(sequence != mapping->truncate_count)) {
pte_unmap_unlock(page_table, ptl);
+ container_rss_release(pcont);
page_cache_release(new_page);
cond_resched();
sequence = mapping->truncate_count;
@@ -2253,10 +2284,10 @@ retry:
if (anon) {
inc_mm_counter(mm, anon_rss);
lru_cache_add_active(new_page);
- page_add_new_anon_rmap(new_page, vma, address);
+ page_add_new_anon_rmap(new_page, vma, address, pcont);
} else {
inc_mm_counter(mm, file_rss);
- page_add_file_rmap(new_page);
+ page_add_file_rmap(new_page, pcont);
if (write_access) {
dirty_page = new_page;
get_page(dirty_page);
@@ -2264,6 +2295,7 @@ retry:
}
} else {
/* One of our sibling threads was faster, back out. */
+ container_rss_release(pcont);
page_cache_release(new_page);
goto unlock;
}
diff -upr linux-2.6.20.orig/mm/migrate.c linux-2.6.20-0/mm/migrate.c
--- linux-2.6.20.orig/mm/migrate.c 2007-02-04 21:44:54.000000000 +0300
+++ linux-2.6.20-0/mm/migrate.c 2007-03-06 13:33:28.000000000 +0300
@@ -134,6 +134,7 @@ static void remove_migration_pte(struct
pte_t *ptep, pte;
spinlock_t *ptl;
unsigned long addr = page_address_in_vma(new, vma);
+ struct page_container *pcont;

if (addr == -EFAULT)
return;
@@ ...

[ Show the rest of the message ]

Report message to a moderator

[Message index]
 
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Read Message
Previous Topic: Re: [ckrm-tech] [PATCH 7/7] containers (V7): Container interface to nsproxy subsystem
Next Topic: Linux-VServer example results for sharing vs. separate mappings ...
Goto Forum:
  

[ Syndicate this forum (XML) ] [ RSS ] [ PDF ]

Current Time: Thu Jul 11 15:38:31 GMT 2024

Total time taken to generate the page: 0.02428 seconds
Powered by FUDforum Powered by Virtuozzo Containers