| Message ID | 20191004160632.30251-1-richardw.yang@linux.intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | mm/rmap.c: reuse mergeable anon_vma as parent when fork | expand |
On Sat, Oct 05, 2019 at 12:06:32AM +0800, Wei Yang wrote:
> After this change, kernel build test reduces 20% anon_vma allocation.
But does it have any effect on elapsed time or peak memory consumption?
On 04/10/2019 19.06, Wei Yang wrote: > In function __anon_vma_prepare(), we will try to find anon_vma if it is > possible to reuse it. While on fork, the logic is different. > > Since commit 5beb49305251 ("mm: change anon_vma linking to fix > multi-process server scalability issue"), function anon_vma_clone() > tries to allocate new anon_vma for child process. But the logic here > will allocate a new anon_vma for each vma, even in parent this vma > is mergeable and share the same anon_vma with its sibling. This may do > better for scalability issue, while it is not necessary to do so > especially after interval tree is used. > > Commit 7a3ef208e662 ("mm: prevent endless growth of anon_vma hierarchy") > tries to reuse some anon_vma by counting child anon_vma and attached > vmas. While for those mergeable anon_vmas, we can just reuse it and not > necessary to go through the logic. > > After this change, kernel build test reduces 20% anon_vma allocation. > Makes sense. This might have much bigger effect for scenarios when task unmaps holes in huge vma as red-zones between allocations and then forks. Acked-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru> > Signed-off-by: Wei Yang <richardw.yang@linux.intel.com> > --- > mm/rmap.c | 11 +++++++++++ > 1 file changed, 11 insertions(+) > > diff --git a/mm/rmap.c b/mm/rmap.c > index d9a23bb773bf..12f6c3d7fd9d 100644 > --- a/mm/rmap.c > +++ b/mm/rmap.c > @@ -262,6 +262,17 @@ int anon_vma_clone(struct vm_area_struct *dst, struct vm_area_struct *src) > { > struct anon_vma_chain *avc, *pavc; > struct anon_vma *root = NULL; > + struct vm_area_struct *prev = dst->vm_prev, *pprev = src->vm_prev; > + > + /* > + * If parent share anon_vma with its vm_prev, keep this sharing in in > + * child. > + * > + * 1. Parent has vm_prev, which implies we have vm_prev. > + * 2. Parent and its vm_prev have the same anon_vma. > + */ > + if (pprev && pprev->anon_vma == src->anon_vma) > + dst->anon_vma = prev->anon_vma; > > list_for_each_entry_reverse(pavc, &src->anon_vma_chain, same_vma) { > struct anon_vma *anon_vma; >
On Fri, Oct 04, 2019 at 09:11:20AM -0700, Matthew Wilcox wrote: >On Sat, Oct 05, 2019 at 12:06:32AM +0800, Wei Yang wrote: >> After this change, kernel build test reduces 20% anon_vma allocation. > >But does it have any effect on elapsed time or peak memory consumption? I didn't evaluate these aspects. BTW, how to evaluate peak memory consumption? This looks a transient value.
On Fri, Oct 04, 2019 at 07:33:53PM +0300, Konstantin Khlebnikov wrote: >On 04/10/2019 19.06, Wei Yang wrote: >> In function __anon_vma_prepare(), we will try to find anon_vma if it is >> possible to reuse it. While on fork, the logic is different. >> >> Since commit 5beb49305251 ("mm: change anon_vma linking to fix >> multi-process server scalability issue"), function anon_vma_clone() >> tries to allocate new anon_vma for child process. But the logic here >> will allocate a new anon_vma for each vma, even in parent this vma >> is mergeable and share the same anon_vma with its sibling. This may do >> better for scalability issue, while it is not necessary to do so >> especially after interval tree is used. >> >> Commit 7a3ef208e662 ("mm: prevent endless growth of anon_vma hierarchy") >> tries to reuse some anon_vma by counting child anon_vma and attached >> vmas. While for those mergeable anon_vmas, we can just reuse it and not >> necessary to go through the logic. >> >> After this change, kernel build test reduces 20% anon_vma allocation. >> > >Makes sense. This might have much bigger effect for scenarios when task >unmaps holes in huge vma as red-zones between allocations and then forks. > >Acked-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru> > Thanks
On Sat, 2019-10-05 at 00:06 +0800, Wei Yang wrote: > In function __anon_vma_prepare(), we will try to find anon_vma if it > is > possible to reuse it. While on fork, the logic is different. > > Since commit 5beb49305251 ("mm: change anon_vma linking to fix > multi-process server scalability issue"), function anon_vma_clone() > tries to allocate new anon_vma for child process. But the logic here > will allocate a new anon_vma for each vma, even in parent this vma > is mergeable and share the same anon_vma with its sibling. This may > do > better for scalability issue, while it is not necessary to do so > especially after interval tree is used. > > Commit 7a3ef208e662 ("mm: prevent endless growth of anon_vma > hierarchy") > tries to reuse some anon_vma by counting child anon_vma and attached > vmas. While for those mergeable anon_vmas, we can just reuse it and > not > necessary to go through the logic. > > After this change, kernel build test reduces 20% anon_vma allocation. > > Signed-off-by: Wei Yang <richardw.yang@linux.intel.com> Acked-by: Rik van Riel <riel@surriel.com>
On Fri, Oct 04, 2019 at 09:11:20AM -0700, Matthew Wilcox wrote: >On Sat, Oct 05, 2019 at 12:06:32AM +0800, Wei Yang wrote: >> After this change, kernel build test reduces 20% anon_vma allocation. > >But does it have any effect on elapsed time or peak memory consumption? Do the same kernel build test and record time: Origin real 2m50.467s user 17m52.002s sys 1m51.953s real 2m48.662s user 17m55.464s sys 1m50.553s real 2m51.143s user 17m59.687s sys 1m53.600s Patched real 2m43.733s user 17m25.705s sys 1m41.791s real 2m47.146s user 17m47.451s sys 1m43.474s real 2m45.763s user 17m38.230s sys 1m42.102s For time in sys, it reduced 8.5%.
On Fri, Oct 04, 2019 at 07:45:26PM -0400, Rik van Riel wrote: >On Sat, 2019-10-05 at 00:06 +0800, Wei Yang wrote: >> In function __anon_vma_prepare(), we will try to find anon_vma if it >> is >> possible to reuse it. While on fork, the logic is different. >> >> Since commit 5beb49305251 ("mm: change anon_vma linking to fix >> multi-process server scalability issue"), function anon_vma_clone() >> tries to allocate new anon_vma for child process. But the logic here >> will allocate a new anon_vma for each vma, even in parent this vma >> is mergeable and share the same anon_vma with its sibling. This may >> do >> better for scalability issue, while it is not necessary to do so >> especially after interval tree is used. >> >> Commit 7a3ef208e662 ("mm: prevent endless growth of anon_vma >> hierarchy") >> tries to reuse some anon_vma by counting child anon_vma and attached >> vmas. While for those mergeable anon_vmas, we can just reuse it and >> not >> necessary to go through the logic. >> >> After this change, kernel build test reduces 20% anon_vma allocation. >> >> Signed-off-by: Wei Yang <richardw.yang@linux.intel.com> > >Acked-by: Rik van Riel <riel@surriel.com> > Thanks >-- >All Rights Reversed.
On Sat, Oct 05, 2019 at 07:48:45AM +0800, Wei Yang wrote: > On Fri, Oct 04, 2019 at 09:11:20AM -0700, Matthew Wilcox wrote: > >On Sat, Oct 05, 2019 at 12:06:32AM +0800, Wei Yang wrote: > >> After this change, kernel build test reduces 20% anon_vma allocation. > > > >But does it have any effect on elapsed time or peak memory consumption? > > Do the same kernel build test and record time: > > > Origin > > real 2m50.467s > user 17m52.002s > sys 1m51.953s > > real 2m48.662s > user 17m55.464s > sys 1m50.553s > > real 2m51.143s > user 17m59.687s > sys 1m53.600s > > > Patched > > real 2m43.733s > user 17m25.705s > sys 1m41.791s > > real 2m47.146s > user 17m47.451s > sys 1m43.474s > > real 2m45.763s > user 17m38.230s > sys 1m42.102s > > > For time in sys, it reduced 8.5%. That's compelling!
On Fri, Oct 04, 2019 at 06:15:54PM -0700, Matthew Wilcox wrote: >On Sat, Oct 05, 2019 at 07:48:45AM +0800, Wei Yang wrote: >> On Fri, Oct 04, 2019 at 09:11:20AM -0700, Matthew Wilcox wrote: >> >On Sat, Oct 05, 2019 at 12:06:32AM +0800, Wei Yang wrote: >> >> After this change, kernel build test reduces 20% anon_vma allocation. >> > >> >But does it have any effect on elapsed time or peak memory consumption? >> >> Do the same kernel build test and record time: >> >> >> Origin >> >> real 2m50.467s >> user 17m52.002s >> sys 1m51.953s >> >> real 2m48.662s >> user 17m55.464s >> sys 1m50.553s >> >> real 2m51.143s >> user 17m59.687s >> sys 1m53.600s >> >> >> Patched >> >> real 2m43.733s >> user 17m25.705s >> sys 1m41.791s >> >> real 2m47.146s >> user 17m47.451s >> sys 1m43.474s >> >> real 2m45.763s >> user 17m38.230s >> sys 1m42.102s >> >> >> For time in sys, it reduced 8.5%. > >That's compelling! Yep, looks good :-)
diff --git a/mm/rmap.c b/mm/rmap.c index d9a23bb773bf..12f6c3d7fd9d 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -262,6 +262,17 @@ int anon_vma_clone(struct vm_area_struct *dst, struct vm_area_struct *src) { struct anon_vma_chain *avc, *pavc; struct anon_vma *root = NULL; + struct vm_area_struct *prev = dst->vm_prev, *pprev = src->vm_prev; + + /* + * If parent share anon_vma with its vm_prev, keep this sharing in in + * child. + * + * 1. Parent has vm_prev, which implies we have vm_prev. + * 2. Parent and its vm_prev have the same anon_vma. + */ + if (pprev && pprev->anon_vma == src->anon_vma) + dst->anon_vma = prev->anon_vma; list_for_each_entry_reverse(pavc, &src->anon_vma_chain, same_vma) { struct anon_vma *anon_vma;
In function __anon_vma_prepare(), we will try to find anon_vma if it is possible to reuse it. While on fork, the logic is different. Since commit 5beb49305251 ("mm: change anon_vma linking to fix multi-process server scalability issue"), function anon_vma_clone() tries to allocate new anon_vma for child process. But the logic here will allocate a new anon_vma for each vma, even in parent this vma is mergeable and share the same anon_vma with its sibling. This may do better for scalability issue, while it is not necessary to do so especially after interval tree is used. Commit 7a3ef208e662 ("mm: prevent endless growth of anon_vma hierarchy") tries to reuse some anon_vma by counting child anon_vma and attached vmas. While for those mergeable anon_vmas, we can just reuse it and not necessary to go through the logic. After this change, kernel build test reduces 20% anon_vma allocation. Signed-off-by: Wei Yang <richardw.yang@linux.intel.com> --- mm/rmap.c | 11 +++++++++++ 1 file changed, 11 insertions(+)