Patchwork [2/2] migration: do not overwrite zero pages

login
register
mail settings
Submitter Peter Lieven
Date June 6, 2013, 8:36 a.m.
Message ID <1370507765-2806-3-git-send-email-pl@kamp.de>
Download mbox | patch
Permalink /patch/249335/
State New
Headers show

Comments

Peter Lieven - June 6, 2013, 8:36 a.m.
on incoming migration do not memset pages to zero if they already read as zero.
this will allocate a new zero page and consume memory unnecessarily. even
if we madvise a MADV_DONTNEED later this will only deallocate the memory
asynchronously.

Signed-off-by: Peter Lieven <pl@kamp.de>
---
 arch_init.c |   14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)
Paolo Bonzini - June 19, 2013, 10:10 a.m.
Il 06/06/2013 10:36, Peter Lieven ha scritto:
> on incoming migration do not memset pages to zero if they already read as zero.
> this will allocate a new zero page and consume memory unnecessarily. even
> if we madvise a MADV_DONTNEED later this will only deallocate the memory
> asynchronously.
> 
> Signed-off-by: Peter Lieven <pl@kamp.de>
> ---
>  arch_init.c |   14 ++++++++------
>  1 file changed, 8 insertions(+), 6 deletions(-)
> 
> diff --git a/arch_init.c b/arch_init.c
> index 283081d..642f241 100644
> --- a/arch_init.c
> +++ b/arch_init.c
> @@ -821,14 +821,16 @@ static int ram_load(QEMUFile *f, void *opaque, int version_id)
>              }
>  
>              ch = qemu_get_byte(f);
> -            memset(host, ch, TARGET_PAGE_SIZE);
> +            if (ch != 0 || !is_zero_page(host)) {
> +                memset(host, ch, TARGET_PAGE_SIZE);
>  #ifndef _WIN32
> -            if (ch == 0 &&
> -                (!kvm_enabled() || kvm_has_sync_mmu()) &&
> -                getpagesize() <= TARGET_PAGE_SIZE) {
> -                qemu_madvise(host, TARGET_PAGE_SIZE, QEMU_MADV_DONTNEED);
> -            }
> +                if (ch == 0 &&
> +                    (!kvm_enabled() || kvm_has_sync_mmu()) &&
> +                    getpagesize() <= TARGET_PAGE_SIZE) {
> +                    qemu_madvise(host, TARGET_PAGE_SIZE, QEMU_MADV_DONTNEED);
> +                }
>  #endif
> +            }
>          } else if (flags & RAM_SAVE_FLAG_PAGE) {
>              void *host;
>  
> 

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Alexey Kardashevskiy - June 24, 2013, 2:59 p.m.
On 06/19/2013 08:10 PM, Paolo Bonzini wrote:
> Il 06/06/2013 10:36, Peter Lieven ha scritto:
>> on incoming migration do not memset pages to zero if they already read as zero.
>> this will allocate a new zero page and consume memory unnecessarily. even
>> if we madvise a MADV_DONTNEED later this will only deallocate the memory
>> asynchronously.
>>
>> Signed-off-by: Peter Lieven <pl@kamp.de>
>> ---
>>  arch_init.c |   14 ++++++++------
>>  1 file changed, 8 insertions(+), 6 deletions(-)
>>
>> diff --git a/arch_init.c b/arch_init.c
>> index 283081d..642f241 100644
>> --- a/arch_init.c
>> +++ b/arch_init.c
>> @@ -821,14 +821,16 @@ static int ram_load(QEMUFile *f, void *opaque, int version_id)
>>              }
>>  
>>              ch = qemu_get_byte(f);
>> -            memset(host, ch, TARGET_PAGE_SIZE);
>> +            if (ch != 0 || !is_zero_page(host)) {
>> +                memset(host, ch, TARGET_PAGE_SIZE);
>>  #ifndef _WIN32
>> -            if (ch == 0 &&
>> -                (!kvm_enabled() || kvm_has_sync_mmu()) &&
>> -                getpagesize() <= TARGET_PAGE_SIZE) {
>> -                qemu_madvise(host, TARGET_PAGE_SIZE, QEMU_MADV_DONTNEED);
>> -            }
>> +                if (ch == 0 &&
>> +                    (!kvm_enabled() || kvm_has_sync_mmu()) &&
>> +                    getpagesize() <= TARGET_PAGE_SIZE) {
>> +                    qemu_madvise(host, TARGET_PAGE_SIZE, QEMU_MADV_DONTNEED);
>> +                }
>>  #endif
>> +            }
>>          } else if (flags & RAM_SAVE_FLAG_PAGE) {
>>              void *host;
>>  
>>
> 
> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>

Is it expected to reach upstream any time soon? Thanks.

Patch

diff --git a/arch_init.c b/arch_init.c
index 283081d..642f241 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -821,14 +821,16 @@  static int ram_load(QEMUFile *f, void *opaque, int version_id)
             }
 
             ch = qemu_get_byte(f);
-            memset(host, ch, TARGET_PAGE_SIZE);
+            if (ch != 0 || !is_zero_page(host)) {
+                memset(host, ch, TARGET_PAGE_SIZE);
 #ifndef _WIN32
-            if (ch == 0 &&
-                (!kvm_enabled() || kvm_has_sync_mmu()) &&
-                getpagesize() <= TARGET_PAGE_SIZE) {
-                qemu_madvise(host, TARGET_PAGE_SIZE, QEMU_MADV_DONTNEED);
-            }
+                if (ch == 0 &&
+                    (!kvm_enabled() || kvm_has_sync_mmu()) &&
+                    getpagesize() <= TARGET_PAGE_SIZE) {
+                    qemu_madvise(host, TARGET_PAGE_SIZE, QEMU_MADV_DONTNEED);
+                }
 #endif
+            }
         } else if (flags & RAM_SAVE_FLAG_PAGE) {
             void *host;