Skip to contentSkip to conversation

debugfs: remove unused variable

PATCH In review Eric Engestrom 3 participants 8 messages Sep 20, 2016 – Sep 21, 2016 Driver core

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 debugfs: remove unused variable

---
 fs/debugfs/file.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/debugfs/file.c +1 -2
index 592059f..04eca0b 100644
@@ -195,7 +195,6 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	const struct dentry *dentry = F_DENTRY(filp);
 	const struct file_operations *real_fops = REAL_FOPS_DEREF(dentry);
 	const struct file_operations *proxy_fops = filp->f_op;
-	int r = 0;
 
 	/*
 	 * We must not protect this against removal races here: the
@@ -204,7 +203,7 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	 * ->i_private is still being meaningful here.
 	 */
 	if (real_fops->release)
-		r = real_fops->release(inode, filp);
+		real_fops->release(inode, filp);
 
 	replace_fops(filp, d_inode(dentry)->i_fop);
 	kfree((void *)proxy_fops);
Signature · 2 lines
Cheers,
  Eric
#1 lore raw
Quoted 18 lines · Eric Engestrom
---
 fs/debugfs/file.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/debugfs/file.c +0 -1
index 592059f..04eca0b 100644
@@ -195,7 +195,6 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	const struct dentry *dentry = F_DENTRY(filp);
 	const struct file_operations *real_fops = REAL_FOPS_DEREF(dentry);
 	const struct file_operations *proxy_fops = filp->f_op;
-	int r = 0;
 
 	/*
 	 * We must not protect this against removal races here: the

fs/debugfs/file.c static int full_proxy_release(struct inode *inode, struct file *filp) · line 206

Quoted 5 lines
fs/debugfs/file.c +1 -1
@@ -204,7 +203,7 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	 * ->i_private is still being meaningful here.
 	 */
 	if (real_fops->release)
-		r = real_fops->release(inode, filp);
+		real_fops->release(inode, filp);

Hm, shouldn't we be propagating the result back up the call chain?

thanks,

greg k-h

#2 lore raw
Quoted 16 lines · Greg Kroah-Hartman
Quoted 15 lines · Eric Engestrom
---
 fs/debugfs/file.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/debugfs/file.c +0 -1
index 592059f..04eca0b 100644
@@ -195,7 +195,6 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	const struct dentry *dentry = F_DENTRY(filp);
 	const struct file_operations *real_fops = REAL_FOPS_DEREF(dentry);
 	const struct file_operations *proxy_fops = filp->f_op;
-	int r = 0;
 
Quoted 10 lines

fs/debugfs/file.c static int full_proxy_release(struct inode *inode, struct file *filp) · line 206

Quoted 8 lines
fs/debugfs/file.c +1 -1
 	/*
 	 * We must not protect this against removal races here: the
@@ -204,7 +203,7 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	 * ->i_private is still being meaningful here.
 	 */
 	if (real_fops->release)
-		r = real_fops->release(inode, filp);
+		real_fops->release(inode, filp);

Hm, shouldn't we be propagating the result back up the call chain?

You're right, sorry, I wasn't thinking. Correct fix incoming :)

Cheers,
Eric

Quoted 4 lines

thanks,

greg k-h

#3 lore raw

PATCH 1/1 debugfs: propagate release() call result

In reply to Greg Kroah-Hartman
---
 fs/debugfs/file.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/debugfs/file.c +1 -1
index 592059f..40e586f 100644
@@ -209,7 +209,7 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	replace_fops(filp, d_inode(dentry)->i_fop);
 	kfree((void *)proxy_fops);
 	fops_put(real_fops);
-	return 0;
+	return r;
 }
 
 static void __full_proxy_fops_init(struct file_operations *proxy_fops,
Signature · 2 lines
Cheers,
  Eric
#5 lore raw

PATCH v2 1/1 debugfs: propagate release() call result

The result was being ignored and 0 was always returned. Return the actual result instead.

---
v2: add a commit message

---
 fs/debugfs/file.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/debugfs/file.c +1 -1
index 592059f..40e586f 100644
@@ -209,7 +209,7 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	replace_fops(filp, d_inode(dentry)->i_fop);
 	kfree((void *)proxy_fops);
 	fops_put(real_fops);
-	return 0;
+	return r;
 }
 
 static void __full_proxy_fops_init(struct file_operations *proxy_fops,
Signature · 2 lines
Cheers,
  Eric
Reviewed-by
#6 lore raw
In reply to Greg Kroah-Hartman
Quoted 16 lines · Greg Kroah-Hartman
Quoted 15 lines · Eric Engestrom
---
 fs/debugfs/file.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/debugfs/file.c +0 -1
index 592059f..04eca0b 100644
@@ -195,7 +195,6 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	const struct dentry *dentry = F_DENTRY(filp);
 	const struct file_operations *real_fops = REAL_FOPS_DEREF(dentry);
 	const struct file_operations *proxy_fops = filp->f_op;
-	int r = 0;
 
Quoted 10 lines

fs/debugfs/file.c static int full_proxy_release(struct inode *inode, struct file *filp) · line 206

Quoted 8 lines
fs/debugfs/file.c +1 -1
 	/*
 	 * We must not protect this against removal races here: the
@@ -204,7 +203,7 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	 * ->i_private is still being meaningful here.
 	 */
 	if (real_fops->release)
-		r = real_fops->release(inode, filp);
+		real_fops->release(inode, filp);

Hm, shouldn't we be propagating the result back up the call chain?

AFAICS, the VFS layer doesn't ever evaluate the return value of ->release(), c.f. __fput() in fs/file_table.c .

OTOH, propagating that value back to caller also wouldn't hurt. But this would be a matter of taste/coding style.

I can't remember whether I left this unused int r there on purpose. I doubt not. Eric, did you run your patch through sparse and Coccinelle?

If so,

for the diff. (This patch lacks a description though.)

Thanks,

Nicolai

#7 lore raw
Quoted 29 lines · Nicolai Stange
Quoted 16 lines · Greg Kroah-Hartman
Quoted 15 lines · Eric Engestrom
---
 fs/debugfs/file.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/debugfs/file.c +0 -1
index 592059f..04eca0b 100644
@@ -195,7 +195,6 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	const struct dentry *dentry = F_DENTRY(filp);
 	const struct file_operations *real_fops = REAL_FOPS_DEREF(dentry);
 	const struct file_operations *proxy_fops = filp->f_op;
-	int r = 0;
 
Quoted 10 lines

fs/debugfs/file.c static int full_proxy_release(struct inode *inode, struct file *filp) · line 206

Quoted 8 lines
fs/debugfs/file.c +1 -1
 	/*
 	 * We must not protect this against removal races here: the
@@ -204,7 +203,7 @@ static int full_proxy_release(struct inode *inode, struct file *filp)
 	 * ->i_private is still being meaningful here.
 	 */
 	if (real_fops->release)
-		r = real_fops->release(inode, filp);
+		real_fops->release(inode, filp);

Hm, shouldn't we be propagating the result back up the call chain?

Quoted 5 lines

AFAICS, the VFS layer doesn't ever evaluate the return value of ->release(), c.f. __fput() in fs/file_table.c .

OTOH, propagating that value back to caller also wouldn't hurt. But this would be a matter of taste/coding style.

I actually sent an updated fix [1] about an hour ago, which propagates the result instead (which is better IMO, I don't know why I didn't do that the first time around).

[1] http://marc.info/?m=147444718118891 (lkml.org is down?)

Quoted 3 lines

I can't remember whether I left this unused int r there on purpose. I doubt not. Eric, did you run your patch through sparse and Coccinelle?

I didn't; how do I do that? I know these tools, but not how to use them in this context.

Cheers,
Eric

Quoted 11 lines

If so,

for the diff. (This patch lacks a description though.)

Thanks,

Nicolai