* Vlastimil Babka <vbabka@xxxxxxx> [220121 06:52]:
> On 12/1/21 15:30, Liam Howlett wrote:
> > From: "Liam R. Howlett" <Liam.Howlett@xxxxxxxxxx>
> >
> > Remove references to mm_struct linked list and highest_vm_end for when
> > they are removed
> >
> > Signed-off-by: Liam R. Howlett <Liam.Howlett@xxxxxxxxxx>
> > ---
> > fs/proc/internal.h | 2 +-
> > fs/proc/task_mmu.c | 73 ++++++++++++++++++++++++++--------------------
> > 2 files changed, 42 insertions(+), 33 deletions(-)
> >
> > diff --git a/fs/proc/internal.h b/fs/proc/internal.h
> > index 03415f3fb3a8..45b132c609ff 100644
> > --- a/fs/proc/internal.h
> > +++ b/fs/proc/internal.h
> > @@ -290,7 +290,7 @@ struct proc_maps_private {
> > struct task_struct *task;
> > struct mm_struct *mm;
> > #ifdef CONFIG_MMU
> > - struct vm_area_struct *tail_vma;
> > + struct vma_iterator iter;
> > #endif
> > #ifdef CONFIG_NUMA
> > struct mempolicy *task_mempolicy;
> > diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c
> > index 300911d6575f..7cc97cdb88c2 100644
> > --- a/fs/proc/task_mmu.c
> > +++ b/fs/proc/task_mmu.c
> > @@ -122,12 +122,26 @@ static void release_task_mempolicy(struct proc_maps_private *priv)
> > }
> > #endif
> >
> > +static struct vm_area_struct *proc_get_vma(struct proc_maps_private *priv,
> > + loff_t *ppos)
> > +{
> > + struct vm_area_struct *vma = vma_next(&priv->iter);
>
> This advances the iterator.
Maybe. vma_next() will call vma_find() which calls mas_find().
mas_find() will return the VMA at that address (or the next VMA) on the
first call.
>
> > +
> > + if (vma) {
> > + *ppos = vma->vm_start;
>
> This advances *ppos.
If the vma_next() returned the next vma, yes. If it returned the one at
vmi->mas->index, then no.
>
> > + } else {
> > + *ppos = -2UL;
> > + vma = get_gate_vma(priv->mm);
> > + }
> > +
> > + return vma;
> > +}
> > +
> > static void *m_start(struct seq_file *m, loff_t *ppos)
> > {
> > struct proc_maps_private *priv = m->private;
> > unsigned long last_addr = *ppos;
> > struct mm_struct *mm;
> > - struct vm_area_struct *vma;
> >
> > /* See m_next(). Zero at the start or after lseek. */
> > if (last_addr == -1UL)
> > @@ -151,31 +165,21 @@ static void *m_start(struct seq_file *m, loff_t *ppos)
> > return ERR_PTR(-EINTR);
> > }
> >
> > + vma_iter_init(&priv->iter, mm, last_addr);
> > hold_task_mempolicy(priv);
> > - priv->tail_vma = get_gate_vma(mm);
> > -
> > - vma = find_vma(mm, last_addr);
> > - if (vma)
> > - return vma;
> > + if (last_addr == -2UL)
> > + return get_gate_vma(mm);
> >
> > - return priv->tail_vma;
> > + return proc_get_vma(priv, ppos);
>
> So here we advance those as part of m_start(), which I think is wrong in the
> seqfile API. See seq_read_iter() in fs/seq_file.c how it handles a full
> buffer, around the comment "// need a bigger buffer" it will do a stop() and
> start() again and that's supposed to get the same vma.
> seqfile is tricky, part #220121
I've built, booted, and tested this as follows:
root@dev0:~# dd if=/proc/1/maps of=one bs=512; dd if=/proc/1/maps of=two bs=4096; cmp one two
23+1 records in
23+1 records out
12188 bytes (12 kB, 12 KiB) copied, 0.000114377 s, 107 MB/s
0+3 records in
0+3 records out
12188 bytes (12 kB, 12 KiB) copied, 6.9177e-05 s, 176 MB/s
root@dev0:~# dd if=/proc/1/maps of=one bs=3; dd if=/proc/1/maps of=two bs=10; cmp one two
4062+1 records in
4062+1 records out
12188 bytes (12 kB, 12 KiB) copied, 0.0184962 s, 659 kB/s
1218+1 records in
1218+1 records out
12188 bytes (12 kB, 12 KiB) copied, 0.0062038 s, 2.0 MB/s
root@dev0:~#
>
> > }
> >
> > static void *m_next(struct seq_file *m, void *v, loff_t *ppos)
> > {
> > - struct proc_maps_private *priv = m->private;
> > - struct vm_area_struct *next, *vma = v;
> > -
> > - if (vma == priv->tail_vma)
> > - next = NULL;
> > - else if (vma->vm_next)
> > - next = vma->vm_next;
> > - else
> > - next = priv->tail_vma;
> > -
> > - *ppos = next ? next->vm_start : -1UL;
> > -
> > - return next;
> > + if (*ppos == -2UL) {
> > + *ppos = -1UL;
> > + return NULL;
> > + }
> > + return proc_get_vma(m->private, ppos);
> > }
> >
> > static void m_stop(struct seq_file *m, void *v)
> > @@ -843,16 +847,16 @@ static int show_smaps_rollup(struct seq_file *m, void *v)
> > {
> > struct proc_maps_private *priv = m->private;
> > struct mem_size_stats mss;
> > - struct mm_struct *mm;
> > + struct mm_struct *mm = priv->mm;
> > struct vm_area_struct *vma;
> > - unsigned long last_vma_end = 0;
> > + unsigned long vma_start = 0, last_vma_end = 0;
> > int ret = 0;
> > + MA_STATE(mas, &mm->mm_mt, 0, 0);
> >
> > priv->task = get_proc_task(priv->inode);
> > if (!priv->task)
> > return -ESRCH;
> >
> > - mm = priv->mm;
> > if (!mm || !mmget_not_zero(mm)) {
> > ret = -ESRCH;
> > goto out_put_task;
> > @@ -865,8 +869,13 @@ static int show_smaps_rollup(struct seq_file *m, void *v)
> > goto out_put_mm;
> >
> > hold_task_mempolicy(priv);
> > + vma = mas_find(&mas, 0);
> > +
> > + if (unlikely(!vma))
> > + goto empty_set;
> >
> > - for (vma = priv->mm->mmap; vma;) {
> > + vma_start = vma->vm_start;
> > + do {
> > smap_gather_stats(vma, &mss, 0);
> > last_vma_end = vma->vm_end;
> >
> > @@ -875,6 +884,7 @@ static int show_smaps_rollup(struct seq_file *m, void *v)
> > * access it for write request.
> > */
> > if (mmap_lock_is_contended(mm)) {
> > + mas_pause(&mas);
> > mmap_read_unlock(mm);
> > ret = mmap_read_lock_killable(mm);
> > if (ret) {
> > @@ -918,7 +928,7 @@ static int show_smaps_rollup(struct seq_file *m, void *v)
> > * contains last_vma_end.
> > * Iterate VMA' from last_vma_end.
> > */
> > - vma = find_vma(mm, last_vma_end - 1);
> > + vma = mas_find(&mas, ULONG_MAX);
> > /* Case 3 above */
> > if (!vma)
> > break;
> > @@ -932,11 +942,10 @@ static int show_smaps_rollup(struct seq_file *m, void *v)
> > smap_gather_stats(vma, &mss, last_vma_end);
> > }
> > /* Case 2 above */
> > - vma = vma->vm_next;
> > - }
> > + } while ((vma = mas_find(&mas, ULONG_MAX)) != NULL);
> >
> > - show_vma_header_prefix(m, priv->mm->mmap->vm_start,
> > - last_vma_end, 0, 0, 0, 0);
> > +empty_set:
> > + show_vma_header_prefix(m, vma_start, last_vma_end, 0, 0, 0, 0);
> > seq_pad(m, ' ');
> > seq_puts(m, "[rollup]\n");
> >
> > @@ -1229,6 +1238,7 @@ static ssize_t clear_refs_write(struct file *file, const char __user *buf,
> > return -ESRCH;
> > mm = get_task_mm(task);
> > if (mm) {
> > + MA_STATE(mas, &mm->mm_mt, 0, 0);
> > struct mmu_notifier_range range;
> > struct clear_refs_private cp = {
> > .type = type,
> > @@ -1248,7 +1258,7 @@ static ssize_t clear_refs_write(struct file *file, const char __user *buf,
> > }
> >
> > if (type == CLEAR_REFS_SOFT_DIRTY) {
> > - for (vma = mm->mmap; vma; vma = vma->vm_next) {
> > + mas_for_each(&mas, vma, ULONG_MAX) {
> > if (!(vma->vm_flags & VM_SOFTDIRTY))
> > continue;
> > vma->vm_flags &= ~VM_SOFTDIRTY;
> > @@ -1260,8 +1270,7 @@ static ssize_t clear_refs_write(struct file *file, const char __user *buf,
> > 0, NULL, mm, 0, -1UL);
> > mmu_notifier_invalidate_range_start(&range);
> > }
> > - walk_page_range(mm, 0, mm->highest_vm_end, &clear_refs_walk_ops,
> > - &cp);
> > + walk_page_range(mm, 0, -1, &clear_refs_walk_ops, &cp);
> > if (type == CLEAR_REFS_SOFT_DIRTY) {
> > mmu_notifier_invalidate_range_end(&range);
> > flush_tlb_mm(mm);
>
