I think we can remove "BUG_ON(start_pfn >= end_pfn)" in __offline_pages(),
because in memory_block_action() "nr_pages = PAGES_PER_SECTION * sections_per_block"
is always greater than 0.
memory_block_action()
offline_pages()
__offline_pages()
BUG_ON(start_pfn >= end_pfn)
Signed-off-by: Xishi Qiu <[email protected]>
---
mm/memory_hotplug.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
index ca1dd3a..8e333f9 100644
--- a/mm/memory_hotplug.c
+++ b/mm/memory_hotplug.c
@@ -1472,7 +1472,6 @@ static int __ref __offline_pages(unsigned long start_pfn,
struct zone *zone;
struct memory_notify arg;
- BUG_ON(start_pfn >= end_pfn);
/* at least, alignment against pageblock is necessary */
if (!IS_ALIGNED(start_pfn, pageblock_nr_pages))
return -EINVAL;
--
1.8.2.2
On Tue, 2013-07-30 at 14:49 +0800, Xishi Qiu wrote:
> I think we can remove "BUG_ON(start_pfn >= end_pfn)" in __offline_pages(),
> because in memory_block_action() "nr_pages = PAGES_PER_SECTION * sections_per_block"
> is always greater than 0.
BUG_ON() is used for checking a condition that should never happen,
unless there is a bug. So, to me, what you described seems to match
with the use of BUG_ON() to prevent a potential bug in the caller.
Thanks,
-Toshi
> memory_block_action()
> offline_pages()
> __offline_pages()
> BUG_ON(start_pfn >= end_pfn)
>
> Signed-off-by: Xishi Qiu <[email protected]>
> ---
> mm/memory_hotplug.c | 1 -
> 1 file changed, 1 deletion(-)
>
> diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c
> index ca1dd3a..8e333f9 100644
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -1472,7 +1472,6 @@ static int __ref __offline_pages(unsigned long start_pfn,
> struct zone *zone;
> struct memory_notify arg;
>
> - BUG_ON(start_pfn >= end_pfn);
> /* at least, alignment against pageblock is necessary */
> if (!IS_ALIGNED(start_pfn, pageblock_nr_pages))
> return -EINVAL;
On 07/29/2013 11:49 PM, Xishi Qiu wrote:
> I think we can remove "BUG_ON(start_pfn >= end_pfn)" in __offline_pages(),
> because in memory_block_action() "nr_pages = PAGES_PER_SECTION * sections_per_block"
> is always greater than 0.
...
> --- a/mm/memory_hotplug.c
> +++ b/mm/memory_hotplug.c
> @@ -1472,7 +1472,6 @@ static int __ref __offline_pages(unsigned long start_pfn,
> struct zone *zone;
> struct memory_notify arg;
>
> - BUG_ON(start_pfn >= end_pfn);
> /* at least, alignment against pageblock is necessary */
> if (!IS_ALIGNED(start_pfn, pageblock_nr_pages))
> return -EINVAL;
I think you're saying that you don't see a way to hit this BUG_ON() in
practice. That does appear to be true, unless sections_per_block ended
up 0 or negative. The odds of getting in to this code if
'sections_per_block' was bogus are pretty small.
Or, is this a theoretical thing that folks might run in to when adding
new features or developing? It's in a cold path and the cost of the
check is miniscule. The original author (cc'd) also saw a need to put
this in probably because he actually ran in to this.
In any case, it looks fairly safe to me:
Reviewed-by: Dave Hansen <[email protected]>
On 2013/8/1 0:55, Dave Hansen wrote:
> On 07/29/2013 11:49 PM, Xishi Qiu wrote:
>> I think we can remove "BUG_ON(start_pfn >= end_pfn)" in __offline_pages(),
>> because in memory_block_action() "nr_pages = PAGES_PER_SECTION * sections_per_block"
>> is always greater than 0.
> ...
>> --- a/mm/memory_hotplug.c
>> +++ b/mm/memory_hotplug.c
>> @@ -1472,7 +1472,6 @@ static int __ref __offline_pages(unsigned long start_pfn,
>> struct zone *zone;
>> struct memory_notify arg;
>>
>> - BUG_ON(start_pfn >= end_pfn);
>> /* at least, alignment against pageblock is necessary */
>> if (!IS_ALIGNED(start_pfn, pageblock_nr_pages))
>> return -EINVAL;
>
> I think you're saying that you don't see a way to hit this BUG_ON() in
> practice. That does appear to be true, unless sections_per_block ended
> up 0 or negative. The odds of getting in to this code if
> 'sections_per_block' was bogus are pretty small.
>
Yes, I find there is an only to hit this BUG_ON() in v3.11, and "sections_per_block"
seems to be always greater than 0.
> Or, is this a theoretical thing that folks might run in to when adding
> new features or developing? It's in a cold path and the cost of the
> check is miniscule. The original author (cc'd) also saw a need to put
> this in probably because he actually ran in to this.
>
In v2.6.32, If info->length==0, this way may hit this BUG_ON().
acpi_memory_disable_device()
remove_memory(info->start_addr, info->length)
offline_pages()
Later Fujitsu's patch rename this function and the BUG_ON() is unnecessary.
Thanks,
Xishi Qiu
> In any case, it looks fairly safe to me:
>
> Reviewed-by: Dave Hansen <[email protected]>
> .
>