Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1433840 > unrolled thread

[PATCH] x86/xen: Use DIV_ROUND_UP

Started byAmitoj Kaur Chawla <amitoj1606@gmail.com>
First post2016-06-29 17:10 +0200
Last post2016-06-29 17:50 +0200
Articles 4 — 4 participants

Back to article view | Back to linux.kernel

Contents

  [PATCH] x86/xen: Use DIV_ROUND_UP Amitoj Kaur Chawla <amitoj1606@gmail.com> - 2016-06-29 17:10 +0200
    Re: [Xen-devel] [PATCH] x86/xen: Use DIV_ROUND_UP Juergen Gross <jgross@suse.com> - 2016-06-29 17:40 +0200
    Re: [Xen-devel] [PATCH] x86/xen: Use DIV_ROUND_UP "Jan Beulich" <JBeulich@suse.com> - 2016-06-29 17:40 +0200
    Re: [Xen-devel] [PATCH] x86/xen: Use DIV_ROUND_UP David Vrabel <david.vrabel@citrix.com> - 2016-06-29 17:50 +0200

#1433840 — [PATCH] x86/xen: Use DIV_ROUND_UP

FromAmitoj Kaur Chawla <amitoj1606@gmail.com>
Date2016-06-29 17:10 +0200
Subject[PATCH] x86/xen: Use DIV_ROUND_UP
Message-ID<rPphL-5KQ-15@gated-at.bofh.it>
The kernel.h macro DIV_ROUND_UP performs the computation
(((n) + (d) - 1) /(d)) but is perhaps more readable.

The Coccinelle script used to make this change is as follows:
@haskernel@
@@

#include <linux/kernel.h>

@depends on haskernel@
expression n,d;
@@

(
- (n + d - 1) / d
+ DIV_ROUND_UP(n,d)
|
- (n + (d - 1)) / d
+ DIV_ROUND_UP(n,d)
)

Signed-off-by: Amitoj Kaur Chawla <amitoj1606@gmail.com>
---
 arch/x86/xen/enlighten.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index 880862c..6847512 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -591,7 +591,7 @@ static void xen_load_gdt(const struct desc_ptr *dtr)
 {
 	unsigned long va = dtr->address;
 	unsigned int size = dtr->size + 1;
-	unsigned pages = (size + PAGE_SIZE - 1) / PAGE_SIZE;
+	unsigned pages = DIV_ROUND_UP(size, PAGE_SIZE);
 	unsigned long frames[pages];
 	int f;
 
@@ -640,7 +640,7 @@ static void __init xen_load_gdt_boot(const struct desc_ptr *dtr)
 {
 	unsigned long va = dtr->address;
 	unsigned int size = dtr->size + 1;
-	unsigned pages = (size + PAGE_SIZE - 1) / PAGE_SIZE;
+	unsigned pages = DIV_ROUND_UP(size, PAGE_SIZE);
 	unsigned long frames[pages];
 	int f;
 
-- 
1.9.1

[toc] | [next] | [standalone]

#1433858 — Re: [Xen-devel] [PATCH] x86/xen: Use DIV_ROUND_UP

FromJuergen Gross <jgross@suse.com>
Date2016-06-29 17:40 +0200
SubjectRe: [Xen-devel] [PATCH] x86/xen: Use DIV_ROUND_UP
Message-ID<rPpKO-5V4-11@gated-at.bofh.it>
In reply to#1433840
On 29/06/16 17:34, Jan Beulich wrote:
>>>> On 29.06.16 at 17:00, <amitoj1606@gmail.com> wrote:
>> --- a/arch/x86/xen/enlighten.c
>> +++ b/arch/x86/xen/enlighten.c
>> @@ -591,7 +591,7 @@ static void xen_load_gdt(const struct desc_ptr *dtr)
>>  {
>>  	unsigned long va = dtr->address;
>>  	unsigned int size = dtr->size + 1;
>> -	unsigned pages = (size + PAGE_SIZE - 1) / PAGE_SIZE;
>> +	unsigned pages = DIV_ROUND_UP(size, PAGE_SIZE);
>>  	unsigned long frames[pages];
>>  	int f;
>>  
>> @@ -640,7 +640,7 @@ static void __init xen_load_gdt_boot(const struct desc_ptr *dtr)
>>  {
>>  	unsigned long va = dtr->address;
>>  	unsigned int size = dtr->size + 1;
>> -	unsigned pages = (size + PAGE_SIZE - 1) / PAGE_SIZE;
>> +	unsigned pages = DIV_ROUND_UP(size, PAGE_SIZE);
>>  	unsigned long frames[pages];
>>  	int f;
>>  
> 
> Perhaps even more readable would be PFN_DOWN()?

Or PFN_UP() to be correct?


Juergen

[toc] | [prev] | [next] | [standalone]

#1433863 — Re: [Xen-devel] [PATCH] x86/xen: Use DIV_ROUND_UP

From"Jan Beulich" <JBeulich@suse.com>
Date2016-06-29 17:40 +0200
SubjectRe: [Xen-devel] [PATCH] x86/xen: Use DIV_ROUND_UP
Message-ID<rPpKO-5V4-31@gated-at.bofh.it>
In reply to#1433840
>>> On 29.06.16 at 17:00, <amitoj1606@gmail.com> wrote:
> --- a/arch/x86/xen/enlighten.c
> +++ b/arch/x86/xen/enlighten.c
> @@ -591,7 +591,7 @@ static void xen_load_gdt(const struct desc_ptr *dtr)
>  {
>  	unsigned long va = dtr->address;
>  	unsigned int size = dtr->size + 1;
> -	unsigned pages = (size + PAGE_SIZE - 1) / PAGE_SIZE;
> +	unsigned pages = DIV_ROUND_UP(size, PAGE_SIZE);
>  	unsigned long frames[pages];
>  	int f;
>  
> @@ -640,7 +640,7 @@ static void __init xen_load_gdt_boot(const struct desc_ptr *dtr)
>  {
>  	unsigned long va = dtr->address;
>  	unsigned int size = dtr->size + 1;
> -	unsigned pages = (size + PAGE_SIZE - 1) / PAGE_SIZE;
> +	unsigned pages = DIV_ROUND_UP(size, PAGE_SIZE);
>  	unsigned long frames[pages];
>  	int f;
>  

Perhaps even more readable would be PFN_DOWN()?

Jan

[toc] | [prev] | [next] | [standalone]

#1433866 — Re: [Xen-devel] [PATCH] x86/xen: Use DIV_ROUND_UP

FromDavid Vrabel <david.vrabel@citrix.com>
Date2016-06-29 17:50 +0200
SubjectRe: [Xen-devel] [PATCH] x86/xen: Use DIV_ROUND_UP
Message-ID<rPpUu-5Yx-7@gated-at.bofh.it>
In reply to#1433840
On 29/06/16 16:00, Amitoj Kaur Chawla wrote:
> The kernel.h macro DIV_ROUND_UP performs the computation
> (((n) + (d) - 1) /(d)) but is perhaps more readable.
> 
> The Coccinelle script used to make this change is as follows:
> @haskernel@
> @@
> 
> #include <linux/kernel.h>
> 
> @depends on haskernel@
> expression n,d;
> @@
> 
> (
> - (n + d - 1) / d
> + DIV_ROUND_UP(n,d)
> |
> - (n + (d - 1)) / d
> + DIV_ROUND_UP(n,d)
> )

Applied to for-linus-4.8, thanks.

PFN_UP/DOWN() are for converting addresses to PFNs.  DIV_ROUND_UP() is
clearer when converting sizes to numbers of pages (as demonstrated by
the incorrect suggestion to use PFN_DOWN()).

David

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web