Re: [PATCH] mm/vmalloc: Remove WARN_ON_ONCE related to adjust_va_to_fit_type

[Uladzislau Rezki]

On Wed, Sep 27, 2023 at 09:33:04PM +0800, bhe@xxxxxxxxxx wrote:
> On 09/27/23 at 01:49pm, Uladzislau Rezki wrote:
> > > > Yes, but GFP_NOWAIT-alloc-error can easily occur for low-memory device.
> > > >
> > > Agree. You are really in a low memory condition. We end up here only if
> > > pre-loading also has not succeeded, i.e. GFP_KERNEL also fails.
> > > 
> > > But i agree with you, we should "improve the warning" because we drain
> > > and repeat.
> > > 
> > > > How about changing fix as below?:
> > > > 
> > > > <snip>
> > > > --- a/mm/vmalloc.c
> > > > +++ b/mm/vmalloc.c
> > > > @@ -1468,6 +1468,7 @@ adjust_va_to_fit_type(struct rb_root *root, struct list_head *head,
> > > >                  */
> > > >                 va->va_start = nva_start_addr + size;
> > > >         } else {
> > > > +               WARN_ON_ONCE(1);
> > > >                 return -1;
> > > >         }
> > > >  
> > > > @@ -1522,7 +1523,7 @@ __alloc_vmap_area(struct rb_root *root, struct list_head *head,
> > > >  
> > > >         /* Update the free vmap_area. */
> > > >         ret = adjust_va_to_fit_type(root, head, va, nva_start_addr, size);
> > > > -       if (WARN_ON_ONCE(ret))
> > > > +       if (ret)
> > > >                 return vend;
> > > >  
> > > >  #if DEBUG_AUGMENT_LOWEST_MATCH_CHECK
> > > > @@ -4143,7 +4144,7 @@ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets,
> > > >                 ret = adjust_va_to_fit_type(&free_vmap_area_root,
> > > >                                             &free_vmap_area_list,
> > > >                                             va, start, size);
> > > > -               if (WARN_ON_ONCE(unlikely(ret)))
> > > > +               if (unlikely(ret))
> > > >                         /* It is a BUG(), but trigger recovery instead. */
> > > >                         goto recovery;
> > > >  
> > > > <snip>
> > > > It will WARN_ONCE_ONCE() only if classify_va_fit_type() is "(type == NOTHING_FIT)".
> > > > 
> > > This is good but i think it should be improved further. We need to
> > > understand from the warning when no memory and when there is no a
> > > vmap space, so:
> > > 
> > > - if NOTHING_FIT, we should WARN() for sure;
> > > - Second place in the pcpu_get_vm_area(), we do not use NE_FIT. Only in
> > >   the begging after boot, but potentially we can trigger -ENOMEM and we
> > >   should warn in this case. Otherwise you just hide it;
> > > - And last one if after repeating we still do not manage to allocate.
> > > 
> > 
> > We should understand a reason of failing. I think error handling should
> > be improved. Something like:
> 
> This looks good to me, while the parameter 'error' looks a little ugly.
> How about this?
> 
Agree. Embedding an error to "addr" is much more better way of handling
it, so we do not need an extra storage for errors. This is good :)

Jaeseon Sim - do you agree on it? If so please re-spin the patch if you
do not want please let us know.

> diff --git a/mm/vmalloc.c b/mm/vmalloc.c
> index ef8599d394fd..32805c82373b 100644
> --- a/mm/vmalloc.c
> +++ b/mm/vmalloc.c
> @@ -1454,7 +1454,7 @@ adjust_va_to_fit_type(struct rb_root *root, struct list_head *head,
>  			 */
>  			lva = kmem_cache_alloc(vmap_area_cachep, GFP_NOWAIT);
>  			if (!lva)
> -				return -1;
> +				return -ENOMEM;
>  		}
>  
>  		/*
> @@ -1468,7 +1468,7 @@ adjust_va_to_fit_type(struct rb_root *root, struct list_head *head,
>  		 */
>  		va->va_start = nva_start_addr + size;
>  	} else {
> -		return -1;
> +		return -EINVAL;
>  	}
>  
>  	if (type != FL_FIT_TYPE) {
> @@ -1509,7 +1509,7 @@ __alloc_vmap_area(struct rb_root *root, struct list_head *head,
>  
>  	va = find_vmap_lowest_match(root, size, align, vstart, adjust_search_size);
>  	if (unlikely(!va))
> -		return vend;
> +		return -ENOENT;
>  
>  	if (va->va_start > vstart)
>  		nva_start_addr = ALIGN(va->va_start, align);
> @@ -1518,12 +1518,12 @@ __alloc_vmap_area(struct rb_root *root, struct list_head *head,
>  
>  	/* Check the "vend" restriction. */
>  	if (nva_start_addr + size > vend)
> -		return vend;
> +		return -ERANGE;
>  
>  	/* Update the free vmap_area. */
>  	ret = adjust_va_to_fit_type(root, head, va, nva_start_addr, size);
> -	if (WARN_ON_ONCE(ret))
> -		return vend;
> +	if (ret)
> +		return ret;
>  
>  #if DEBUG_AUGMENT_LOWEST_MATCH_CHECK
>  	find_vmap_lowest_match_check(root, head, size, align);
> @@ -1616,13 +1616,13 @@ static struct vmap_area *alloc_vmap_area(unsigned long size,
>  		size, align, vstart, vend);
>  	spin_unlock(&free_vmap_area_lock);
>  
> -	trace_alloc_vmap_area(addr, size, align, vstart, vend, addr == vend);
> +	trace_alloc_vmap_area(addr, size, align, vstart, vend, IS_ERR(addr));
>  
>  	/*
> -	 * If an allocation fails, the "vend" address is
> +	 * If an allocation fails, the error value is
>  	 * returned. Therefore trigger the overflow path.
>  	 */
> -	if (unlikely(addr == vend))
> +	if (unlikely(IS_ERR(addr)))
>  		goto overflow;
>  
>  	va->va_start = addr;
> @@ -1662,8 +1662,9 @@ static struct vmap_area *alloc_vmap_area(unsigned long size,
>  	}
>  
>  	if (!(gfp_mask & __GFP_NOWARN) && printk_ratelimit())
> -		pr_warn("vmap allocation for size %lu failed: use vmalloc=<size> to increase size\n",
> -			size);
> +		pr_warn("vmap allocation for size %lu failed: "
> +			"use vmalloc=<size> to increase size, errno: (%d)\n",
> +			size, addr);
>  
>  	kmem_cache_free(vmap_area_cachep, va);
>  	return ERR_PTR(-EBUSY);
> @@ -4143,8 +4144,8 @@ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets,
>  		ret = adjust_va_to_fit_type(&free_vmap_area_root,
>  					    &free_vmap_area_list,
>  					    va, start, size);
> -		if (WARN_ON_ONCE(unlikely(ret)))
> -			/* It is a BUG(), but trigger recovery instead. */
> +		if ((unlikely(ret)))
> +			WARN_ONCE(1, "%s error: errno (%d)\n", __func__, ret);
>  			goto recovery;
>  
>  		/* Allocated area. */
> 
> > 
> > <snip>
> > diff --git a/mm/vmalloc.c b/mm/vmalloc.c
> > index ef8599d394fd..03a36921a3fc 100644
> > --- a/mm/vmalloc.c
> > +++ b/mm/vmalloc.c
> > @@ -1454,7 +1454,7 @@ adjust_va_to_fit_type(struct rb_root *root, struct list_head *head,
> >  			 */
> >  			lva = kmem_cache_alloc(vmap_area_cachep, GFP_NOWAIT);
> >  			if (!lva)
> > -				return -1;
> > +				return -ENOMEM;
> >  		}
> >  
> >  		/*
> > @@ -1468,7 +1468,7 @@ adjust_va_to_fit_type(struct rb_root *root, struct list_head *head,
> >  		 */
> >  		va->va_start = nva_start_addr + size;
> >  	} else {
> > -		return -1;
> > +		return -EINVAL;
> >  	}
> >  
> >  	if (type != FL_FIT_TYPE) {
> > @@ -1488,7 +1488,8 @@ adjust_va_to_fit_type(struct rb_root *root, struct list_head *head,
> >  static __always_inline unsigned long
> >  __alloc_vmap_area(struct rb_root *root, struct list_head *head,
> >  	unsigned long size, unsigned long align,
> > -	unsigned long vstart, unsigned long vend)
> > +	unsigned long vstart, unsigned long vend,
> > +	int *error)
> >  {
> >  	bool adjust_search_size = true;
> >  	unsigned long nva_start_addr;
> > @@ -1508,8 +1509,10 @@ __alloc_vmap_area(struct rb_root *root, struct list_head *head,
> >  		adjust_search_size = false;
> >  
> >  	va = find_vmap_lowest_match(root, size, align, vstart, adjust_search_size);
> > -	if (unlikely(!va))
> > +	if (unlikely(!va)) {
> > +		*error = -ENOENT;
> >  		return vend;
> > +	}
> >  
> >  	if (va->va_start > vstart)
> >  		nva_start_addr = ALIGN(va->va_start, align);
> > @@ -1517,13 +1520,17 @@ __alloc_vmap_area(struct rb_root *root, struct list_head *head,
> >  		nva_start_addr = ALIGN(vstart, align);
> >  
> >  	/* Check the "vend" restriction. */
> > -	if (nva_start_addr + size > vend)
> > +	if (nva_start_addr + size > vend) {
> > +		*error = -ERANGE;
> >  		return vend;
> > +	}
> >  
> >  	/* Update the free vmap_area. */
> >  	ret = adjust_va_to_fit_type(root, head, va, nva_start_addr, size);
> > -	if (WARN_ON_ONCE(ret))
> > +	if (ret) {
> > +		*error = ret;
> >  		return vend;
> > +	}
> >  
> >  #if DEBUG_AUGMENT_LOWEST_MATCH_CHECK
> >  	find_vmap_lowest_match_check(root, head, size, align);
> > @@ -1589,7 +1596,7 @@ static struct vmap_area *alloc_vmap_area(unsigned long size,
> >  	unsigned long freed;
> >  	unsigned long addr;
> >  	int purged = 0;
> > -	int ret;
> > +	int ret, error;
> >  
> >  	if (unlikely(!size || offset_in_page(size) || !is_power_of_2(align)))
> >  		return ERR_PTR(-EINVAL);
> > @@ -1613,7 +1620,7 @@ static struct vmap_area *alloc_vmap_area(unsigned long size,
> >  retry:
> >  	preload_this_cpu_lock(&free_vmap_area_lock, gfp_mask, node);
> >  	addr = __alloc_vmap_area(&free_vmap_area_root, &free_vmap_area_list,
> > -		size, align, vstart, vend);
> > +		size, align, vstart, vend, &error);
> >  	spin_unlock(&free_vmap_area_lock);
> >  
> >  	trace_alloc_vmap_area(addr, size, align, vstart, vend, addr == vend);
> > @@ -1662,8 +1669,9 @@ static struct vmap_area *alloc_vmap_area(unsigned long size,
> >  	}
> >  
> >  	if (!(gfp_mask & __GFP_NOWARN) && printk_ratelimit())
> > -		pr_warn("vmap allocation for size %lu failed: use vmalloc=<size> to increase size\n",
> > -			size);
> > +		pr_warn("vmap allocation for size %lu failed: "
> > +			"use vmalloc=<size> to increase size, errno: (%d)\n",
> > +			size, error);
> >  
> >  	kmem_cache_free(vmap_area_cachep, va);
> >  	return ERR_PTR(-EBUSY);
> > @@ -4143,9 +4151,10 @@ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets,
> >  		ret = adjust_va_to_fit_type(&free_vmap_area_root,
> >  					    &free_vmap_area_list,
> >  					    va, start, size);
> > -		if (WARN_ON_ONCE(unlikely(ret)))
> > -			/* It is a BUG(), but trigger recovery instead. */
> > +		if (unlikely(ret)) {
> > +			WARN_ONCE(1, "%s error: errno (%d)\n", __func__, ret);
> >  			goto recovery;
> > +		}
> >  
> >  		/* Allocated area. */
> >  		va = vas[area];
> > <snip>
> > 
> > Any thoughts?
> 

--
Uladzislau Rezki