[PATCH] mm/cma: print total and used pages in cma_alloc()

[PATCH] mm/cma: print total and used pages in cma_alloc()

[Xiang Gao]

From: gaoxiang17 <gaoxiang17@xiaomi.com>

This makes cma info more intuitive during debugging.

before:
[   24.407814] cma: cma_alloc(cma (____ptrval____), name: reserved, count 1, align 0)
[   24.413397] cma: cma_alloc(cma (____ptrval____), name: reserved, count 1, align 0)
[   24.415886] cma: cma_alloc(cma (____ptrval____), name: reserved, count 1, align 0)

after:
[   24.069738] cma: cma_alloc(cma (____ptrval____), name: reserved, total pages: 16384, used pages: 64, request pages: 1, align 0)
[   24.075317] cma: cma_alloc(cma (____ptrval____), name: reserved, total pages: 16384, used pages: 65, request pages: 1, align 0)
[   24.078455] cma: cma_alloc(cma (____ptrval____), name: reserved, total pages: 16384, used pages: 66, request pages: 1, align 0)

Signed-off-by: gaoxiang17 <gaoxiang17@xiaomi.com>
---
 mm/cma.c | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/mm/cma.c b/mm/cma.c
index 2ffa4befb99a..46cc98e7f587 100644
--- a/mm/cma.c
+++ b/mm/cma.c
@@ -776,6 +776,17 @@ static void cma_debug_show_areas(struct cma *cma)
 	spin_unlock_irq(&cma->lock);
 }
 
+static unsigned long cma_get_used_pages(struct cma *cma)
+{
+	unsigned long used;
+
+	spin_lock_irq(&cma->lock);
+	used = bitmap_weight(cma->bitmap, (int)cma_bitmap_maxno(cma));
+	spin_unlock_irq(&cma->lock);
+
+	return used << cma->order_per_bit;
+}
+
 static int cma_range_alloc(struct cma *cma, struct cma_memrange *cmr,
 				unsigned long count, unsigned int align,
 				struct page **pagep, gfp_t gfp)
@@ -858,8 +869,8 @@ static struct page *__cma_alloc(struct cma *cma, unsigned long count,
 	if (!cma || !cma->count)
 		return page;
 
-	pr_debug("%s(cma %p, name: %s, count %lu, align %d)\n", __func__,
-		(void *)cma, cma->name, count, align);
+	pr_debug("%s(cma %p, name: %s, total pages: %lu, used pages: %lu, request pages: %lu, align %d)\n",
+		__func__, (void *)cma, cma->name, cma->count, cma_get_used_pages(cma), count, align);
 
 	if (!count)
 		return page;
-- 
2.34.1

Re: [PATCH] mm/cma: print total and used pages in cma_alloc()

[David Hildenbrand]

On 16.08.25 06:28, Xiang Gao wrote:
> From: gaoxiang17 <gaoxiang17@xiaomi.com>
> 
> This makes cma info more intuitive during debugging.
> 
> before:
> [   24.407814] cma: cma_alloc(cma (____ptrval____), name: reserved, count 1, align 0)
> [   24.413397] cma: cma_alloc(cma (____ptrval____), name: reserved, count 1, align 0)
> [   24.415886] cma: cma_alloc(cma (____ptrval____), name: reserved, count 1, align 0)
> 
> after:
> [   24.069738] cma: cma_alloc(cma (____ptrval____), name: reserved, total pages: 16384, used pages: 64, request pages: 1, align 0)
> [   24.075317] cma: cma_alloc(cma (____ptrval____), name: reserved, total pages: 16384, used pages: 65, request pages: 1, align 0)
> [   24.078455] cma: cma_alloc(cma (____ptrval____), name: reserved, total pages: 16384, used pages: 66, request pages: 1, align 0)
> 
> Signed-off-by: gaoxiang17 <gaoxiang17@xiaomi.com>
> ---
>   mm/cma.c | 15 +++++++++++++--
>   1 file changed, 13 insertions(+), 2 deletions(-)
> 
> diff --git a/mm/cma.c b/mm/cma.c
> index 2ffa4befb99a..46cc98e7f587 100644
> --- a/mm/cma.c
> +++ b/mm/cma.c
> @@ -776,6 +776,17 @@ static void cma_debug_show_areas(struct cma *cma)
>   	spin_unlock_irq(&cma->lock);
>   }
>   
> +static unsigned long cma_get_used_pages(struct cma *cma)
> +{
> +	unsigned long used;
> +
> +	spin_lock_irq(&cma->lock);
> +	used = bitmap_weight(cma->bitmap, (int)cma_bitmap_maxno(cma));
> +	spin_unlock_irq(&cma->lock);
> +
> +	return used << cma->order_per_bit;
> +}
> +
>   static int cma_range_alloc(struct cma *cma, struct cma_memrange *cmr,
>   				unsigned long count, unsigned int align,
>   				struct page **pagep, gfp_t gfp)
> @@ -858,8 +869,8 @@ static struct page *__cma_alloc(struct cma *cma, unsigned long count,
>   	if (!cma || !cma->count)
>   		return page;
>   
> -	pr_debug("%s(cma %p, name: %s, count %lu, align %d)\n", __func__,
> -		(void *)cma, cma->name, count, align);
> +	pr_debug("%s(cma %p, name: %s, total pages: %lu, used pages: %lu, request pages: %lu, align %d)\n",
> +		__func__, (void *)cma, cma->name, cma->count, cma_get_used_pages(cma), count, align);

		^ one space missing for proper indentation.

But doing another spinlock cycle just for debugging purposes? That does 
not feel right, sorry.

-- 
Cheers

David / dhildenb

Re: [PATCH] mm/cma: print total and used pages in cma_alloc()

[Giorgi Tchankvetadze]

What about using tracepoints?
Add a trace_cma_alloc() event that only runs when tracing is enabled. No 
lock unless someone is actively monitoring

On 8/16/2025 10:27 AM, David Hildenbrand wrote:
> On 16.08.25 06:28, Xiang Gao wrote:
>> From: gaoxiang17 <gaoxiang17@xiaomi.com>
>>
>> This makes cma info more intuitive during debugging.
>>
>> before:
>> [   24.407814] cma: cma_alloc(cma (____ptrval____), name: reserved, 
>> count 1, align 0)
>> [   24.413397] cma: cma_alloc(cma (____ptrval____), name: reserved, 
>> count 1, align 0)
>> [   24.415886] cma: cma_alloc(cma (____ptrval____), name: reserved, 
>> count 1, align 0)
>>
>> after:
>> [   24.069738] cma: cma_alloc(cma (____ptrval____), name: reserved, 
>> total pages: 16384, used pages: 64, request pages: 1, align 0)
>> [   24.075317] cma: cma_alloc(cma (____ptrval____), name: reserved, 
>> total pages: 16384, used pages: 65, request pages: 1, align 0)
>> [   24.078455] cma: cma_alloc(cma (____ptrval____), name: reserved, 
>> total pages: 16384, used pages: 66, request pages: 1, align 0)
>>
>> Signed-off-by: gaoxiang17 <gaoxiang17@xiaomi.com>
>> ---
>>   mm/cma.c | 15 +++++++++++++--
>>   1 file changed, 13 insertions(+), 2 deletions(-)
>>
>> diff --git a/mm/cma.c b/mm/cma.c
>> index 2ffa4befb99a..46cc98e7f587 100644
>> --- a/mm/cma.c
>> +++ b/mm/cma.c
>> @@ -776,6 +776,17 @@ static void cma_debug_show_areas(struct cma *cma)
>>       spin_unlock_irq(&cma->lock);
>>   }
>> +static unsigned long cma_get_used_pages(struct cma *cma)
>> +{
>> +    unsigned long used;
>> +
>> +    spin_lock_irq(&cma->lock);
>> +    used = bitmap_weight(cma->bitmap, (int)cma_bitmap_maxno(cma));
>> +    spin_unlock_irq(&cma->lock);
>> +
>> +    return used << cma->order_per_bit;
>> +}
>> +
>>   static int cma_range_alloc(struct cma *cma, struct cma_memrange *cmr,
>>                   unsigned long count, unsigned int align,
>>                   struct page **pagep, gfp_t gfp)
>> @@ -858,8 +869,8 @@ static struct page *__cma_alloc(struct cma *cma, 
>> unsigned long count,
>>       if (!cma || !cma->count)
>>           return page;
>> -    pr_debug("%s(cma %p, name: %s, count %lu, align %d)\n", __func__,
>> -        (void *)cma, cma->name, count, align);
>> +    pr_debug("%s(cma %p, name: %s, total pages: %lu, used pages: %lu, 
>> request pages: %lu, align %d)\n",
>> +        __func__, (void *)cma, cma->name, cma->count, 
>> cma_get_used_pages(cma), count, align);
> 
>          ^ one space missing for proper indentation.
> 
> But doing another spinlock cycle just for debugging purposes? That does 
> not feel right, sorry.
> 

Re: [PATCH] mm/cma: print total and used pages in cma_alloc()

[David Hildenbrand]

On 16.08.25 08:42, Giorgi Tchankvetadze wrote:
> What about using tracepoints?
> Add a trace_cma_alloc() event that only runs when tracing is enabled. No
> lock unless someone is actively monitoring

Tracing in general sounds like a much better approach here than the two 
pr_debug() in this function.

-- 
Cheers

David / dhildenb

Re: [PATCH] mm/cma: print total and used pages in cma_alloc()

[Andrew Morton]

On Sat, 16 Aug 2025 08:27:39 +0200 David Hildenbrand <david@redhat.com> wrote:

> > @@ -858,8 +869,8 @@ static struct page *__cma_alloc(struct cma *cma, unsigned long count,
> >   	if (!cma || !cma->count)
> >   		return page;
> >   
> > -	pr_debug("%s(cma %p, name: %s, count %lu, align %d)\n", __func__,
> > -		(void *)cma, cma->name, count, align);
> > +	pr_debug("%s(cma %p, name: %s, total pages: %lu, used pages: %lu, request pages: %lu, align %d)\n",
> > +		__func__, (void *)cma, cma->name, cma->count, cma_get_used_pages(cma), count, align);
> 
> 		^ one space missing for proper indentation.
> 
> But doing another spinlock cycle just for debugging purposes? That does 
> not feel right, sorry.

If we're calling pr_debug() frequently enough for this to matter, we
have other problems!

Re: [PATCH] mm/cma: print total and used pages in cma_alloc()

[David Hildenbrand]

On 16.08.25 08:45, Andrew Morton wrote:
> On Sat, 16 Aug 2025 08:27:39 +0200 David Hildenbrand <david@redhat.com> wrote:
> 
>>> @@ -858,8 +869,8 @@ static struct page *__cma_alloc(struct cma *cma, unsigned long count,
>>>    	if (!cma || !cma->count)
>>>    		return page;
>>>    
>>> -	pr_debug("%s(cma %p, name: %s, count %lu, align %d)\n", __func__,
>>> -		(void *)cma, cma->name, count, align);
>>> +	pr_debug("%s(cma %p, name: %s, total pages: %lu, used pages: %lu, request pages: %lu, align %d)\n",
>>> +		__func__, (void *)cma, cma->name, cma->count, cma_get_used_pages(cma), count, align);
>>
>> 		^ one space missing for proper indentation.
>>
>> But doing another spinlock cycle just for debugging purposes? That does
>> not feel right, sorry.
> 
> If we're calling pr_debug() frequently enough for this to matter, we
> have other problems!

We call it for each and every actual CMA allocation? I really don't see 
why we want to just randomly make CMA allocation latency worse.

Is the existing pr_debug() a problem? Maybe. But who actually has debug 
messages enabled in any sane setup?

-- 
Cheers

David / dhildenb

Re: [PATCH] mm/cma: print total and used pages in cma_alloc()

[Andrew Morton]

On Sat, 16 Aug 2025 08:56:47 +0200 David Hildenbrand <david@redhat.com> wrote:

> On 16.08.25 08:45, Andrew Morton wrote:
> > On Sat, 16 Aug 2025 08:27:39 +0200 David Hildenbrand <david@redhat.com> wrote:
> > 
> >>> @@ -858,8 +869,8 @@ static struct page *__cma_alloc(struct cma *cma, unsigned long count,
> >>>    	if (!cma || !cma->count)
> >>>    		return page;
> >>>    
> >>> -	pr_debug("%s(cma %p, name: %s, count %lu, align %d)\n", __func__,
> >>> -		(void *)cma, cma->name, count, align);
> >>> +	pr_debug("%s(cma %p, name: %s, total pages: %lu, used pages: %lu, request pages: %lu, align %d)\n",
> >>> +		__func__, (void *)cma, cma->name, cma->count, cma_get_used_pages(cma), count, align);
> >>
> >> 		^ one space missing for proper indentation.
> >>
> >> But doing another spinlock cycle just for debugging purposes? That does
> >> not feel right, sorry.
> > 
> > If we're calling pr_debug() frequently enough for this to matter, we
> > have other problems!
> 
> We call it for each and every actual CMA allocation? I really don't see 
> why we want to just randomly make CMA allocation latency worse.

pr_debug() is 12 million times more expensive than a spin_lock()!

> Is the existing pr_debug() a problem? Maybe. But who actually has debug 
> messages enabled in any sane setup?

Nobody, clearly.  If anyone enabled pr_debug() in here, they'd
immediately have to remove those statements to get any work done.  Kill
it.

Re: [PATCH] mm/cma: print total and used pages in cma_alloc()

[David Hildenbrand]

On 16.08.25 09:34, Andrew Morton wrote:
> On Sat, 16 Aug 2025 08:56:47 +0200 David Hildenbrand <david@redhat.com> wrote:
> 
>> On 16.08.25 08:45, Andrew Morton wrote:
>>> On Sat, 16 Aug 2025 08:27:39 +0200 David Hildenbrand <david@redhat.com> wrote:
>>>
>>>>> @@ -858,8 +869,8 @@ static struct page *__cma_alloc(struct cma *cma, unsigned long count,
>>>>>     	if (!cma || !cma->count)
>>>>>     		return page;
>>>>>     
>>>>> -	pr_debug("%s(cma %p, name: %s, count %lu, align %d)\n", __func__,
>>>>> -		(void *)cma, cma->name, count, align);
>>>>> +	pr_debug("%s(cma %p, name: %s, total pages: %lu, used pages: %lu, request pages: %lu, align %d)\n",
>>>>> +		__func__, (void *)cma, cma->name, cma->count, cma_get_used_pages(cma), count, align);
>>>>
>>>> 		^ one space missing for proper indentation.
>>>>
>>>> But doing another spinlock cycle just for debugging purposes? That does
>>>> not feel right, sorry.
>>>
>>> If we're calling pr_debug() frequently enough for this to matter, we
>>> have other problems!
>>
>> We call it for each and every actual CMA allocation? I really don't see
>> why we want to just randomly make CMA allocation latency worse.
> 
> pr_debug() is 12 million times more expensive than a spin_lock()!
> 
>> Is the existing pr_debug() a problem? Maybe. But who actually has debug
>> messages enabled in any sane setup?
> 
> Nobody, clearly.  If anyone enabled pr_debug() in here, they'd
> immediately have to remove those statements to get any work done.  Kill
> it.

I just learned that pr_debug() on a !CONFIG_DEBUG kernel translates to 
no_printk(), which is just a mostly-empty macro that doesn't really use 
any of the parameters.

I would assume the cma_get_used_pages() would get completely optimized 
out in that case.

So, I don't care, but ... moving to tracing seems much more reasonable.

-- 
Cheers

David / dhildenb

答复: [External Mail]Re: [PATCH] mm/cma: print total and used pages in cma_alloc()

[高翔]

[-- Attachment #1: Type: text/plain, Size: 2580 bytes --]

From this perspective, it seems better to add it to the trace. Then shall I make a patch to add to the trace?


________________________________
发件人: David Hildenbrand <david@redhat.com>
发送时间: 2025年8月16日 15:49:52
收件人: Andrew Morton
抄送: Xiang Gao; lorenzo.stoakes@oracle.com; Liam.Howlett@oracle.com; vbabka@suse.cz; rppt@kernel.org; surenb@google.com; mhocko@suse.com; linux-mm@kvack.org; linux-kernel@vger.kernel.org; 高翔
主题: [External Mail]Re: [PATCH] mm/cma: print total and used pages in cma_alloc()

[外部邮件] 此邮件来源于小米公司外部，请谨慎处理。若对邮件安全性存疑，请将邮件转发给misec@xiaomi.com进行反馈

On 16.08.25 09:34, Andrew Morton wrote:
> On Sat, 16 Aug 2025 08:56:47 +0200 David Hildenbrand <david@redhat.com> wrote:
>
>> On 16.08.25 08:45, Andrew Morton wrote:
>>> On Sat, 16 Aug 2025 08:27:39 +0200 David Hildenbrand <david@redhat.com> wrote:
>>>
>>>>> @@ -858,8 +869,8 @@ static struct page *__cma_alloc(struct cma *cma, unsigned long count,
>>>>>           if (!cma || !cma->count)
>>>>>                   return page;
>>>>>
>>>>> - pr_debug("%s(cma %p, name: %s, count %lu, align %d)\n", __func__,
>>>>> -         (void *)cma, cma->name, count, align);
>>>>> + pr_debug("%s(cma %p, name: %s, total pages: %lu, used pages: %lu, request pages: %lu, align %d)\n",
>>>>> +         __func__, (void *)cma, cma->name, cma->count, cma_get_used_pages(cma), count, align);
>>>>
>>>>            ^ one space missing for proper indentation.
>>>>
>>>> But doing another spinlock cycle just for debugging purposes? That does
>>>> not feel right, sorry.
>>>
>>> If we're calling pr_debug() frequently enough for this to matter, we
>>> have other problems!
>>
>> We call it for each and every actual CMA allocation? I really don't see
>> why we want to just randomly make CMA allocation latency worse.
>
> pr_debug() is 12 million times more expensive than a spin_lock()!
>
>> Is the existing pr_debug() a problem? Maybe. But who actually has debug
>> messages enabled in any sane setup?
>
> Nobody, clearly.  If anyone enabled pr_debug() in here, they'd
> immediately have to remove those statements to get any work done.  Kill
> it.

I just learned that pr_debug() on a !CONFIG_DEBUG kernel translates to
no_printk(), which is just a mostly-empty macro that doesn't really use
any of the parameters.

I would assume the cma_get_used_pages() would get completely optimized
out in that case.

So, I don't care, but ... moving to tracing seems much more reasonable.

--
Cheers

David / dhildenb


[-- Attachment #2: Type: text/html, Size: 4444 bytes --]

Re: [PATCH] mm/cma: print total and used pages in cma_alloc()

[kernel test robot]

Hi Xiang,

kernel test robot noticed the following build errors:

[auto build test ERROR on akpm-mm/mm-everything]

url:    https://github.com/intel-lab-lkp/linux/commits/Xiang-Gao/mm-cma-print-total-and-used-pages-in-cma_alloc/20250816-122940
base:   https://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git mm-everything
patch link:    https://lore.kernel.org/r/20250816042842.3959315-1-gxxa03070307%40gmail.com
patch subject: [PATCH] mm/cma: print total and used pages in cma_alloc()
config: arm-randconfig-002-20250816 (https://download.01.org/0day-ci/archive/20250817/202508170014.PK57XSd7-lkp@intel.com/config)
compiler: arm-linux-gnueabi-gcc (GCC) 10.5.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250817/202508170014.PK57XSd7-lkp@intel.com/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202508170014.PK57XSd7-lkp@intel.com/

All errors (new ones prefixed by >>):

   mm/cma.c: In function 'cma_get_used_pages':
>> mm/cma.c:784:26: error: 'struct cma' has no member named 'bitmap'
     784 |  used = bitmap_weight(cma->bitmap, (int)cma_bitmap_maxno(cma));
         |                          ^~
>> mm/cma.c:784:41: error: too few arguments to function 'cma_bitmap_maxno'
     784 |  used = bitmap_weight(cma->bitmap, (int)cma_bitmap_maxno(cma));
         |                                         ^~~~~~~~~~~~~~~~
   In file included from mm/cma.c:34:
   mm/cma.h:77:29: note: declared here
      77 | static inline unsigned long cma_bitmap_maxno(struct cma *cma,
         |                             ^~~~~~~~~~~~~~~~


vim +784 mm/cma.c

   778	
   779	static unsigned long cma_get_used_pages(struct cma *cma)
   780	{
   781		unsigned long used;
   782	
   783		spin_lock_irq(&cma->lock);
 > 784		used = bitmap_weight(cma->bitmap, (int)cma_bitmap_maxno(cma));
   785		spin_unlock_irq(&cma->lock);
   786	
   787		return used << cma->order_per_bit;
   788	}
   789	

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki