* Re: [PATCH 1/3] ksm: add ksm involvement information for each process
[not found] <20240901043525.16831>
@ 2024-09-03 4:27 ` xu xin
2024-09-03 8:04 ` David Hildenbrand
0 siblings, 1 reply; 2+ messages in thread
From: xu xin @ 2024-09-03 4:27 UTC (permalink / raw)
To: david; +Cc: linux-kernel, akpm, shr, hughd, xu.xin16
> > In /proc/<pid>/ksm_stat, Add two extra ksm involvement items including
> > KSM_mergeable and KSM_merge_any. It helps administrators to
> > better know the system's KSM behavior at process level.
> >
> > KSM_mergeable: yes/no
> > whether any VMAs of the process'mm are currently applicable to KSM.
> >
> > KSM_merge_any: yes/no
> > whether the process'mm is added by prctl() into the candidate list
> > of KSM or not, and fully enabled at process level.
> >
> > Signed-off-by: xu xin <xu.xin16@zte.com.cn>
> > ---
> > Changelog
> > =========
> > v2 -> v3:
> > Update the KSM_mergeable getting method: loop up if any vma is
> > mergeable to KSM.
> > https://lore.kernel.org/all/bc0e1cdd-2d9d-437c-8fc9-4df0e13c48c0@redhat.com/
> >
> > v1 -> v2:
> > replace the internal flag names with straightforward strings.
> > * MMF_VM_MERGEABLE -> KSM_mergeable
> > * MMF_VM_MERGE_ANY -> KSM_merge_any
> >
> > fs/proc/base.c | 4 ++++
> > include/linux/ksm.h | 1 +
> > mm/ksm.c | 16 ++++++++++++++++
> > 3 files changed, 21 insertions(+)
> >
> > diff --git a/fs/proc/base.c b/fs/proc/base.c
> > index 18550c071d71..45e12560e698 100644
> > --- a/fs/proc/base.c
> > +++ b/fs/proc/base.c
> > @@ -3217,6 +3217,10 @@ static int proc_pid_ksm_stat(struct seq_file *m, struct pid_namespace *ns,
> > seq_printf(m, "ksm_zero_pages %lu\n", mm->ksm_zero_pages);
> > seq_printf(m, "ksm_merging_pages %lu\n", mm->ksm_merging_pages);
> > seq_printf(m, "ksm_process_profit %ld\n", ksm_process_profit(mm));
> > + seq_printf(m, "KSM_mergeable: %s\n",
> > + ksm_process_mergeable(mm) ? "yes" : "no");
> > + seq_printf(m, "KSM_merge_any: %s\n",
> > + test_bit(MMF_VM_MERGE_ANY, &mm->flags) ? "yes" : "no");
>
> Inconsistent "KSM" casing .
Excuse me, could you be more specific? I didn't get it
>
> > mmput(mm);
> > }
> >
> > diff --git a/include/linux/ksm.h b/include/linux/ksm.h
> > index 52c63a9c5a9c..5286b84964d3 100644
> > --- a/include/linux/ksm.h
> > +++ b/include/linux/ksm.h
> > @@ -84,6 +84,7 @@ void folio_migrate_ksm(struct folio *newfolio, struct folio *folio);
> > void collect_procs_ksm(struct folio *folio, struct page *page,
> > struct list_head *to_kill, int force_early);
> > long ksm_process_profit(struct mm_struct *);
> > +bool ksm_process_mergeable(struct mm_struct *mm);
> >
> > #else /* !CONFIG_KSM */
> >
> > diff --git a/mm/ksm.c b/mm/ksm.c
> > index f5138f43f0d2..6647f2ef27ca 100644
> > --- a/mm/ksm.c
> > +++ b/mm/ksm.c
> > @@ -3373,6 +3373,22 @@ static void wait_while_offlining(void)
> > #endif /* CONFIG_MEMORY_HOTREMOVE */
> >
> > #ifdef CONFIG_PROC_FS
> > +/*
> > + * The process is mergeable only if any VMA (and which) is currently
>
> confusing "(and which)"
Oh yes, I'll correct it
>
> > + * applicable to KSM.
> > + */
> > +bool ksm_process_mergeable(struct mm_struct *mm)
> > +{
> > + struct vm_area_struct *vma;
> > +
> > + VMA_ITERATOR(vmi, mm, 0);
> > + for_each_vma(vmi, vma)
> > + if (vma->vm_flags & VM_MERGEABLE)
> > + return true;
> > +
>
> Are we holding the mmap lock here? I only see a mmput() in
> proc_pid_ksm_stat() above, which might mean that we are not holding the
> mmap lock.
No, we'are missing the read lock of mmap. I'll fix it.
Thanks a lot.
^ permalink raw reply [flat|nested] 2+ messages in thread