| Message ID | 20201207220949.830352-1-yuzhao@google.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | mm: lru related cleanups | expand |
Hi Yu, btw, after this patchset, to do cacheline alignment on each of lru lists are possible, so did you try that to see performance changes? Thanks Alex 在 2020/12/8 上午6:09, Yu Zhao 写道: > The cleanups are intended to reduce the verbosity in lru list > operations and make them less error-prone. A typical example > would be how the patches change __activate_page(): > > static void __activate_page(struct page *page, struct lruvec *lruvec) > { > if (!PageActive(page) && !PageUnevictable(page)) { > - int lru = page_lru_base_type(page); > int nr_pages = thp_nr_pages(page); > > - del_page_from_lru_list(page, lruvec, lru); > + del_page_from_lru_list(page, lruvec); > SetPageActive(page); > - lru += LRU_ACTIVE; > - add_page_to_lru_list(page, lruvec, lru); > + add_page_to_lru_list(page, lruvec); > trace_mm_lru_activate(page); > > There are a few more places like __activate_page() and they are > unnecessarily repetitive in terms of figuring out which list a page > should be added onto or deleted from. And with the duplicated code > removed, they are easier to read, IMO. > > Patch 1 to 5 basically cover the above. Patch 6 and 7 make code more > robust by improving bug reporting. Patch 8, 9 and 10 take care of > some dangling helpers left in header files. Patch 11 isn't strictly a > clean-up patch, but it seems still relevant to include it here. > > Yu Zhao (11): > mm: use add_page_to_lru_list() > mm: shuffle lru list addition and deletion functions > mm: don't pass "enum lru_list" to lru list addition functions > mm: don't pass "enum lru_list" to trace_mm_lru_insertion() > mm: don't pass "enum lru_list" to del_page_from_lru_list() > mm: add __clear_page_lru_flags() to replace page_off_lru() > mm: VM_BUG_ON lru page flags > mm: fold page_lru_base_type() into its sole caller > mm: fold __update_lru_size() into its sole caller > mm: make lruvec_lru_size() static > mm: enlarge the "int nr_pages" parameter of update_lru_size() > > include/linux/memcontrol.h | 10 +-- > include/linux/mm_inline.h | 115 ++++++++++++++------------------- > include/linux/mmzone.h | 2 - > include/linux/vmstat.h | 6 +- > include/trace/events/pagemap.h | 11 ++-- > mm/compaction.c | 2 +- > mm/memcontrol.c | 10 +-- > mm/mlock.c | 3 +- > mm/swap.c | 50 ++++++-------- > mm/vmscan.c | 21 ++---- > 10 files changed, 91 insertions(+), 139 deletions(-) >
On Thu, Dec 10, 2020 at 05:28:08PM +0800, Alex Shi wrote: > Hi Yu, > > btw, after this patchset, to do cacheline alignment on each of lru lists > are possible, so did you try that to see performance changes? I ran a Chrome-based performance benchmark without memcg and with one memcg many times. The good news is I didn't see any regressions for these basic cases. But I can't say how much improvements there would be with hundreds of memcgs.
The cleanups are intended to reduce the verbosity in lru list operations and make them less error-prone. A typical example would be how the patches change __activate_page(): static void __activate_page(struct page *page, struct lruvec *lruvec) { if (!PageActive(page) && !PageUnevictable(page)) { - int lru = page_lru_base_type(page); int nr_pages = thp_nr_pages(page); - del_page_from_lru_list(page, lruvec, lru); + del_page_from_lru_list(page, lruvec); SetPageActive(page); - lru += LRU_ACTIVE; - add_page_to_lru_list(page, lruvec, lru); + add_page_to_lru_list(page, lruvec); trace_mm_lru_activate(page); There are a few more places like __activate_page() and they are unnecessarily repetitive in terms of figuring out which list a page should be added onto or deleted from. And with the duplicated code removed, they are easier to read, IMO. Patch 1 to 5 basically cover the above. Patch 6 and 7 make code more robust by improving bug reporting. Patch 8, 9 and 10 take care of some dangling helpers left in header files. Patch 11 isn't strictly a clean-up patch, but it seems still relevant to include it here. Yu Zhao (11): mm: use add_page_to_lru_list() mm: shuffle lru list addition and deletion functions mm: don't pass "enum lru_list" to lru list addition functions mm: don't pass "enum lru_list" to trace_mm_lru_insertion() mm: don't pass "enum lru_list" to del_page_from_lru_list() mm: add __clear_page_lru_flags() to replace page_off_lru() mm: VM_BUG_ON lru page flags mm: fold page_lru_base_type() into its sole caller mm: fold __update_lru_size() into its sole caller mm: make lruvec_lru_size() static mm: enlarge the "int nr_pages" parameter of update_lru_size() include/linux/memcontrol.h | 10 +-- include/linux/mm_inline.h | 115 ++++++++++++++------------------- include/linux/mmzone.h | 2 - include/linux/vmstat.h | 6 +- include/trace/events/pagemap.h | 11 ++-- mm/compaction.c | 2 +- mm/memcontrol.c | 10 +-- mm/mlock.c | 3 +- mm/swap.c | 50 ++++++-------- mm/vmscan.c | 21 ++---- 10 files changed, 91 insertions(+), 139 deletions(-)