GFS2: Replace PTR_RET with PTR_ERR_OR_ZERO
PATCH 1/1 GFS2: Replace PTR_RET with PTR_ERR_OR_ZERO
PTR_RET is now deprecated. Use PTR_ERR_OR_ZERO instead.
---
Compile tested and based on the following tree:git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux.git (PTR_RET)
Dependent on [1]
[1] http://lkml.indiana.edu/hypermail/linux/kernel/1306.2/00010.html
--- fs/gfs2/inode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)
fs/gfs2/inode.c +2 -1
@@ -19,6 +19,7 @@#include <linux/crc32.h> #include <linux/fiemap.h> #include <linux/security.h> +#include <linux/err.h> #include <asm/uaccess.h> #include "gfs2.h"
@@ -594,7 +595,7 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, }gfs2_glock_dq_uninit (ghs);if (IS_ERR (d)) -return PTR_RET (d); +return PTR_ERR_OR_ZERO (d);return error; }else if (error != -ENOENT ) { goto fail_gunlock;
Signature · 1 line
1.7.9.5
PATCH 1/1 GFS2: Replace PTR_RET with PTR_ERR_OR_ZERO
Hi,
Quoted 26 lines · Sachin Kamat
PTR_RETis now deprecated. UsePTR_ERR_OR_ZEROinstead.Signed-off-by Sachin Kamat <[email protected]>---
Compile tested and based on the following tree:git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux.git(PTR_RET)Dependent on [1]
[1] http://lkml.indiana.edu/hypermail/linux/kernel/1306.2/00010.html--- fs/gfs2/inode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)fs/gfs2/inode.c +1
-0@@ -19,6 +19,7 @@#include <linux/crc32.h> #include <linux/fiemap.h> #include <linux/security.h> +#include <linux/err.h> #include <asm/uaccess.h> #include "gfs2.h"
fs/gfs2/inode.c static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, · line 598
Quoted 5 lines
fs/gfs2/inode.c +1
-1@@ -594,7 +595,7 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, }gfs2_glock_dq_uninit (ghs);if (IS_ERR (d)) -return PTR_RET (d); +return PTR_ERR_OR_ZERO (d);
I'm not sure I follow what this is supposed to be doing... what is the reason for this change? This macro/function doesn't seem to be defined in the current kernel, so I assume that it is "coming soon" but the thread pointed to above wasn't very enlightening,
Steve.
Quoted 3 lines
fs/gfs2/inode.c +0
-0return error; }else if (error != -ENOENT ) { goto fail_gunlock;
PATCH 1/1 GFS2: Replace PTR_RET with PTR_ERR_OR_ZERO
Quoted 27 lines · Steven Whitehouse
Hi,
Quoted 24 lines · Sachin Kamat
PTR_RETis now deprecated. UsePTR_ERR_OR_ZEROinstead.Signed-off-by Sachin Kamat <[email protected]>---
Compile tested and based on the following tree:git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux.git(PTR_RET)Dependent on [1]
[1] http://lkml.indiana.edu/hypermail/linux/kernel/1306.2/00010.html--- fs/gfs2/inode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)fs/gfs2/inode.c +1
-0@@ -19,6 +19,7 @@#include <linux/crc32.h> #include <linux/fiemap.h> #include <linux/security.h> +#include <linux/err.h> #include <asm/uaccess.h>
Quoted 12 lines
fs/gfs2/inode.c static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, · line 598
Quoted 7 lines
fs/gfs2/inode.c +1
-1#include "gfs2.h" @@ -594,7 +595,7 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, }gfs2_glock_dq_uninit (ghs);if (IS_ERR (d)) -return PTR_RET (d); +return PTR_ERR_OR_ZERO (d);I'm not sure I follow what this is supposed to be doing... what is the reason for this change? This macro/function doesn't seem to be defined in the current kernel, so I assume that it is "coming soon" but the thread pointed to above wasn't very enlightening,
It's a clarification rename.
But this fix is wrong, it should just be changed to PTR_ERR(d). It never needed PTR_RET() in the first place.
Thanks,
Rusty.
PATCH 1/1 GFS2: Replace PTR_RET with PTR_ERR_OR_ZERO
Hi,
Quoted 41 lines · Rusty Russell
Quoted 27 lines · Steven Whitehouse
Hi,
Quoted 24 lines · Sachin Kamat
PTR_RETis now deprecated. UsePTR_ERR_OR_ZEROinstead.Signed-off-by Sachin Kamat <[email protected]>---
Compile tested and based on the following tree:git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux.git(PTR_RET)Dependent on [1]
[1] http://lkml.indiana.edu/hypermail/linux/kernel/1306.2/00010.html--- fs/gfs2/inode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)fs/gfs2/inode.c +1
-0@@ -19,6 +19,7 @@#include <linux/crc32.h> #include <linux/fiemap.h> #include <linux/security.h> +#include <linux/err.h> #include <asm/uaccess.h> Quoted 12 lines
fs/gfs2/inode.c static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, · line 598
Quoted 7 lines
fs/gfs2/inode.c +1
-1#include "gfs2.h" @@ -594,7 +595,7 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, }gfs2_glock_dq_uninit (ghs);if (IS_ERR (d)) -return PTR_RET (d); +return PTR_ERR_OR_ZERO (d);I'm not sure I follow what this is supposed to be doing... what is the reason for this change? This macro/function doesn't seem to be defined in the current kernel, so I assume that it is "coming soon" but the thread pointed to above wasn't very enlightening,
Quoted 7 lines
It's a clarification rename.
But this fix is wrong, it should just be changed to
PTR_ERR(d). It never neededPTR_RET()in the first place.Thanks,
Rusty.
Ok, thanks for clarifying. I've sorted out a patch, attached below, which I'll put in my tree unless there are any objections. I don't think that it should affect your patch series, but let me know if there is a problem,
Steve.
From 1067f2a5b96d11c2c3dd5ba83e3969cc5ed51b50 Mon Sep 17 00:00:00 2001
From: Steven Whitehouse <[email protected]>
Date: Wed, 17 Jul 2013 08:11:32 +0100
Subject: [PATCH] GFS2: Fix typo in gfs2_create_inode()
PTR_RET should be PTR_ERR
fs/gfs2/inode.c +1 -1
@@ -594,7 +594,7 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, }gfs2_glock_dq_uninit (ghs);if (IS_ERR (d)) -return PTR_RET (d); +return PTR_ERR (d);return error; }else if (error != -ENOENT ) { goto fail_gunlock;
Signature · 1 line
1.7.4
Rusty Russell
AppliedPATCH 1/1 GFS2: Replace PTR_RET with PTR_ERR_OR_ZERO
Quoted 85 lines · Steven Whitehouse
Hi,
Quoted 41 lines · Rusty Russell
Quoted 27 lines · Steven Whitehouse
Hi,
Quoted 24 lines · Sachin Kamat
PTR_RETis now deprecated. UsePTR_ERR_OR_ZEROinstead.Signed-off-by Sachin Kamat <[email protected]>---
Compile tested and based on the following tree:git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux.git(PTR_RET)Dependent on [1]
[1] http://lkml.indiana.edu/hypermail/linux/kernel/1306.2/00010.html--- fs/gfs2/inode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)fs/gfs2/inode.c +1
-0@@ -19,6 +19,7 @@#include <linux/crc32.h> #include <linux/fiemap.h> #include <linux/security.h> +#include <linux/err.h> #include <asm/uaccess.h> Quoted 12 lines
fs/gfs2/inode.c static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, · line 598
Quoted 7 lines
fs/gfs2/inode.c +1
-1#include "gfs2.h" @@ -594,7 +595,7 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, }gfs2_glock_dq_uninit (ghs);if (IS_ERR (d)) -return PTR_RET (d); +return PTR_ERR_OR_ZERO (d);I'm not sure I follow what this is supposed to be doing... what is the reason for this change? This macro/function doesn't seem to be defined in the current kernel, so I assume that it is "coming soon" but the thread pointed to above wasn't very enlightening,
Quoted 7 lines
It's a clarification rename.
But this fix is wrong, it should just be changed to
PTR_ERR(d). It never neededPTR_RET()in the first place.Thanks,
Rusty.Ok, thanks for clarifying. I've sorted out a patch, attached below, which I'll put in my tree unless there are any objections. I don't think that it should affect your patch series, but let me know if there is a problem,
Steve.
Quoted 1 line
From
1067f2a5b96d11c2c3dd5ba83e3969cc5ed51b50Mon Sep 17 00:00:00 2001From: Steven Whitehouse <[email protected]>
Date: Wed, 17 Jul 2013 08:11:32 +0100
Subject: [PATCH] GFS2: Fix typo ingfs2_create_inode()
PTR_RETshould bePTR_ERRReported-by Sachin Kamat <[email protected]>Cc Rusty Russell <[email protected]>Signed-off-by Steven Whitehouse <[email protected]>fs/gfs2/inode.c +1
-1@@ -594,7 +594,7 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, }gfs2_glock_dq_uninit (ghs);if (IS_ERR (d)) -return PTR_RET (d); +return PTR_ERR (d);return error; }else if (error != -ENOENT ) { goto fail_gunlock;Signature · 1 line
1.7.4
Thanks, applied to my PTR_RET branch, but you could just put it in your tree too.
Cheers,
Rusty.
PATCH 1/1 GFS2: Replace PTR_RET with PTR_ERR_OR_ZERO
In reply to Steven WhitehouseQuoted 27 lines · Steven Whitehouse
Hi,
Quoted 24 lines · Sachin Kamat
PTR_RETis now deprecated. UsePTR_ERR_OR_ZEROinstead.Signed-off-by Sachin Kamat <[email protected]>---
Compile tested and based on the following tree:git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux.git(PTR_RET)Dependent on [1]
[1] http://lkml.indiana.edu/hypermail/linux/kernel/1306.2/00010.html--- fs/gfs2/inode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)fs/gfs2/inode.c +1
-0@@ -19,6 +19,7 @@#include <linux/crc32.h> #include <linux/fiemap.h> #include <linux/security.h> +#include <linux/err.h> #include <asm/uaccess.h>
Quoted 10 lines
fs/gfs2/inode.c static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, · line 598
Quoted 7 lines
fs/gfs2/inode.c +1
-1#include "gfs2.h" @@ -594,7 +595,7 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, }gfs2_glock_dq_uninit (ghs);if (IS_ERR (d)) -return PTR_RET (d); +return PTR_ERR_OR_ZERO (d);I'm not sure I follow what this is supposed to be doing... what is the reason for this change?
As far as I understand there is no change in the functionality of the code. The function has been renamed to better
reflect its functionality.
Quoted 3 lines
This macro/function doesn't seem to be defined
in the current kernel, so I assume that it is "coming soon" but the thread pointed to above wasn't very enlightening,
Yes, this is currently available in Rusty's tree [1] and now merged into linux-next. Since he has done this replacement in most of the places, i guess he
will pick this patch too with your ack. Rusty?
[1] git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux.git
(branch: PTR_RET)
Signature · 2 lines
With warm regards, Sachin
PATCH 1/1 GFS2: Replace PTR_RET with PTR_ERR_OR_ZERO
Hi,
Quoted 39 lines · Sachin Kamat
Quoted 27 lines · Steven Whitehouse
Hi,
Quoted 24 lines · Sachin Kamat
PTR_RETis now deprecated. UsePTR_ERR_OR_ZEROinstead.Signed-off-by Sachin Kamat <[email protected]>---
Compile tested and based on the following tree:git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux.git(PTR_RET)Dependent on [1]
[1] http://lkml.indiana.edu/hypermail/linux/kernel/1306.2/00010.html--- fs/gfs2/inode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)fs/gfs2/inode.c +1
-0@@ -19,6 +19,7 @@#include <linux/crc32.h> #include <linux/fiemap.h> #include <linux/security.h> +#include <linux/err.h> #include <asm/uaccess.h> Quoted 10 lines
fs/gfs2/inode.c static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, · line 598
Quoted 7 lines
fs/gfs2/inode.c +1
-1#include "gfs2.h" @@ -594,7 +595,7 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, }gfs2_glock_dq_uninit (ghs);if (IS_ERR (d)) -return PTR_RET (d); +return PTR_ERR_OR_ZERO (d);I'm not sure I follow what this is supposed to be doing... what is the reason for this change?
Quoted 4 lines
As far as I understand there is no change in the functionality of the code. The function has been renamed to better
reflect its functionality.
I think that the problem is that PTR_RET is the wrong function in the first place. It should be PTR_ERR rather than PTR_RET, since we already know that the value is an error value,
Steve.
Quoted 12 lines
Quoted 3 lines
This macro/function doesn't seem to be defined
in the current kernel, so I assume that it is "coming soon" but the thread pointed to above wasn't very enlightening,Yes, this is currently available in Rusty's tree [1] and now merged into linux-next. Since he has done this replacement in most of the places, i guess he
will pick this patch too with your ack. Rusty?[1]
git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux.git
(branch:PTR_RET)
PATCH 1/1 GFS2: Replace PTR_RET with PTR_ERR_OR_ZERO
Quoted 42 lines · Steven Whitehouse
Hi,
Quoted 39 lines · Sachin Kamat
Quoted 37 lines · Steven Whitehouse
Hi,
Quoted 26 lines · Sachin Kamat
PTR_RETis now deprecated. UsePTR_ERR_OR_ZEROinstead.Signed-off-by Sachin Kamat <[email protected]>---
Compile tested and based on the following tree:git://git.kernel.org/pub/scm/linux/kernel/git/rusty/linux.git(PTR_RET)Dependent on [1]
[1] http://lkml.indiana.edu/hypermail/linux/kernel/1306.2/00010.html--- fs/gfs2/inode.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)fs/gfs2/inode.c +1
-0@@ -19,6 +19,7 @@#include <linux/crc32.h> #include <linux/fiemap.h> #include <linux/security.h> +#include <linux/err.h> #include <asm/uaccess.h> #include "gfs2.h" fs/gfs2/inode.c static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, · line 598
Quoted 5 lines
fs/gfs2/inode.c +1
-1@@ -594,7 +595,7 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, }gfs2_glock_dq_uninit (ghs);if (IS_ERR (d)) -return PTR_RET (d); +return PTR_ERR_OR_ZERO (d);I'm not sure I follow what this is supposed to be doing... what is the reason for this change?
Quoted 7 lines
Quoted 4 lines
As far as I understand there is no change in the functionality of the code. The function has been renamed to better
reflect its functionality.I think that the problem is that
PTR_RETis the wrong function in the first place. It should bePTR_ERRrather thanPTR_RET, since we already know that the value is an error value,
Yes you are right. I somehow overlooked this. It should just be PTR_ERR(). Please ignore this patch.
Signature · 2 lines
With warm regards, Sachin