* [PATCH 0/6] fs: fix up error handlers for insert_inode_locked
@ 2011-12-06 22:55 Eric Sandeen
2011-12-06 22:57 ` [PATCH 1/6] ext2: fix up error handling " Eric Sandeen
` (6 more replies)
0 siblings, 7 replies; 12+ messages in thread
From: Eric Sandeen @ 2011-12-06 22:55 UTC (permalink / raw)
To: linux-fsdevel@vger.kernel.org
after 250df6ed274d767da844a5d9f05720b804240197
(fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
no longer returns the inode with I_NEW set. However, error handlers
still call unlock_new_inode() on failure, which does a WARN_ON if
I_NEW is not set, so any failure spews a lot of warnings.
Patches follow to skip the call to unlock_new_inode() if
insert_inode_locked() fails.
-Eric
^ permalink raw reply [flat|nested] 12+ messages in thread
* [PATCH 1/6] ext2: fix up error handling for insert_inode_locked
2011-12-06 22:55 [PATCH 0/6] fs: fix up error handlers for insert_inode_locked Eric Sandeen
@ 2011-12-06 22:57 ` Eric Sandeen
2011-12-06 23:18 ` [PATCH 1/6 v2] " Eric Sandeen
2011-12-06 22:59 ` [PATCH 2/6] ext3: " Eric Sandeen
` (5 subsequent siblings)
6 siblings, 1 reply; 12+ messages in thread
From: Eric Sandeen @ 2011-12-06 22:57 UTC (permalink / raw)
To: linux-fsdevel@vger.kernel.org; +Cc: ext4 development
after 250df6ed274d767da844a5d9f05720b804240197
(fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
no longer returns the inode with I_NEW set. However, the error
handler still calls unlock_new_inode() on failure, which does a
WARN_ON if I_NEW is not set, so any failure spews a lot of
warnings.
(We also were doing dquot_drop, etc before we had initialized
the inode, that gets skipped as well)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
---
diff --git a/fs/ext2/ialloc.c b/fs/ext2/ialloc.c
index c4e81df..769351a 100644
--- a/fs/ext2/ialloc.c
+++ b/fs/ext2/ialloc.c
@@ -574,7 +574,7 @@ got:
spin_unlock(&sbi->s_next_gen_lock);
if (insert_inode_locked(inode) < 0) {
err = -EINVAL;
- goto fail_drop;
+ goto fail_put;
}
dquot_initialize(inode);
@@ -601,8 +601,9 @@ fail_free_drop:
fail_drop:
dquot_drop(inode);
inode->i_flags |= S_NOQUOTA;
- clear_nlink(inode);
unlock_new_inode(inode);
+fail_put:
+ clear_nlink(inode);
iput(inode);
return ERR_PTR(err);
^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH 2/6] ext3: fix up error handling for insert_inode_locked
2011-12-06 22:55 [PATCH 0/6] fs: fix up error handlers for insert_inode_locked Eric Sandeen
2011-12-06 22:57 ` [PATCH 1/6] ext2: fix up error handling " Eric Sandeen
@ 2011-12-06 22:59 ` Eric Sandeen
2011-12-06 23:19 ` [PATCH 2/6 V2] " Eric Sandeen
2011-12-06 23:02 ` [PATCH 3/6] ext4: " Eric Sandeen
` (4 subsequent siblings)
6 siblings, 1 reply; 12+ messages in thread
From: Eric Sandeen @ 2011-12-06 22:59 UTC (permalink / raw)
To: linux-fsdevel@vger.kernel.org; +Cc: ext4 development
after 250df6ed274d767da844a5d9f05720b804240197
(fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
no longer returns the inode with I_NEW set. However, the error
handler still calls unlock_new_inode() on failure, which does a
WARN_ON if I_NEW is not set, so any failure spews a lot of
warnings.
(We also were doing dquot_drop, etc before we had initialized
the quota, that gets skipped as well)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
---
diff --git a/fs/ext3/ialloc.c b/fs/ext3/ialloc.c
index 5c866e0..7d0493b 100644
--- a/fs/ext3/ialloc.c
+++ b/fs/ext3/ialloc.c
@@ -526,7 +526,7 @@ got:
handle->h_sync = 1;
if (insert_inode_locked(inode) < 0) {
err = -EINVAL;
- goto fail_drop;
+ goto fail_put;
}
spin_lock(&sbi->s_next_gen_lock);
inode->i_generation = sbi->s_next_generation++;
@@ -582,8 +582,9 @@ fail_free_drop:
fail_drop:
dquot_drop(inode);
inode->i_flags |= S_NOQUOTA;
- clear_nlink(inode);
unlock_new_inode(inode);
+fail_put:
+ clear_nlink(inode);
iput(inode);
brelse(bitmap_bh);
return ERR_PTR(err);
^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH 3/6] ext4: fix up error handling for insert_inode_locked
2011-12-06 22:55 [PATCH 0/6] fs: fix up error handlers for insert_inode_locked Eric Sandeen
2011-12-06 22:57 ` [PATCH 1/6] ext2: fix up error handling " Eric Sandeen
2011-12-06 22:59 ` [PATCH 2/6] ext3: " Eric Sandeen
@ 2011-12-06 23:02 ` Eric Sandeen
2011-12-06 23:19 ` [PATCH 3/6 V2] " Eric Sandeen
2011-12-06 23:06 ` [PATCH 4/6] jffs2: " Eric Sandeen
` (3 subsequent siblings)
6 siblings, 1 reply; 12+ messages in thread
From: Eric Sandeen @ 2011-12-06 23:02 UTC (permalink / raw)
To: linux-fsdevel@vger.kernel.org; +Cc: ext4 development
after 250df6ed274d767da844a5d9f05720b804240197
(fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
no longer returns the inode with I_NEW set on failure. However,
the error handler still calls unlock_new_inode() on failure,
which does a WARN_ON if I_NEW is not set, so any failure spews
a lot of warnings.
(We also were doing dquot_drop, etc before we had initialized
the quota, that gets skipped as well)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
---
diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index 00beb4f..f00c19a 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -886,7 +886,7 @@ got:
ext4_handle_sync(handle);
if (insert_inode_locked(inode) < 0) {
err = -EINVAL;
- goto fail_drop;
+ goto fail_put;
}
spin_lock(&sbi->s_next_gen_lock);
inode->i_generation = sbi->s_next_generation++;
@@ -948,8 +948,9 @@ fail_free_drop:
fail_drop:
dquot_drop(inode);
inode->i_flags |= S_NOQUOTA;
- clear_nlink(inode);
unlock_new_inode(inode);
+fail_put:
+ clear_nlink(inode);
iput(inode);
brelse(inode_bitmap_bh);
return ERR_PTR(err);
^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH 4/6] jffs2: fix up error handling for insert_inode_locked
2011-12-06 22:55 [PATCH 0/6] fs: fix up error handlers for insert_inode_locked Eric Sandeen
` (2 preceding siblings ...)
2011-12-06 23:02 ` [PATCH 3/6] ext4: " Eric Sandeen
@ 2011-12-06 23:06 ` Eric Sandeen
2011-12-08 21:54 ` Artem Bityutskiy
2011-12-06 23:10 ` [PATCH 5/6] jfs: " Eric Sandeen
` (2 subsequent siblings)
6 siblings, 1 reply; 12+ messages in thread
From: Eric Sandeen @ 2011-12-06 23:06 UTC (permalink / raw)
To: linux-fsdevel@vger.kernel.org; +Cc: linux-mtd
after 250df6ed274d767da844a5d9f05720b804240197
(fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
no longer returns the inode with I_NEW set on failure. However,
the error handler still calls unlock_new_inode() on failure,
which does a WARN_ON if I_NEW is not set, so any failure spews
a lot of warnings.
We can just drop the unlock_new_inode() if insert_inode_locked()
fails here.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
---
diff --git a/fs/jffs2/fs.c b/fs/jffs2/fs.c
index 4b8afe3..2e01238 100644
--- a/fs/jffs2/fs.c
+++ b/fs/jffs2/fs.c
@@ -466,7 +466,6 @@ struct inode *jffs2_new_inode (struct inode *dir_i, umode_t mode, struct jffs2_r
if (insert_inode_locked(inode) < 0) {
make_bad_inode(inode);
- unlock_new_inode(inode);
iput(inode);
return ERR_PTR(-EINVAL);
}
^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH 5/6] jfs: fix up error handling for insert_inode_locked
2011-12-06 22:55 [PATCH 0/6] fs: fix up error handlers for insert_inode_locked Eric Sandeen
` (3 preceding siblings ...)
2011-12-06 23:06 ` [PATCH 4/6] jffs2: " Eric Sandeen
@ 2011-12-06 23:10 ` Eric Sandeen
2011-12-06 23:12 ` [PATCH 6/6] reiserfs: fix up error handling for insert_inode_locked4 Eric Sandeen
2011-12-06 23:13 ` [PATCH 0/6] fs: fix up error handlers for insert_inode_locked Eric Sandeen
6 siblings, 0 replies; 12+ messages in thread
From: Eric Sandeen @ 2011-12-06 23:10 UTC (permalink / raw)
To: linux-fsdevel@vger.kernel.org; +Cc: jfs-discussion
after 250df6ed274d767da844a5d9f05720b804240197
(fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
no longer returns the inode with I_NEW set on failure. However,
the error handler still calls unlock_new_inode() on failure,
which does a WARN_ON if I_NEW is not set, so any failure spews
a lot of warnings.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
---
diff --git a/fs/jfs/jfs_inode.c b/fs/jfs/jfs_inode.c
index c1a3e60..7f464c5 100644
--- a/fs/jfs/jfs_inode.c
+++ b/fs/jfs/jfs_inode.c
@@ -95,7 +95,7 @@ struct inode *ialloc(struct inode *parent, umode_t mode)
if (insert_inode_locked(inode) < 0) {
rc = -EINVAL;
- goto fail_unlock;
+ goto fail_put;
}
inode_init_owner(inode, parent, mode);
@@ -156,7 +156,6 @@ struct inode *ialloc(struct inode *parent, umode_t mode)
fail_drop:
dquot_drop(inode);
inode->i_flags |= S_NOQUOTA;
-fail_unlock:
clear_nlink(inode);
unlock_new_inode(inode);
fail_put:
^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH 6/6] reiserfs: fix up error handling for insert_inode_locked4
2011-12-06 22:55 [PATCH 0/6] fs: fix up error handlers for insert_inode_locked Eric Sandeen
` (4 preceding siblings ...)
2011-12-06 23:10 ` [PATCH 5/6] jfs: " Eric Sandeen
@ 2011-12-06 23:12 ` Eric Sandeen
2011-12-06 23:13 ` [PATCH 0/6] fs: fix up error handlers for insert_inode_locked Eric Sandeen
6 siblings, 0 replies; 12+ messages in thread
From: Eric Sandeen @ 2011-12-06 23:12 UTC (permalink / raw)
To: linux-fsdevel@vger.kernel.org; +Cc: reiserfs-devel
after 250df6ed274d767da844a5d9f05720b804240197
(fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
no longer returns the inode with I_NEW set on failure. However,
the error handler still calls unlock_new_inode() on failure,
which does a WARN_ON if I_NEW is not set, so any failure spews
a lot of warnings.
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
---
(I had a hard time unraveling the gotos, so punted with an "inode_locked"
var... prettier replacements are fine with me!)
diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index 950f13a..b0cab7c 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -1779,6 +1779,7 @@ int reiserfs_new_inode(struct reiserfs_transaction_handle *th,
struct cpu_key key;
struct item_head ih;
struct stat_data sd;
+ int inode_locked = 0;
int retval;
int err;
@@ -1816,6 +1817,7 @@ int reiserfs_new_inode(struct reiserfs_transaction_handle *th,
err = -EINVAL;
goto out_bad_inode;
}
+ inode_locked = 1;
if (old_format_only(sb))
/* not a perfect generation count, as object ids can be reused, but
** this is as good as reiserfs can do right now.
@@ -1989,7 +1991,8 @@ int reiserfs_new_inode(struct reiserfs_transaction_handle *th,
out_inserted_sd:
clear_nlink(inode);
th->t_trans_id = 0; /* so the caller can't use this handle later */
- unlock_new_inode(inode); /* OK to do even if we hadn't locked it */
+ if (inode_locked)
+ unlock_new_inode(inode);
iput(inode);
return err;
}
^ permalink raw reply related [flat|nested] 12+ messages in thread
* Re: [PATCH 0/6] fs: fix up error handlers for insert_inode_locked
2011-12-06 22:55 [PATCH 0/6] fs: fix up error handlers for insert_inode_locked Eric Sandeen
` (5 preceding siblings ...)
2011-12-06 23:12 ` [PATCH 6/6] reiserfs: fix up error handling for insert_inode_locked4 Eric Sandeen
@ 2011-12-06 23:13 ` Eric Sandeen
6 siblings, 0 replies; 12+ messages in thread
From: Eric Sandeen @ 2011-12-06 23:13 UTC (permalink / raw)
To: linux-fsdevel@vger.kernel.org
On 12/6/11 4:55 PM, Eric Sandeen wrote:
> after 250df6ed274d767da844a5d9f05720b804240197
> (fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
> no longer returns the inode with I_NEW set. However, error handlers
Whoops that should say "with I_NEW set on failure"
Will resend for the fs patches I copied that text to :(
-Eric
> still call unlock_new_inode() on failure, which does a WARN_ON if
> I_NEW is not set, so any failure spews a lot of warnings.
>
> Patches follow to skip the call to unlock_new_inode() if
> insert_inode_locked() fails.
>
> -Eric
> --
> To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
^ permalink raw reply [flat|nested] 12+ messages in thread
* [PATCH 1/6 v2] ext2: fix up error handling for insert_inode_locked
2011-12-06 22:57 ` [PATCH 1/6] ext2: fix up error handling " Eric Sandeen
@ 2011-12-06 23:18 ` Eric Sandeen
0 siblings, 0 replies; 12+ messages in thread
From: Eric Sandeen @ 2011-12-06 23:18 UTC (permalink / raw)
Cc: linux-fsdevel@vger.kernel.org, ext4 development
after 250df6ed274d767da844a5d9f05720b804240197
(fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
no longer returns the inode with I_NEW set on failure. However,
the error handler still calls unlock_new_inode() on failure,
which does a WARN_ON if I_NEW is not set, so any failure spews
a lot of warnings.
(We also were doing dquot_drop, etc before we had initialized
the quota, that gets skipped as well)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
---
V2: don't rearrange clear_nlink, and edit commit message.
diff --git a/fs/ext2/ialloc.c b/fs/ext2/ialloc.c
index c4e81df..06cbc22 100644
--- a/fs/ext2/ialloc.c
+++ b/fs/ext2/ialloc.c
@@ -574,7 +574,7 @@ got:
spin_unlock(&sbi->s_next_gen_lock);
if (insert_inode_locked(inode) < 0) {
err = -EINVAL;
- goto fail_drop;
+ goto fail_put;
}
dquot_initialize(inode);
@@ -603,6 +603,7 @@ fail_drop:
inode->i_flags |= S_NOQUOTA;
clear_nlink(inode);
unlock_new_inode(inode);
+fail_put:
iput(inode);
return ERR_PTR(err);
^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH 2/6 V2] ext3: fix up error handling for insert_inode_locked
2011-12-06 22:59 ` [PATCH 2/6] ext3: " Eric Sandeen
@ 2011-12-06 23:19 ` Eric Sandeen
0 siblings, 0 replies; 12+ messages in thread
From: Eric Sandeen @ 2011-12-06 23:19 UTC (permalink / raw)
To: linux-fsdevel@vger.kernel.org; +Cc: ext4 development
after 250df6ed274d767da844a5d9f05720b804240197
(fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
no longer returns the inode with I_NEW set on failure. However,
the error handler still calls unlock_new_inode() on failure,
which does a WARN_ON if I_NEW is not set, so any failure spews
a lot of warnings.
(We also were doing dquot_drop, etc before we had initialized
the quota, that gets skipped as well)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
---
V2: don't rearrange clear_nlink, and edit commit message.
diff --git a/fs/ext3/ialloc.c b/fs/ext3/ialloc.c
index 5c866e0..5563d99 100644
--- a/fs/ext3/ialloc.c
+++ b/fs/ext3/ialloc.c
@@ -526,7 +526,7 @@ got:
handle->h_sync = 1;
if (insert_inode_locked(inode) < 0) {
err = -EINVAL;
- goto fail_drop;
+ goto fail_put;
}
spin_lock(&sbi->s_next_gen_lock);
inode->i_generation = sbi->s_next_generation++;
@@ -584,6 +584,7 @@ fail_drop:
inode->i_flags |= S_NOQUOTA;
clear_nlink(inode);
unlock_new_inode(inode);
+fail_put:
iput(inode);
brelse(bitmap_bh);
return ERR_PTR(err);
^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH 3/6 V2] ext4: fix up error handling for insert_inode_locked
2011-12-06 23:02 ` [PATCH 3/6] ext4: " Eric Sandeen
@ 2011-12-06 23:19 ` Eric Sandeen
0 siblings, 0 replies; 12+ messages in thread
From: Eric Sandeen @ 2011-12-06 23:19 UTC (permalink / raw)
To: linux-fsdevel@vger.kernel.org; +Cc: ext4 development
after 250df6ed274d767da844a5d9f05720b804240197
(fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
no longer returns the inode with I_NEW set on failure. However,
the error handler still calls unlock_new_inode() on failure,
which does a WARN_ON if I_NEW is not set, so any failure spews
a lot of warnings.
(We also were doing dquot_drop, etc before we had initialized
the quota, that gets skipped as well)
Signed-off-by: Eric Sandeen <sandeen@redhat.com>
---
V2: don't rearrange clear_nlink, and edit commit message.
diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index 00beb4f..1e0c3bf 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -886,7 +886,7 @@ got:
ext4_handle_sync(handle);
if (insert_inode_locked(inode) < 0) {
err = -EINVAL;
- goto fail_drop;
+ goto fail_put;
}
spin_lock(&sbi->s_next_gen_lock);
inode->i_generation = sbi->s_next_generation++;
@@ -950,6 +950,7 @@ fail_drop:
inode->i_flags |= S_NOQUOTA;
clear_nlink(inode);
unlock_new_inode(inode);
+fail_put:
iput(inode);
brelse(inode_bitmap_bh);
return ERR_PTR(err);
^ permalink raw reply related [flat|nested] 12+ messages in thread
* Re: [PATCH 4/6] jffs2: fix up error handling for insert_inode_locked
2011-12-06 23:06 ` [PATCH 4/6] jffs2: " Eric Sandeen
@ 2011-12-08 21:54 ` Artem Bityutskiy
0 siblings, 0 replies; 12+ messages in thread
From: Artem Bityutskiy @ 2011-12-08 21:54 UTC (permalink / raw)
To: Eric Sandeen; +Cc: linux-fsdevel@vger.kernel.org, linux-mtd
On Tue, 2011-12-06 at 17:06 -0600, Eric Sandeen wrote:
> after 250df6ed274d767da844a5d9f05720b804240197
> (fs: protect inode->i_state with inode->i_lock), insert_inode_locked()
> no longer returns the inode with I_NEW set on failure. However,
> the error handler still calls unlock_new_inode() on failure,
> which does a WARN_ON if I_NEW is not set, so any failure spews
> a lot of warnings.
>
> We can just drop the unlock_new_inode() if insert_inode_locked()
> fails here.
>
> Signed-off-by: Eric Sandeen <sandeen@redhat.com>
Pushed to l2-mtd-2.6.git tree, thanks!
Artem.
^ permalink raw reply [flat|nested] 12+ messages in thread
end of thread, other threads:[~2011-12-08 21:54 UTC | newest]
Thread overview: 12+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2011-12-06 22:55 [PATCH 0/6] fs: fix up error handlers for insert_inode_locked Eric Sandeen
2011-12-06 22:57 ` [PATCH 1/6] ext2: fix up error handling " Eric Sandeen
2011-12-06 23:18 ` [PATCH 1/6 v2] " Eric Sandeen
2011-12-06 22:59 ` [PATCH 2/6] ext3: " Eric Sandeen
2011-12-06 23:19 ` [PATCH 2/6 V2] " Eric Sandeen
2011-12-06 23:02 ` [PATCH 3/6] ext4: " Eric Sandeen
2011-12-06 23:19 ` [PATCH 3/6 V2] " Eric Sandeen
2011-12-06 23:06 ` [PATCH 4/6] jffs2: " Eric Sandeen
2011-12-08 21:54 ` Artem Bityutskiy
2011-12-06 23:10 ` [PATCH 5/6] jfs: " Eric Sandeen
2011-12-06 23:12 ` [PATCH 6/6] reiserfs: fix up error handling for insert_inode_locked4 Eric Sandeen
2011-12-06 23:13 ` [PATCH 0/6] fs: fix up error handlers for insert_inode_locked Eric Sandeen
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).