* [PATCH 1/3] mm/page_alloc: Use dump_page() directly
@ 2021-05-12 3:10 Kefeng Wang
2021-05-12 3:10 ` [PATCH 2/3] mm: Make __dump_page() static Kefeng Wang
` (2 more replies)
0 siblings, 3 replies; 6+ messages in thread
From: Kefeng Wang @ 2021-05-12 3:10 UTC (permalink / raw)
To: Andrew Morton; +Cc: willy, linux-mm, linux-kernel, Kefeng Wang
Simplfy use dump_page().
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
mm/page_alloc.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index a2fe714aed93..f23702e7c564 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -658,8 +658,7 @@ static void bad_page(struct page *page, const char *reason)
pr_alert("BUG: Bad page state in process %s pfn:%05lx\n",
current->comm, page_to_pfn(page));
- __dump_page(page, reason);
- dump_page_owner(page);
+ dump_page(page, reason);
print_modules();
dump_stack();
--
2.26.2
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH 2/3] mm: Make __dump_page() static
2021-05-12 3:10 [PATCH 1/3] mm/page_alloc: Use dump_page() directly Kefeng Wang
@ 2021-05-12 3:10 ` Kefeng Wang
2021-05-12 3:10 ` [PATCH 3/3] mm/debug: Check page poisoned firstly in dump_page() Kefeng Wang
2021-05-12 3:14 ` [PATCH 1/3] mm/page_alloc: Use dump_page() directly Matthew Wilcox
2 siblings, 0 replies; 6+ messages in thread
From: Kefeng Wang @ 2021-05-12 3:10 UTC (permalink / raw)
To: Andrew Morton; +Cc: willy, linux-mm, linux-kernel, Kefeng Wang
No __dump_page() caller, so make it static.
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
include/linux/mmdebug.h | 1 -
mm/debug.c | 2 +-
2 files changed, 1 insertion(+), 2 deletions(-)
diff --git a/include/linux/mmdebug.h b/include/linux/mmdebug.h
index 5d0767cb424a..9086157b46f3 100644
--- a/include/linux/mmdebug.h
+++ b/include/linux/mmdebug.h
@@ -10,7 +10,6 @@ struct vm_area_struct;
struct mm_struct;
extern void dump_page(struct page *page, const char *reason);
-extern void __dump_page(struct page *page, const char *reason);
void dump_vma(const struct vm_area_struct *vma);
void dump_mm(const struct mm_struct *mm);
diff --git a/mm/debug.c b/mm/debug.c
index 0bdda8407f71..84cdcd0f7bd3 100644
--- a/mm/debug.c
+++ b/mm/debug.c
@@ -42,7 +42,7 @@ const struct trace_print_flags vmaflag_names[] = {
{0, NULL}
};
-void __dump_page(struct page *page, const char *reason)
+static void __dump_page(struct page *page, const char *reason)
{
struct page *head = compound_head(page);
struct address_space *mapping;
--
2.26.2
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH 3/3] mm/debug: Check page poisoned firstly in dump_page()
2021-05-12 3:10 [PATCH 1/3] mm/page_alloc: Use dump_page() directly Kefeng Wang
2021-05-12 3:10 ` [PATCH 2/3] mm: Make __dump_page() static Kefeng Wang
@ 2021-05-12 3:10 ` Kefeng Wang
2021-05-12 3:15 ` Matthew Wilcox
2021-05-12 3:14 ` [PATCH 1/3] mm/page_alloc: Use dump_page() directly Matthew Wilcox
2 siblings, 1 reply; 6+ messages in thread
From: Kefeng Wang @ 2021-05-12 3:10 UTC (permalink / raw)
To: Andrew Morton; +Cc: willy, linux-mm, linux-kernel, Kefeng Wang
If page is poisoned, it will crash when we call some page related
functions, so must check whether the page is poisoned or not firstly.
Fixes: 6197ab984b41 ("mm: improve dump_page() for compound pages")
Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
---
mm/debug.c | 15 +++++++++------
1 file changed, 9 insertions(+), 6 deletions(-)
diff --git a/mm/debug.c b/mm/debug.c
index 84cdcd0f7bd3..cf84cd9df527 100644
--- a/mm/debug.c
+++ b/mm/debug.c
@@ -44,20 +44,19 @@ const struct trace_print_flags vmaflag_names[] = {
static void __dump_page(struct page *page, const char *reason)
{
- struct page *head = compound_head(page);
+ struct page *head = NULL;
struct address_space *mapping;
- bool page_poisoned = PagePoisoned(page);
- bool compound = PageCompound(page);
+ bool compound;
/*
* Accessing the pageblock without the zone lock. It could change to
* "isolate" again in the meantime, but since we are just dumping the
* state for debugging, it should be fine to accept a bit of
* inaccuracy here due to racing.
*/
- bool page_cma = is_migrate_cma_page(page);
+ bool page_cma;
int mapcount;
char *type = "";
-
+ bool page_poisoned = PagePoisoned(page);
/*
* If struct page is poisoned don't access Page*() functions as that
* leads to recursive loop. Page*() check for poisoned pages, and calls
@@ -68,6 +67,10 @@ static void __dump_page(struct page *page, const char *reason)
goto hex_only;
}
+ head = compound_head(page);
+ page_poisoned = PagePoisoned(page);
+ page_cma = is_migrate_cma_page(page);
+
if (page < head || (page >= head + MAX_ORDER_NR_PAGES)) {
/*
* Corrupt page, so we cannot call page_mapping. Instead, do a
@@ -178,7 +181,7 @@ static void __dump_page(struct page *page, const char *reason)
print_hex_dump(KERN_WARNING, "raw: ", DUMP_PREFIX_NONE, 32,
sizeof(unsigned long), page,
sizeof(struct page), false);
- if (head != page)
+ if (head && head != page)
print_hex_dump(KERN_WARNING, "head: ", DUMP_PREFIX_NONE, 32,
sizeof(unsigned long), head,
sizeof(struct page), false);
--
2.26.2
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH 1/3] mm/page_alloc: Use dump_page() directly
2021-05-12 3:10 [PATCH 1/3] mm/page_alloc: Use dump_page() directly Kefeng Wang
2021-05-12 3:10 ` [PATCH 2/3] mm: Make __dump_page() static Kefeng Wang
2021-05-12 3:10 ` [PATCH 3/3] mm/debug: Check page poisoned firstly in dump_page() Kefeng Wang
@ 2021-05-12 3:14 ` Matthew Wilcox
2 siblings, 0 replies; 6+ messages in thread
From: Matthew Wilcox @ 2021-05-12 3:14 UTC (permalink / raw)
To: Kefeng Wang; +Cc: Andrew Morton, linux-mm, linux-kernel
On Wed, May 12, 2021 at 11:10:55AM +0800, Kefeng Wang wrote:
> Simplfy use dump_page().
https://lore.kernel.org/linux-mm/20210430145549.2662354-3-willy@infradead.org/
> Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
> ---
> mm/page_alloc.c | 3 +--
> 1 file changed, 1 insertion(+), 2 deletions(-)
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index a2fe714aed93..f23702e7c564 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -658,8 +658,7 @@ static void bad_page(struct page *page, const char *reason)
>
> pr_alert("BUG: Bad page state in process %s pfn:%05lx\n",
> current->comm, page_to_pfn(page));
> - __dump_page(page, reason);
> - dump_page_owner(page);
> + dump_page(page, reason);
>
> print_modules();
> dump_stack();
> --
> 2.26.2
>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH 3/3] mm/debug: Check page poisoned firstly in dump_page()
2021-05-12 3:10 ` [PATCH 3/3] mm/debug: Check page poisoned firstly in dump_page() Kefeng Wang
@ 2021-05-12 3:15 ` Matthew Wilcox
2021-05-12 3:23 ` Kefeng Wang
0 siblings, 1 reply; 6+ messages in thread
From: Matthew Wilcox @ 2021-05-12 3:15 UTC (permalink / raw)
To: Kefeng Wang; +Cc: Andrew Morton, linux-mm, linux-kernel
On Wed, May 12, 2021 at 11:10:57AM +0800, Kefeng Wang wrote:
> If page is poisoned, it will crash when we call some page related
> functions, so must check whether the page is poisoned or not firstly.
https://lore.kernel.org/linux-mm/20210430145549.2662354-4-willy@infradead.org/
> Fixes: 6197ab984b41 ("mm: improve dump_page() for compound pages")
> Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
> ---
> mm/debug.c | 15 +++++++++------
> 1 file changed, 9 insertions(+), 6 deletions(-)
>
> diff --git a/mm/debug.c b/mm/debug.c
> index 84cdcd0f7bd3..cf84cd9df527 100644
> --- a/mm/debug.c
> +++ b/mm/debug.c
> @@ -44,20 +44,19 @@ const struct trace_print_flags vmaflag_names[] = {
>
> static void __dump_page(struct page *page, const char *reason)
> {
> - struct page *head = compound_head(page);
> + struct page *head = NULL;
> struct address_space *mapping;
> - bool page_poisoned = PagePoisoned(page);
> - bool compound = PageCompound(page);
> + bool compound;
> /*
> * Accessing the pageblock without the zone lock. It could change to
> * "isolate" again in the meantime, but since we are just dumping the
> * state for debugging, it should be fine to accept a bit of
> * inaccuracy here due to racing.
> */
> - bool page_cma = is_migrate_cma_page(page);
> + bool page_cma;
> int mapcount;
> char *type = "";
> -
> + bool page_poisoned = PagePoisoned(page);
> /*
> * If struct page is poisoned don't access Page*() functions as that
> * leads to recursive loop. Page*() check for poisoned pages, and calls
> @@ -68,6 +67,10 @@ static void __dump_page(struct page *page, const char *reason)
> goto hex_only;
> }
>
> + head = compound_head(page);
> + page_poisoned = PagePoisoned(page);
> + page_cma = is_migrate_cma_page(page);
> +
> if (page < head || (page >= head + MAX_ORDER_NR_PAGES)) {
> /*
> * Corrupt page, so we cannot call page_mapping. Instead, do a
> @@ -178,7 +181,7 @@ static void __dump_page(struct page *page, const char *reason)
> print_hex_dump(KERN_WARNING, "raw: ", DUMP_PREFIX_NONE, 32,
> sizeof(unsigned long), page,
> sizeof(struct page), false);
> - if (head != page)
> + if (head && head != page)
> print_hex_dump(KERN_WARNING, "head: ", DUMP_PREFIX_NONE, 32,
> sizeof(unsigned long), head,
> sizeof(struct page), false);
> --
> 2.26.2
>
>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH 3/3] mm/debug: Check page poisoned firstly in dump_page()
2021-05-12 3:15 ` Matthew Wilcox
@ 2021-05-12 3:23 ` Kefeng Wang
0 siblings, 0 replies; 6+ messages in thread
From: Kefeng Wang @ 2021-05-12 3:23 UTC (permalink / raw)
To: Matthew Wilcox; +Cc: Andrew Morton, linux-mm, linux-kernel
On 2021/5/12 11:15, Matthew Wilcox wrote:
> On Wed, May 12, 2021 at 11:10:57AM +0800, Kefeng Wang wrote:
>> If page is poisoned, it will crash when we call some page related
>> functions, so must check whether the page is poisoned or not firstly.
>
> https://lore.kernel.org/linux-mm/20210430145549.2662354-4-willy@infradead.org/
OK, please ignore them, I found this when debug arm issue in lts5.10,
thanks.
>
>> Fixes: 6197ab984b41 ("mm: improve dump_page() for compound pages")
>> Signed-off-by: Kefeng Wang <wangkefeng.wang@huawei.com>
>> ---
>> mm/debug.c | 15 +++++++++------
>> 1 file changed, 9 insertions(+), 6 deletions(-)
>>
>> diff --git a/mm/debug.c b/mm/debug.c
>> index 84cdcd0f7bd3..cf84cd9df527 100644
>> --- a/mm/debug.c
>> +++ b/mm/debug.c
>> @@ -44,20 +44,19 @@ const struct trace_print_flags vmaflag_names[] = {
>>
>> static void __dump_page(struct page *page, const char *reason)
>> {
>> - struct page *head = compound_head(page);
>> + struct page *head = NULL;
>> struct address_space *mapping;
>> - bool page_poisoned = PagePoisoned(page);
>> - bool compound = PageCompound(page);
>> + bool compound;
>> /*
>> * Accessing the pageblock without the zone lock. It could change to
>> * "isolate" again in the meantime, but since we are just dumping the
>> * state for debugging, it should be fine to accept a bit of
>> * inaccuracy here due to racing.
>> */
>> - bool page_cma = is_migrate_cma_page(page);
>> + bool page_cma;
>> int mapcount;
>> char *type = "";
>> -
>> + bool page_poisoned = PagePoisoned(page);
>> /*
>> * If struct page is poisoned don't access Page*() functions as that
>> * leads to recursive loop. Page*() check for poisoned pages, and calls
>> @@ -68,6 +67,10 @@ static void __dump_page(struct page *page, const char *reason)
>> goto hex_only;
>> }
>>
>> + head = compound_head(page);
>> + page_poisoned = PagePoisoned(page);
>> + page_cma = is_migrate_cma_page(page);
>> +
>> if (page < head || (page >= head + MAX_ORDER_NR_PAGES)) {
>> /*
>> * Corrupt page, so we cannot call page_mapping. Instead, do a
>> @@ -178,7 +181,7 @@ static void __dump_page(struct page *page, const char *reason)
>> print_hex_dump(KERN_WARNING, "raw: ", DUMP_PREFIX_NONE, 32,
>> sizeof(unsigned long), page,
>> sizeof(struct page), false);
>> - if (head != page)
>> + if (head && head != page)
>> print_hex_dump(KERN_WARNING, "head: ", DUMP_PREFIX_NONE, 32,
>> sizeof(unsigned long), head,
>> sizeof(struct page), false);
>> --
>> 2.26.2
>>
>>
> .
>
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2021-05-12 3:23 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-05-12 3:10 [PATCH 1/3] mm/page_alloc: Use dump_page() directly Kefeng Wang
2021-05-12 3:10 ` [PATCH 2/3] mm: Make __dump_page() static Kefeng Wang
2021-05-12 3:10 ` [PATCH 3/3] mm/debug: Check page poisoned firstly in dump_page() Kefeng Wang
2021-05-12 3:15 ` Matthew Wilcox
2021-05-12 3:23 ` Kefeng Wang
2021-05-12 3:14 ` [PATCH 1/3] mm/page_alloc: Use dump_page() directly Matthew Wilcox
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).