| Message ID | 20210204035043.36609-9-songmuchun@bytedance.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Free some vmemmap pages of HugeTLB page | expand |
On 2021/2/4 11:50, Muchun Song wrote: > We cannot optimize if a "struct page" crosses page boundaries. If > it is true, we can optimize the code with the help of a compiler. > When free_vmemmap_pages_per_hpage() returns zero, most functions are > optimized by the compiler. > > Signed-off-by: Muchun Song <songmuchun@bytedance.com> I like it. Thanks. Reviewed-by: Miaohe Lin <linmiaohe@huawei.com> > --- > include/linux/hugetlb.h | 3 ++- > mm/hugetlb_vmemmap.c | 13 +++++++++++++ > 2 files changed, 15 insertions(+), 1 deletion(-) > > diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h > index 822ab2f5542a..7bfb06e16298 100644 > --- a/include/linux/hugetlb.h > +++ b/include/linux/hugetlb.h > @@ -878,7 +878,8 @@ extern bool hugetlb_free_vmemmap_enabled; > > static inline bool is_hugetlb_free_vmemmap_enabled(void) > { > - return hugetlb_free_vmemmap_enabled; > + return hugetlb_free_vmemmap_enabled && > + is_power_of_2(sizeof(struct page)); > } > #else > static inline bool is_hugetlb_free_vmemmap_enabled(void) > diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c > index 8efad9978821..068d0e0cebc8 100644 > --- a/mm/hugetlb_vmemmap.c > +++ b/mm/hugetlb_vmemmap.c > @@ -211,6 +211,12 @@ early_param("hugetlb_free_vmemmap", early_hugetlb_free_vmemmap_param); > */ > static inline unsigned int free_vmemmap_pages_per_hpage(struct hstate *h) > { > + /* > + * This check aims to let the compiler help us optimize the code as > + * much as possible. > + */ > + if (!is_power_of_2(sizeof(struct page))) > + return 0; > return h->nr_free_vmemmap_pages; > } > > @@ -280,6 +286,13 @@ void __init hugetlb_vmemmap_init(struct hstate *h) > BUILD_BUG_ON(NR_USED_SUBPAGE >= > RESERVE_VMEMMAP_SIZE / sizeof(struct page)); > > + /* > + * The compiler can help us to optimize this function to null > + * when the size of the struct page is not power of 2. > + */ > + if (!is_power_of_2(sizeof(struct page))) > + return; > + > if (!hugetlb_free_vmemmap_enabled) > return; > >
On Thu, Feb 04, 2021 at 11:50:43AM +0800, Muchun Song wrote: > We cannot optimize if a "struct page" crosses page boundaries. If > it is true, we can optimize the code with the help of a compiler. > When free_vmemmap_pages_per_hpage() returns zero, most functions are > optimized by the compiler. "When the "struct page size" crosses page boundaries we cannot make use of this feature. Let free_vmemmap_pages_per_hpage() return zero if that is the case, most of the functions can be optimized away." I think the above is more clear, but just a suggestion. > Signed-off-by: Muchun Song <songmuchun@bytedance.com> Reviewed-by: Oscar Salvador <osalvador@suse.de> > --- > include/linux/hugetlb.h | 3 ++- > mm/hugetlb_vmemmap.c | 13 +++++++++++++ > 2 files changed, 15 insertions(+), 1 deletion(-) > > diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h > index 822ab2f5542a..7bfb06e16298 100644 > --- a/include/linux/hugetlb.h > +++ b/include/linux/hugetlb.h > @@ -878,7 +878,8 @@ extern bool hugetlb_free_vmemmap_enabled; > > static inline bool is_hugetlb_free_vmemmap_enabled(void) > { > - return hugetlb_free_vmemmap_enabled; > + return hugetlb_free_vmemmap_enabled && > + is_power_of_2(sizeof(struct page)); > } > #else > static inline bool is_hugetlb_free_vmemmap_enabled(void) > diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c > index 8efad9978821..068d0e0cebc8 100644 > --- a/mm/hugetlb_vmemmap.c > +++ b/mm/hugetlb_vmemmap.c > @@ -211,6 +211,12 @@ early_param("hugetlb_free_vmemmap", early_hugetlb_free_vmemmap_param); > */ > static inline unsigned int free_vmemmap_pages_per_hpage(struct hstate *h) > { > + /* > + * This check aims to let the compiler help us optimize the code as > + * much as possible. > + */ > + if (!is_power_of_2(sizeof(struct page))) > + return 0; > return h->nr_free_vmemmap_pages; > } > > @@ -280,6 +286,13 @@ void __init hugetlb_vmemmap_init(struct hstate *h) > BUILD_BUG_ON(NR_USED_SUBPAGE >= > RESERVE_VMEMMAP_SIZE / sizeof(struct page)); > > + /* > + * The compiler can help us to optimize this function to null > + * when the size of the struct page is not power of 2. > + */ > + if (!is_power_of_2(sizeof(struct page))) > + return; > + > if (!hugetlb_free_vmemmap_enabled) > return; > > -- > 2.11.0 > >
On Fri, Feb 5, 2021 at 5:09 PM Oscar Salvador <osalvador@suse.de> wrote: > > On Thu, Feb 04, 2021 at 11:50:43AM +0800, Muchun Song wrote: > > We cannot optimize if a "struct page" crosses page boundaries. If > > it is true, we can optimize the code with the help of a compiler. > > When free_vmemmap_pages_per_hpage() returns zero, most functions are > > optimized by the compiler. > > "When the "struct page size" crosses page boundaries we cannot > make use of this feature. > Let free_vmemmap_pages_per_hpage() return zero if that is the case, > most of the functions can be optimized away." > > I think the above is more clear, but just a suggestion. OK. I will use it. Thanks for your suggestions. > > > > Signed-off-by: Muchun Song <songmuchun@bytedance.com> > > Reviewed-by: Oscar Salvador <osalvador@suse.de> Thanks. > > > --- > > include/linux/hugetlb.h | 3 ++- > > mm/hugetlb_vmemmap.c | 13 +++++++++++++ > > 2 files changed, 15 insertions(+), 1 deletion(-) > > > > diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h > > index 822ab2f5542a..7bfb06e16298 100644 > > --- a/include/linux/hugetlb.h > > +++ b/include/linux/hugetlb.h > > @@ -878,7 +878,8 @@ extern bool hugetlb_free_vmemmap_enabled; > > > > static inline bool is_hugetlb_free_vmemmap_enabled(void) > > { > > - return hugetlb_free_vmemmap_enabled; > > + return hugetlb_free_vmemmap_enabled && > > + is_power_of_2(sizeof(struct page)); > > } > > #else > > static inline bool is_hugetlb_free_vmemmap_enabled(void) > > diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c > > index 8efad9978821..068d0e0cebc8 100644 > > --- a/mm/hugetlb_vmemmap.c > > +++ b/mm/hugetlb_vmemmap.c > > @@ -211,6 +211,12 @@ early_param("hugetlb_free_vmemmap", early_hugetlb_free_vmemmap_param); > > */ > > static inline unsigned int free_vmemmap_pages_per_hpage(struct hstate *h) > > { > > + /* > > + * This check aims to let the compiler help us optimize the code as > > + * much as possible. > > + */ > > + if (!is_power_of_2(sizeof(struct page))) > > + return 0; > > return h->nr_free_vmemmap_pages; > > } > > > > @@ -280,6 +286,13 @@ void __init hugetlb_vmemmap_init(struct hstate *h) > > BUILD_BUG_ON(NR_USED_SUBPAGE >= > > RESERVE_VMEMMAP_SIZE / sizeof(struct page)); > > > > + /* > > + * The compiler can help us to optimize this function to null > > + * when the size of the struct page is not power of 2. > > + */ > > + if (!is_power_of_2(sizeof(struct page))) > > + return; > > + > > if (!hugetlb_free_vmemmap_enabled) > > return; > > > > -- > > 2.11.0 > > > > > > -- > Oscar Salvador > SUSE L3
diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index 822ab2f5542a..7bfb06e16298 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -878,7 +878,8 @@ extern bool hugetlb_free_vmemmap_enabled; static inline bool is_hugetlb_free_vmemmap_enabled(void) { - return hugetlb_free_vmemmap_enabled; + return hugetlb_free_vmemmap_enabled && + is_power_of_2(sizeof(struct page)); } #else static inline bool is_hugetlb_free_vmemmap_enabled(void) diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c index 8efad9978821..068d0e0cebc8 100644 --- a/mm/hugetlb_vmemmap.c +++ b/mm/hugetlb_vmemmap.c @@ -211,6 +211,12 @@ early_param("hugetlb_free_vmemmap", early_hugetlb_free_vmemmap_param); */ static inline unsigned int free_vmemmap_pages_per_hpage(struct hstate *h) { + /* + * This check aims to let the compiler help us optimize the code as + * much as possible. + */ + if (!is_power_of_2(sizeof(struct page))) + return 0; return h->nr_free_vmemmap_pages; } @@ -280,6 +286,13 @@ void __init hugetlb_vmemmap_init(struct hstate *h) BUILD_BUG_ON(NR_USED_SUBPAGE >= RESERVE_VMEMMAP_SIZE / sizeof(struct page)); + /* + * The compiler can help us to optimize this function to null + * when the size of the struct page is not power of 2. + */ + if (!is_power_of_2(sizeof(struct page))) + return; + if (!hugetlb_free_vmemmap_enabled) return;
We cannot optimize if a "struct page" crosses page boundaries. If it is true, we can optimize the code with the help of a compiler. When free_vmemmap_pages_per_hpage() returns zero, most functions are optimized by the compiler. Signed-off-by: Muchun Song <songmuchun@bytedance.com> --- include/linux/hugetlb.h | 3 ++- mm/hugetlb_vmemmap.c | 13 +++++++++++++ 2 files changed, 15 insertions(+), 1 deletion(-)