From: Zhen Lei <[email protected]>
The local variable 'page_poisoning_requested' is assigned true at only
one point. It can be eliminated by moving the code that depends on it
to the location where it is assigned true. This also make the moved
code to be compiled only if CONFIG_PAGE_POISONING is set.
Signed-off-by: Zhen Lei <[email protected]>
---
mm/mm_init.c | 17 +++++++----------
1 file changed, 7 insertions(+), 10 deletions(-)
diff --git a/mm/mm_init.c b/mm/mm_init.c
index 549e76af8f82a8e..3eb217130bcb2b5 100644
--- a/mm/mm_init.c
+++ b/mm/mm_init.c
@@ -2614,7 +2614,6 @@ DEFINE_STATIC_KEY_MAYBE(CONFIG_DEBUG_VM, check_pages_enabled);
*/
static void __init mem_debugging_and_hardening_init(void)
{
- bool page_poisoning_requested = false;
bool want_check_pages = false;
#ifdef CONFIG_PAGE_POISONING
@@ -2626,18 +2625,16 @@ static void __init mem_debugging_and_hardening_init(void)
(!IS_ENABLED(CONFIG_ARCH_SUPPORTS_DEBUG_PAGEALLOC) &&
debug_pagealloc_enabled())) {
static_branch_enable(&_page_poisoning_enabled);
- page_poisoning_requested = true;
want_check_pages = true;
- }
-#endif
- if ((_init_on_alloc_enabled_early || _init_on_free_enabled_early) &&
- page_poisoning_requested) {
- pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
- "will take precedence over init_on_alloc and init_on_free\n");
- _init_on_alloc_enabled_early = false;
- _init_on_free_enabled_early = false;
+ if (_init_on_alloc_enabled_early || _init_on_free_enabled_early) {
+ pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
+ "will take precedence over init_on_alloc and init_on_free\n");
+ _init_on_alloc_enabled_early = false;
+ _init_on_free_enabled_early = false;
+ }
}
+#endif
if (_init_on_alloc_enabled_early) {
want_check_pages = true;
--
2.34.1
On Mon, Mar 18, 2024 at 09:57:14PM +0800, [email protected] wrote:
> From: Zhen Lei <[email protected]>
>
> The local variable 'page_poisoning_requested' is assigned true at only
> one point. It can be eliminated by moving the code that depends on it
> to the location where it is assigned true. This also make the moved
> code to be compiled only if CONFIG_PAGE_POISONING is set.
I don't see it as much of an improvement and code readability becomes worse
IMO.
> Signed-off-by: Zhen Lei <[email protected]>
> ---
> mm/mm_init.c | 17 +++++++----------
> 1 file changed, 7 insertions(+), 10 deletions(-)
>
> diff --git a/mm/mm_init.c b/mm/mm_init.c
> index 549e76af8f82a8e..3eb217130bcb2b5 100644
> --- a/mm/mm_init.c
> +++ b/mm/mm_init.c
> @@ -2614,7 +2614,6 @@ DEFINE_STATIC_KEY_MAYBE(CONFIG_DEBUG_VM, check_pages_enabled);
> */
> static void __init mem_debugging_and_hardening_init(void)
> {
> - bool page_poisoning_requested = false;
> bool want_check_pages = false;
>
> #ifdef CONFIG_PAGE_POISONING
> @@ -2626,18 +2625,16 @@ static void __init mem_debugging_and_hardening_init(void)
> (!IS_ENABLED(CONFIG_ARCH_SUPPORTS_DEBUG_PAGEALLOC) &&
> debug_pagealloc_enabled())) {
> static_branch_enable(&_page_poisoning_enabled);
> - page_poisoning_requested = true;
> want_check_pages = true;
> - }
> -#endif
>
> - if ((_init_on_alloc_enabled_early || _init_on_free_enabled_early) &&
> - page_poisoning_requested) {
> - pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
> - "will take precedence over init_on_alloc and init_on_free\n");
> - _init_on_alloc_enabled_early = false;
> - _init_on_free_enabled_early = false;
> + if (_init_on_alloc_enabled_early || _init_on_free_enabled_early) {
> + pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
> + "will take precedence over init_on_alloc and init_on_free\n");
> + _init_on_alloc_enabled_early = false;
> + _init_on_free_enabled_early = false;
> + }
> }
> +#endif
>
> if (_init_on_alloc_enabled_early) {
> want_check_pages = true;
> --
> 2.34.1
>
--
Sincerely yours,
Mike.
On 2024/3/19 0:33, Mike Rapoport wrote:
> On Mon, Mar 18, 2024 at 09:57:14PM +0800, [email protected] wrote:
>> From: Zhen Lei <[email protected]>
>>
>> The local variable 'page_poisoning_requested' is assigned true at only
>> one point. It can be eliminated by moving the code that depends on it
>> to the location where it is assigned true. This also make the moved
>> code to be compiled only if CONFIG_PAGE_POISONING is set.
>
> I don't see it as much of an improvement and code readability becomes worse
> IMO.
Yes, the moved branch will be optimized by the compiler if CONFIG_PAGE_POISONING is
not set. But for a reader, he can simply skip over that moved branch when
CONFIG_PAGE_POISONING is not set.
>
>> Signed-off-by: Zhen Lei <[email protected]>
>> ---
>> mm/mm_init.c | 17 +++++++----------
>> 1 file changed, 7 insertions(+), 10 deletions(-)
>>
>> diff --git a/mm/mm_init.c b/mm/mm_init.c
>> index 549e76af8f82a8e..3eb217130bcb2b5 100644
>> --- a/mm/mm_init.c
>> +++ b/mm/mm_init.c
>> @@ -2614,7 +2614,6 @@ DEFINE_STATIC_KEY_MAYBE(CONFIG_DEBUG_VM, check_pages_enabled);
>> */
>> static void __init mem_debugging_and_hardening_init(void)
>> {
>> - bool page_poisoning_requested = false;
>> bool want_check_pages = false;
>>
>> #ifdef CONFIG_PAGE_POISONING
>> @@ -2626,18 +2625,16 @@ static void __init mem_debugging_and_hardening_init(void)
>> (!IS_ENABLED(CONFIG_ARCH_SUPPORTS_DEBUG_PAGEALLOC) &&
>> debug_pagealloc_enabled())) {
>> static_branch_enable(&_page_poisoning_enabled);
>> - page_poisoning_requested = true;
>> want_check_pages = true;
>> - }
>> -#endif
>>
>> - if ((_init_on_alloc_enabled_early || _init_on_free_enabled_early) &&
>> - page_poisoning_requested) {
>> - pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
>> - "will take precedence over init_on_alloc and init_on_free\n");
>> - _init_on_alloc_enabled_early = false;
>> - _init_on_free_enabled_early = false;
>> + if (_init_on_alloc_enabled_early || _init_on_free_enabled_early) {
>> + pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
>> + "will take precedence over init_on_alloc and init_on_free\n");
>> + _init_on_alloc_enabled_early = false;
>> + _init_on_free_enabled_early = false;
>> + }
>> }
>> +#endif
>>
>> if (_init_on_alloc_enabled_early) {
>> want_check_pages = true;
>> --
>> 2.34.1
>>
>
--
Regards,
Zhen Lei
On 3/18/24 22:03, Mike Rapoport wrote:
> On Mon, Mar 18, 2024 at 09:57:14PM +0800, [email protected] wrote:
>> From: Zhen Lei <[email protected]>
>>
>> The local variable 'page_poisoning_requested' is assigned true at only
>> one point. It can be eliminated by moving the code that depends on it
>> to the location where it is assigned true. This also make the moved
>> code to be compiled only if CONFIG_PAGE_POISONING is set.
>
> I don't see it as much of an improvement and code readability becomes worse
> IMO.
+1 agreed.
>
>> Signed-off-by: Zhen Lei <[email protected]>
>> ---
>> mm/mm_init.c | 17 +++++++----------
>> 1 file changed, 7 insertions(+), 10 deletions(-)
>>
>> diff --git a/mm/mm_init.c b/mm/mm_init.c
>> index 549e76af8f82a8e..3eb217130bcb2b5 100644
>> --- a/mm/mm_init.c
>> +++ b/mm/mm_init.c
>> @@ -2614,7 +2614,6 @@ DEFINE_STATIC_KEY_MAYBE(CONFIG_DEBUG_VM, check_pages_enabled);
>> */
>> static void __init mem_debugging_and_hardening_init(void)
>> {
>> - bool page_poisoning_requested = false;
>> bool want_check_pages = false;
>>
>> #ifdef CONFIG_PAGE_POISONING
>> @@ -2626,18 +2625,16 @@ static void __init mem_debugging_and_hardening_init(void)
>> (!IS_ENABLED(CONFIG_ARCH_SUPPORTS_DEBUG_PAGEALLOC) &&
>> debug_pagealloc_enabled())) {
>> static_branch_enable(&_page_poisoning_enabled);
>> - page_poisoning_requested = true;
>> want_check_pages = true;
>> - }
>> -#endif
>>
>> - if ((_init_on_alloc_enabled_early || _init_on_free_enabled_early) &&
>> - page_poisoning_requested) {
>> - pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
>> - "will take precedence over init_on_alloc and init_on_free\n");
>> - _init_on_alloc_enabled_early = false;
>> - _init_on_free_enabled_early = false;
>> + if (_init_on_alloc_enabled_early || _init_on_free_enabled_early) {
>> + pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
>> + "will take precedence over init_on_alloc and init_on_free\n");
>> + _init_on_alloc_enabled_early = false;
>> + _init_on_free_enabled_early = false;
>> + }
>> }
>> +#endif
>>
>> if (_init_on_alloc_enabled_early) {
>> want_check_pages = true;
>> --
>> 2.34.1
>>
>
On Tue, Mar 19, 2024 at 09:22:03AM +0800, Leizhen (ThunderTown) wrote:
>
>
> On 2024/3/19 0:33, Mike Rapoport wrote:
> > On Mon, Mar 18, 2024 at 09:57:14PM +0800, [email protected] wrote:
> >> From: Zhen Lei <[email protected]>
> >>
> >> The local variable 'page_poisoning_requested' is assigned true at only
> >> one point. It can be eliminated by moving the code that depends on it
> >> to the location where it is assigned true. This also make the moved
> >> code to be compiled only if CONFIG_PAGE_POISONING is set.
> >
> > I don't see it as much of an improvement and code readability becomes worse
> > IMO.
>
> Yes, the moved branch will be optimized by the compiler if CONFIG_PAGE_POISONING is
> not set. But for a reader, he can simply skip over that moved branch when
> CONFIG_PAGE_POISONING is not set.
Saving one branch at init does not justify the churn and reduced
readability.
> >> Signed-off-by: Zhen Lei <[email protected]>
> >> ---
> >> mm/mm_init.c | 17 +++++++----------
> >> 1 file changed, 7 insertions(+), 10 deletions(-)
> >>
> >> diff --git a/mm/mm_init.c b/mm/mm_init.c
> >> index 549e76af8f82a8e..3eb217130bcb2b5 100644
> >> --- a/mm/mm_init.c
> >> +++ b/mm/mm_init.c
> >> @@ -2614,7 +2614,6 @@ DEFINE_STATIC_KEY_MAYBE(CONFIG_DEBUG_VM, check_pages_enabled);
> >> */
> >> static void __init mem_debugging_and_hardening_init(void)
> >> {
> >> - bool page_poisoning_requested = false;
> >> bool want_check_pages = false;
> >>
> >> #ifdef CONFIG_PAGE_POISONING
> >> @@ -2626,18 +2625,16 @@ static void __init mem_debugging_and_hardening_init(void)
> >> (!IS_ENABLED(CONFIG_ARCH_SUPPORTS_DEBUG_PAGEALLOC) &&
> >> debug_pagealloc_enabled())) {
> >> static_branch_enable(&_page_poisoning_enabled);
> >> - page_poisoning_requested = true;
> >> want_check_pages = true;
> >> - }
> >> -#endif
> >>
> >> - if ((_init_on_alloc_enabled_early || _init_on_free_enabled_early) &&
> >> - page_poisoning_requested) {
> >> - pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
> >> - "will take precedence over init_on_alloc and init_on_free\n");
> >> - _init_on_alloc_enabled_early = false;
> >> - _init_on_free_enabled_early = false;
> >> + if (_init_on_alloc_enabled_early || _init_on_free_enabled_early) {
> >> + pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
> >> + "will take precedence over init_on_alloc and init_on_free\n");
> >> + _init_on_alloc_enabled_early = false;
> >> + _init_on_free_enabled_early = false;
> >> + }
> >> }
> >> +#endif
> >>
> >> if (_init_on_alloc_enabled_early) {
> >> want_check_pages = true;
> >> --
> >> 2.34.1
> >>
> >
>
> --
> Regards,
> Zhen Lei
>
--
Sincerely yours,
Mike.
On 2024/3/19 23:01, Mike Rapoport wrote:
> On Tue, Mar 19, 2024 at 09:22:03AM +0800, Leizhen (ThunderTown) wrote:
>>
>>
>> On 2024/3/19 0:33, Mike Rapoport wrote:
>>> On Mon, Mar 18, 2024 at 09:57:14PM +0800, [email protected] wrote:
>>>> From: Zhen Lei <[email protected]>
>>>>
>>>> The local variable 'page_poisoning_requested' is assigned true at only
>>>> one point. It can be eliminated by moving the code that depends on it
>>>> to the location where it is assigned true. This also make the moved
>>>> code to be compiled only if CONFIG_PAGE_POISONING is set.
>>>
>>> I don't see it as much of an improvement and code readability becomes worse
>>> IMO.
>>
>> Yes, the moved branch will be optimized by the compiler if CONFIG_PAGE_POISONING is
>> not set. But for a reader, he can simply skip over that moved branch when
>> CONFIG_PAGE_POISONING is not set.
>
> Saving one branch at init does not justify the churn and reduced
> readability.
OK, I got it.
>
>>>> Signed-off-by: Zhen Lei <[email protected]>
>>>> ---
>>>> mm/mm_init.c | 17 +++++++----------
>>>> 1 file changed, 7 insertions(+), 10 deletions(-)
>>>>
>>>> diff --git a/mm/mm_init.c b/mm/mm_init.c
>>>> index 549e76af8f82a8e..3eb217130bcb2b5 100644
>>>> --- a/mm/mm_init.c
>>>> +++ b/mm/mm_init.c
>>>> @@ -2614,7 +2614,6 @@ DEFINE_STATIC_KEY_MAYBE(CONFIG_DEBUG_VM, check_pages_enabled);
>>>> */
>>>> static void __init mem_debugging_and_hardening_init(void)
>>>> {
>>>> - bool page_poisoning_requested = false;
>>>> bool want_check_pages = false;
>>>>
>>>> #ifdef CONFIG_PAGE_POISONING
>>>> @@ -2626,18 +2625,16 @@ static void __init mem_debugging_and_hardening_init(void)
>>>> (!IS_ENABLED(CONFIG_ARCH_SUPPORTS_DEBUG_PAGEALLOC) &&
>>>> debug_pagealloc_enabled())) {
>>>> static_branch_enable(&_page_poisoning_enabled);
>>>> - page_poisoning_requested = true;
>>>> want_check_pages = true;
>>>> - }
>>>> -#endif
>>>>
>>>> - if ((_init_on_alloc_enabled_early || _init_on_free_enabled_early) &&
>>>> - page_poisoning_requested) {
>>>> - pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
>>>> - "will take precedence over init_on_alloc and init_on_free\n");
>>>> - _init_on_alloc_enabled_early = false;
>>>> - _init_on_free_enabled_early = false;
>>>> + if (_init_on_alloc_enabled_early || _init_on_free_enabled_early) {
>>>> + pr_info("mem auto-init: CONFIG_PAGE_POISONING is on, "
>>>> + "will take precedence over init_on_alloc and init_on_free\n");
>>>> + _init_on_alloc_enabled_early = false;
>>>> + _init_on_free_enabled_early = false;
>>>> + }
>>>> }
>>>> +#endif
>>>>
>>>> if (_init_on_alloc_enabled_early) {
>>>> want_check_pages = true;
>>>> --
>>>> 2.34.1
>>>>
>>>
>>
>> --
>> Regards,
>> Zhen Lei
>>
>
--
Regards,
Zhen Lei